t h e ox f o r d h a n d b o o k o f

PL ATO
the oxford handbook of

PLATO
second edition

Edited by
GAIL FINE

1
 1
Oxford University Press is a department of the University of Oxford. It furthers
the University’s objective of excellence in research, scholarship, and education
by publishing worldwide. Oxford is a registered trade mark of Oxford University
Press in the UK and certain other countries.

Published in the United States of America by Oxford University Press

198 Madison Avenue, New York, NY 10016, United States of America.

© Oxford University Press 2019

All rights reserved. No part of this publication may be reproduced, stored in

a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by license, or under terms agreed with the appropriate reproduction
rights organization. Inquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above.

You must not circulate this work in any other form

and you must impose this same condition on any acquirer.

Library of Congress Control Number: 2019948403

ISBN 978–0–19–063973–0

1 3 5 7 9 8 6 4 2
Printed by Sheridan Books, Inc., United States of America
 Preface to the Second Edition

This second edition of the Oxford Handbook of Plato differs in two main ways from
the first edition. First, six leading scholars of ancient philosophy have contributed
entirely new chapters: Hugh Benson on the Apology, Crito, and Euthyphro; James Warren
on the Protagoras and Gorgias; Lindsay Judson on the Meno; Luca Castagnoli on the
Phaedo; Susan Sauvé Meyer on the Laws; and David Sedley on Plato’s theology. The
second edition therefore covers both dialogues and topics in more depth than the first
edition did. Second, most of the original chapters have been revised and updated, some
in small, others in large, ways. The Introduction has been revised to reflect these changes
from the first edition. The Bibliography has also been updated.
I am grateful to Peter Ohlin for inviting me to do a second edition. I also grateful to all
the authors for their contributions and their collegiality. Thanks too to Peter Osorio for
help with the indexes.
 Table of Contents

Contributorsxi

1. Introduction 1
Gail Fine

2. Plato in His Time and Place 41

Malcolm Schofield

3. The Platonic Corpus 69

T. H. Irwin

4. Plato’s Ways of Writing 93

Mary Margaret McCabe

5. Euthyphro, Apology, and Crito: The Examined and Virtuous Life 119
Hugh H. Benson

6. The Protagoras and Gorgias141

James Warren

7. The Meno161
Lindsay Judson

8. The Phaedo on Philosophy and the Soul 183

Luca Castagnoli

9. The Republic207
Dominic Scott

10. Plato’s Parmenides: A Reconsideration of Forms 231

Sandra Peterson

11. The Theaetetus261

Mi-Kyoung Lee
viii   table of contents

12. The Timaeus on the Principles of Cosmology 287

Thomas Kjeller Johansen

13. The Sophist on Statements, Predication, and Falsehood 309

Lesley Brown

14. The Philebus337

Constance C. Meinwald

15. Plato’s Laws359

Susan Sauvé Meyer

16. The Epistemology and Metaphysics of Socrates 379

Gareth B. Matthews

17. Socratic Ethics and Moral Psychology 403

Daniel Devereux

18. Plato’s Epistemology 429

C. C. W. Taylor

19. Plato’s Metaphysics 455

Verity Harte

20. Plato’s Philosophy of Language 481

Paolo Crivelli

21. Plato on the Soul 507

Hendrik Lorenz

22. Plato’s Ethics 531

Julia Annas

23. Plato on Love 551

Richard Kraut

24. Plato’s Politics 575

Christopher Bobonich

25. Plato on Education and Art 605

Rachana Kamtekar

26. Plato’s Theology 627

David Sedley
 table of contents   ix

27. Plato and Aristotle in the Academy 645

Christopher Shields

28. Plato and Platonism 669

Charles Brittain

Bibliography697
Index Locorum727
Index Nominum756
General Index764
 Contributors

Julia Annas is Regents Professor of Philosophy at the University of Arizona. Her many
books include Aristotle’s Metaphysics M and N (Clarendon, 1976), An Introduction to
Plato’s Republic (Clarendon, 1981; 2nd ed., 1984), The Morality of Happiness (Clarendon,
1993), Platonic Ethics Old and New (Cornell University Press, 1999), A Very Short
Introduction to Plato (Oxford University Press, 2003), and Virtue and Law in Plato and
Beyond (Oxford University Press, 2017).
Hugh H. Benson is Emeritus George Lynn Cross Research Professor, Samuel Roberts
Noble Presidential Professor, and former Chair of the Department of Philosophy at the
University of Oklahoma. He is the editor of Essays on the Philosophy of Socrates (Oxford
University Press, 1992) and Blackwell Companion to Plato (Blackwell, 2006), and the
author of Socratic Wisdom (Oxford University Press, 2000), Clitophon’s Challenge
(Oxford University Press, 2015), and various articles on the philosophy of Socrates,
Plato, and Aristotle.
Christopher Bobonich is C.I. Lewis Professor of Philosophy at Stanford University. He
is the author of Plato’s Utopia Recast: His Later Ethics and Politics (Clarendon, 2002), as
well as of various articles in ancient ethics, political philosophy, and psychology. He is
also co-editor (with Pierre Destrée) of Akrasia in Greek Philosophy: From Socrates to
Plotinus (Brill, 2007), and editor of Plato’s Laws: A Critical Guide (Cambridge University
Press, 2010) and of The Cambridge Companion to Ancient Ethics (Cambridge University
Press, 2017).
Charles Brittain is Susan Linn Sage Professor of Philosophy and Humane Letters at
Cornell University. He is the author of Philo of Larissa (Clarendon, 2001) and Cicero:
On Academic Scepticism (Hackett, 2006); co-translator (with Tad Brennan) of Simpli-
cius: On Epictetus’ Handbook (Duckworth, 2002); and co-editor (with Rachel Barney
and Tad Brennan) of Plato and the Divided Self (Cambridge University Press, 2012).
Lesley Brown is Fellow in Philosophy Emeritus at Somerville College, Oxford. She has
published several articles on Plato’s Sophist, as well as papers on ancient philosophy of
language and on moral and political philosophy in Plato and Aristotle.
Luca Castagnoli is Associate Professor of Ancient Greek Philosophy in the University
of Oxford and Stavros Niarchos Foundation Clarendon Fellow at Oriel College. He is
the author of Ancient Self-Refutation (Cambridge University Press, 2010) and of a num-
ber of articles on ancient philosophy, especially logic and epistemology.
xii   contributors

Paolo Crivelli (MA, University of Florence; Perfezionamento (PhD), Scuola Normale

Superiore of Pisa) is Ordinary Professor at the University of Geneva. He has also been
a lecturer and senior lecturer at the University of Edinburgh, and a Tutorial Fellow at
New College, Oxford. He is the author of Aristotle on Truth (Cambridge University
Press, 2004), of Plato’s Account of Falsehood. A Study of the Sophist (Cambridge Univer-
sity Press, 2012), and of many articles in ancient philosophy.
Daniel Devereux is Emeritus Professor of Philosophy at the University of Virginia.
He has published many articles on Plato and Aristotle, especially on their ethics and
metaphysics.
Gail Fine is Professor of Philosophy Emerita at Cornell University, Senior Research
Fellow at Merton College, Oxford, and Visiting Professor of Ancient Philosophy at the
University of Oxford. She is the author of On Ideas: Aristotle’s Criticism of Plato’s Theory
of Forms (Clarendon, 1993), Plato on Knowledge and Forms: Selected Essays (Clarendon,
2003), and The Possibility of Inquiry: Meno’s Paradox from Socrates to Sextus (Oxford
University Press, 2014), and the editor of Plato 1: Metaphysics and Epistemology and of
Plato 2: Ethics, Politics, Religion, and the Soul (Oxford University Press, 1999), both in
the Oxford Readings in Philosophy series.
Verity Harte is George A. Saden Professor of Philosophy and Classics at Yale University.
She is the author of Plato on Parts and Wholes: the Metaphysics of Structure (Clarendon,
2002) and of the various articles on ancient philosophy. She is also co-editor (with MM
McCabe, Robert W. Sharples and Anne Sheppard) of Aristotle and the Stoics Reading
Plato (BICS, 2010); (with Melissa Lane) of Politeia in Greek and Roman Philosophy
(Cambridge University Press, 2013); and (with Raphael Woolf) of Rereading Ancient
Philosophy: Old Chestnuts and Sacred Cows (Cambridge University Press, 2017).
T. H. Irwin is Professor of Ancient Philosophy Emeritus in the University of Oxford
and an Emeritus Fellow of Keble College. From 1975 to 2006 he taught at Cornell Uni-
versity. His many books include Plato’s Moral Theory: The Early and Middle Dialogues
(Clarendon, 1977), Plato’s Gorgias (translation and notes) (Clarendon, 1979); Aristotle’s
First Principles (Clarendon, 1988); Classical Thought (Oxford University Press, 1989);
Aristotle’s Nicomachean Ethics (translation and notes) (Hackett, 1999); Plato’s Ethics
(Oxford University Press, 1995); and The Development of Ethics, 3 vols. (Oxford Univer-
sity Press, 2007–2009).
Thomas K. Johansen is Professor of Philosophy at the University of Oslo. He previously
taught at the University of Oxford, the University of Edinburgh, and Bristol University.
He is the author of Aristotle on the Sense-Organs (Cambridge University Press, 1998),
Plato’s Natural Philosophy: A Study of the Timaeus-Critias (Cambridge University Press,
2004), and Powers of Aristotle’s Soul (Oxford University Press, 2012).
Lindsay Judson is Associate Professor of Philosophy in the University of Oxford and an
Official Student of Christ Church. He is the General Editor of the Clarendon Aristotle
Series and of the Oxford Aristotle Studies. He is the author of Aristotle: Metaphysics Λ:
 contributors   xiii

A Translation and Commentary (Clarendon, 2018), the editor of Aristotle’s Physics:

A Collection of Essays (Clarendon, 1991), and the co-editor (with Vassilis Karasmanis)
of Remembering Socrates: Philosophical Essays (Clarendon, 2006). He is currently working
on a book on Plato’s Euthyphro.
Rachana Kamtekar is Professor of Philosophy and Classics at Cornell University. She
works in ancient philosophy and moral psychology. Her book Plato’s Moral Psychology:
Intellectualism, the Divided Soul, and the Desire for Good was published by Oxford in 2017.
She is currently working on a project on human agency and cause from Aristotle to
Alexander.
Richard Kraut is Charles and Emma Morrison Professor in the Humanities at North-
western University. His many books include Socrates and the State (Princeton University
Press, 1984), Aristotle on the Human Good (Princeton University Press, 1989), Aristotle
Politics Books VII and VIII, translation with commentary (Clarendon, 1997), Aristotle:
Political Philosophy (Oxford University Press, 2002), What Is Good and Why: The Ethics
of Well-Being (Harvard University Press, 2007), and Against Absolute Goodness (Oxford
University Press, 2011). He also edited The Cambridge Companion to Plato (Cambridge
University Press, 1992), Critical Essays on Plato’s Republic (Rowman & Littlefield, 1997),
and The Blackwell Guide to Aristotle’s Nicomachean Ethics (Blackwell, 2006).
Mi-Kyoung Lee is Associate Professor of Philosophy at the University of Colorado
at Boulder. She is the author of Epistemology after Protagoras: Responses to Relativism
in Plato, Aristotle, and Democritus (Oxford University Press, 2005), and the editor of
Strategies of Argument: Essays in Ancient Ethics, Epistemology, and Logic (Oxford Uni-
versity Press, 2014).
Hendrik Lorenz is Professor of Philosophy at Princeton University. He is the author of
The Brute Within: Appetitive Desire in Plato and Aristotle (Clarendon, 2006) and of
­several articles on Plato and Aristotle.
Gareth B. Matthews was Professor of Philosophy at the University of Massachusetts at
Amherst from 1969 to 2005, and Professor of Philosophy Emeritus at the University of
Massachusetts at Amherst from 2005 until his death in 2011. He previously taught at the
University of Virginia and the University of Minnesota. He is the author of Thought’s
Ego: Augustine and Descartes (Blackwell, 1992), Socratic Perplexity and the Nature of
Philosophy (Oxford University Press, 1999), and Augustine (Wiley, 2005).
Mary Margaret Mccabe is Professor of Ancient Philosophy Emerita at King’s College
London. She is the author of Plato’s Individuals (Princeton University Press, 1994),
Plato and His Predecessors: The Dramatisation of Reason (Cambridge University Press,
1996) and Platonic Conversations (Oxford University Press, 2015). She is co-editor (with
C. Gill) of Form and Argument in Late Plato (Oxford University Press, 1996) and (with
Verity Harte, Robert W. Sharples, and Anne Sheppard) of Aristotle and the Stoics Read-
ing Plato (BICS, 2010). She is also the general editor of the series Cambridge Studies in
the Dialogues of Plato.
xiv   contributors

Constance C. Meinwald is Professor of Philosophy at the University of Illinois at

­Chicago. She is the author of Plato’s Parmenides (Oxford University Press, 1991) and of
Plato (Routledge, 2016), as well as of a number of articles on ancient philosophy.
Susan Sauvé Meyer is Professor of Philosophy at the University of Pennsylvania.
­Educated at the University of Toronto (BA 1982) and Cornell University (PhD 1987), she
taught at Harvard University before joining the faculty at the University of Pennsylvania
in 1994. She is the author of Aristotle on Moral Responsibility (Oxford University Press,
1993; reissued in 2011), Ancient Ethics (Routledge, 2008), and Plato: Laws 1 and 2 in the
Clarendon Plato Series (2015). She is currently an editor of the journal Archiv für
Geschichte der Philosophie.
Sandra Peterson is retired from her position as Professor of Philosophy at the Univer-
sity of Minnesota, Twin Cities. She is the author of Socrates and Philosophy in the
­Dialogues of Plato (Cambridge University Press, 2011) and of several articles on ancient
philosophy. Her essay on the Parmenides for this volume is her seventh essay on that
dialogue.
Malcolm Schofield is Emeritus Professor of Ancient Philosophy in the University of
Cambridge, where he taught in the Classics Faculty for close on 40 years. He is a Fellow
of St John’s, and a Fellow of the British Academy. He has worked in many areas of the
subject, mostly over the last two or three decades on ancient political philosophy, where
with Christopher Rowe he co-edited The Cambridge History of Greek and Roman
Political Thought (Cambridge University Press, 2000). He is the author of Plato: Politi-
cal Philosophy (Oxford University Press, 2006). He is currently working on Cicero’s
political philosophy.
Dominic Scott is Professor of Philosophy at the University of Oxford and a fellow of
Lady Margaret Hall. He previously taught at Cambridge University and the University
of Virginia. He is the author of Recollection and Experience: Plato’s Theory of Learning
and Its Successors (Cambridge University Press, 1995), Plato’s Meno (Cambridge Uni-
versity Press, 2006), and Levels of Argument: A Comparative Study of Plato’s Republic
and Aristotle’s Nicomachean Ethics (Oxford University Press, 2015), and the editor of
Maieusis: Essays in Honour of Myles Burnyeat (Oxford University Press, 2007) and of
The Pseudo-Platonic Seventh Letter, by Myles Burnyeat and Michael Frede (Oxford
University Press, 2015).
David Sedley (born 1947; BA Trinity College Oxford, 1969; PhD University College
London, 1974) taught 1975 to 2014 at the University of Cambridge, where he was
­Laurence Professor of Ancient Philosophy from 2000 and remains a Fellow of Christ’s
College. He has edited The Classical Quarterly (1986–1992) and Oxford Studies in
Ancient Philosophy (1998–2007). His books include Plato’s Cratylus (Cambridge
University Press, 2003), The Midwife of Platonism: Text and Subtext in Plato’s Theaetetus
(Oxford University Press, 2004), and Creationism and Its Critics in Antiquity (University
of California Press, 2007).
 contributors   xv

Christopher Shields is Shuster Professor of Philosophy at the University of Notre

Dame, and an Honorary Research Fellow of Lady Margaret Hall, University of Oxford.
He is the author of Order in Multiplicity: Homonymy in the Philosophy of Aristotle (Oxford
University Press, 1999); Classical Philosophy: A Contemporary Introduction (Routledge,
2003); Aristotle (Routledge, 2007); Ancient Philosophy: A Contemporary Introduction
(Routledge, 2011); with Robert Pasnau, The Philosophy of Thomas Aquinas (Westview,
2003; 2nd rev. ed. Oxford University Press, 2015); and Aristotle’s De Anima, Translated
with Introduction and Commentary (Oxford University Press, 2016). He is the editor of
The Blackwell Guide to Ancient Philosophy (Blackwell, 2002), The Oxford Handbook of
Aristotle (Oxford University Press, 2012), and general editor (with Becko Copenhaver)
of the six-volume The History of the Philosophy of Mind (Routledge, 2018).
C. C. W. Taylor is Emeritus Professor of Philosophy in the University of Oxford and
an Emeritus Fellow of Corpus Christi College. He is the author of Plato, Protagoras,
translated with notes (Clarendon, 1976, 2nd ed. 1991); The Greeks on Pleasure (with
J. C. B. Gosling) (Clarendon, 1982); Socrates (Oxford University Press, 1998); The Atom-
ists: Leucippus and Democritus: Fragments, a text and translation with a commentary
(University of Toronto Press, 1999); Aristotle, Nicomachean Ethics II–IV, translated
with commentary (Clarendon, 2006), and Pleasure, Mind and Soul Selected Papers in
Ancient Philosophy (Oxford University Press, 2008). His translation of S. Luria’s edition
of the fragments and testimonia of Democritus is available online at https://oxford.
academia.edu/ChristopherTaylor.
James Warren is Professor of Ancient Philosophy at the University of Cambridge and a
Fellow of Corpus Christi College. He is the author of Epicurus and Democritean Ethics:
An Archaeology of Ataraxia (Cambridge University Press, 2002), Facing Death: Epicu-
rus and His Critics (Oxford University Press, 2004), Presocratics (Routledge, 2007), The
Pleasures of Reason in Plato, Aristotle, and the Hellenistic Hedonists (Cambridge Univer-
sity Press, 2014), the editor of The Cambridge Companion to Epicureanism (Cambridge
University Press, 2009, and the co-editor (with Frisbee Sheffield) of The Routledge
Companion to Ancient Philosophy (Routledge, 2014).
 chapter 1

I n troduction

Gail Fine

1. Introduction

This volume falls into four parts. Chapters 2–4 discuss preliminaries to the philosophical
study of Plato. Chapters 5–15 discuss individual dialogues. Chapters 16–26 consider
central themes in Plato’s work. Chapters 27–28 explore Plato’s legacy.
A handbook on Plato could be organized in different ways. One might have chapters
just on individual dialogues, just on particular topics, or chapters of both sorts. This
volume favors the last of these three options. This makes the volume richer and more
varied than it would otherwise have been, providing different angles from which to view
Plato’s multi-textured thought. Each dialogue is an integral whole and should be read as
such, with proper attention to and appreciation of its overall structure and the intercon-
nections among its various themes and arguments; one also needs to pay attention to the
dialectical and dramatic context. If one focuses just on what is said on a given topic,
abstracting it from its context, one runs the risk of misinterpretation.1 On the other
hand, Plato discusses the same topics in many dialogues. Some dialogues seem to have
the same, or similar, views; by considering them together, we can paint a fuller picture of
Plato’s thought.
For example, the Phaedo, Symposium, Republic, and Phaedrus (all of which are generally
classified as middle dialogues2) discuss forms (e.g., the form of equal, the form of beauty)
in roughly similar ways, though each of these dialogues also says something that adds to
what is said in the others. When these discussions are read together, we gain deeper
insight into what Plato might have had in mind. Other dialogues sometimes seem to

1 Of course, some dialogues are more tightly interwoven than others. Nonetheless, the default
assumption should be that dialogues should be read as wholes, not merely piecemeal. For an
example of how the dialogues interweave different themes, see the beginning of Section 8 of this
chapter, on the Republic.
2 On the chronology of the dialogues, see Irwin, chapter 3 of this volume.
2   Gail Fine

express different views about forms, ones that are sometimes thought to be incompatible
with the middle dialogues’ view of them. This might be evidence of Plato’s development,
though whether his views develop and, if they do, how they do so, are matters of contro-
versy. For example, the Parmenides criticizes a theory of forms. Some commentators
think it criticizes views about forms to be found in the middle dialogues: views that,
according to some commentators, do not reappear after the Parmenides. On another
interpretation, the Parmenides deliberately distorts the middle dialogues’ views about
forms, so that we can see how not to understand them. Others argue that the middle
dialogues don’t have a fully determinate theory of forms; the Parmenides guards against
one way of making their views more determinate.3 Similar remarks apply, mutatis
mutandis, to what the dialogues say about, for example, epistemology, ethics, politics,
and the soul.
Though this volume contains chapters on both dialogues and topics, not every dialogue
or topic receives its own chapter: doing that would have made an already long volume
far too long. However, some dialogues that don’t receive their own chapters are consid-
ered in the topical chapters. For example, Devereux and Annas discuss the Euthydemus;
Kraut discusses the Symposium; Crivelli discusses the Cratylus. Some dialogues both
receive their own chapters and are also considered, often from different points of view
and with a different focus, in the topical chapters. For example, the Protagoras and
Gorgias are discussed not only in Warren’s chapter on those two dialogues, but also in
Devereux’s more general chapter on Socratic ethics and moral psychology. The Meno is
discussed not only in Judson’s chapter on that dialogue, but also by Matthews and Taylor.
The Phaedo is discussed not only in Castagnoli’s chapter on that dialogue, but also by
Taylor (epistemology), Harte (metaphysics), and Lorenz (the soul). The Republic is dis-
cussed not only in Scott’s chapter on that dialogue, but also in many other chapters. Lee’s
chapter focuses on the Theaetetus; Taylor discusses the Theaetetus in the broader context
of Plato’s epistemology in general. Brown’s chapter is devoted to the Sophist, a dialogue
that is also discussed by Crivelli in his more general chapter on Plato’s philosophy of
language. The Laws is discussed not only in its own chapter, by Meyer, but also by
Bobonich (political theory), Kamtekar (education), and Sedley (theology). I hope that
the fact that some chapters discuss the same dialogues and issues, sometimes from
different points of view, or within different contexts, or by focusing on different parts,
will afford the reader a deeper insight into Plato’s thought than would be possible in a
volume that included chapters only on topics or only on individual dialogues.

3 Here the Timaeus is one relevant dialogue, but there’s dispute about whether it’s a middle or late
(i.e., post-Parmenides) dialogue. Its views about forms are also disputed. On one interpretation, it has
the same views about forms as those expressed in the middle dialogues. If this view is right, and if the
Timaeus is a late dialogue, then at least one late dialogue doesn’t express new views about forms.
However, some think the Timaeus has a different view of forms from those to be found in the middle
dialogues, one that is responsive to the Parmenides’ criticisms.
 Introduction   3

In the rest of this introduction, I provide an overview of the dialogues and topics
discussed in subsequent chapters.4

2. Preliminaries

In chapter 2, Schofield locates Plato in his place and time. Plato was influenced by earlier
philosophers (the Presocratics, or early Greek philosophers), Greek drama, historians
and historical events, the medical writers, and more. Schofield discusses some of the
philosophical and non-philosophical influences on Plato; he also discusses Plato’s life.
In chapter 3, Irwin discusses various features of the Platonic corpus: how the dialogues
survived from Plato’s time to our own, how the earliest (Academic and Alexandrian)
editions of his work came into existence, how and when the dialogues came to be arranged
in tetralogies, and the order of the dialogues. According to the ­standard view, the dialogues
may be divided into early (or Socratic), middle, and late dialogues. Though the standard
view has been challenged, Irwin defends it. However, as he notes, the relative dates of
some dialogues are more controversial than those of others.
Acceptance of the standard view of the chronology of Plato’s works is neutral as
between “developmentalism” and “unitarianism”: to suggest a given order of the dialogues
says nothing about how, if at all, Plato’s thought develops. However, on one familiar
view, Plato’s early dialogues represent the thought of the historical Socrates (as well as
Plato’s own first thoughts), whereas the middle and late dialogues develop Plato’s more
independent views.5 This is so in two ways. First, the middle and late dialogues engage
in systematic discussion of issues that the early dialogues do not discuss in detail, such
as metaphysics and epistemology. Second, in some cases they defend different views
from those to be found in the early dialogues. For example, from the Gorgias and
Republic on, Plato countenances non-rational desires, which are not countenanced in
earlier dialogues.6 Or again, the middle dialogues articulate views about forms that are

4 However, given the limitations of space, I discuss some dialogues and topics in more detail than
others; nor do I always follow the order of the chapters. The reader is warned that I sometimes defend
views at odds with those defended in one or another chapter. Further, some authors disagree with one
another. I have not noted all such disagreements. In places I have adapted the introductions to my Plato
(Oxford: Oxford University Press, 2000), in the Oxford Readings in Philosophy series. This is a
one-volume version of Plato 1: Metaphysics and Epistemology (1999) and Plato 2: Ethics, Politics,
Religion, and the Soul (1999).
5 Hence some commentators contrast Socrates’ and Plato’s views, meaning thereby the views the
character Socrates expresses in the early dialogues (which some think Plato also accepted at this time),
on the one hand, and the views the character Socrates (or the leading character: Socrates is not the
main speaker in all of the late dialogues) in the middle and late dialogues, on the other hand.
6 This assumes that the Gorgias is written after the Protagoras, a common but not universally
accepted view. The view that the early dialogues don’t countenance non-rational desires, though
common, is disputed; see e.g. Devereux, chapter 6 of this volume. The view that the Gorgias
countenances non-rational desires is also controversial; see e.g. Warren, chapter 6 of this volume.
4   Gail Fine

at least not explicit in earlier dialogues. The late dialogues, in turn, are often thought to
suggest yet a different view of forms.
In chapter 4, McCabe discusses Plato’s various ways of writing. She asks to what extent
the pictures in the dialogues (what is said) are affected by their frames (the setting in
which something is said). She argues that even if one is primarily concerned with
philosophical argument, one can’t afford to ignore the frames, for they are actually part
of the pictures. Her argument is developed not just in the abstract but also by attention
to literary and dramatic details that influence our understanding of particular arguments.
She suggests that, while the dialogues are not dogmatic in the sense of claiming to
present the final truth, subject to no revision, neither are they merely exploratory in the
sense of articulating views to which Plato is not at all committed; nor are they intended
to convey a hidden message. However, though McCabe thinks Plato’s lack of dogmatism
explains some aspects of his ways of writing—such as the puzzling nature of the endings
of some of the dialogues, and the tentative way in which Socrates sometimes declares his
commitments7—she also argues that it does not fully account for the multifarious forms
of the dialogues, for which she provides yet further explanations.

3. Apology, Crito, Euthyphro

In chapter 5, Benson discusses the Apology, Crito, and Euthyphro. These three dialogues
are generally thought to be among Plato’s earliest. They are linked both chronologically
and dramatically. In the Euthyphro, Socrates is on his way to trial for impiety and cor-
rupting the young; the Apology describes his trial and conviction; the Crito describes
one of his days in prison after he was convicted.8 (The Phaedo describes his last day and
his death. Despite the dramatic continuity, the Phaedo is usually thought to be from
Plato’s middle period.)
As is well known, Socrates often disclaims knowledge. For example, at Apology 21d he
says that “neither of us knows (eidenai) anything fine and good (kalon kagathon).”
However, sometimes he claims to have knowledge, including, it seems, knowledge of
what’s fine and good. For example, at Apology 29b he says: “I know (oida), however, that
it is wicked and shameful to do wrong, to disobey one’s superior, be he man or god.”
Socrates’ account of his cognitive condition therefore seems to be inconsistent.
There are alternatives to the view that he is inconsistent: for example, that he is insincere
(or ironical) in denying that he has knowledge; that he claims to know in one sense of the
term but not in another; that he claims to have one kind of knowledge but to lack another
kind of knowledge, in a single sense of the term; that he thinks he knows some things but

7 See also Castagnoli, who emphasizes how tentative Socrates often is in the Phaedo.
8 The fact that these three dialogues have this dramatic order doesn’t imply that they were written in
that order. For a dating of the dialogues based on the dramatic events they depict, see D. Nails, The
People of Plato (Indianapolis: Hackett, 2002).
 Introduction   5

not others, in a single sense of the term.9 On one version of the last of these views, he
takes himself to know some particular truths, including some moral truths, but thinks
he lacks knowledge of what, for example, virtue is: he can’t provide satisfactory answers
to his characteristic “What is F?” question (the nature of which is discussed below).
As against this, however, it has been argued that Socrates accepts the Priority of
Knowledge of a Definition (PKD), according to which one cannot know anything about
F unless one knows what F is.10 If he accepts PKD in the Apology, he can’t consistently
maintain there both that he knows some particular moral truths and also that he doesn’t
know what virtue is. In chapter 16, however, Matthews argues that Socrates is not com-
mitted to PKD in the early dialogues. If Matthews is right (a matter about which there’s
dispute), Socrates’ failure to know what virtue is doesn’t imply that he doesn’t know any
particular moral truths.
Benson and Taylor favor a different view, according to which Socrates recognizes two
kinds of knowledge: high- and low-level (Benson), or expert and non-expert (Taylor);
Socrates disclaims the first sort of knowledge but takes himself to have some of the second
sort of knowledge. But, Benson argues, though this solution allows us to accommodate
many of Socrates’ otherwise seemingly contradictory avowals and disavowals, it doesn’t
clearly accommodate all of them. Be that as it may, and contrary to what is sometimes
said, Socrates at least isn’t committed to the claim that he knows that he knows nothing,
where that amounts to a contradiction. Either he knows, in a low-level way, that he lacks
all high-level knowledge; or else he is aware, in a way that falls short of knowing, that he
lacks high-level knowledge.
Despite claiming to lack knowledge, Socrates claims to have “human wisdom”
(Ap. 20d6-e3). This too has been understood in various ways. On one view, his human
­wisdom consists in his not thinking he knows something when he doesn’t (Ap. 21d2–8;
29a5–b5).
The Euthyphro is a classic “elenctic” dialogue: Socrates cross-examines (elenchein) an
interlocutor in an effort to find an answer to a “What is F?” question (in this case, what is
the pious, or the holy: to hosion). A satisfactory answer will specify not only necessary and
sufficient conditions for something’s being F but also the form (eidos, 6d11; idea, 6e1, 4)
of F, which is that by which all F things are F; it is the essence of F-ness, what F-ness
really is.11 Various answers are considered and then rejected. For example, at one point

9 For one classic discussion, see G. Vlastos, “Socrates’ Disavowal of Knowledge,” Philosophical
Quarterly 35 (1985), 1–31, reprinted in Fine (ed.), Plato 1, ch. 2. I give my own account in “Does Socrates
Claim to Know That He Knows Nothing?”, Oxford Studies in Ancient Philosophy 35 (2008), 49–88.
10 Some think PKD is less extensive than it is on my formulation. On one view, for example, Plato
requires knowledge of a definition of F, not for having any other knowledge about F at all, but only for
knowing about difficult cases. It’s important to be clear that the relevant sort of definition, whatever it is
needed for, is a real rather than a nominal definition: it explains what F-ness really is, not our
conventional understanding of F-ness or the ordinary meaning of the term “F.” This explains why
Socrates thinks it is so difficult to acquire knowledge of what F is: though every competent speaker of a
language grasps nominal definitions, they don’t all grasp real definitions.
For a detailed discussion of PKD, see H. Benson, “The Priority of Definition and the Socratic
Elenchus,” Oxford Studies in Ancient Philosophy 8 (1990), 19–65; a revised version is in Benson, Socratic
Wisdom (New York: Oxford University Press, 2000), ch. 6.
11 For discussion of the metaphysics of the early dialogues, see Matthews, chapter 16 of this volume.
6   Gail Fine

Euthyphro suggests that piety is what all the gods love (9e1–3). Socrates doesn’t dispute
that being loved by all the gods is necessary and sufficient for being pious. But he argues
that “being loved by the gods” is not the correct answer to the question “What is piety?.”
For what is pious isn’t pious because the gods love it; rather, the gods love what is pious
because it is pious. The fact that the gods love piety therefore doesn’t tell us what piety is
and so it doesn’t answer the question “What is piety?” (101–11b5). It is tempting to infer
that Socrates also rejects the broader view that moral properties should be defined in
theological terms and, more broadly still, that he defends the autonomy of ethics.12
The Euthyphro is just one of many dialogues in which Plato discusses theology (theo-
logia, a term that first occurs in Republic 2). In chapter 26, Sedley explores Plato’s theol-
ogy, focusing on the Republic, Timaeus, and Laws. He explains that Plato’s commitment
is not so much to monotheism as to the essential unity of the gods; and that fundamen-
tally, god, for Plato, is intelligence (nous), which loves imposing order. Sedley explores
the Timaeus’ explanation of how god does so. But since god, for Plato, is essentially
good, he can’t be the source of evil. How, then, does Plato explain the existence of evil?
Sedley also discusses, among other things, the Laws’ argument for the existence of god;
and the relation between Plato’s scientific theology and traditional views of the gods.
In attempting to answer “What is F?” questions, Socrates practices what has come to
be known as the “elenctic method.” It takes various forms but, in one standard version,
an interlocutor proposes an answer to a “What is F?” question (for example, that piety is
prosecuting the wrongdoer: Eu. 5d8–6e2); Socrates then cross-examines him by asking
about various purported examples (for example, whether there are other cases or kinds
of piety: Eu. 6d6–8) and general principles (for example, whether piety is some one
thing, the same in all cases: Eu. 5c4–d4; 6d9–e1). Eventually the interlocutor is caught in
a contradiction (as Euthyphro is: for he defines piety as prosecuting the wrongdoer,
agrees that there are further cases or kinds of piety, and that piety is some one thing, the
same in all cases). Then, usually, the interlocutor rejects the definition and tries another
one, whereupon the same pattern is generally repeated.13
It’s sometimes thought that Socrates thinks that if one practices this method often
enough, one can gain knowledge of what for example piety is. Yet it seems that all the
elenchus can do is to uncover contradictions among various propositions, or among a
given interlocutor’s beliefs. If an interlocutor has inconsistent beliefs, not all of them can
be true; but that doesn’t, by itself, allow us to know which of them is false. Why, then, is
one proposition, or belief, generally rejected and the others retained? That would be

12 For this view, see S. Marc Cohen, “Socrates on the Definition of Piety: Euthyphro 10a–11b,” Journal
of the History of Philosophy 9 (1971), 1–13; reprinted in G. Vlastos (ed.), The Philosophy of Socrates
(Garden City, NY.: Anchor Books, 1971), ch. 8. Discussion of the “Euthyphro problem” has a long
history. For some discussion, see T. H. Irwin, “Socrates and Euthyphro: The Argument and Its Revival,”
in L. Judson and V. Karasmanis (eds.), Remembering Socrates (Oxford: Clarendon Press, 2006), 58–71.
13 There are variations on this general pattern. For example, in the Laches, when a contradiction is
uncovered among an interlocutor’s beliefs, he rejects a purported example (that lions, despite lacking
knowledge, are brave) rather than the definition (that bravery is knowledge of what is and is not to be
feared): 196e1–197c4. Or again, the Crito practices the elenctic method, not in order to find an answer
to a “What is F?” question but in order to decide a practical question (whether Socrates should flee).
 Introduction   7

justified if they knew that the retained propositions, or beliefs, were true. But given
Socrates’ denial of knowledge, it seems unlikely that he knows, or thinks he knows, which
of them are true.14 What, then, justifies rejecting a particular proposition, or belief? And
how can repeated practice of the method yield knowledge? This is one version of “the
problem of the elenchus.”15 It is essentially what Matthews calls the stronger version:
how can the elenchus be used to acquire knowledge of what F-ness is, when all it seems
capable of doing is to uncover inconsistencies among a given set of propositions, or beliefs?
According to one view, Socrates doesn’t use the elenchus to do more than uncover
inconsistencies; hence the problem doesn’t arise. On another view, he thinks we begin
with knowledge of examples of things that are F, and we then use the elenchus to acquire
knowledge of what F-ness is. This view assumes that Socrates rejects PKD, allowing us to
have some knowledge about F-ness even if we don’t know what F-ness is. On yet another
view, even if we lack knowledge at the outset, the elenchus can help us acquire it, if we
have and rely on enough relevant true beliefs.
In the Crito, the personified Laws explain why Socrates should not flee from prison,
though he is urged to do so by his friend Crito and apparently could easily have done so.
On one view, the dialogue urges absolute submission to the Laws. Yet Socrates says
that one should never act unjustly (48–9). But surely the Laws could require one to do
something unjust? Further, in the Apology, Socrates seems to describe cases where he
did disobey a law or, at least, an order. One solution is to say that the Laws present a
prima facie case for always obeying them—a case that can in principle be overturned,
though in his particular case Socrates doesn’t think it should be.
But why does he think this? After all we’ve seen that he disavows knowledge. He
doesn’t claim to know what justice is, yet he wants to do the just thing: Is it just for him to
flee, or not? He makes the decision to stay, but on what basis, given his lack of knowledge?
Benson addresses these questions.

4. Socratic ethics

Socrates takes an answer to the “What is F?” question to be not only of epistemological
but also of moral importance. For, in his view, knowledge is necessary for virtue: if one
does not know what virtue is, one cannot be a virtuous person.16 Hence failure to answer
the “What is F?” question indicates not just an epistemological but also a moral failing.

14 Interlocutors often begin by thinking they know the answer; but they are quickly shown that they
lack the knowledge they thought they had.
15 G. Vlastos, “The Socratic Elenchus,” Oxford Studies in Ancient Philosophy 1 (1983), 27–58;
reprinted in Fine (ed.), Plato 1, ch. 1.
16 In this section, in speaking of Socrates I mean the Socrates of the early dialogues. For the view
that knowledge is necessary for virtue, see, for example, Eu. 15d4–8; La. 196e; Ch. 176a–b; and
Ly. 212a1–7. It’s sometimes but not always thought that beginning with the Meno, the view that
knowledge is necessary for virtue is rejected in favor of the view that true belief (but not knowledge) is
necessary for virtue. See Judson, chapter 7 of this volume.
8   Gail Fine

Socrates is generally thought to hold that knowledge is not only necessary but also
sufficient for virtue.17 Indeed, he seems to think that virtue just is knowledge—in
particular, knowledge of good and bad.18 This is one of two so-called Socratic paradoxes.
The other is that no one does wrong willingly or voluntarily; I discuss it in the following
paragraph.
If virtue is good and bad, it seems that anyone who knows what it is good—that is,
best overall for oneself—to do will do it (given that the virtuous person acts virtuously).19
Hence anyone who does what is bad must not have known what it was best to do. Indeed,
Socrates holds an even stronger view: not only will anyone who knows that it is best to do
x, do x; if one even believes that it is best to do x, one will do it. This is the second Socratic
paradox: that no one does wrong willingly or voluntarily. There is, then, no such thing as
akrasia (incontinence or weakness of will): no such thing, that is, as knowing, or believing,
that it is better to do x than y, but doing y instead.20 Yet it is often assumed that there is such
a phenomenon. Socrates therefore owes us an explanation of his denial of its possibility.
He offers one in the Protagoras, a dialogue discussed by both Warren and Devereux.21
Before considering it, it will be helpful to lay some groundwork.
Let us say that a rational desire is good-dependent: it is a desire one has because one
believes it is in one’s best overall interest to pursue a given course of action.22 What is
best for one overall is to be eudaimon, which is conventionally translated into English as
“happy.” This translation is liable to mislead, since to a modern ear it suggests feeling
pleased or content, whereas eudaimonia is doing well (eu prattein). Eudaimonia is a
property of a life, not something fleeting; the eudaimon life is the best life possible for
a human being, whatever that turns out to be.
Socrates assumes that a rational desire is one that ultimately aim at one’s own happi-
ness. Since he thinks that all desires ultimately aim at one’s own happiness, he thinks that
all desires are rational. Because he takes all desires to be rational, and so to aim at one’s

17 See, for example, La. 192c2–d11, where Laches suggests that courage is wise endurance; but the
reference to endurance then drops out, though the view that courage is wisdom is retained. Claiming
that knowledge is sufficient for virtue makes it sound as though Socrates thinks that a purely cognitive
condition can be action guiding, contrary to a common view according to which desire is also
necessary for action. However, in the Protagoras (358c–d; cf. Eud. 278e) Socrates says that it is “not in
human nature” to choose what one thinks is bad rather than what is good. Hence his view is that
knowledge is sufficient for virtue given the basic human desire to secure what is good (sc. for oneself).
In what follows, I take this point for granted.
18 See, for example, La. 199d–e.
19 Or, at any rate, she will try to do the virtuous thing. But she might be prevented from doing so by
factors beyond her control.
20 This is just one way of characterizing akrasia. I shall not here distinguish between knowledge and
belief akrasia. For discussion of how Socrates might distinguish them, see Devereux, chapter 17 of this
volume.
21 See also Lorenz’s discussion in chapter 21 of this volume. It’s worth noting that, though Socrates
argues against the possibility of this phenomenon in the Protagoras, he doesn’t there use the term
akrasia. He asks, instead, whether someone can be “overcome by pleasure” (352e6–353a1).
22 The term “good-dependent” (like the term “good-independent,” used later on in this chapter in
connection with the division of the soul in the Republic) is due to T. H. Irwin, Plato’s Moral Theory
(Oxford: Clarendon Press, 1977), 192.
 Introduction   9

own happiness, he—like virtually all Greek moral philosophers—is a eudaimonist.23

Devereux discusses the eudaimonist framework of the early dialogues, and Annas discusses
it as it figures in Plato’s dialogues more generally.
Here it is useful to distinguish between rational and psychological eudaimonism.24
According to rational eudaimonism, happiness provides the only ultimate justifying
reason for doing something: it is the ultimate answer to the question, “Why is x worth
pursuing?.” According to psychological eudaimonism, happiness is the only ultimate
explanatory reason for doing something: it is the ultimate answer to the question, “Why
did you do x?.” The early dialogues assume both sorts of eudaimonism. They assume, that
is, not only that it is rational for me to do something only to the extent that it contributes
to my happiness but also that whatever I do, I do because I believe it will most contribute to
my happiness.
It follows from psychological eudaimonism that no one ever acts against what he
believes it is best for him to do.25 This is sometimes called the prudential version of the
second Socratic paradox. There is also a moral version, according to which no one ever
does moral wrong willingly.26 This version is secured if we add Socrates’ beliefs that
(1) acting unjustly or immorally is bad for the agent, and (2) we only want what is good
for us. It follows from (1) and (2) that if we act unjustly or immorally, we do not want to
act that way, and we therefore do so unwillingly or involuntarily (see Gorgias 509e).27
Though Socrates takes psychological eudaimonism to be a fundamental fact about us,
he doesn’t just leave it there. Rather, he argues that akrasia is impossible. Hence, his own
alternative (that no one does wrong willingly) is either implied or at least rendered more
plausible. In the Protagoras, the argument against akrasia assumes hedonism, the view
that the good—that is, happiness—is the same as pleasure (353c–354e); it also assumes
that we always choose what we take to be the maximum amount of pleasure. If the good
consists in pleasure, and if we always choose what we think will yield the good—which,
according to the sort of hedonism at issue here, is the maximum amount of pleasure—
then, in choosing y over x, we must believe that y will yield more pleasure than x. But in a
case of akrasia, we choose what we take to be less good, though more pleasant. So, in this
alleged case of akrasia, we must think x is better than y. But then, given that the good just

23 The Cyrenaics may be an exception. Further, as we shall see later on in this chapter, it’s sometimes
thought that the Republic is not altogether eudaimonist. Though ancient moral philosophers generally
agree that eudaimonia is in some sense the ultimate end, they disagree about how to achieve it and
about what it consists in.
24 For this terminology, see T. H. Irwin, Plato’s Ethics (Oxford: Oxford University Press, 1995), sect. 36.
25 We’ve seen that Socrates takes virtue to be knowledge of good and bad. We can now say, more
exactly, that it is knowledge of what is good and bad for oneself. Since everyone most desires what is
good for oneself, this makes it clear why anyone who knows what it is best (for oneself overall) to do,
will (try to) do it.
26 For the distinction between the prudential and moral versions of the second Socratic paradox, see
G. Santas, Socrates: Philosophy in Plato’s Early Dialogues (London: Routledge and Kegan Paul, 1979),
183–94.
27 The highly controversial assumption that it is in one’s best overall interest to be just is defended in
detail in the Republic.
10   Gail Fine

is pleasure, it’s as though we chose y over x, thinking y more pleasant than x, but also
thinking x more pleasant than y. This seems to involve inconsistent beliefs: we think
both that y is more pleasant than x and that x is more pleasant than y. We can avoid this
unpalatable result if we assume that, if we choose y over x, it isn’t because of akrasia but
because of a mistaken belief about what would yield the most pleasure. The phenomenon
that some describe as akrasia therefore really just involves false beliefs (e.g., 358c1–5):
there is a purely cognitive failure, not weakness of the will.
The argument against the possibility of akrasia is open to objection. There is also dis-
pute about whether, in the Protagoras, Socrates accepts the hedonism that his argument
against akrasia rests on, or whether the argument is purely ad hominem.28
We saw above that the moral version of the second Socratic paradox assumes that
what is best for one is being morally virtuous. More strongly, Socrates thinks that virtue
is sufficient for happiness. This, too, is a highly controversial claim, one Socrates defends
in the Euthydemus, where he argues that x is either a part of, or necessary for, happiness
if and only if virtue secures it.
The claim that virtue is sufficient for happiness has been understood in two different
ways. On one view, it means that virtue, all by itself, is sufficient for happiness. On another,
weaker, view, it means that virtue is sufficient for happiness only given a sufficient (modest)
amount of certain other goods, such as health. Annas argues that both views sit side by
side throughout the corpus, without ever being clearly distinguished from one another.29
Suppose, however, that virtue is literally sufficient for happiness. We can then ask
whether virtue is sufficient for happiness by being its sole component or by being an
infallible means to it. An analogy will illustrate the difference between these two views.
Milk, flour, and eggs are parts of, ingredients in, a cake; going to the store to buy these
ingredients is an instrumental means of making the cake, but it is not part of the cake.
It is sometimes thought that the fact that Socrates takes virtue to be a craft (technê,
also translated as “skill” or “art”) supports the instrumental view: just as shoemaking is a
craft with the distinct product of shoes, so virtue is a craft with the distinct product of
happiness. But it has been argued that not all crafts have distinct products; music and
dance have been thought to be counterexamples. If this is right, then the mere fact that
Socrates takes virtue to be a craft does not imply that he takes it to be merely an instru-
mental means to happiness. We need to know what sort of craft he takes it to be: one that
has, or lacks, an independent product. Devereux discusses this issue.

28 For the view that Socrates accepts hedonism in the Protagoras, see C. C. W. Taylor, Plato’s
Protagoras, 2nd ed. (Oxford: Clarendon Press, 1991), 162–70; Irwin, Plato’s Moral Theory, ch. 4, and
Plato’s Ethics, ch. 6. Annas doesn’t think Plato is committed to it. Warren remains neutral.
29 On the second, weaker, view, virtue is not literally sufficient for happiness (even given human
nature); rather, it is just especially important to securing happiness. Hence it is better to call the second
view an alternative to the sufficiency thesis rather than a version of it. For the view that virtue is
sufficient for happiness only given a sufficient (modest) amount of other goods such as health, see
G. Vlastos, “Happiness and Virtue in Socrates’ Moral Theory,” in his Socrates, Ironist and Moral
Philosopher (Ithaca, NY: Cornell University Press, 1991), 200–32; and in Fine (ed.), Plato 2, ch. 4. For
the view that virtue is literally sufficient for happiness (given human nature), see Irwin, Plato’s Moral
Theory, esp. ch. 3, and Plato’s Ethics, esp. chs. 3–4.
 Introduction   11

If virtue is the sole component of happiness, we have an account of what Socrates

takes happiness to consist in. If, however, virtue is merely an instrumental means to
happiness, we still lack such an account. If he endorses the hedonism described in the
Protagoras, that would provide such an account; but, as we have seen, it is disputed
whether he endorses it there. Whether or not he endorses hedonism in the Protagoras,
he firmly rejects it by the time of the Gorgias as well as in subsequent dialogues, where he
argues that, though some pleasures—the good ones—are part of the happy life, happiness
does not consist just in pleasure.30

5. Meno

In chapter 7 Judson discusses the Meno,31 a dialogue that is often thought to begin a new
phase in Plato’s thought: the early dialogues are primarily devoted to ethical questions;
though the Meno (like most of Plato’s dialogues) also discusses such questions, it devotes
more attention than earlier dialogues do to questions in epistemology and metaphysics.
It also explicitly addresses, and provides an answer to, the problem of the elenchus
(discussed in Section 3 of this chapter), an answer that goes beyond anything to be found
in earlier dialogues.
The dialogue opens with Meno asking Socrates’ whether virtue can be taught. It is not
entirely clear what Socrates’ answer to this question is. For, as Judson explains, later in
the dialogue Socrates argues both that virtue is teachable and that it isn’t (87c1–96b10),
and there’s dispute about which of these arguments (or conclusions) he favors.32
But, Socrates says at the outset of the dialogue, unless one knows what virtue is, one
can’t know whether it is teachable or, indeed, anything at all about it (71a4–b8). This is an
instance of PKD, which we discussed in Section 3 of this chapter, in connection with
the early dialogues. Socrates claims not to know what virtue is, yet knowing what it is,
is prior to all other knowledge of virtue; hence they spend some time asking what it
is. Though Meno initially thinks he knows the answer, it emerges that he doesn’t;
hence he doesn’t know anything at all about virtue. How, then, can he inquire into it?
This leads Meno to formulate “Meno’s paradox” (80d5–8), which Socrates then refor-
mulates (80e1–5).

30 There is dispute about whether the version of hedonism described in the Protagoras is the same as
the version(s) rejected in other dialogues. For discussion, see Annas; and J. C. B. Gosling and
C. C. W. Taylor, The Greeks on Pleasure (Oxford: Clarendon Press, 1982).
31 The Meno is also discussed by Matthews, chapter 16 and Taylor, chapter 18 of this volume.
32 Though he presumably thinks that virtue either is, or isn’t, teachable (and so accepts the
conclusion of one of his arguments), he could reject both arguments. (Alternatively, one might argue
that in the Meno, he suspends judgment either way.) The question of whether virtue is teachable is also
discussed in the Prot.
12   Gail Fine

Meno is sometimes thought to raise two problems: the problem of inquiry and the
problem of discovery.33 The first asks how, if one lacks knowledge, one can begin an
inquiry; the second asks how, if one lacks knowledge, one can complete an inquiry.
Socrates reformulates Meno’s questions into a constructive dilemma: whether one
knows or doesn’t know, one can’t inquire. There’s dispute about the precise connection
between Meno’s questions and Socrates’ reformulation. There’s also dispute about how
Socrates replies to Meno’s questions and to his own reformulation of them.
Socrates’ reply is in three stages. In the first stage he describes the theory of recollec-
tion, according to which we had prenatal knowledge, and what’s called learning is really
recollection of that knowledge. When Meno professes not to understand, Socrates says
he’ll explain. He does so in the second stage. Here he cross-examines a slave about a
geometry problem whose answer the slave doesn’t know either at the beginning or at the
end of their discussion. He begins with a false belief about the answer (though he has
some related true beliefs). He eventually acquires a true belief about the answer. Socrates
says that the slave still doesn’t know the answer (85c2), though he could come to know it
if he were questioned further (85c10–d1). In the third stage Socrates reiterates the theory
of recollection.34
How should we understand Socrates’ reply? On one view, favored by Matthews and
Taylor, it distinguishes latent innate knowledge from manifest knowledge. We all have
the former but lack the latter; inquiry consists in making our latent innate knowledge
manifest.
It is clear that the theory of recollection posits prenatal knowledge (knowledge we
had before birth), but that doesn’t imply that we have innate knowledge (knowledge
we have when we are born). For one might lose one’s prenatal knowledge on being born,
in such a way that one no longer knows and so lacks innate knowledge.35 And, on one
view, Plato thinks we are born without any knowledge. Nonetheless, we eventually acquire
true beliefs; and by relying on relevant true beliefs, along with our tendency to favor
truths over falsehoods, we are able to acquire knowledge. That’s the point of the discussion
with the slave, who, as we’ve seen, eventually acquires a true belief about the answer to a
geometry problem, even though he doesn’t yet know it (nor does he have other relevant

33 For this view, see D. Scott, Plato’s Meno (Cambridge: Cambridge University Press, 2006), part 2;
and his Recollection and Experience (Cambridge: Cambridge University Press, 1995), 24–52. I discuss
my own view of the Meno in “Inquiry in the Meno,” in R. Kraut (ed.), Cambridge Companion to Plato
(Cambridge: Cambridge University Press, 1992), 220–26, reprinted, with minor modifications, in my
Plato on Forms and Knowledge: Selected Essays (Oxford: Clarendon Press, 2003), ch. 2; and in The
Possibility of Inquiry (Oxford: Oxford University Press, 2014), part I.
34 There’s dispute about whether recollection is described in exactly the same way in the first and
third stages.
35 The Meno doesn’t explicitly say that we lose our prenatal knowledge when we are born, but the
Phaedo does so. In chapter 8, Castagnoli distinguishes different ways of losing something, and he
argues that the way in which Plato thinks we lose knowledge when we are born is compatible with our
still having knowledge in a way; see also Scott, Plato’s Meno and Recollection and Experience. However,
neither in the Meno nor in the Phaedo does Plato explicitly say this, though in both dialogues he does
explicitly say that we lack knowledge.
 Introduction   13

knowledge), though he can come to know it. This is the true-belief response, according
to which Plato argues that one doesn’t need any prior knowledge (in this life) in order to
inquire or discover; having and relying on relevant true beliefs will do.36
Judson argues against this view and in favor of the view that what’s crucial, and
suggested by the passages on recollection, is our ability to recognize the correct answer
when we find it. Since recognition implies prior knowledge, we already know in a way;
but insofar as inquiry involves articulating the answer (which we can’t always do), we
don’t already know.37
The true-belief and recognitional responses are alternative solutions not just to
Meno’s paradox, but also to the problem of the elenchus, which, as we saw in Section 3,
asks (among other things) how we can acquire knowledge when we don’t already have it.
According to the true-belief response, our reliance on relevant true beliefs, coupled with
our tendency to favor truths over falsehoods, enables us to acquire yet further true beliefs
that we can eventually convert into knowledge (when, for example, we can interrelate a
sufficient number of true beliefs into an explanatory whole). According to the recognitional
response, even if we in some sense lack knowledge now, we had it once; when we come
upon the right answer, we’ll recognize it as such and thereby know it.
However that may be, Plato undoubtedly distinguishes knowledge from true belief:
he does so in saying that the slave has true belief but not knowledge. Then, at 97a, he
distinguishes someone who knows the way to Larisa from someone who has a mere
true belief about it. Hence Plato allows us to have knowledge and beliefs about at least
some of the same things: knowledge and belief aren’t individuated by their distinct
objects. He also countenances empirical knowledge.38
How exactly do knowledge and true belief differ? Plato doesn’t provide an explicit
account of true belief or of belief. But at 98a he says that knowledge is true belief tied
down with reasoning about the explanation (aitias logismos). One can know that p is so
only if one knows why p is so; all knowledge requires an account of the reason why what
one knows is true. There’s dispute about whether this is a version of a justified-true-
belief account of knowledge. It isn’t if, according to such an account, any old justification
is sufficient for turning a true belief into knowledge. For Plato thinks knowledge requires,
not just any old justification for believing that p is true, but an explanation of why p is
true. However, if one takes the justified-true-belief account of knowledge to say that, to
know that p, p must be true, one must believe that p is true, and one must have a justification

36 However, as we’ve seen, the theory of recollection posits prenatal knowledge. Why does Plato do
so, according to the true-belief response? Perhaps he thinks that our tendency to favor truths over
falsehoods can be explained only in terms of prior knowledge. Since we don’t have the relevant prior
knowledge in this life, we must have had it prenatally. Though we lose this knowledge at birth, we
retain the tendency it conferred.
37 Judson remains neutral on the question of whether this prior knowledge is innate.
38 This assumes that he mentions the way to Larisa as a literal example of something one might
know. However, it is sometimes thought that he mentions it only as an analogy, designed to explain
how knowledge differs from true belief. Nor is it clear exactly how he distinguishes knowing from
having a mere true belief about the way to Larisa. See Judson, chapter 7 and Taylor, chapter 18 in this
volume for discussion.
14   Gail Fine

for one’s belief that is sufficient for turning it into knowledge, then Plato does have a
version of a justified-true-belief account of knowledge.39

6. Phaedo: soul

In chapter 8, Castagnoli discusses the Phaedo, a dialogue that is generally thought to

belong to Plato’s middle period. The Phaedo is also discussed by Taylor (epistemology),
Harte (metaphysics), and Lorenz (the soul). The dialogue describes Socrates’ last day
and his death. His friends are gathered around him in prison, distressed at the prospect
of his imminent death. To reassure them that they need not be distressed (though the
end of the dialogue suggests he hasn’t fully convinced them), he engages them in discussion
of the immortality of the soul. For if, as he believes, the soul is immortal, then he won’t
die (115).40 And if he’s lived a properly philosophical life, as he has striven to do, it’s
reasonable to think he will have a good afterlife. Several arguments are considered: the
Cyclical Argument (69e6–72e1), the Argument from Recollection (72e3–78b3), the Affinity
Argument (78b4–84b8), and the Final Argument (102a10–107b10).41 Not all of these
arguments claim to prove immortality. The Recollection Argument, for example, aims
to prove just the preexistence of the soul (76; this contrasts with the Meno, where it is
said to prove immortality: 86b). According to the Affinity Argument, the soul is more
like everlasting things than it is like sensibles, and so the soul is likely to be immortal; but
that doesn’t imply that it is immortal. Socrates takes at most the Final Argument to show
this, though even here he remains open-minded. Castagnoli and Lorenz discuss some
of Socrates’ views about the soul in the Phaedo. Castagnoli also discusses Socrates’
attitude to arguments and philosophy, arguing that he is more tentative than is some-
times thought.

7. Phaedo: forms

The other main topic of the Phaedo, apart from the soul, is the theory of forms.42 The
Euthyphro and Meno discuss forms, though only briefly. The Phaedo says much more
about them, and the dialogue is often thought to describe a new view of them, sometimes
called the middle or classical theory of forms.

39 I discuss this issue in “Knowledge and True Belief in the Meno,” Oxford Studies in Ancient
Philosophy 27 (2004), 41–81. For criticism, see Judson chapter 7 of this volume; and W. Schwab,
“Explanation in the Epistemology of the Meno,” Oxford Studies in Ancient Philosophy 48 (2015), 1–36.
40 This assumes that Socrates just is his soul. See Castagnoli chapter 8 of this volume for discussion.
41 They also consider the attunement theory, which aims to show that the soul is not immortal. It is
eventually rejected on the ground that it is incompatible with the theory of recollection, which they
think is better established than the attunement theory: see 87a–88b.
42 F. M. Cornford calls the “immortality and divinity of the rational soul” and the existence of forms
the “two pillars” of Platonism (Plato’s Theory of Knowledge [London: Routledge and Kegan Paul, 1935], 2).
 Introduction   15

We’ve seen that in the early dialogues, a satisfactory answer to the “What is F?” question
says what it is to be F. The correct answer to the question “What is justice?,” for example,
says what justice itself—the nature or real essence of justice—is. In the early dialogues,
Socrates sometimes calls justice, piety, and so on—the referents of correct answers to
“What is F?” questions—forms.43 The form of piety, for example, is that “by which (hô(i))
all pious things are pious” (Eu. 6d10–11), something that “is the same in every [pious]
action” (5d1).
Since a form is some one thing, the same in all cases, it seems to be a universal, in the
Aristotelian sense of being a one over many (De Int. 17a38–b1). Aristotle, however, says,
or comes close to saying, that forms are both universals and particulars (see, e.g., Met. 13. 9).
In chapter 19, Harte suggests that it is difficult to decide about this, partly because Plato
lacks technical terms for “particular” and “universal,” and partly because the contrast
between universal and particular is not one of his central concerns. For example, he often
contrasts forms and sensibles. But this is not the contrast between all universals, on the
one hand, and all particulars, on the other. For there are nonsensible particulars, such as
god and individual souls. There are also sensible or perceivable universals, such as redness
and being 3 inches, which are universals in the sense that they are repeatable, or can be
had or shared by many things: there are many red things and many three-inch-long
things. Plato’s primary concern is not to distinguish universals from particulars but to
argue for the existence of a certain sort of nonsensible entity, the forms. This, by itself,
doesn’t allow us to know whether forms are universals, particulars, or both. In chapter 27,
Shields asks why Aristotle nonetheless claims that forms are, or come close to being, both
universals and particulars.44
Why does Plato posit forms? As Harte and Shields explain, forms have various func-
tional and explanatory roles. Aristotle, for example, says that Plato introduced forms as the
(basic) objects of knowledge and definition because he thought that entities in the sensible
world are in flux or change, which disqualifies them from being (the basic) items of
knowledge and definition. Hence there must be stable objects that can so serve, and
these are the forms (Met. 1.6, 13.4, 13.9).
There is dispute about what, if any, sort of flux or change Plato appeals to in arguing
that there are forms. Plato takes both the compresence and the succession of opposites
to be kinds of flux or change. Compresence obtains when something is both F and not F
at the same time:45 for example, Helen is both beautiful (insofar as she is more beautiful
than other women) and ugly (insofar as she is less beautiful than Aphrodite); bright
color is both beautiful (in this painting) and ugly (in that one). The former is compresence
in a particular (Helen); the latter is compresence in a property or type (bright color).
Succession obtains when something is F at t1, and then ceases to be F, and becomes not-F,
at a later time t2. For example, Helen is first short and then becomes taller as she grows

43 Although eidos is usually rendered as “form” and idea as “idea,” the latter is misleading insofar as
“idea” nowadays suggests some sort of mind-dependence, whereas forms are objectively existing,
mind-independent entities.
44 See also Section 18 of this chapter.
45 Compresence can also be described in temporal terms: Helen is beautiful at t1 (when compared
with me), but ugly at t2 (when compared with Aphrodite).
16   Gail Fine

older. There are also more radical sorts of succession, as encapsulated in Heraclitus’
alleged remark that one can’t step into the same river twice (DK B91), the idea being that
it changes so rapidly that it doesn’t persist over time.
On one view, in the middle dialogues Plato takes the sensible world to undergo the
most radical sort of succession of opposites, according to which each sensible is, at every
moment, changing in every respect. But it has also been argued that Plato consistently
rejects the view that anything, whether sensible or nonsensible, changes in this radical
way. Moreover, though sensibles undoubtedly undergo some sorts of succession—if of
more orderly sorts—it is not clear that that is what motivates the introduction of forms.
Rather, as Harte argues, it is the compresence of opposites, especially in properties or
types, that does so. For Plato accepts the oneness condition: he thinks that beauty, for
example, is some one thing, the same in all cases. Beauty cannot be identified with any
single sensible property; since beauty must be a single, non-disjunctive property, it must
be a nonsensible one, and this is the form of beauty.
Though compresence in sensible Fs is a sufficient reason for positing the existence of a
form of F, it is not clearly necessary. Plato sometimes seems to suggest that there are forms in
any case where perception is inadequate for answering the “What is F?” question, and this
might yield a broader range of forms than is licensed by compresence. Be that as it may,
forms are, at any rate, unobservable entities. This explains why, in the Phaedo, Plato argues
that we can’t acquire knowledge (or, at least, some knowledge) if we rely solely on percep-
tion. For perception has access only to sensibles; but attention solely to them can’t confer
knowledge, since knowledge requires a grasp of forms, which aren’t perceptible (65–67).
Socrates returns to the importance of forms later in the Phaedo, especially in the
passage on aitia (causes or explanations) (95d4–102a9), which begins with a discussion
of his earlier views. This part of the passage is sometimes said to be Socrates’, or Plato’s,
“intellectual autobiography,” though how genuinely autobiographical it is remains open
to dispute. Socrates says that he used to think one could explain various phenomena in
material terms. For example, human beings grow by eating and drinking; one person is
taller than another by a head (96c–e). But he eventually decided that such purported
explanations are not genuine explanations. He would have liked to find teleological
explanations for all phenomena: an explanation of why it’s best that each thing comes to be
as it does and is at is. But he was unable to find such accounts on a broad scale (97c–99d).
Hence he settled on a deuteros plous (99c9–d1), a second sailing, that is, a second way of
finding explanations (including teleological ones). This is the “safe aitia,” according to
which x is F if and only if, and because, it participates in the form of F (if there is one)
(99b–101d). For example, something is beautiful if and only if it participates in the form
of beauty. Socrates also proposes the “clever” (or “sophisticated”) aitia, according to which
x is F if G brings F-ness to x. For example, something is cold if snow brings the form of
coldness to it (102b–105c). Both aitiai invoke forms, but the clever aitia also makes use of
“intermediary” entities such as snow (the physical stuff: not the form of snow).46

46 For discussion of the aitia passage, see G. Vlastos, “Reasons and Causes in the Phaedo,”
Philosophical Review 78 (1969), 291–325; reprinted in Vlastos (ed.), Plato 1, 322–66; J. Annas, “Aristotle
on Inefficient Causes,” Philosophical Quarterly 32 (1982), 311–26; and G. Fine, “Forms as Causes: Plato
 Introduction   17

The view that forms are, or are parts of, (at least some) aitiai, recalls the Meno’s claim
that knowledge is true belief tied down with an aitias logismos. The Phaedo adds that the
relevant aitiai involve forms. Hence knowledge (or, at least knowledge-why, in at least
some cases) requires knowledge of forms. It doesn’t follow, however, that only forms can
be known. Rather, the aitia passage suggests, we know sensibles when we can explain
why they are as they are in the light of forms. Just as the Phaedo allows knowledge of
sensibles, so it also allows mere belief about forms. Hence it is not committed to the
so-called Two Worlds Theory, according to which there is knowledge but not belief
about forms, and belief but not knowledge about sensibles.47
Forms are sometimes thought to be necessary for the possibility not only of (at least
some) knowledge but also of language. On one view, Plato thinks that grasping the
meaning of a general term requires grasping a form; it is also sometimes thought that
forms are the meanings of general terms. If not every meaningful general term has a
corresponding form, then forms are not the meanings of (all) general terms. Nonetheless,
they might be central to thought and language in a different way. In chapter 20, Crivelli
asks whether there is a linguistic dimension to the theory of forms. He also considers
Plato’s views about language more broadly, focusing on the Cratylus and Sophist.
To say that forms are nonsensible properties is to say that they are different from (non-
identical to) sensibles. Are they also separate from them? That is, can they exist whether
or not sensibles do? Difference and separation are quite different: the latter implies the
former but not conversely. For example, I am different from oxygen but couldn’t exist
without it. Aristotle thinks Plato is committed to separation, and he argues that this is
responsible for various difficulties in the theory of forms.48
It is sometimes thought that if forms are separate, they cannot exist in sensibles.
However, separation implies only that it’s possible for forms to exist whether or not any
sensibles have them. That doesn’t preclude immanence. But how can forms be imma-
nent? Is either the whole, or a part, of the form of F in each sensible particular? In the
first part of the Parmenides, Plato considers problems for both options. But, as Peterson
explains in chapter 10, the problems arise because he treats immanence in crudely
physicalistic terms. Perhaps he means us to infer that, on other interpretations, there is
no difficulty in forms being immanent. For example, perhaps forms are in things by
being properties of them. Aristotle canvasses a number of ways in which one thing can
be in another in Physics 4.3; Cat. 1a24–25; and Met. 5.23.
Plato’s metaphysics is not exhausted by his views about forms or by his view that some
sensible properties and particu­­lars suffer compresence, as well as some sorts of succession,
of opposites. For example, we’ve seen that in the Phaedo he sought, but failed to find (at least

and Aristotle,” in A. Graeser (ed.), Mathematics and Metaphysics in Aristotle (Bern: Haupt, 1987),
69–112, reprinted in Plato on Knowledge and Forms, ch. 14. These articles also discuss ways in which
Plato’s account of aitiai anticipate Aristotle’s doctrine of the four causes: the material, formal, final or
teleological, and efficient.
47 I defend this view in “The ‘Two Worlds’ Theory in the Phaedo” in British Journal of the History of
Philosophy 24 (2016), 557–72. See also Taylor, chapter 18 of this volume.
48 For some discussion, see Shields, chapter 27 of this volume, and Section 18 of this chapter.
18   Gail Fine

on a broad scale), teleological explanations. The Republic, however, posits the form of the
good as the basic explanatory principle. This reaffirms Plato’s commitment to teleology,
including natural teleology; but the Republic doesn’t supply many details. This gap is
filled primarily by the Timaeus, which provides a detailed account of the principles that
govern the coming to be of the sensible world; it also describes its nature once it exists.
Insofar as Plato thinks the sensible world and its coming into being can be explained, he
does not think it completely eludes our grasp, as it would if it were in the most extreme
sort of Heraclitean flux described previously. In chapter 13, Johansen explores these and
other aspects of the Timaeus.

8. Republic: ethics

In chapter 9, Scott discusses the Republic, aspects of which are also discussed in many
other chapters. The main ostensible topic of the Republic is the question “What is justice
(dikaiosunê)?,” but the dialogue also discusses many other topics, including the soul,
politics, art, education, knowledge and belief, and forms. These issues are intimately
connected. For example, the dialogue argues that the best polis49—city or state—should
be governed by the best people. The best people are those who are virtuous. Virtue
requires knowledge, and one can have moral knowledge only if one knows forms. Only
philosophers have this knowledge; hence only they should rule. Plato’s political views
therefore rely on his views about ethics, which, in turn, rely on his epistemological and
metaphysical views. Or again, he argues that justice is the dominant component of
happiness (eudaimonia): that is, though it isn’t sufficient for happiness, it makes the
single greatest contribution to our happiness, outweighing every other combination of
goods.50 What happiness for humans consists in depends on what we are like: on the
nature of our souls. So Plato also discusses the nature of the soul. Further, he thinks we can
come closer to happiness by improving ourselves in various ways: by acquiring more
true beliefs and by training our desires. Here education and the arts play important
roles; hence Plato discusses them as well.
We have seen that the early dialogues assume rational eudaimonism: one has reason to
do something only insofar as it contributes to one’s happiness. One therefore has reason
to be just only if that contributes to one’s happiness. The early dialogues assume that
being just contributes to one’s happiness.51 But the assumption is controversial. For justice

49 Polis is variously translated as “city,” “state,” and “city-state.”

50 For this view, see Irwin, Plato’s Moral Theory, ch. 7; and Plato’s Ethics, e.g. sects. 140 and 176. On
this view, Plato rejects the view, often thought to be maintained in the early dialogues, that virtue is
sufficient for happiness. By contrast, in chapter 22 of this volume, Annas argues that the dominant
component view and the sufficiency view sit side by side throughout the Republic, and indeed
throughout Plato’s work.
51 In the Crito, for example, Socrates says that the good life is the same as the just life (48b). What
about wisdom, courage, and temperance? Aren’t they also part of the good life? Yes, but Socrates thinks
 Introduction   19

seems to be other-regarding, in the sense that my justice seems to benefit you. As

Thrasymachus says in Book 1, justice is “another’s good” (343c): that is, my being just
is good for you but it harms me. What if any reason, then, does a rational eudaimonist
have for being just? This is one of the main questions that Plato considers in the Republic.
It is his version of the fundamental, and perennially fascinating, question: “Why be
moral?”52
At the beginning of Republic 2, Glaucon challenges Socrates to show that justice is
good not only for its consequences (it is agreed on all hands that it is good in this way)
but also in itself, a view that, Glaucon says, is generally denied. To develop this point, he
describes a thought-experiment: Gyges had a ring that made him invisible and so
enabled him to commit injustice with impunity. Glaucon thinks that if we reflect on
Gyges’s situation, we will see that everyone would act as Gyges did: we would all behave
unjustly if we could get away with it. We practice justice reluctantly, because we lack
such a ring.53
But what is justice? As ordinarily conceived, it seems to involve both not harming
others and also benefiting them. But this is not yet to say exactly what justice is. And
Glaucon admits that most people might have an incorrect understanding of what it is;
hence he asks Socrates to provide his own account. At the end of Book 4, Socrates sug-
gests that for a person to be just is for her soul to be in a certain sort of psychic harmony,
with each part fulfilling its proper function.54 It has been argued that in explaining what
it is for a person to be just this way, Plato commits the fallacy of irrelevance:55 he was
asked to explain why I have reason not to harm, and to benefit, others; instead he explains

either that all the virtues are identical or that one can have one virtue if and only if one has them all.
Hence if one is just, one has all the other virtues as well.
52 Plato’s version of this question is sometimes thought to differ from another version of it. The
question is sometimes taken to mean: Should one be moral for self-interested reasons, or for other
reasons? Plato, by contrast, assumes that one has reason to be moral only if that promotes self-interest;
his question is whether it does so. The difference between these two ways of understanding the
question is part of what led H. Prichard to think that Plato’s approach rests on a mistake: see his “Does
Moral Philosophy Rest on a Mistake?,” Moral Obligation and Duty and Interest (Oxford: Oxford
University Press, 1968), ch. 1.
53 It is an overstatement to say that each of us would act unjustly whenever we could get away with
it. First, at 366c–d it is said that anyone who really knows what justice is will be just willingly. Second,
even on the ordinary view of justice that Glaucon describes, the social contract that arises when each of
us agrees not to harm others in exchange for not being harmed ourselves is in our interests. Hence,
even if I could be guaranteed to get away with doing an unjust act, I would not do it if doing it would
undermine the system that protects and benefits me. But most unjust actions would not have such
serious consequences, or so we might think.
54 In addition to saying what it is for a person to be just, Plato also says what it is for a state to be
just; I discuss this later on in the chapter. It is disputed whether he also provides at least a partial
account of what justice as such is (not just what it is for a person, and state, to be just). If he does,
presumably the account says that justice is a certain sort of harmony; he then explains what sort of
harmony is involved in a person’s being just, and what sort is involved in a state’s being just.
55 See G. Grote, Plato and the Other Companions of Socrates, vol. 4, new ed. (New York: Burt
Franklin, 1973; reprint of the 1888 ed.), 99–106; and D. Sachs, “A Fallacy in Plato’s Republic?,”
Philosophical Review 72 (1963), 141–58, reprinted in Vlastos (ed.), Plato 2, ch. 2.
20   Gail Fine

why I have reason to promote my own psychic harmony. The fallacy of irrelevance
alleges that the latter account of justice is too far removed from the ordinary, other-
regarding, understanding of it to provide an explanation of why one should be just, as
the question was originally intended. Glaucon wanted to know why one should, for
example, honor one’s commitments and parents and not steal. Socrates explains why
one has reason to want one’s soul to be well ordered. What is the connection between the
original question and Socrates’ answer?
Scott canvasses various replies. The basic strategy is to argue that one can achieve and
maintain one’s psychic harmony only if one doesn’t harm, and indeed benefits, others,
perhaps only if one benefits them for their own sakes; hence benefiting others turns out
to be part of one’s good. But there are many different versions of this general strategy.
Scott considers both psychological and metaphysical defenses. According to the psy-
chological defense, conventionally unjust behavior is motivated by desires a person who
has achieved Platonic justice will lack; for example, someone who has Platonic justice
lacks the motives to steal that someone else might have. Hence the Platonically just
person can be relied on to act, at least by and large, as a conventionally just person
would. There is therefore a significant overlap between conventional and Platonic justice;
the fallacy of irrelevance is therefore avoided. According to the metaphysical defense,
the Platonically just person will both understand and love the form of the good; and this,
in turn, will lead him to behave, at least by and large, as a conventionally just person
would, if for different reasons (because of his love of the form of the good, rather than,
for example, out of a fear of being caught). In Scott’s view, Plato explicitly relies on the
psychological defense. Though he may also intend the metaphysical defense, he does
not explicitly offer it; commentators who appeal to it are, in Scott’s view, engaged in
“rational reconstruction.”
In considering how this rational reconstruction would go, it is worthwhile to consider
Plato’s views on love, which Kraut explores (though not primarily in connection with
Plato’s reply to Glaucon) in chapter 23. If one loves something, one wants to be surrounded
by it. Philosophers—who alone are Platonically, hence truly, just—love goodness; hence
they want to be surrounded by goodness. Hence they have reason to benefit others, by
making them as good as their natures allow them to be. I return to this sort of considera-
tion later on in this chapter.
Plato’s views about love have been criticized. For example, it has been argued that he
thinks we do, or at any rate should, love others just for their admirable traits. This might
seem to imply that we do not love others as the distinctive individuals they are: we love
their admirable traits, not the people who have them.56 Kraut however, argues that Plato
leaves room for loving people as the people they are.

56 For this view, see G. Vlastos, “The Individual as Object of Love,” in his Platonic Studies (Princeton,
NJ: Princeton University Press (1973), 3–34; reprinted in Fine (ed.), Plato 2, ch. 5.
 Introduction   21

9. Republic: division of the soul

Plato describes Platonic justice in terms of his division of the soul into three “parts” or
“kinds” (435e–441c): the rational (to logistikon), the spirited (to thumoeides), and the
appetitive (to epithumêtikon). A person is just when each of these parts fulfills its proper
function, and when they are in the right sort of harmony with one another. In chapter 21,
Lorenz discusses Plato’s division of the soul and compares it with Plato’s view of the soul
in the Protagoras, Gorgias, and Phaedo.
On one view, the rational part is all reasoning, the appetitive all desire; on this view,
there doesn’t seem to be room for the third, spirited, part.57 On another view, Plato’s
distinction is between three irreducibly different sorts of desires or motivating factors
or, alternatively, between the subjects of those desires.58 On this view, the rational part
of the soul either consists of one’s rational desires or is the subject for such desires; these
are good-dependent desires (see Section 4 of this chapter). I have a rational desire to
drink milk, for example, if I desire to do so because I think that’s best for me, all things
considered—because, say, I believe that it will promote my health, which I think is good
for me. Appetitive desires, by contrast, are good-independent: they do not consider what
is best for me overall. I have an appetitive desire to drink milk, for example, if I just feel
like drinking it. As this example makes clear, one can have an appetitive and a rational
desire for at least some of the same things, if for different reasons. But these desires can
also conflict: I might want to drink milk because I think it will contribute to my health,
but I might want not to drink it because I don’t like its taste.
This way of conceiving of the parts of the soul—as types of desires, or as subjects for
types of desires—leaves room for a third part, since the division between good-dependent
and good-independent desires is not exhaustive. There are, however, different ways of
conceiving of the spirited part. Lorenz, for example, suggests the spirited part involves
the desire to distinguish oneself and to be esteemed and respected by others, as well as
an awareness of one’s social position and of one’s merits. This explains the spirited person’s
sensitivity to slights and insults; it also explains why Plato associates spirit with anger.
In acknowledging the existence of spirited and appetitive desires, Plato rejects the
view, often attributed to the early dialogues, that all desires are rational; he therefore rejects
psychological (but not rational) eudaimonism.59 In allowing that appetitive desires can

57 For this view, see T. Penner, “Thought and Desire in Plato,” in Vlastos (ed.), Plato 2, 96–118.
58 Just as each part of the soul has its own kind of desire, so each part has its own distinctive kind of
cognition. There is dispute about what kind of cognition each part has. Lorenz argues that only the
rational part is capable of means-end reasoning; contrast C. Bobonich, Plato’s Utopia Recast (Oxford:
Clarendon Press, 2002), 244.
59 Rep. 438a calls attention to this point.
22   Gail Fine

not only conflict with but also overcome rational desires, he acknowledges the possibility of
akrasia, again in contrast to the early dialogues.60

10. Republic: city and soul

Just as Plato divides the soul into three parts, so he divides the ideally just polis into three
occupationally defined classes: the guardians or rulers, the auxiliaries or military class,
and the workers or productive class. And just as he argues that justice for an individual
consists in the harmony of the three parts of the individual’s soul, with each part fulfill-
ing its proper function, so he argues that justice for a city consists in the proper harmony
of its three parts (the three classes), with each part (class) fulfilling its function.
There is dispute about how to interpret Plato’s elaborate analogy between justice in a
soul and in a city. According to the Whole-Part account, a city is just if and only if all
or most of its members are just; but this view leads to considerable difficulties. For
example, Plato thinks that even in the ideally just city, most people aren’t just; only the
guardians are.61 For they alone know what justice is, and one can be just only if one has
this knowledge.
This suggests the Macro-Micro account, according to which there is a structural
isomorphism between the justice of a person and a city. For a person to be just is for the
parts of her soul to be in a particular sort of harmony, and for each part to fulfill its function.
For a city to be just is for its occupationally defined classes to be in structurally the same
harmony, and for each of its parts to fulfill its function. On this view, a city can be just
even if not all or most of its citizens are just. All that is required is that each class fulfill its
proper function, and that the classes stand in the appropriate harmonious relations to
one another.62

60 At least, he allows that one can believe that it is better for one to do x than y, yet do y instead. It is
less clear whether he thinks that someone with full knowledge of what is best for one could ever act
against that knowledge.
61 For the view that, in the ideal city, only the philosophers are just, see J. Neu, “Plato’s Analogy of
State and Individual,” Philosophy 24 (1971), 238–54. For the view that not only are they just, see R. Kraut,
“Reason and Justice in Plato’s Republic,” in E. N. Lee, A. P. D. Mourelatos, and R. M. Rorty (eds.),
Exegesis and Argument (Assen: Van Gorcum, 1973), 207–24.
62 For discussion of the Whole-Part interpretation, see B. A. O. Williams, “The Analogy of City and
Soul in Plato’s Republic,” in Lee, Mourelatos, and Rorty (eds.), Exegesis and Argument, 196–206,
reprinted in Fine (ed.), Plato 2, ch. 10. For a defense of the Macro-Micro account, see Neu, “Plato’s
Analogy of State and Individual”; and G. F. R. Ferrari, City and Soul in Plato’s Republic (Sankt Augustin:
Academia Verlag, 2003, and Chicago: University of Chicago Press, 2005). At 435e, Plato says that
characteristics of communities are derived from those of its members. This might seem to favor the
Whole-Part account. However, the passage commits Plato only to the weaker view that some features of
a community are in some way derived from features of its members. It doesn’t commit him to the more
specific view that a city is F if and only if all or most of its citizens are F.
 Introduction   23

11. Republic: knowledge and belief

In Republic Book 5, Plato introduces three “waves of paradox,” the third and largest of
which is that philosophers should rule (473d; cf. 484d). One reason he favors this view is
that he thinks the virtuous should rule, but only philosophers have the knowledge
needed for virtue. What sort of knowledge do they need to have? And how can Socrates
persuade others of this audacious view? Part of his answer involves developing his epis-
temological—and metaphysical—views further, especially in a difficult argument at the
end of Book 5, and in the famous images of the Sun, Line, and Cave in Books 6 and 7.63
These passages are discussed by Taylor.
At the end of Book 5, Plato claims that knowledge is of what is, whereas belief is of
what is and is not.64 On one interpretation, “is” is predicative (is F, e.g. is blue). On one
familiar interpretation, the point is that one can know only entities that are F (and not
also not F); these are the forms, which escape compresence. By contrast, beliefs are only
about entities that are both F and not F; these are sensibles, which suffer compresence.
Hence, one can’t know sensibles or have beliefs about forms. This is a version of the
Two Worlds Theory, discussed briefly in Section 7 of this chapter in connection with
the Phaedo.65
On an alternative interpretation, “is” is veridical (is true). The point then is that
knowledge is of what is in the sense that it implies truth; whereas belief is of what is and
is not in the sense that belief doesn’t imply either truth or falsity, since there are both
true, and false, beliefs. This says nothing about what objects knowledge and belief can be
of or about, and so it doesn’t involve commitment to the Two Worlds Theory. It’s true that
Plato argues that, to have any knowledge (or, at least, to know what, e.g., beauty is), one
needs to know forms. But that doesn’t imply that one can’t also know sensibles or that
one can’t have mere beliefs about forms.

63 Book 10 also considers some relevant issues. I discuss the argument in Republic 5 in “Knowledge
and Belief in Republic 5,” in Archiv für Geschichte der Philosohie 60 (1978), 121–39; and in “Knowledge and
Belief in Republic 5–7”, in S. Everson (ed.), Epistemology (Cambridge Companions to Ancient Thought
1: Cambridge: Cambridge University Press, 1990), 85–115. The second of these two papers also
discusses the Sun, Line, and Cave. Both papers are reprinted in my Plato on Knowledge and Forms,
chs. 3 and 4, respectively.
64 Einai (“to be”) can be used in various ways. There’s considerable dispute about how it’s used in
Republic 5. See Taylor, chapter 18 of this volume, for discussion.
65 Some commentators argue that even if “is” is predicative, in the claims that knowledge is of what
is and that belief is of what is and is not, that doesn’t imply the Two Worlds Theory. In their view, the
claims concern, not what knowledge and belief are restricted to, but what they are for or typically
about. See, for example, V. Harte, “Knowing and Believing in Republic 5,” in V. Harte and R. Woolf
(eds.), Rereading Ancient Philosophy (Cambridge: Cambridge University Press, 2017), 141–62. “Is” is
clearly predicative in 479a–c, where Plato says that the many beautifuls are and are not: that is, each of
them is both beautiful and ugly. This adverts to the familiar compresence of opposites that we explored
previously. It doesn’t follow that “is” is also used predicatively when he says that knowledge is of what
is, and that belief is of what is and is not.
24   Gail Fine

Even if Republic 5 is not committed to the Two Worlds Theory, Books 6 and 7 might
be. However, Socrates introduces the simile of the Sun because, he says, he has belief, but
not knowledge, about the form of the good (506c). Hence, contrary to the Two Worlds
Theory, he admits beliefs about forms. The simile of the Sun describes, among other
things, the crucial role of the form of the good in explaining other phenomena. As Taylor
explains, Plato suggests that just as the sun generates and illuminates visible things, so
the form of the good explains the existence, nature, and knowability of forms: forms exist,
and are as they are, because it is best that things be that way; and we know forms when
we understand how and why that’s so. Here Plato reasserts the teleology mentioned but
despaired of in the Phaedo, though he doesn’t explain it in detail. However, Taylor suggests
that, for Plato, goodness involves order and proportion, which is to be understood
mathematically. This partly explains the importance Plato places on mathematics in the
education of the guardians.66
In the Line, Plato describes two kinds of belief and two kinds of knowledge. On one
interpretation, each of these cognitive conditions is individuated in terms of a certain
sort of object, such that one has the lowest sort of belief, for example, if and only if one
looks at shadows. This interpretation is congenial to the Two Worlds Theory. On an
alternative interpretation, on which Plato rejects Two Worlds Theory, what cognitive
condition one is in is determined not by the objects one considers but by how one reasons
about them. More precisely, if, for example, one is restricted to sensibles and doesn’t
acknowledge the existence of forms, one can have at best belief; whereas, if one counte-
nances forms, one can have knowledge that, however, is not restricted to forms, for one
can know sensibles in the light of forms.
In the allegory of the Cave, Plato explains how one can move from the lowest sort of
belief to the best sort of knowledge, which requires knowing the form of the good. That’s
the upward path: one emerges from the cave (the sensible world) that most of us are
confined to, into the light of the sun (the form of the good) outside the cave. Plato
also describes the downward path, whereby those who have attained the best sort of
knowledge return to the cave. After a period of adjustment, they will be able to see the
things there better than the prisoners, who have never left the cave, can do. Indeed,
unlike the prisoners, they will know sensibles (520c). This counts against the Two Worlds
theory, since Socrates countenances knowledge of sensibles.
What is involved in knowing forms? On one view, one grasps isolated individual forms
through particular acts of non-propositional acquaintance. On an alternative view,
Plato doesn’t take knowledge of forms to consist in non-propositional acquaintance.
Rather, to know a form is to know what it is, which is propositional knowledge: to know
what the form of beauty is, for example, is to know that it is thus and so. Nor does Plato
think one can know just a single form all on its own. Rather, he is an epistemological
holist in the sense that he thinks that knowing any form requires knowing its place in a
broader system, which requires knowing other forms.

66 See e.g. 522b2–531d4. For discussion, see M. Burnyeat, “Plato on Why Mathematics Is Good for
the Soul,” Proceedings of the British Academy 103 (2000), 1–81.
 Introduction   25

12. Republic: political theory

At the end of Section 10 of this chapter, we saw that Plato thinks that, in the best polis,
each class will fulfill its proper function. This requires the members of each class to
devote themselves to just one type of work; this is mandated by the principle of special-
ization laid down in Book 2. Those in the productive class will spend their time making
shoes, producing food, and so on; the auxiliaries will devote themselves to defending the
city from both external and internal enemies; and the guardians will contemplate forms
and rule in the light of the knowledge that confers.
In restricting ruling to the guardians, Plato rejects democracy; the guardians are the
only ones who have a say in how the city will be run. Hence most people are deprived of
political autonomy. Their personal autonomy is also severely limited. For example,
someone who is most suited to be in the working class cannot be an auxiliary, even if she
wants to be. Both political and personal autonomy are often thought to be important
goods; in depriving most members of the city of much of their personal and political
autonomy, isn’t Plato making them less happy than they could otherwise be? We might
also ask, as Glaucon does at 519e, whether, in requiring philosophers, at least temporarily,
to forgo contemplation of forms in order to rule, Plato is making them less happy than
they could be; for contemplating forms is a greater good than ruling.67
In reply to the worry about the happiness of philosophers, Socrates says that he isn’t
aiming at the happiness of any one class but at the happiness of the whole (420b–421c;
519e–520a). According to Popper, Plato accepts the “organic theory,” which involves both a
metaphysical and a political component. The metaphysical component says that the
state is an entity in its own right, distinct from its parts. The political component involves
the view that individual citizens must sacrifice their interests for those of the city.
Scott argues, against Popper, that the city-soul analogy does not imply the metaphysical
component. Rather, it implies only that the city and the soul are structurally similar.
Though Scott also rejects Popper’s version of the political component, he agrees that
Plato thinks that the interests of individual citizens are subordinate to the greatest good
of the state, which is its unity.68 On an alternative view, in saying that he aims at the
happiness of the whole, Plato means that he wants each citizen to be as happy as he can
be. To say this is not to sacrifice individual happiness to the happiness of a distinct entity,
the city; nor is it to give priority to the interests of the city over those of its citizens.69

67 At 419a, Adeimantus raises a further question about the guardians’ happiness: Doesn’t the fact
that they live in relatively austere conditions make them less happy than they could be? Plato’s answer is
that thinking this involves overvaluing material possessions; the guardians have all they need for
happiness, and they will know this.
68 See also L. Brown, “How Totalitarian Is Plato’s Republic?,” in E. Ostenfeld (ed.), Essays on Plato’s
Republic (Aarhus: Aarhus University Press, 1998), 13–27.
69 Nor is it to endorse utilitarianism. Plato’s concern is to describe a city that allows each person to
achieve the greatest amount of happiness they are capable of. In contrast to utilitarianism, he doesn’t
think it legitimate to sacrifice the happiness of a few in order to produce a greater overall aggregate
amount of happiness.
26   Gail Fine

How does Plato defend the view that he aims to make ordinary citizens as happy as
possible, given that they have so little personal and political autonomy? And how does
he defend the view that guardians are happy, when they are made to abandon the greater
good of contemplating forms for the lesser good of ruling?
On one view, Plato argues that the guardians must return to the cave to rule, despite
the fact that it is not in their self-interest, because it is good to do so: not good for them,
but impersonally good.70 If this is right, then at this stage Plato abandons rational eudai-
monism and does not show, as he undertook to do in Book 2, that it is always in one’s
interest to be just.
On an alternative view, he retains rational eudaimonism and argues that it is in the
guardians’ best overall interest to rule.71 To be sure, contemplating forms is a greater
good than ruling. But it doesn’t follow that the philosopher who has been trained in the
ideal city would be better off, all things considered, if she continued to philosophize
(and thereby violated the just requirement that she rule) than if she spent some time
ruling (thereby fulfilling a just requirement). Still, one wants to know why the philosopher
is better off occasionally engaging in a less good activity than she would be if she more
single-mindedly devoted herself to contemplating the forms.
One possibility is that ruling is instrumentally good for philosophers: if they don’t rule,
the city will be less stable than it would otherwise be, and that would harm them. By way
of analogy, if philosophers don’t eat, they will die and so could not contemplate the forms;
hence even though eating is less good than contemplation, philosophers will spend
some time eating, and for their own sakes.72 Similarly, even if ruling is less good than
contemplating, it may be in the philosophers’ best interest to spend some time ruling.
One might also argue that ruling is not merely instrumentally good. For example, as
we saw previously, given their love of goodness, philosophers want, for their own sakes,
to be surrounded by as much goodness as possible. In ruling, they make others as good
as possible, which is in the intrinsic interests of those others. But in benefiting others for
their own sakes, the guardians also benefit themselves; achieving the good of others is
therefore part of their own happiness.
As to non-guardians, Plato thinks they will come as close to being happy as their
natures allow only if they are ruled by the guardians, since only the guardians know what
is truly good. In Plato’s view, political and personal autonomy are less important to one’s
happiness than they are sometimes taken to be. Living in a stable, well-ordered city
in which one devotes oneself to the task for which one is best suited contributes more
to one’s happiness than having more autonomy would. We may not agree with Plato’s

70 Alternatively, one might argue that they must do so because, though it is not good for them, it is
good for the city. For versions of the view that it is not in the philosopher’s best overall interest to rule,
see J. Cooper, “The Psychology of Justice in Plato,” American Philosophical Quarterly 14 (1977), 151–57;
J. Annas, An Introduction to Plato’s Republic (Oxford: Clarendon Press, 1982); and N. P. White, “Plato’s
Concept of Good,” in H. Benson (ed.), A Companion to Plato (Oxford: Blackwell, 2006), ch. 24.
71 For discussion, see R. Kraut, “Return to the Cave: Republic 519–21,” in Fine (ed.), Plato 2, ch. 9, to
which my discussion is indebted.
72 This analogy is imperfect, since one can free ride on others’ ruling, but not on others’ eating.
 Introduction   27

low estimation of the importance of autonomy or with his defense of paternalism. It is

nonetheless important to see that he limits personal and political autonomy not in sacrifice
to a superorganism or the state but to enable each individual to come as close to being
happy as possible.

13. Political theory

For the ideally just city to come into existence, people’s attitudes need to be radically
transformed. For example, they need to learn what is most valuable: virtue, not material
goods. They also need to learn to value true philosophers who, in turn, need to undergo
the proper training so as to allow their natures to flourish. In chapter 25, Kamtekar
considers the sort of education this transformation requires. She also considers Plato’s
views on art, which are intimately connected to his views on education since, for example,
attending to certain sorts of art can inhibit proper development by arousing and encour-
aging inappropriate emotions.
Though Plato seems to think his ideally just city is possible,73 it certainly isn’t actual.
We might then wonder what the best city we can hope for is, given people as they actually
are. In chapter 24, Christopher Bobonich suggests that different dialogues defend different
views. In his view, in the early dialogues Socrates thinks no one has the knowledge that
is required for virtue; hence no one is qualified to rule in the way the guardians of the
Republic are. Does Plato nonetheless think there should be absolute rulers, albeit less-
qualified ones? Or does he think the city should be run in a different way? Presumably,
he thinks the best city we can hope for, taking us more or less as we are, is one run
according to his moral principles: for example, that it is better to suffer than to commit
injustice; that if one commits injustice, it is better to be punished than to escape punish-
ment; and so on. What would a city look like if it embodied these and other Socratic
principles, but taking us more or less as we are? Such a city would need to impose some
sanctions, which, in turn, might require a fair amount of coercion. What legitimates such
coercion? Would all citizens benefit equally from living in a city founded on Socratic
principles? How stable would such a city be? Does political activity, in such a city, compete
with developing one’s own virtue? Because the ethical views of the early dialogues are so
underdeveloped, it is difficult to answer these questions in their case. But, Bobonich
suggests, some questions receive fuller—and different—answers, beginning in the Phaedo.
Bobonich traces Plato’s answers to the various questions just mentioned, from the earliest
through the latest dialogues. He also explores Plato’s changing views of the nature of the
ideally best city. He argues that the early, middle, and late dialogues espouse different
ethical, epistemological, and metaphysical views, which, in turn, lead to different views

73 See, for example, M. F. Burnyeat, “Utopia and Fantasy: The Practicability of Plato’s Ideally Just
City,” in J. Hopkins and A. Savile (eds.), Psychoanalysis, Mind and Art (Oxford: Blackwell, 1992), 175–87,
reprinted in Fine (ed.), Plato 2, ch. 13.
28   Gail Fine

about the best city, both for people as they are and for people who have undergone the
transformation needed for a more radical change in society.
For example, according to Bobonich, the early dialogues deny the existence of non-
rational desires, whereas later dialogues admit their existence. Hence, according to the early
dialogues, one can persuade people to change their lives only by changing their beliefs.
By contrast, admitting the existence of non-rational desires opens up the possibility that
people can be trained to care about the right things not, or not only, by changing their
beliefs but also by training their non-rational desires in such a way that they come to care
about the right things, whether or not they can appreciate their true value.
According to Bobonich, Plato’s views about the ideally just city, and about the best
state for us more or less as we are, change again in the Statesman. For example, it seems
to have more demanding qualifications for citizenship, and citizenship has greater ethical
significance. In the Republic, the members of all three classes are citizens, though only
the philosophers are just; in the Statesman, only just people can be citizens. However,
this would allow non-philosophers to be citizens if they are just, and Bobonich thinks
that is Plato’s view by the time of the Statesman. This could be so for one of two reasons:
either Plato now thinks that non-philosophers have knowledge, or he no longer requires
knowledge for virtue.
The Laws has different views again. For example, it revises the Republic’s view of the
nature of the ideally just city. It also allows, contrary to the Republic, that non-philosophers
can be educated so as to have a reasoned grasp of basic ethical and political truths. Hence
non-philosophers are fit to rule; accordingly, instead of philosopher rulers, Plato now
posits an Assembly open to all citizens.
The Laws also takes up other questions considered in earlier dialogues. For example,
as Meyer explains in chapter 15, it considers the various virtues (such as wisdom, justice,
moderation, and courage) and asks about the relations among them. Does Plato still
believe, as he is often thought to believe in earlier dialogues, that one can have one of these
virtues if and only if he has the others? Does he think they are hierarchically arranged
and if so, how and on what basis? Another topic taken up in the Laws is theology, a topic
that is also of central concern in the Timaeus. Sedley explores this in chapter 26.

14. Parmenides

In chapter 10, Sandra Peterson discusses the Parmenides, which is sometimes thought to
criticize the middle dialogues’ views about forms. This dialogue falls into two sharply
distinct parts. The first part discusses and criticizes a theory of forms. The second part
conducts an exercise designed to help one resolve some of the problems broached in the
first part.
At the beginning of the dialogue, Socrates says that forms are introduced to solve a
puzzle raised by the fact that things are both one and many, like and unlike: Simmias,
for example, is one man with many limbs; he is like some things and unlike others.
 Introduction   29

We can understand how this can be so only by grasping that he participates in both the
form of one and the form of many, and in the forms of likeness and unlikeness. As in the
middle dialogues, compresence in sensibles is explained by reference to forms.74
Socrates is also tempted to posit forms of man, fire, and water, yet sensible men, fire,
and water do not in any obvious way suffer compresence: Socrates is not both a man and
not a man. Perhaps, then, as suggested in Section 7 of this chapter, compresence is suffi-
cient but not necessary for positing forms.
Socrates denies that there are forms of mud, dirt, and hair. One of his reasons is that
they are just what we see them to be. Perhaps this supports the suggestion, mentioned
in Section 7 of this chapter, that his general concern is with the limits of perception. So
far, then, the Parmenides seems to capture at least one central line of thought in the
middle dialogues.
The middle dialogues also seem to suggest that forms are Self-Predicative: any form
of F is itself F; the form of beauty, for example, is itself beautiful. Self-Predication can
seem absurd: how, for example, could the form of large be large? For, one might think,
something can be large only if it has a size; yet forms are incorporeal. Self-Predication
would indeed be absurd if it required the form of F to be F in the very same way in which
sensible particulars are F, such that something can be large, for example, only if it has a
size. However, as Peterson explains, Self-Predication can be understood as a much more
plausible thesis. Indeed, in her view, if we were to reject it, we would also have to give up
many ordinary statements we routinely accept, such as the biblical statement that “charity
suffereth long” and encyclopedia statements such as “the tiger is a carnivore.” She suggests
that there are various accounts of the semantics of Self-Predications on which they are
true, including the one Plato accepts; and she shows how some of his arguments (such as
the celebrated Third Man Argument (132a-b) and the Greatest Difficulty (133b-134e))
require no more than a version of Self-Predication on which it is true.
Strenuous efforts have been made to argue either that Plato was never committed to
Self-Predication or that, even if he was committed to it in the middle dialogues, he aban-
doned it because of the Parmenides’ criticisms. But if Self-Predication is arguably true, it
is less clear that we should attempt to extricate Plato from it. If, rightly or wrongly, he
remains committed to Self-Predication, he needs a different escape route from some of
the arguments leveled in the first part of the Parmenides.75
Perhaps the most famous of these arguments is the Third Man Argument.76 According
to it, the, or a, theory of forms is vulnerable to a vicious infinite regress: if there is
even one form of F, there are infinitely many of them. This violates Plato’s Uniqueness

74 See Section 7 of this chapter for discussion of compresence in the middle dialogues.
75 I say “some of the arguments,” because it is not clear that all of them assume Self-Predication.
76 So-called because Aristotle describes an argument that he calls the Third Man, and it is generally
thought to be the same argument, from the logical point of view, as the argument Plato describes. See
Aristotle, Peri Ideôn; SE 22; and Met. 7.13. However, whereas Aristotle describes a regress of forms of man,
Plato describes a regress of forms of large. The classic discussion of the Third Man Argument in recent
times is G. Vlastos, “The Third Man Argument in the Parmenides,” Philosophical Review 63 (1954), 319–49.
30   Gail Fine

assumption, according to which there is at most one form for any given predicate.77 The
regress goes roughly as follows. Each form is a one over many: that is, whenever many
things are F, there is one form in virtue of which they are F.78 Consider the set of sensible
large things. According to the One over Many assumption, there is one form of large—
call it the form1 of large—over them. Since forms are Self-Predicative, we may posit a
new set of large things, one consisting of the members of the original set, along with the
form1 of large. The One over Many assumption tells us that there is one form of large over
this set. This can’t be the form in the set (= the form1 of large). For, or so the Third Man
Argument assumes, nothing is F in virtue of itself: this is the so-called Non-Identity
assumption. Hence there must be another form of large—call it the form2 of large—
which is the form of large in virtue of which the members of our new set of large things
are large. By Self-Predication, the form2 of large is large. We can now posit yet another
set of large things, which consists of the members of the previous set along with the
form2 of large. By the One over Many assumption, there must be a form of large over
this set, which, by Non-Identity, must be non-identical with anything in the set—and so
on ad infinitum, and in violation of Uniqueness.79
The Third Man Argument validly generates a regress.80 Hence Plato can avoid the
argument only if he is not committed to all its premises. If Self-Predication is arguably
true, it is not a likely candidate for rejection. Plato also accepts a one over many assumption,
as well as the view that forms are different from, and perhaps separate from, sensibles.
But these latter two views are not enough to commit him to either the One over Many
assumption or the Non-Identity assumption, that are at work in the Third Man
Argument. It is debated whether he is committed to these assumptions for other reasons.
It is sometimes thought that in the difficult second part of the Parmenides, Plato provides
clues about how to answer at least some of the puzzles in the first part of the dialogue.81
But it remains a matter of controversy whether—and if so, how—Plato revises his views
about forms either there or in subsequent dialogues. The Theaetetus, Timaeus, Sophist,
and Philebus all either mention entities called forms, or describe entities that seem similar,
in at least some ways, to forms as they are described in the middle dialogues. But there is
dispute about the precise connection between these entities and the forms countenanced
in the middle dialogues.

77 I put it this way because Plato doesn’t think there is a form corresponding to every predicate.
There is a form of justice but not of barbarian (Pol. 262a–e; cf. Phdrs. 265e1–2).
78 If there is a form in this case at all: see preceding note.
79 This way of reading the argument is challenged by, among others, M. L. Gill; see her “Problems
for Forms” in H. Benson (ed.), A Companion to Plato (Oxford: Blackwell, 2006), ch. 13. She suggests
that one difficulty for it is that it is unclear what motivates the inference from positing mia idea (one
form or idea) of largeness, to the claim that hen to mega (the large is one) (193). The answer is that the
inference is from a one over many assumption to Uniqueness: from the claim that a given group of F
things has just one form over it, to the claim that there is just one form of F simpliciter.
80 For a particularly lucid account, see. S. M. Cohen, “The Logic of the Third Man,” Philosophical
Review 80 (1971), 448–75, reprinted in Fine (ed.), Plato 1, ch. 10.
81 For one version of this view, one Peterson is broadly sympathetic to, see C. Meinwald, Plato’s
Parmenides (Oxford: Oxford University Press, 1991).
 Introduction   31

15. Theaetetus

The Theaetetus is Plato’s longest systematic discussion of knowledge. In chapter 11, Lee
discusses the dialogue as a whole, focusing on two related issues: whether the dialogue
espouses the view that epistemology can be done without metaphysics, and what if
anything it suggests about the nature and existence of forms.82 She argues that some of
the accounts of knowledge considered in the dialogue are supported with metaphysical
theories that are incompatible with the existence of forms; but, since Plato rejects those
accounts of knowledge, he is not committed to the metaphysics used to support them.
Though this falls short of positing forms, she suggests that Plato hints at requirements
for knowledge that we could satisfy by positing forms.
The first and longest part of the dialogue discusses the view that knowledge is percep-
tion. That view is linked both to Protagoras’s measure doctrine, according to which things
are (to one) as they appear to one, and to a Heraclitean flux doctrine. The refutation of
Protagoras’s measure doctrine is sometimes thought to be a refutation (or an attempted
refutation) of relativism, though it is disputed whether Protagoras is a relativist and, if
he is, in what sense he is.83 It is also disputed whether Plato’s refutation of a Heraclitean
flux doctrine in 181–83 refutes any sort of Heracliteanism that he himself accepts in earlier
dialogues,84 or whether he is just refuting the flux doctrine that is needed to support
Theaetetus’ suggestion that knowledge is perception (or, more precisely, that suggestion
when it is interpreted along Protagorean lines).
In 184–86, Plato presents his final refutation of the claim that knowledge is perception.
On one view, which Lee is sympathetic to, he argues that when perception is conceived as
being below the propositional and conceptual threshold, it cannot constitute knowledge;
indeed, when perception is so conceived, it does not even get as far as belief.85 This is
compatible with allowing—although it does not imply—that perceiving that something
is so (which, in contrast to “pure” perception, involves identifying what one sees as being
something or other) can be a case of knowledge.
In the second part of the dialogue, Plato asks whether knowledge is true belief. He
eventually argues that it is not, on the ground that the members of a jury might have a
true belief about who committed a crime but, not having been eyewitnesses, they do not

82 In chapter 18 of this volume, Taylor discusses it in the context of Plato’s epistemology in general.
83 For the view that he is a relativist, see M. F. Burnyeat, “Protagoras and Self-Refutation,” Philosophical
Review 85 (1976), 44–69. For the view that he is not a relativist, see G. Fine, Plato on Knowledge and
Forms, chs. 6–8. Cf. M. Lee, Epistemology after Protagoras (Oxford: Clarendon Press, 2005), esp. ch. 3.
84 See Section 7 of this chapter for some discussion of what if any sort of flux doctrine the middle
dialogues hold.
85 For this interpretation, see also M. F. Burnyeat, “Plato on the Grammar of Perceiving,” Classical
Quarterly 26 (1976), 29–51; M. Frede, “Observations on Perception in Plato’s Later Dialogues,” in his Essays
in Ancient Philosophy (Oxford: Clarendon Press, 1987), 3–8, reprinted in Fine (ed.), Plato 1, 377–83. For
different views, see J. Cooper, “Plato on Sense-Perception and Knowledge (Theaetetus 184–6),” Phronesis
(1970), 123–46, both reprinted in Fine (ed.), Plato 1, 355–76; and Taylor in chapter 18 of this volume.
32   Gail Fine

know who committed it; hence true belief is not sufficient for knowledge. Before directly
rejecting the view that knowledge is true belief, he asks about the possibility of false
belief. He proposes five explanations; but each of them seems to fail.
It has been argued that Plato’s failure to explain false belief stems from his alleged
unclarity about being and not being, or about the distinction between naming and stating;
according to some commentators, it is only in the Sophist that he attains clarity on these
issues and so is able to explain the nature of false statement and belief. Lee suggests,
however, that the failure is due instead to the dialectical context. In particular, it is an
indirect indictment of the definition of knowledge as true belief. For in order to have
a false belief about something, one must succeed in thinking about it, in which case one
must have a true belief about it. If knowledge is true belief, it follows that whenever
one succeeds in thinking about something, one thereby knows it. But, according to Plato,
one cannot both know and not know the same thing. Hence if thinking about something
involves knowing it, one cannot also have a false belief about it.86
In the third and final part of the dialogue, Plato asks whether knowledge is true belief
plus an account (logos). He explores this issue partly in terms of a dream Socrates says he
has had, according to which there are basic elements that can be perceived and named
but that have no account and so are unknowable. Lee suggests that Plato deliberately
leaves open the question of precisely what these elements are because he wants to focus
on abstract questions about ontology and language that a more determinate account of
the elements might obscure.
On one view, Plato accepts the dream theory’s claim that some things can be known
without an account. If he accepts it, he rejects the Meno’s claim that knowledge is true
belief tied down with an aitias logismos, as well as the claim made in the Phaedo (76b) and
Republic (531e, 534b) that knowledge requires a logos. On another view, he believes both
that knowledge requires an account and that elements are knowable; and so he rejects
the view that they lack accounts.
In order to adjudicate between these and other options, we need to know what an
account is. Plato considers three possibilities, but appears to reject all of them. On one
view, however, he hints that elements can be seen to have accounts once we realize that
accounts need not consist in listing a thing’s elements but can also consist in describing
something’s place in the larger whole of which it is a part. This leads to the sort of holism,
or interrelation model of knowledge, that is explicit not only in such late dialogues as the
Statesman but also, according to some commentators, in earlier dialogues.87

86 For a different version of the view that the failure to explain false belief is due to the dialectical
context, see my “False Belief in the Theaetetus,” Phronesis (1979), 27–80, reprinted, with minor
modifications, in Fine, Plato on Knowledge and Forms, ch. 9. Cf. M. F. Burnyeat, Introduction to the
Theaetetus of Plato (Indianapolis: Hackett, 1990), 65–123.
87 For epistemological holism in the Republic, see the end of Section 11 of this chapter. For my own
view of the Theaetetus on this issue, see my “Knowledge and Logos in the Theaetetus,” Philosophical
Review 88 (1979), 366–97, reprinted, with minor modifications, in Fine, Plato on Knowledge and Forms,
ch. 10. For criticism and alternative views, see D. Bostock, Plato’s Theaetetus (Oxford: Clarendon Press,
1988); Burnyeat, Introduction to the Theaetetus.
 Introduction   33

16. Sophist

The ostensible main topic of the Sophist is the definition of a sophist. Seven definitions
are considered, but each seems to fail. In exploring the seventh definition, Plato broaches
various issues about being and not being. Among the many issues he considers within
that broad compass is the nature of predication and correct speaking. The Late-Learners
deny that one can predicate one thing of another. It is often thought that Plato can solve
their problem only if he distinguishes between the identity and predication senses, or
uses, of “is” (e.g., “Cicero is Tully”; “snow is white”). Commentators who share this
assumption divide into optimists, who think Plato succeeds in distinguishing identity
and predication senses or uses of “is,” and pessimists, who think he doesn’t. In chapter 12,
Brown denies their common assumption, arguing, with the pessimists, that Plato does
not distinguish different senses or uses of “is,” but resisting their conclusion that he
thereby fails to solve the Late-Learners’ problem. For in her view, distinguishing different
senses or uses of “is” is not necessary for solving their problem. It can be solved by dis-
tinguishing between identity and predication statements. Plato does this by considering
the “communion of kinds.” This allows us to see how, for instance, Change is both the
same and not the same. For to say that Change is the same, is to predicate “the same” of
it; and to say that Change is not the same, is to say that it is not identical with the kind,
Sameness. Likewise, Change is both different (from other things) and also not different,
in that it is not (identical with) the kind, Different.88
Another, related issue taken up in the “middle part” (as opposed to the “outer part”)
of the dialogue is the possibility of false statement (and belief). Showing that kinds mix
is part of the solution, but it is not the whole of it. In addition, Plato provides an account
of what a statement (logos) is: it involves interweaving a name (onoma) and a verb (rhêma)
in such a way that one names something, and then says something about it. Hence Plato
distinguishes naming from stating.89 This, in turn, allows him to explain the nature of both
true and false statements: a true statement says of things that are, that they are; a false
statement says different things from the things that are. Both a true statement, such as
“Theaetetus sits,” and a false statement, such as “Theaetetus flies,” name Theaetetus and
also say something about him; hence both count as statements. But the true statement
says, concerning Theaetetus, that things that are, while the false one says different things
from the things that are. The precise interpretation of this account of false statement is

88 In explaining this, I’ve used “is” in a way that might encourage the view that Plato does
distinguish senses or uses of “is.” However, as Brown explains, in the Greek not all the crucial sentences
contain any form of einai (to be); this is one reason she suggests that the crucial distinction is between
identity and predication statements, whether or not they include some form of einai.
89 It is sometimes thought that he failed to do so in earlier dialogues—including the Theaetetus—and
thereby ran into difficulties. On an alternative view, the distinction is observed when he is speaking in his
own right—as opposed to engaging in dialectical discussion with an opponent, as in the puzzles of false
belief in the Theaetetus—even if it is not, until the Sophist, laid out in the same clear and explicit way.
34   Gail Fine

disputed. Both Brown and Crivelli canvass and criticize a number of options; in the end,
they favor different accounts.

17. Philebus

The main topic of the Philebus is the good human life: Is it pleasure, intelligence, or some
combination thereof? In ways that anticipate Aristotle’s Nicomachean Ethics, Plato argues
that the good (i.e., the best) human life—that is, the happy (eudaimon) life—doesn’t consist
just in pleasure or just in intelligence but in a combination of them; intelligence is the
more important component, but some pleasures (the good but not the bad ones) are also
part of happiness. In rejecting the view that happiness consists in pleasure, Plato rejects
hedonism, as he also does in the earlier Gorgias and Republic.90 In rejecting the view that
happiness consists in intelligence alone, he rejects a purely contemplative view of happi-
ness that, according to some but not others, he favors in the Republic and Symposium.
In chapter 14, Meinwald discusses these issues about the Philebus, placing special
emphasis on its treatment of method and metaphysics, an understanding of which is
necessary if we are to understand how Plato arrives at his final view of the constituents of
happiness. In doing so, Plato introduces the Promethean Method, which is based on the
view that there is both limit (peras) and the unlimited (to apeiron) in things. This method
involves dividing subjects into subkinds and knowing how they combine with each other.
In addition to the Promethean Method, Plato also describes a fourfold division of things
into limit (peras), the unlimited (to apeiron), what is mixed from them, and the cause
of the mixture. Pleasure is put in the category of the unlimited, and mind in the category of
cause; this helps explain their place in the good human life.
The method of division, adverted to in the Promethean Method, is also described in
other dialogues. But the Philebus is unique in linking it to limit and the unlimited. There
are disputes about how to understand these notions. Indeed, it is not even clear that they
have the same sense or reference in the Promethean Method and in the fourfold division,
though Meinwald argues that they do. In her view, the unlimited is best understood as a
blurred condition in which kinds run together with no significant demarcations. For
example, below the level of specific vowels there is a continuum of sounds; below the
lowest division into kinds of cats, there is indefinite variation in softness, fur, and so on,
all at the level of types. On this view, unlimited things (apeira) are not, as is sometimes
thought, particulars (this cat and that one) but types considered independently of their
division into determinate kinds, which is the realm of limit (peras). Limit involves pro-
portion, which, in turn, involves explaining forms or kinds mathematically. The Philebus
is not unique in emphasizing the fundamentally mathematical nature of things. As we’ve

90 As we’ve seen, it is disputed whether he endorses it in the Protagoras. In chapter 22 of this volume,
Annas discusses Plato’s views on pleasure.
 Introduction   35

seen, the Republic also does so. The importance of mathematics for understanding reality
is also emphasized in the Timaeus.
What are the kinds (eidê) that Plato discusses in the Philebus? Meinwald suggests that
they are given by genus-species trees. Here it is worth asking (as one might also do about
the “greatest kinds” in the Sophist) how these kinds compare with the forms described in
the middle dialogues.

18. Aristotle’s criticism of

Plato; Platonism

Aristotle was Plato’s student, or associate, in the Academy for nearly 20 years, and he is
an important source of information about Plato, though there are disputes about his
reliability.91 In chapter 27, Shields explores some of Aristotle’s criticisms of Plato, espe-
cially his claim that Platonic forms turn out to be both particulars and universals, a claim
we mentioned briefly in Section 7 of this chapter.92 Shields also explores a closely related,
but possibly importantly different, claim: that “universals and particulars are practically
the same natures” (Met. 1086b10–11). Shields argues that if Aristotle means to argue that
forms are both universals and particulars, where these are taken to be exclusive catego-
ries of being, his arguments fail; whereas, on some ways of understanding the claim that
universals and particulars are practically the same natures, he has a more challenging
criticism, one he himself needs to grapple with.
As Shields explains, Aristotle sometimes seems to suggest that the separation of forms
makes them particulars. This is a curious claim for Aristotle to make. First, we’ve seen
that to say that the form of F is separate is to say that it can exist whether or not there are
any F sensibles. That is to say, they can exist uninstantiated (by sensibles, at any rate). But
to say that forms can exist uninstantiated doesn’t seem to make them into particulars:
Why can’t there be uninstantiated universals? Further, as Shields notes, Aristotle him-
self, in at least some phases of his career, admits universal substances (ousiai), as in the
Categories, where the species and genera of individual substances count as secondary
substances; yet species and genera are universals, not particulars.
Shields also explores Plato’s account of participation in terms of mimeticism (or imi-
tation), asking whether, as Aristotle may believe, it commits Plato to the view that forms
are particulars. Shields concludes that it does not do so. For these and other reasons,
Shields argues, Aristotle has no easy route to the conclusion that forms are particulars.
However, Shields argues, Aristotle also makes the good point that Plato seems to overtax
forms, giving them too many roles to play, roles no single sort of entity could obviously play.
Shields suggests, for example, that, according to Aristotle, Plato posits forms in order to

91 For some discussion, see Irwin, chapter 3 of this volume.

92 For my own account of Aristotle’s criticisms of Plato’s views about forms, see my On Ideas:
Aristotle’s Criticism of Plato’s Theory of Forms (Oxford: Clarendon Press, 1993).
36   Gail Fine

explain both the knowability and the unity (both synchronic and diachronic) of sensible
particulars. Insofar as they explain knowability, they must, in Aristotle’s view, be universals;
but insofar as they are principles of the synchronic and diachronic unity of particulars,
it seems they must, according to Aristotle, be particulars. Shields suggests that this
Aristotelian line of criticism is more promising than one that attempts to argue that
forms are, impossibly, both universals and particulars.
More generally, Shields suggests that, though Aristotle does not succeed in delivering
a knockout blow to Plato, he raises important criticisms that are well worth considering.
In any case, we can understand both Plato’s and Aristotle’s views better by considering
both Aristotle’s objections and Plato’s resources in the face of them.
Whether or not Plato was the first Platonist, he was certainly not the last. In chapter 28,
Brittain describes the fascinating—though difficult and complex, and still relatively
unexplored—history of how later Platonists appropriated, or claimed to appropriate,
Plato’s views. Though “Platonism” is often taken to involve a single, unified body of
thought, Brittain shows how heterogeneous the Platonic tradition is. Nonetheless, he
identifies three generally shared commitments: (1) to the authoritative status of Plato’s
work; (2) to the assumption that experience is an inadequate basis for understanding the
world and that there are various primary immaterial principles, including forms, souls,
and a transcendent god, that do explain it; and (3) to an increasing interest in a range of
religious practices and concerns. As Brittain notes, the results of these commitments are
likely to strike modern readers as remote from Plato’s text, at least at first glance. However,
as he also notes, even if we do not share the three commitments just mentioned, we can
benefit greatly by reading the work of the philosophers who made them—not just because
that work is intrinsically interesting but also because it sheds light on Plato, by providing
a range of imaginative solutions to interpretative difficulties that are still with us. Just as
exploring Aristotle’s criticisms of Plato allows us to gain further insight into both Plato’s
and Aristotle’s views, so comparing Plato’s Platonism with later Platonism promises to
shed light on both.93

Bibliography
Annas, J. “Aristotle on Inefficient Causes,” Philosophical Quarterly 32 (1982), 311–26.
Annas, J. An Introduction to Plato’s Republic (Oxford: Clarendon Press, 1982).
Benson. H. “The Priority of Definition and the Socratic Elenchus,” Oxford Studies in Ancient
Philosophy 8 (1990), 19–65.
Benson, H. Socratic Wisdom (New York: Oxford University Press, 2000).
Bobonich, C. Plato’s Utopia Recast (Oxford: Clarendon Press, 2002).
Bostock, D. Plato’s Phaedo (Oxford: Clarendon Press, 1986).
Bostock, D. Plato’s Theaetetus (Oxford: Clarendon Press, 1988).

93 Thanks to Chris Bobonich, Lesley Brown, Dan Devereux, Terry Irwin, Christopher Shields, and
Christopher Taylor for helpful comments and/or discussion.
 Introduction   37

Brown, L. “How Totalitarian Is Plato’s Republic?,” in E. Ostenfeld (ed.), Essays on Plato’s

Republic (Aarhus: Aarhus University Press, 1998), 13–27.
Burnyeat, M. F. Introduction to the Theaetetus of Plato (Indianapolis: Hackett, 1990).
Burnyeat, M. F. “Plato on the Grammar of Perceiving,” Classical Quarterly 26 (1976), 29–51.
Burnyeat, M. F. “Protagoras and Self-Refutation,” Philosophical Review 85 (1976), 172–95.
Burnyeat, M. F. “Utopia and Fantasy: The Practicability of Plato’s Ideally Just City,” in J. Hopkins
and A. Savile (eds.), Psychoanalysis, Mind and Art (Oxford: Blackwell, 1992), 175–87. Reprinted
in Fine, Plato 2, ch. 13.
Burnyeat, M. F. “Plato on Why Mathematics Is Good for the Soul,” Proceedings of the British
Academy 103 (2000), 1–81.
Cohen, S. M. “The Logic of the Third Man,” Philosophical Review 80 (1971), 448–75. Reprinted
in Fine (ed.), Plato 1, ch. 10.
Cohen, S. M. “Socrates on the Definition of Piety: Euthyphro 10a–11b,” Journal of the History of
Philosophy 9 (1971), 1–13. Reprinted in G. Vlastos (ed.), The Philosophy of Socrates (Garden
City, NY: Anchor Books, 1971), ch. 8.
Cooper, J. “Plato on Sense-Perception and Knowledge (Theaetetus 184–6),” Phronesis (1970),
123–46. Reprinted in Fine (ed.), Plato 1, ch. 13.
Cooper, J. “The Psychology of Justice in Plato,” American Philosophical Quarterly 14 (1977),
151–57.
Dancy, R. Plato’s Introduction of Forms (Cambridge: Cambridge University Press, 2004).
Ferrari, G. F. R. City and Soul in Plato’s Republic (Sankt Augustin: Academia Verlag, 2003; and
Chicago: University of Chicago Press, 2005).
Fine, G. “Knowledge and Belief in Republic 5,” in Archiv für Geschichte der Philosohie 60 (1978),
121–39. Reprinted in Fine, Plato on Knowledge and Forms, ch. 3.
Fine, G. “False Belief in the Theaetetus,” Phronesis 24 (1979), 70–80. Reprinted, with minor
modifications, in Fine, Plato on Knowledge and Forms, ch. 9.
Fine, G. “Knowledge and Logos in the Theaetetus,” Philosophical Review 88 (1979), 366–97.
Reprinted, with minor modifications, in Fine, Plato on Knowledge and Forms, ch. 10.
Fine, G. “Forms as Causes: Plato and Aristotle,” in A. Graeser (ed.), Mathematics and
Metaphysics in Aristotle (Bern: Haupt, 1987), 69–112. Reprinted in Fine, Plato on Knowledge
and Forms, ch. 14.
Fine, G. “Knowledge and Belief in Republic 5–7,” in S. Everson (ed.), Epistemology (Cambridge
Companions to Ancient Thought 1: Cambridge: Cambridge University Press, 1990), 85–115.
Reprinted in Fine, Plato on Knowledge and Forms, ch. 4.
Fine, G. “Inquiry in the Meno,” in R. Kraut (ed.), Cambridge Companion to Plato (Cambridge:
Cambridge University Press, 1992), 200–26. Reprinted, with minor modifications in Fine,
Plato on Knowledge and Forms, ch. 2.
Fine, G. “Knowledge and True Belief in the Meno,” Oxford Studies in Ancient Philosophy __ 27
(2004), 41–81.
Fine, G. “Does Socrates Claim to Know That He Knows Nothing?,” Oxford Studies in Ancient
Philosophy 35 (2008), 49–88.
Fine, G. “The ‘Two Worlds’ Theory in the Phaedo,” British Journal for the History of Philosophy
24 (2016), 557–72.
Fine, G. On Ideas: Aristotle’s Criticism of Plato’s Theory of Forms (Oxford: Clarendon Press,
1993).
Fine, G. (ed.) Plato (Oxford: Oxford University Press, 2000).
Fine, G. (ed.) Plato 1: Metaphysics and Epistemology (Oxford: Oxford University Press, 1999).
38   Gail Fine

Fine, G. (ed.) Plato 2: Ethics, Politics, Religion, and the Soul (Oxford: Oxford University Press,
1999).
Fine, G. Plato on Knowledge and Forms: Selected Essays (Oxford: Clarendon Press, 2003).
Fine, G., The Possibility of Inquiry: Meno’s Paradox from Socrates to Sextus (Oxford: Oxford
University Press, 2014).
Gill, M. L. “Problems for Forms,” in H. Benson (ed.), A Companion to Plato (Oxford: Blackwell,
2006), ch. 13.
Gosling, J. C. B., and Taylor, C. C. W. The Greeks on Pleasure (Oxford: Clarendon Press, 1982).
Grote, G. Plato and the Other Companions of Socrates, vol. 4, new ed. (New York: Burt Franklin,
1973; reprint of the 1888 ed.).
Harte, V. “Knowing and Believing in Republic 5” in V. Harte and R. Woolf (eds.), Rereading
Ancient Philosophy (Cambridge: Cambridge University Press, 2017), 141–62.
Irwin, T. H., “Socrates and Euthyphro: The Argument and Its Revival,” in L. Judson and
V. Karasmanis (eds.), Remembering Socrates (Oxford: Clarendon Press, 2006), 58–71.
Irwin, T. H. Plato’s Moral Theory (Oxford: Clarendon Press, 1977).
Irwin, T. H. Plato’s Ethics (Oxford: Oxford University Press, 1995).
Kraut, R. “Reason and Justice in Plato’s Republic,” in E. N. Lee, A. P. D. Mourelatos, and
R. M. Rorty (eds.), Exegesis and Argument (Assen: Van Gorcum, 1973), 207–24.
Kraut, R. “Return to the Cave: Republic 519–21,” in Fine (ed.), Plato 2, ch. 9.
Lee, M. Epistemology after Protagoras (Oxford: Clarendon Press, 2005).
Meinwald, C. Plato’s Parmenides (Oxford: Oxford University Press, 1991).
Nails, D. The People of Plato (Indianapolis: Hackett, 2002).
Neu, J. “Plato’s Analogy of State and Individual,” Philosophy 24 (1971), 238–54.
Penner, T. “Thought and Desire in Plato,” in Vlastos (ed.), Plato 2, ch. 6.
Prichard, H. “Does Moral Philosophy Rest on a Mistake?,” Moral Obligation and Duty and
Interest (Oxford: Oxford University Press, 1968), ch. 1.
Sachs, D. “A Fallacy in Plato’s Republic?” Philosophical Review 72 (1963), 141–58; reprinted in
Vlastos (ed.), Plato 2, ch. 2.
Santas, G. Socrates: Philosophy in Plato’s Early Dialogues (London: Routledge and Kegan Paul,
1979).
Scott, D. Recollection and Experience (Cambridge: Cambridge University Press, 1995).
Scott, D. Plato’s Meno (Cambridge: Cambridge University Press, 2006).
Schwab, W. “Explanation in the Epistemology of the Meno,” Oxford Studies in Ancient Philosophy
48 (2015), 1–36.
Schwab, W. “Understanding epistêmê in Plato’s Republic,” Oxford Studies in Ancient Philosophy
51 (2016), 41–86.
Taylor, C. C. W. Plato’s Protagoras, 2nd ed. (Oxford: Clarendon Press, 1991), 162–70.
Vlastos, G. “The Third Man Argument in the Parmenides,” Philosophical Review 63 (1954),
319–49.
Vlastos, G. “Reasons and Causes in the Phaedo,” Philosophical Review 78 (1969), 291–325.
Reprinted in Vlastos (ed.), Plato 1, 322–66.
Vlastos, G. “Happiness and Virtue in Socrates’ Moral Theory,” in Vlastos, Socrates, 200–32;
and in Fine (ed.), Plato 2, ch. 4.
Vlastos, G. “The Individual as Object of Love in Plato,” Platonic Studies (Princeton, NJ: Princeton
University Press (1973), 3–34. Reprinted in Fine (ed.), Plato 2, ch. 5.
 Introduction   39

Vlastos, G. “The Socratic Elenchus,” Oxford Studies in Ancient Philosophy 1 (1983), 27–58.
Reprinted in Fine (ed.), Plato 1, ch. 1.
Vlastos, G. “Socrates’ Disavowal of Knowledge,” Philosophical Quarterly 35 (1985), 1–31. Reprinted
in Fine (ed.), Plato 1, ch. 2.
Vlastos, G. Socrates, Ironist and Moral Philosopher (Ithaca, NY: Cornell University Press, 1991).
White, N. P. “Plato’s Concept of Good,” in H. Benson (ed.), A Companion to Plato (Oxford:
Blackwell, 2006), ch. 24.
Williams, B. A. O. “The Analogy of City and Soul in Plato’s Republic,” in E. N. Lee,
A. P. D. Mourelatos, and R. M. Rorty (eds.), Exegesis and Argument (Assen: Van Gorcum,
1973), 196–206. Reprinted in Fine (ed.), Plato 2, ch. 10.
 chapter 2

Pl ato i n his Ti m e
a n d Pl ace

Malcolm Schofield

1. Socrates and the Fifth-Century

Enlightenment

As a young man—perhaps as a teenager—Plato became a member of Socrates’ intimate

circle, something he is careful to indicate himself in both Apology (34A) and Phaedo
(59b). After Socrates’ death, a number of those who had belonged to the group started
writing fictional dialogues (just how soon we don’t know) designed to illustrate his
character and personality, along with the distinctive themes and methods of Socratic
conversation: Phaedo, Euclides, Aeschines, for example, and probably Antisthenes,
already an intellectual heavyweight and prolific author, represented by Xenophon as
quite inseparable from Socrates (Mem. 3.11.17). Plato was to compose many more
“Socratic discourses” (Sôkratikoi logoi, as Aristotle calls them: Poet. 2, 1147b11) than
anyone, over a much longer time span—about 40 years, if we date Ion to the late 390s
and Philebus to the late 350s. Socrates was with Plato a continuing and dominating
obsession. Nearly the entire output of the most powerful and fertile thinker in the entire
tradition of Western philosophy is conceived as an homage to Socrates and in re-creation
of his philosophizing.1
At the very end of the Phaedo and its famous death scene, Plato has Phaedo say: “That
was the way our friend met his end, Echecrates—a man, as we would say, who of all in
his time that we had experience of was the best, and certainly wisest and most just”

1 A full discussion on Sôkratikoi logoi and their authors is in C. H. Kahn, Plato and the Socratic
Dialogue (Cambridge, 1996), ch. 1. Brief information on the Socratic authors listed here (and on the
many other thinkers mentioned in this chapter) is in D. Zeyl (ed.), Encyclopedia of Classical
Philosophy (Westport, Conn., 1997). Also useful for the chapter in general are G. C. Field, Plato and His
Contemporaries (London, 1930), and D. Nails, The People of Plato (Indianapolis, Ind, 2002).
42   Malcolm Schofield

(Phd. 118a). Those attributes as Plato saw it were the keynotes of Socrates’ life as they
were of his philosophical conversation. The harmony between the two was evidently
what made him irresistibly charismatic for those he captivated. In Plato’s case, it was
precisely those Socratic preoccupations—justice, the good, knowledge—that formed
his notion of the philosophical life. Doing philosophy meant trying to understand how
to live the life of a just person: getting rid of illusions about what we know or what we
think we want, and coming to see what living well really consists in. That is the mani-
festo Socrates enunciates in his speech to the jurors in the Apology (Ap. 28a–33c). That is
the theme Plato makes him elaborate and defend on a massive scale in the Republic,
longest and most complex of all his Sôkratikoi logoi.
Fundamental in what he took from Socrates is the idea that philosophy is an inquiry,
and inquiry best pursued in conversation with someone else.2 The conversation can be
of different sorts and can accommodate flights of fancy, as well as close questioning of an
interlocutor about the entailments of any views he may have advanced. Yet even when
Plato’s Socrates has ideas of his own to propound, they are expressly put forward for
others to consider—for acceptance, qualification, or rejection—not as teaching
imparted to those in need of instruction by someone secure in the knowledge of truth.3
No doubt the young Plato was so much in thrall to Socrates that it never seriously
occurred to him to think that philosophy ought to be more didactic or authoritarian, or
a system of doctrines rather than an activity. In various of the early dialogues, his
awareness of an alternative model of what education should be is nonetheless made
crystal clear. Over and again, Socrates is represented as clashing with those who take
education to be a matter of absorbing a mathêma, or body of knowledge, from someone
who commands the relevant technê, or expertise—as though acquisition of moral
understanding could be like learning medicine from a doctor or going to a sculptor to
pick up his craft.4
It is not that the Platonic Socrates rejects the conception of knowledge as technê. That
conception is omnipresent in the dialogues, from the earliest (like Ion and Hippias
Minor) to late works in which Socrates scarcely figures (notably, Sophist and Statesman).
Indeed, he introduces it into discussion in contexts where a modern reader would be
surprised to find it figuring at all. Take, for example, the idea expressed early in book I of
the Republic that justice is giving each individual his due, explicated as what is appropriate
for him (Rep. 1.331e–332d). Socrates compares the technê that is called medicine: What
does it give that is due and appropriate, and to what or whom does it give it? Having
obtained an answer to that question and an analogous one about cookery, he frames a
parallel question about justice: What would a technê have to deliver, and to what or
whom, if it were to deserve the name “justice”? Socrates doesn’t here necessarily assume

2 See now M. M. McCabe, Platonic Conversations (Oxford, 2015).

3 A classic example is Socrates’ introduction of his opinion about the Good in the Republic
(Rep. 6.506b–507a).
4 This issue is a major preoccupation in, for example, Laches, Protagoras, and Meno. Socrates’ classic
disavowal of any claim to be a teacher is in Ap. 33a–c (he practices examination of those who think they
know when they don’t).
 Plato in his Time and Place   43

himself that justice is a form of expertise comparable with medicine or cookery. His
question is hypothetical in form. But it could not have been articulated as it is, except in
an intellectual world where there was (1) a strong inclination to suppose that the value in
any valuable activity must derive from its being practiced knowledgeably, and (2) an
assumption that such knowledge must constitute a technê, or form of expertise.
Just such a world came into existence in ancient Greece in the last decades of the fifth
century—in other words, precisely during the period in which Plato sets the conversa-
tions that take place in his Socratic dialogues. The second half of the century saw an
explosion of prose writing on all manner of technical topics, from horsemanship to
perspective in painting for the dramatic stage. To this period belong the first medical
surviving treatises in the Hippocratic corpus5 and the first attempt we know of to
articulate elements of geometry, by Hippocrates of Chios.6 In a famous passage of the
Prometheus Bound ascribed to Aeschylus, the Titan catalogues the skills and crafts he
has taught mankind, from astronomy, numerical calculation, and writing to housing,
animal husbandry, navigation and medicine, divination and sacrifice, and the knowledge
and use of metals. “In one short word you may know all at once,” he concludes (PV 506).
“All tekhnai men owe to Prometheus.” This text—perhaps from the 440s—is not the only
piece of writing in this period to celebrate the range of tekhnai commanded by humans.
It reflects growing confidence in human ability to make discoveries and master nature.7
The Hippocratic author of On Ancient Medicine, for example, explains in his opening
chapters that medicine has made the discoveries he claims for it by following a principle
and a procedure; in fact, in essence, this is just the same procedure that has been
­followed for generations, as people have gradually learned better what sort of food and
drink prepared in what ways suit what sorts of constitution—not usually recognized as a
technê, to be sure, but a technê nonetheless.8 By the time Plato was writing, such
self-consciousness about what it is for a technê to be a technê had evidently become
commonplace. At the beginning of the Gorgias he parodies the mannerisms of writers of
guides to this or that technê by having Polus (author of such a work—on rhetoric: 462b)
declaim (448c): “Chaerephon, many forms of expertise among people have been

5 On the Hippocratic corpus, see G. E. R. Lloyd (ed.), Hippocratic Writings (Harmondsworth, 1978),
with useful bibliography. For general discussion, see G. E. R. Lloyd, Magic, Reason and Experience
(Cambridge, 1979), ch. 1, and J. Jouanna, Hippocrates (Baltimore, 1999).
6 On Hippocrates of Chios, see Lloyd, Magic, Reason and Experience, 102–15, with more general
discussion and bibliographical orientation in Lloyd, Demystifying Mentalities (Cambridge, 1990), ch. 3
7 A good survey of these fifth-century developments is in M. J. O’Brien, The Socratic Paradoxes and
the Greek Mind (Chapel Hill, N.C., 1967), ch. 2. See also Lloyd, Magic, Reason and Experience, ch. 3. The
Prometheus passage, together with a similar passage from Sophocles’s Antigone (332–71), is presented in
translation by W. K. C. Guthrie, A History of Greek Philosophy, Vol. 3 (Cambridge, 1969), 79–80. The
authenticity and date of Prometheus are disputed: see M. Griffith, The Authenticity of Prometheus Bound
(Cambridge, 1977).
8 There is a major edition of this treatise: M. J. Schiefsky, Hippocrates On Ancient Medicine (Leiden,
2005). Its defense of medicine as a technê has been much discussed. The topic is treated in two studies
published at the same time (which refer to previous bibliography): J. Barton, “Hippocratic
Explanations,” and F. Dunn, “On Ancient Medicine and Its Intellectual Context,” both in P. J. van der
Eijk (ed.), Hippocrates in Context (Leiden, 2005), 29–47 and 49–67.
44   Malcolm Schofield

­ iscovered by experience from experiences. Experience is what makes our life proceed
d
on the basis of expertise, inexperience on that of chance.”
The period of intellectual revolution I have been describing is often referred to as the
age of the Sophists.9 The polymath Hippias, treated by Plato as one of the leading figures
among the Sophists, certainly epitomized something of its spirit in his own person.
Astronomy seems to have been his favorite subject, but he was prepared to teach
­virtually anything, from mathematics, grammar, and music to what we might call
antiquarian subjects—although he has some claim to be considered the first historian of
philosophy (Hi. Ma. 285b–e; cf. Prot. 318d–e). The word “sophist” originally signified
(in George Grote’s magisterial formulation) “a wise man—a clever man—one who stood
prominently before the public as distinguished for intellect or talent of some kind.”10
Thus Herodotus in the fifth century bc calls the lawgiver Solon, the religious thinker
Pythagoras, and the Homeric seer Melampus all sophists (Histories 1.29, 2.49, 4.95).
“Sophist” never quite lost this general connotation, but in the pages of Plato, Xenophon,
and Isocrates, it has come to have a more specific meaning: an expert who would teach
you his subject for a fee.11 Thinkers such as Protagoras of Abdera, Prodicus of Keos, and
Hippias (from Elis in the Peloponnese), who traveled the Greek world to do just that,
were evidently salient presences in Athens around the time of the beginning of the
Peloponnesian War (431 bc).
One of Plato’s most elaborate dramatic masterpieces—the Protagoras—imagines
them all assembled together in Athens shortly before the outbreak of the war. He pits
Socrates in debate with Protagoras, initially on the subject of Protagoras’s educational
manifesto: the Sophist undertook to teach good decision-making, whether in running a
household or in the public sphere, where those he taught were to be equipped with an
exceptional capacity for the conduct and discussion of the affairs of the city (Prot.
318e–319a). How such a grandiose promise was to be honored is not clear.12 Much of the
Sophists’ teaching seems to have been conveyed in sustained set-piece performances.
Plato’s Protagoras gives an impressive demonstration speech in the dialogue, and
Prodicus (whose passion for precise distinctions between near synonyms is frequently
satirized by Plato: for example, Prot. 337a–c, 339d–342a, 358d–e; Charm. 163a–d; Crat.
384a–c) was celebrated for his lecture on the choice of Heracles, portrayed as a

9 The best guide to the Sophists is Guthrie, History of Greek Philosophy, Vol. 3, Part I. Another
view is in G. B. Kerferd, The Sophistic Movement (Cambridge, 1981). Recent treatments are offered by
R. Barney, “The Sophistic Movement,” in M.L. Gill and P. Pellegrin (eds.), A Companion to Ancient
Philosophy (Malden, Oxford, and Carlton, 2006), 77–97; and by C. C. W. Taylor and Mi-Kyoung Lee,
“The Sophists,” The Stanford Encyclopedia of Philosophy (Winter 2016 Edition), E. N. Zalta (ed.),
https://plato.stanford.edu/archives/win2016/entries/sophists/.
10 G. Grote, History of Greece (London, 1850), 8.479.
11 For example, Xen. Mem. 1.6.13; Guthrie, History of Greek Philosophy, Vol. 3, 35–40; and
D. L. Blank, “Socratics versus Sophists on Payment for Teaching,” California Studies in Classical
Antiquity 1 (1985), 1–49.
12 See the quizzical reflections of A. Ford, “Sophists without Rhetoric: The Arts of Speech in
Fifth-Century Athens,” in Y. L. Too (ed.), Education in Greek and Roman Antiquity (Leiden, 2001),
85–109.
 Plato in his Time and Place   45

­ aradigmatic figure at a crossroads in life who wins the struggle of virtue over vice
p
(Xen. Mem. 2.1.21–34).
Most people, says Plato’s Socrates, think that some young men get corrupted by
Sophists, and that there are some unprofessional Sophists who do the corrupting (Rep.
6.492a). His own line is that the whims of the Athenian people—in the assembly or in
the courts, on huge public juries—do much more damage. And while Plato would
probably not disagree that some Sophists harmed some individuals, the tone of the
Protagoras is mostly one of urbane amusement at the antics of the Sophists and their
followers, coupled with respect for Protagoras himself. Elsewhere, he has Socrates argue
that it is just not credible that someone such as Protagoras could have fooled the whole
of Greece and got away with making his students more depraved than they were when
he took them on, for more than 40 years (Meno 91E). In short, the suggestion is that the
Sophists’ reputation for good or ill is much inflated. To be sure, Plato does himself
engage with some of the ideas they generated (for example, Protagoras’s famous slogan:
“Man is the measure of all things,” in the Theaetetus).13 But in the early and middle
dialogues, the one important line of thought with a Sophistic pedigree that he confronts
(in different versions in the Gorgias and book II of the Republic) is the antinomian claim
that justice is a matter of convention (nomos) and will be ignored by anyone strong or
adroit enough to pursue self-interest as nature (phusis) would dictate—although it is not
clear that any Sophist actually advocated such behavior.14 Otherwise, it cannot be said
that the Sophists or their teaching loom that large in the dialogues, certainly by comparison
with the massive presence of Socrates himself. As W. K. C. Guthrie says, Plato “was a
post-war figure writing in an Athens of different intellectual temper. When he put on to
his stage the giants of the Sophistic era, he was recalling them from the dead.”15

2. Philosophy, Politics, and Athens

Socrates’ trial and condemnation by an Athenian court in 399 b.c., on charges of impiety
and immoral influence over young people, was devastating for Plato. It was not just a
personal trauma. In his mind, it constituted a confrontation that crystallized the
­inevitability of conflict between philosophy and politics and their incommensurable
assumptions. That issue, with its Socratic resonances, was to become one of central
significance in Plato’s treatment of the philosopher. It is highlighted at critical junc-
tures in some of the most important dialogues in the corpus. Some particular passages
serve to illustrate the point.

13 See Taylor, chapter 18 in this volume.

14 On Sophistic pedigree, see the surviving fragments of Antiphon’s On Truth, as presented in, for
example, Guthrie, History of Greek Philosophy, Vol. 3, 107–13, or D. J. Furley, Cosmic Problems
(Cambridge, 1989), ch. 6.
15 W. K. C. Guthrie, A History of Greek Philosophy, Vol. 4 (Cambridge, 1975), 6.
46   Malcolm Schofield

Nowhere are the rival claims of politics and philosophy more trenchantly advanced
than in the Gorgias. The dialogue begins with Socrates’ critique of rhetoric, but once
Callicles enters the discussion he reciprocates with a politician’s critique of philosophy.
In his famous monologue he warns Socrates that philosophy makes a person helpless to
defend himself in the public forum: if Socrates were brought before a court and faced
with an unprincipled prosecutor, he would end up dead if the death penalty was what
the prosecutor wanted (Gorg. 485E–486B).16 This thinly veiled prediction of Socrates’
actual fate is then reprised by Socrates himself near the end of the dialogue, where
he imagines himself as a doctor prosecuted by a pastry chef before a jury of children—
on a charge of ruining their health by his medicines and surgical interventions. The
doctor could find nothing to say in such a court in his self-defense (521E–522C). Of
course, Socrates did speak at his trial. Plato’s point is that there was nothing he could
have said then that could have begun to persuade the infantile citizenry of a self-indulgent
democracy.
In the Meno, Plato actually makes Socrates’ chief accuser—Anytus—a participant for
a while in the dialogue. Conversation turns to the question whether, if you want some-
one to acquire virtue, you should send him to a Sophist for training and instruction.
Anytus is outraged at the thought: Sophists corrupt the young—any decent father could
do a better job. But when Socrates points out that the great and the good—statesmen
such as Themistocles, Aristides, Pericles—signally failed to turn out sons of the same
caliber, Anytus advises Socrates to watch his tongue and his back, too. In Athens, as else-
where, it’s easier to do harm than good, he adds for good measure (Meno 94e). On that
note, he leaves. The whole passage is coded commentary on what Plato saw as the inco-
herent malice motivating the charge of corruption brought against Socrates at his trial.
There is further general reflection on the plight of the philosopher in the city, again
evoking the trial, in a famous passage in the Cave analogy in book VII of the Republic.
Socrates imagines a philosopher escaping from the cave, acquiring a true understanding
of reality—and then returning into the darkness once more. Such a person would find it
hard to reacclimatize. People would think he had damaged his eyesight. And if he tried
to free others, they would seize him and kill him if they could. He would make a fool of
himself if before reacclimatization he was forced to compete over shadows or images of
justice, in the law courts or anywhere else, with those who have never seen justice itself
(Rep. 7.516e–517e). The theme is replayed once more, and in very similar accents, in the
digression about the philosopher in the Theaetetus, a passage containing many echoes of
the Cave. Philosophers, says Socrates again, will only make fools of themselves if they
speak in the law courts (Tht. 172c). He goes on to develop—at length and in detail—a
contrast between the truly important and the trivial, then to argue the mutual incom-
prehension with which the philosopher (preoccupied with the one) and the rest of
humanity (mired in the other) view each other. The final words of the dialogue make the

16 Nietzschean affinities of Callicles’s speech are explored in an appendix to the great modern
edition of the dialogue: E. R. Dodds (ed.), Plato: Gorgias (Oxford, 1959).
 Plato in his Time and Place   47

implicit reference to Socrates’ own history all but explicit (210d): “Now I must go to the
King’s Porch to meet the indictment that Meletus has brought against me.”
These Athenian texts—obsessively replaying the demise of an Athenian philosopher
at the hands of the Athenian democracy—illustrate what Plato took to be the funda-
mental problem for all politics. But Athens and its democracy exerted over him a
compelling fascination. He himself was born into the Athenian aristocracy. And his
dialogues communicate an unforgettable sense of the high spirits, variety, and intel-
lectual freedom of Athenian aristocratic life—in the gymnasium, at the symposium, at
Sophistic performances, or just in private conversation—as Plato partly remembered
and partly imagined them during the years in which his philosophical dramas are
represented as being played out—that is, the last third of the fifth century bc. It may
seem paradoxical that such a vigorous aristocratic culture flourished—as, of course, did
Plato’s own writing and thinking—under a democracy.
But, in truth, the Athenian political settlement was always a complex negotiation
between mass and elite.17 In his funeral speech of 429 bc, Pericles, the aristocrat who
was the dominant figure in Athenian democratic politics in the 440s and 430s, remarked
(in the words Thucydides attributes to him) that in Athens’s meritocratic form of
democracy “we have provided for the mind many relaxations from exertions,” and again
“we cultivate beauty with economy and philosophy without enervation” (History 2.38, 40).
Loathe and despise democracy though he did, it is hard to suppose that Plato was alto-
gether unaware that the vitality and range of his own writing owed much to Athenian
intellectual life as he experienced it under the democracy during his formative years.
Plato’s vivid portrait of democracy and the democratic lifestyle in book VIII of the
Republic itself (Rep. 8.557a–564a) exhibits the color, energy, and variety that it is officially
deprecating, and a kind of intimacy, too, all in marked contrast with the external and
chilling account of oligarchy that has just preceded.
One thing Plato certainly communicates is a sense of the precariousness of the world
he describes. The dramatic date of the drinking party the inebriated Alcibiades bursts in
on in the Symposium is deliberately set a few months before the public outcry provoked
in 415 bc by events—the mutilation of the herms and the profanation of the mysteries—
in which he was implicated (along with many others, including, among those present,
Phaedrus, for example, and probably Eryximachus, too), and which were to be the
catalyst for his political downfall.18 The Charmides, set early in the Peloponnesian War,
ends with some menacing words from Charmides (Plato’s uncle) to Socrates. This is
doubtless designed to remind us that Charmides, incidentally someone else implicated
in the outrages of 415, would be involved with the oligarchic junta of the Thirty Tyrants

17 J. Ober, Mass and Elite in Democratic Athens (Princeton, N.J., 1989). Plato’s own stepfather and
guardian, Pyrilampes, was a friend of Pericles, active in the democracy’s public life, and called his own
son Demos.
18 Brief accounts are in S. Price, Religions of the Ancient Greeks (Cambridge, 1999), 82–85, and
P. J. Rhodes, A History of the Classical Greek World 478–323 bc (Oxford, 2006), 157–60. A full treatment
is in W. D. Furley, Andokides and the Herms: A Study of Crisis in Fifth-Century Athenian Religion
(London, 1996).
48   Malcolm Schofield

that seized power briefly in 404 (and took pains to silence Socrates), and that was led by
Critias, portrayed as Charmides’s mentor in the dialogue.
The two generals who figure as main participants in the discussion of the Laches—
Laches and Nicias—were both dead within a few years of its dramatic date. Nicias’s
acceptance in the dialogue of divination as a form of knowledge, and Socrates’ question
about its relation to generalship (Laches 195e–196a, 198e–199a), are clearly meant to
prefigure the disastrous decision that triggered the final debacle of the Sicilian expe-
dition in 413 bc: Nicias took an eclipse of the moon as a portent requiring delay in
departure (Thuc. Hist. 7.50). The Republic, too, is set at some point in the war’s duration.
The first two of those mentioned as accosting Socrates as he is leaving the Piraeus at the
beginning of the dialogue are Polemarchus and Niceratus. Both were to be executed by
the Thirty, who also confiscated the immensely profitable arms factory Polemarchus’s
father Cephalus—Socrates’ first main interlocutor in book I—had built up.19
So one could go on. Plato certainly did not think democracy (with the intellectual
world it sustained at Athens) was the only system of government liable to collapse under
the pressure of its own contradictory dynamic: witness the saga of regime change
sketched brilliantly in book VIII of the Republic. But the fragility of the world of the
dialogues and of the political system at Athens that supported it is surely an insistent
subtext. In the ideal communities delineated in Republic and Laws, life—including
intellectual life—is to be strictly controlled at every point. There will simply be no
potential for development of the exuberant proliferation of viewpoints of every kind,
and of the social structures enabling debate between them, which makes the dialogues
such attractive reading. Presumably, Plato concluded that that was the price that would
have to be paid for a secure political order—in key respects, more reminiscent of
unintellectual Sparta than of Athens—that would promote virtue and happiness. It was
a conclusion perhaps already implicit in the enthusiasm for the Spartan social and
political system fashionable among some Athenian aristocrats in his formative years
and shared by Socrates and his mother’s cousin Critias.

3. Light from the West

“When I first came to Syracuse, being then about forty years of age . . .” So writes the
author of the Seventh Letter (Ep. 7.324a); and whether he really is Plato or not, the letter’s
evidence that Plato made a first visit to Sicily around his fortieth year is more or less
universally regarded as reliable. Coupled with the usual dating of Plato’s birth to 427 b.c.
(D.L. 3.2), it yields a rough date of 387 for the visit, which on any reckoning must belong

19 On these and similar resonances of the Peloponnesian War in Plato, see M. Gifford, “Dramatic
Dialectic in Republic Book 1,” Oxford Studies in Ancient Philosophy 20 (2001), 35–106. Narratives and
analyses of the war are in S. Hornblower, The Greek World 479–323 bc, 3rd ed. (London, 2003), chs. 12
and 13, and Rhodes, History of the Classical Greek World, chs. 8–15. A fuller account is in D. Kagan, The
Peloponnesian War: Athens and Sparta in Savage Conflict, 431–404 bc (London, 2003).
 Plato in his Time and Place   49

somewhere in the early to middle 380s. The letter’s narrative focuses on the friendship
he formed with the young Dion, brother-in-law of the tyrant of Syracuse, Dionysius I, by
way of introduction to Plato’s entanglements in Sicilian court politics 20 years later. He
doesn’t indicate his motivation in making the voyage west, but the most obvious reason
is the one that has been transmitted and often repeated in ancient tradition: Plato was
wanting to make contact with the Pythagorean philosophers in South Italy (probably his
primary destination), and especially with Archytas in Tarentum.20
What was the outcome of that meeting of minds? Here is one way of telling the
story21—which construes the encounter as a decisive moment with extraordinary
impact on the future direction of Plato’s thought. To put it in a nutshell, Plato converted
to Pythagoreanism: to belief in the immortality of the soul; to a fascination with
eschatology and myths of a last judgment; to a conviction that mathematics held the
key to understanding the nature of reality; to the idea that politics might, after all, be
reshaped by philosophy and philosophers; to the resolve to create in Athens his own
community of friends dedicated to the pursuit of philosophy. From the conversion will
have flowed much of the energy and vision that fueled the writing of dialogues such as
Gorgias, Meno, Phaedo, and Republic. Its most practical consequence was to be the
founding of the Academy.
If this book were about a famous philosopher of the modern period, there would
probably be well-documented evidence of known date and in quantity supporting the
interpretation—which might still, of course, be controversial. For ancient Greek think-
ers, biographical facts are in short supply and hard facts almost nonexistent. Diogenes
Laertius tells us it was after returning from his travels abroad that the Academy gymna-
sium and its environs became the seat of his activities (D.L. 3.7). Otherwise everything is
more or less insecure inference. We have no absolute or even relative dates for the four
dialogues listed in the preceding paragraph. Issues relating to the chronology of the
dialogues are discussed elsewhere in this volume: suffice to say here that all modern
scholarship that is prepared to attempt a dating puts this quartet in that order, and (with
some hesitation or disagreement over Gorgias) makes their production subsequent to
Plato’s return from Italy and Sicily. Quite how far Plato’s preoccupation with mathematics
or philosopher rulers or even eschatology has a major Pythagorean dimension could
perhaps be disputed, as could the idea of a Pythagorean pedigree for the foundation of

20 The following all give fairly similar variants of this account: Phld. Acad. Ind. X.5–11; Cic. Rep. 1.16;
Fin. 5.87; V. Max. 8.7 ext. 3; and Olymp. in Alc. 2.86–93. Other variants include Apul. Pl. 1.3; D.L. 3.6
(which mentions not Archytas, but Philolaus and Eurytus); and Hier. Contra Rufinum 3.40. These and
yet further texts on the subject are collected (and translated) in C. A. Huffman, Archytas of Tarentum:
Pythagorean, Philosopher and Mathematician King (Cambridge, 2005), 272–74.
21 For example, Guthrie, History of Greek Philosophy, Vol. 4, 35–38 (with 9 n.1, 24 n.2, 284). But the
effect of Archytas’s personal influence is given greater stress, for example, by Dodds, Plato, 26–27, and
G. Vlastos, Socrates: Ironist and Moral Philosopher (Cambridge, 1991), 128–30. For a more recent
discussion, see C. Huffman, “Archytas,” in The Stanford Encyclopedia of Philosophy (Fall 2016 Edition),
E. N. Zalta (ed.), https://plato.stanford.edu/archives/fall2016/entries/archytas/.
50   Malcolm Schofield

the Academy.22 And we do not know that Plato met Archytas or other Pythagoreans on
his visit, or, consequently, that discussions with them had any effect on his thought at all.
So the hypothesis about Plato’s development sketched previously is undeniably
speculative. Nonetheless, it is reasonable speculation designed to give an economical
explanation of something that certainly calls for explanation. It is a striking fact (here we
can speak of fact) that the four dialogues under consideration share a preoccupation
with mathematics and the ultimate origin and fate of the soul that is entirely absent
from dialogues such as Ion, Hippias Minor, Euthyphro, Laches, and Protagoras (for
example), which are paradigmatically Socratic in method and content. What
accounts for the difference? A simple answer suggests itself: the newly registered
impact on Plato of powerful ideas encountered in an exotic non-Athenian religious
and intellectual environment.
From as early as the eighth century bc, the Greeks had been establishing settlements
on the coasts of South Italy and Sicily, which more or less rapidly achieved political
control over the hinterland and its indigenous inhabitants, with more gradual cultural
penetration. By the fifth century, cities such as Acragas and Syracuse, in Sicily, and
Croton and Tarentum, in South Italy, had become among the richest and most powerful
in the Greek world; “sybaritic” derives from the notoriously luxurious Sybaris, a city in
South Italy already destroyed in 510 bc. In many respects, the cities of these western
Greeks passed through phases of development comparable with those familiar from
mainland Greece. In religion, they were more distinctive. The surviving evidence indicates
a preoccupation with cults concerned with marriage, death, and the afterlife, often
associated with Demeter and Persephone. There are burials indicating that the deceased
were initiates into mysteries designed to achieve purification and a safe passage to a
better life in the hereafter, with “other famous initiates and bacchants,” as one gold plate
of the late fifth century discovered at Hipponion puts it.23
This was the world in which Pythagoras arrived, with Croton his destination, as a refugee
from Samos in the eastern Aegean, perhaps somewhere in the decade 535–25 bc. He
quickly became a charismatic figure whose life, work, and teaching are now the stuff of
impenetrable legend. There is no doubt, however, that the main focus of his teaching
was the soul and its place in the cosmic scheme of things—and the practices needed to
ensure that, after death and judgment, its journey through an inevitable cycle of rein-
carnation will bring it eventually to the isles of the blessed. Not only in Croton, but
elsewhere in South Italy, too, there formed groups of initiates into the austere
Pythagorean way of life, instructed in its doctrines and practices, which encompassed

22 For example, a Pythagorean model for the Academy is dismissed by M. Ostwald and J. P. Lynch,
“The Growth of Schools and the Advance of Knowledge,” in D. M. Lewis, J. Boardman, S. Hornblower,
and M. Ostwald (eds.), The Cambridge Ancient History, Vol. 6, 2nd ed. (Cambridge, 1994), 604.
23 For the text of the Hipponion gold plate (and some discussion), see G. S. Kirk, J. E. Raven, and
M. Schofield, The Presocratic Philosophers, 2nd ed. (Cambridge, 1983), 29–30. On western Greek
religion, see G. Zuntz, Persephone (Oxford, 1971).
 Plato in his Time and Place   51

everything from diet (where abstinence from beans was the most famous prohibition)
to sacrificial and funerary rites.24
At Croton (and probably in other cities), the Pythagoreans acquired considerable
political influence around the late sixth and early fifth centuries, although that dominance
had long since ended by the time of Plato’s visit; according to the fourth-century music
theorist Aristoxenus (who came from Tarentum), Pythagoreanism petered out in South
Italy (I think he means as a political force), with Archytas the one exception he mentions
(Iamb. VP 249–51). Archytas seems to have achieved a prominence in democratic
Tarentum, at the height of its considerable power, comparable with Pericles’s at Athens,
and like Pericles as general—probably seven years in succession, but probably also some
time after Plato’s visit (D.L. 8.79).
The name of Pythagoras is nowadays associated preeminently with a famous geomet-
rical theorem about right-angled triangles. But there is no credible ancient evidence
connecting him with the idea or practice of mathematics or with the identification or
solution of mathematical problems. The pioneers here, as in so many other fields of
inquiry, were the eastern Greeks in Asia Minor and the neighboring islands (the name
of Hippocrates of Chios has already been mentioned).25 What Pythagoras does seem to
have pressed is the idea that number and proportion—particularly, in the fundamental
harmonic ratios of 2:1 (the octave), 3:2 (the fifth), and 4:3 (the fourth)—were in some
symbolic way the key to understanding the universe.
Mysterious generality started to give way to the new style of mathematical inquiry of
the later fifth century in the work of the Pythagorean Philolaus (probably of Croton),
who developed a complete mathematical analysis of the diatonic octave, apparently
in the context of the theory of cosmic harmony.26 Archytas, however, is the first
Pythagorean known to us who was able to stand comparison with other leading
mathematicians of his day. His musical theory was devoted to analysis of scale systems
in terms of different means and proportions (arithmetical, geometrical, harmonic) and
to physical explanation of pitch expressible in terms of ratios. He was famous for his
solution of the problem of finding two mean proportionals to double the cube. Archytas
presented the study of music programmatically as the sister science of arithmetic,
geometry, and astronomy, as Plato seems to be acknowledging when he refers to this as

24 For Pythagoras and the Pythagoreans, see Kirk et al., Presocratic Philosophers, ch. 7. The
fundamental modern treatment is W. Burkert, Lore and Science in Ancient Pythagoreanism
(Cambridge, Mass., 1972); a briefer treatment is in C. H. Kahn, Pythagoras and the Pythagoreans: A
Brief History (Indianapolis, Ind., 2001).
25 On early Greek mathematics, see W. A. Heidel, “The Pythagoreans and Greek Mathematics,”
American Journal of Philology 61 (1940), 1–33; Burkert, Lore and Science in Ancient Pythagoreanism,
ch. 6; W. R. Knorr, “On the Early History of Axiomatics: The Interaction of Mathematics and
Philosophy in Greek Antiquity,” in J. Hintikka, D. Gruender, and A. Agazzi (eds.), Theory Change,
Ancient Axiomatics and Galileo’s Methodology (Dordrecht, 1981), 145–86.
26 On Philolaus, see Kirk et al., Presocratic Philosophers, ch. 11. Philolaus is discussed more fully in
Burkert, Lore and Science in Ancient Pythagoreanism, ch. 3, and the major edition of C. A. Huffman,
Philolaus of Croton: Pythagorean and Presocratic (Cambridge, 1993).
52   Malcolm Schofield

the Pythagorean view in appropriating it in book VII of the Republic (Rep. 7.530d). And
he claimed that “calculation” was the way to promote justice and political harmony.27
So when in a climactic passage of the Gorgias (Gorg. 507e–508a), utterly unlike
anything in Plato before, Socrates reports “the wise” as saying that “heaven and earth
and gods and men are bound together by community, friendship, orderliness, self-control
and justice,” which is “why they call the whole thing a world-order (kosmos),” and when
he invokes at this point the power of “geometrical equality,” his words are best explained
as an echo of the conversations Plato had recently enjoyed with Archytas and other
Pythagoreans. When the dialogue concludes with an eschatological myth about the
contrasting fates of souls who have lived lives of justice or injustice, this new destination
for a Platonic dialogue is again best explained as a reflection of the Pythagoreanism its
author had assimilated on his western travels—however much or little of the detailed
content of the story may owe to Pythagorean models. Most readers sense in Gorgias not
just a shift in philosophical direction but an insistent and radicalized urgency of tone
that was quite novel and perhaps unparalleled in Plato’s work. That cannot all be put
down to the passion of a new convert to Pythagoreanism, but conversion on his travels
might well have been the catalyst.

4. Critique of Rhetoric and Rivalry

with Isocrates

What the Gorgias is most urgent about is the choice between philosophy and politics
(or politics as it is currently conceived and practiced): how radical it is, how much is at
stake in making it. That is just the kind of focus we might have expected if the dialogue is
the most immediate product of Plato’s visit to Italy and Sicily, at any rate, given his reactions
to the hedonistic lifestyle and the conception of happiness he found prevailing there
(according to the Seventh Letter). Gorgias is the first of the dialogues to be preoccupied
with tyranny and the tyrant (as the supreme lawless hedonist). Readers have often
thought that the passage on the difficulty inherent in friendship with a tyrant (509c–511a)
encapsulates Plato’s reflections on Dion’s relationship with Dionysius I.28 But there is a
sense in which the entire dialogue grapples with the problem of tyranny. It is as though
Plato is now viewing Athens—which is foregrounded in the discussion with Callicles, in
particular—through lenses sharpened in Sicily. He looks for tyranny at home, and he
finds it in the ambitions of political rhetoric—which, as he portrays it, seeks not the
good of city or citizens but the manipulation and control of the populace by flattery, as
diagnosed by earlier writers such as Aristophanes and Thucydides. The Seventh Letter

27 Archytas is the subject of a major edition (with introductory essays) by Huffman, Archytas of
Tarentum. A more recent treatment of the debt Plato may have owed to Pythagorean mathematical
speculation is offered by P. S. Horky, Plato and Pythagoreanism (Oxford, 2013).
28 For example, Guthrie, History of Greek Philosophy, Vol. 4, 284 n.4.
 Plato in his Time and Place   53

presents a Plato already primed for comparative political analysis at the time of his stay
in Italy and Sicily. Gorgias shows us comparison in operation, as, for example, quite
explicitly in the long section on the resemblances between the orator and the tyrant in
the conversation with Polus (466a–471d).
The Gorgias attacks the credentials both of those who exercise political power by the
practice of rhetoric and of those who teach it (whether by performance, like Gorgias, or
through handbooks, like Polus). By starting with Gorgias and Polus (the teachers) and
finishing with Callicles (the practitioner),29 Plato makes a point: teaching, even by
someone as apparently benign as Gorgias, and practice, with all its corrupting potential,
form a dangerous continuum. For while Plato treats the Sophists mostly just as intellec-
tual poseurs, he sees rhetoric as a real force for harm. The message is illustrated in some
of the later pages of the dialogue, with what is effectively a counterhistory of Athenian
imperialism. Here the greatest of the orators on its political stage—Themistocles,
Cimon, Pericles—are accused of making Athens bloated and rotten and of deserving the
blame for its eventual downfall (Gorg. 515b–519a). Plato’s anger at the grossness of the
deception and self-deception needed to sustain Athenian democratic rhetoric seems to
have been fierce in the years after his return from Sicily. In the Menexenus, likely to have
been written in about 385 bc, he puts in Socrates’ mouth a pastiche funeral oration, said
to have been learned from Pericles’s mistress Aspasia (Menex. 235e–236c).30 By virtue of
blatant omissions and distortions, this blandly satirical composition paints a picture of
Athens’s entire history since the Persian Wars at the beginning of the fifth century right
down to the ignominious King’s Peace of 386 (some years after Socrates’ death, of
course) as one of noble and mostly successful endeavor. Rhetoric, we are to understand,
is both agent and expression of Athenian political bankruptcy.
There was one specific reason Plato might well have thought it timely to put the case
against rhetoric with all the force he could muster: the Athenian speechwriter Isocrates’s
decision around 390 bc to start taking pupils, marked by publication of his tract Against
the Sophists.31 The Gorgias is probably not a critique of Isocrates or Against the Sophists

29 On Gorgias, Polus, and Callicles as historical figures, see Dodds, Plato, 6–15; see also Guthrie,
History of Greek Philosophy, Vol. 3, 101–7, 192–200, 269–74; Kerferd, Sophistic Movement, ch. 8; and
R. B. B. Wardy, The Birth of Rhetoric (London, 1996), chs. 1–3.
30 For the Menexenus, see Guthrie, History of Greek Philosophy, Vol. 4, 312–23 and M. Schofield and
T. Griffith (ed. and trans.), Plato: Gorgias, Menexenus, Protagoras (Cambridge, 2010), xviii–xxiii; see
also the study of C. H. Kahn, “Plato’s Funeral Oration: The Motive of the Menexenus,” Classical
Philology 58 (1963), 220–34.
31 The writings of Isocrates (436–338 b.c.) survive and are most conveniently consulted in the
three-volume Loeb edition: G. Norlin (ed.), Isocrates, Vols. 1 and 2 (London, 1928 and 1929), and L. van
Hook (ed.), Isocrates, Vol. 3 (London, 1945). More recent translations with good bibliography are
D. Mirhady and Y. L. Too, (trans.), Isocrates I (Austin, Tex., 2000), and T. L. Papillon (trans.), Isocrates
II (Austin, Tex., 2005). Good brief studies are Ostwald and Lynch, “Growth of Schools and Advance of
Knowledge,” 595–602, and G. Kennedy, The Art of Persuasion in Greece (Princeton, N.J., 1963), 174–203.
Interactions between Isocrates and other Athenian thinkers, above all Plato, in the first half of the
fourth century b.c. are the subject of a valuable if often speculative monograph by C. Eucken, Isokrates:
Seine Positionen in der Auseinandersetzung mit den zeitgenössischen Philosophen (Berlin, 1983). Thus
Isocrates’s account of the Egyptian polity in the Busiris is read as a parody of the Republic (see also
N. Livingstone, A Commentary on Isocrates’ Busiris (Leiden, 2001)).
54   Malcolm Schofield

in particular,32 although at least one significant passage (Gorg. 463A, on the psychological
equipment of the orator) seems to turn Isocrates’s specific claims for rhetoric (Against
the Sophists 17) to its discredit. The dialogue is planned on an altogether grander design,
as an assault on rhetoric itself. That is why it is named after Gorgias, the first famous
exponent of rhetoric conceived as a technê, and why it makes this fifth-century figure the
initial target. Plato turns his guns much more narrowly and explicitly on Isocrates in
the Euthydemus, whose date of composition is disputed, but—echoing or pre-echoing
the Republic’s distinction between the geometer and astronomer and the dialectician as
it does—would probably have been written later than Gorgias and Menexenus.
The main body of the Euthydemus is devoted to a Socratic exposé of the logic-chopping
of a later generation of Sophists, here represented by the brothers Euthydemus and
Dionysodorus. But the frame dialogue is a conversation between Socrates and his old
friend Crito, who is represented as having been present for the encounter but out of
earshot (Euthd. 271a). In the final chapter at the end of the dialogue (304b–307c), Crito
relates a conversation he had when leaving with an unnamed person, identified as a
clever speechwriter who never appears in court himself, with a high sense of his own
wisdom. This description fits Isocrates exactly,33 and he is portrayed as confusing logic-
chopping with philosophy—a fair charge against Against the Sophists. Socrates makes a
damning assessment. Someone like that occupies the borderland between philosopher
and politician—neither one thing nor the other, and inferior to both. However such
people have a huge reputation for wisdom, except among real philosophers—whom it is
therefore in their interest (especially when their own pretensions are exposed) to
represent as no more significant than the likes of Euthydemus.
The likeliest reason Plato decided he needed to rebut the insinuation (the Euthydemus
makes the difference between Socratic philosophizing and logic-chopping crystal clear)
is that Isocrates’s school was by this time highly successful in training budding politicians
in oratory, as, indeed, we know it became.34 The dialogue doesn’t advertise the contrasting
merits of the Academy—in fact, it concludes with an injunction to give serious con-
sideration to philosophy itself and not bother with its practitioners, good or bad. But
readers might be expected to draw their own conclusions, alerted, perhaps, by the
reference to geometers and astronomers: geometry and astronomy are what Isocrates, at
any rate, later represented as the distinctive ingredients in the educational program of
the Academy (Antidosis 261–68).

32 But it is sometimes so taken, as in, for example, Ostwald and Lynch, “Growth of Schools and
Advance of Knowledge,” 605.
33 For example, W. H. Thompson, The Phaedrus of Plato (London, 1868), app. 2.
34 Isocrates is not a gripping writer, and his identification of philosophy with training for political
rhetoric (e.g., in Antidosis 270–96) is unlikely to appeal to most readers of this volume. But it seems
clear that as a teacher he was highly effective: see the interesting study of R. Johnson, “Isocrates’
Methods of Teaching,” American Journal of Philology 80 (1959), 25–36. A stimulating general study is
that of Y. L. Too, The Rhetoric of Identity in Isocrates: Text, Power, Pedagogy (Cambridge, 1995). A useful
recent collection of articles considering many aspects of Isocrates’s writings and thought is C. Bouchet
and P. Giovanelli (eds.), Isocrate: entre jeu rhétorique et enjeux politiques (Lyon, 2015).
 Plato in his Time and Place   55

Hard facts about the Academy are unsurprisingly in short supply.35 We should not
conceive of it as a school in any formal sense, with its own property and institutional
structures. However, Plato did acquire a house and garden in the vicinity of the gymna-
sium, where communal meals were probably taken. Did Plato take pupils? If so, not
(like Isocrates) for money. Ancient sources sometimes speak of “companions”
(e.g., Plu. adv. Col. 1126C). Perhaps we should think of a more or less loosely defined
society of friends (recalling the Pythagorean slogan, “friends share what they have”),
with younger adherents learning from the conversation of their seniors. Doubtless,
discussion would often be conducted in Socratic question and answer mode: in his
early Topics, Aristotle—a member of the Academy for 20 years—formulated rules for its
conduct. But mathematicians were among those attracted to the Academy, with
Eudoxus of Cnidos notable among them.36 While we should not assume that the
mathematical curriculum of book VII of the Republic was in any way replicated in its
modus operandi, anecdotes of Plato setting mathematical problems for Eudoxus and
others (e.g., What uniform motions will account for the apparently disorderly motions
of the planets?)37 and the interest in it reflected in the dialogues (e.g., Meno 82b–87c and
Rep. 7.529c–531c) suggest that mathematical questions will, indeed, have figured in the
discussions a good deal.
The rivalry between Isocrates and Plato persisted. Isocrates seems to have responded
to the Euthydemus by granting a distinction between those who (like the Socratic
Antisthenes) deny the possibility of falsehood and contradiction,38 and those who (like
Plato’s Socrates in the Protagoras) claim that all the virtues are a single form of knowledge
(see the beginning of his Helen, which is of uncertain date). But this is a distinction
without a difference: both groups are eristic paradoxmongers. It is much better, he says,
to venture reasonable opinions on useful subjects than to have exact knowledge of
useless ones. Plato, for his part, returned to a reconsideration of rhetoric in the Phaedrus.
His Socrates projects a rhetoric reformed by philosophy. That would, indeed, be a technê,

35 Many accounts of the Academy as an institution in standard works on Plato are rather
speculative: not exempt from the charge is Guthrie, History of Greek Philosophy, Vol. 4, 19–24. For a
corrective, see, e.g., H. Cherniss, The Riddle of the Early Academy (Berkeley, 1945), ch. 3. A lively and
balanced brief treatment is that of J. Dillon, The Heirs of Plato (Oxford, 2003), 1–16.
36 Eudoxus was notable for his development of the general theory of proportion expounded in book
V of Euclid’s Elements and for the elaborate theory of concentric spheres he devised to account for the
apparently irregular motions of the planets. See further W. K. C. Guthrie, A History of Greek Philosophy,
Vol. 5 (Cambridge, 1978), 447–57.
37 For the evidence, see A. Riginos, Platonica: The Anecdotes concerning the Life and Writings of Plato
(Leiden, 1976), 141–45. On the importance of problems in the development of Greek mathematics, see
W. R. Knorr, The Ancient Tradition of Geometric Problems (Boston, 1986).
38 Antisthenes’s intellectual activity spanned the fifth and fourth centuries; his literary output was
huge (D.L. 6.15–19), although nearly all of it is lost. While ethics was the main preoccupation, he also
engaged in Homeric interpretation and theorizing about language. He seems to have fallen under the
influence of both Gorgias and the Sophists, before becoming a devoted Socratic. For him, virtue was
sufficient for happiness—all that was needed was the strength of a Socrates (D.L. 6.11). See further,
M. Schofield, “Antisthenes,” in E. Craig (ed.), Routledge Encyclopedia of Philosophy (London, 1998),
1.314–17.
56   Malcolm Schofield

unlike the rhetoric of current theory and practice: “the art of speech by one who has
gone chasing after opinions, instead of knowing the truth, will be a comical sort of art,
in fact no art at all” (Phdr. 262c). The dialogue ends with some flattering words from
Socrates about the natural powers of the young Isocrates and the promise of philosophy
in him if he should become dissatisfied with his current activity (278e–279b)—a
backhanded compliment, if ever there was one. But the Phaedrus paradoxically begins
to exhibit in its prose style more of the deliberate avoidance of hiatus that had been
Isocrates’s constant trademark—and was clearly beginning to catch on more generally,
with Plato himself a total convert in the late dialogues. Isocrates, in fact, had the last
word, in his late and autobiographical Antidosis (353 bc). Here he makes the patronizing
concession that the sort of “philosophy” practiced by those who occupy themselves with
the exactness of geometry and astronomy is just training for the mind, a preparation
for philosophy—more advanced than what boys do in school but similar in most
respects (Antidosis 266).

5. Parmenides, Heracliteanism,
and the Theory of Forms

In chapter 6 of the first book of the Metaphysics (which surveys earlier thinkers’ views on
the first principles of things), Aristotle presents Plato as close to the Pythagoreans in
making numbers occupy a key place in metaphysical foundations. He is looking at
Platonic ontology through the lens of the late Philebus and Plato’s oral discussions of
the one and the indefinite dyad; more generally, Aristotle’s perspective is informed
by the Pythagorizing approach to metaphysics that prevailed in the Academy during
his membership in it. But unlike Speusippus or Xenocrates, he was intent on stress-
ing that there were important differences between Plato and the Pythagoreans.
Above all, as he saw it, the Pythagoreans assimilate numbers and the contents of the
sensible world, whereas Plato holds that numbers have an existence separate from
sensible things.39 The first main section of the chapter is accordingly devoted to a
narrative explaining how Plato came to “separate” the Forms (how Forms relate to
numbers is deferred until later).

39 There are brief accounts of the work of Speusippus and Xenocrates in Guthrie, History of Greek
Philosophy, Vol. 5, 457–83; more extended treatments are in Dillon, Heirs of Plato. Speusippus, Plato’s
nephew and successor, seemed to have been the more interesting thinker of the two, particularly
notable for his hypothesis of different but analogous pairs of principles explaining successive levels of
reality (e.g., numbers, magnitudes, soul)—on which, see R. M. Dancy, Two Studies in the Early
Academy (Albany, N.Y., 1991), 63–119, 146–78. Xenocrates developed the Symposium’s idea of daimones,
spiritual beings mediating between gods and humans (Symp. 202e–203a). Collections of the evidence
are L. Tarán, Speusippus of Athens (Leiden, 1981), and M. Isnardi Parente, Frammenti: Senocrate,
Ermippo (Naples, 1982).
 Plato in his Time and Place   57

Here is Aristotle’s narrative (Metaph. A6, 987a32–b10):

In his youth he [Plato] had become familiar first of all with Cratylus and with
Heraclitean views to the effect that all perceptible things are always in flux, and there
is no knowledge that relates to them. This is a position he later subscribed to in these
terms. Socrates, on the other hand, engaged in discussion of ethics, and had nothing
to say about the general system of nature. But he was intent on finding out what was
universal in this field, and was the first to fix his thinking on definitions. Plato
followed him in this, and subscribed to the position that definition relates to
something else, and not to the perceptibles—on the kind of grounds indicated: he
thought it impossible for there to be a common definition of any of the perceptibles,
since they were always changing. Plato, then, called these kinds of realities ‘‘ideas,’’
and claimed that the perceptibles were something in addition to them, and were all
spoken of in terms of them—what he said was that by virtue of participation, the
many shared their names with the forms.

The gist of Aristotle’s account is clear. Plato accepts Socrates’ view that knowledge as
articulated in definitions must relate to something universal; takes over the Heraclitean
view that there can be no such knowledge of perceptibles, because they are always
changing—they have no definite or at any rate definable nature; and so posits Forms
separate from perceptibles as the realities to which definitions do apply. But while the
general thrust of the passage is not in doubt, it prompts questions:

1. What exactly is Aristotle claiming about Plato’s relationship with Cratylus?

2. Whatever the claim, is it likely to be true?
3. Was subscription to Heracliteanism really a key component in Plato’s motivation
for positing Forms?

I shall deal summarily with points 1 and 2 and at greater length with point 3.40

1. Aristotle is certainly saying that when Plato was young he got to know Cratylus,
and through him the Heraclitean theory of flux. Nevertheless, it is not claimed in
so many words that Cratylus was his “teacher.” Is it being stated or suggested that
he got acquainted with Cratylus before joining Socrates’ circle? The answer turns
on what is meant by “first of all”: first in temporal sequence, or the first point in
Aristotle’s exposition? I don’t think we can be sure, although the second option
better fits my sense of the flow of the passage. In any case, the issue will not be of
much moment for our purposes.

40 On Cratylus, see D. N. Sedley, Plato’s Cratylus (Cambridge, 2003), 16–23. A preoccupation with
the flux of the perceptible world is not self-evidently what was fundamental in the philosophy of
Heraclitus: for example, Kirk et al., Presocratic Philosophers, ch. 6; C. H. Kahn, The Art and Thought of
Heraclitus (Cambridge, 1979). But Aristotle correctly interprets the way Heracliteanism was construed
by Plato. See further, chapter 11 of this volume.
58   Malcolm Schofield

2. It has been suspected that Aristotle doesn’t really have biographical information
at his disposal but is simply extrapolating from the end of Plato’s Cratylus.41 There
Cratylus at least ends up a Heraclitean, and Socrates argues against a thoroughgoing
Heracliteanism that if there were knowledge, it would have to relate to entities
such as “the beautiful itself,” which is always such as it is and cannot therefore be
in flux (Crat. 439b–440e). This passage is surely what Aristotle uses to interpret
the way Plato came to use the Heracliteanism he learned about from Cratylus.42
But Aristotle knows things about Cratylus—an obscure figure—not to be found
in Plato (e.g., his famous criticism of Heraclitus: you can’t step into the same river
even once [Metaph. G5, 1010a7–15]); he doesn’t actually need the biographical
claim for his main purpose—to explain Plato’s motivation for positing Forms, and
the way he highlights it at the beginning of the passage suggests someone who
thinks he has real news to impart. The verdict must be that Aristotle was told it by
someone he had reason to think reliable—conceivably, Plato himself.
3. The reason for mentioning Plato’s early familiarity with Cratylus and
Heracliteanism is evidently its significance in the light of what later transpired.
Plato had got to know the Heraclitean theory of flux and to understand its
consequences for knowledge when young. But it was only later, when reflecting
on Socrates’ search for definitions (the focus of the early dialogues), and when
puzzling about the nature of knowledge on his own account in consequence, that
he came to put the Heracliteanism he had imbibed from Cratylus to philosophical
work—in concluding that Forms, not perceptibles, must be the object of knowledge
and definition. The key question is whether this is a believable account of the
origin of the theory.

Nobody doubts that Plato posited Forms as a consequence of reflection on Socrates’

definitional “What is X?” questions. One of his preferred locutions for referring to a
Form is to have Socrates speak of “the very thing that X is.” That formula is not itself a
definition. Instead, it specifies what it is that would be captured by an adequate defini-
tion of X if we could find it. In other words, there has to be “the very thing that X is” if
there is to be the kind of definitional knowledge of what all Xs have in common that
Socrates was looking for. That, then, prompts the further question: What kind of thing is
the very thing that X is?
Well-known texts in the so-called middle dialogues address themselves to this issue
(notably, Symp. 211a–b; Phd. 65d–66a, 74a–c, 78c–79e). In doing so, they avail them-
selves of the radical thought—at the core of the argument of the great metaphysical
poem of the Presocratic Parmenides of Elea (“much of Plato’s philosophy,” especially his

41 For example, G. S. Kirk, “The Problem of Cratylus,” American Journal of Philology 72 (1951),
225–53, and Kahn, Plato and the Socratic Dialogue, 80–83.
42 Although if he did, he read somewhat more into Plato’s text (even supplemented by, e.g., Phd.
78c–e) than it literally contains: see T. H. Irwin, “The Theory of Forms,” in G. Fine (ed.), Plato 1:
Metaphysics and Epistemology (Oxford, 1999), 149–52—a study that explores different ways in which
Heraclitean flux is understood by Plato.
 Plato in his Time and Place   59

later philosophy, “is unimaginable without the towering figure of Parmenides”)43—that

if reality is to be the object of knowledge, then insofar as it is knowable, it must be what it
is without qualification: it cannot not be in any respect at all (Rep. 5.476e–477a). Aristotle
does not mention Parmenides in his account of the rationale underlying the theory of
Forms (he confines himself to figures with whom Plato had significant early encounters).
But Plato’s encounter with Parmenides’s thought was clearly decisive in this context.
Parmenides’s central argument is that only what is is an intelligible object of thought—
use of the expression “is not” attempts something impossible: specifying not a some-
thing but nothing (Fr.2; cf. Fr.3, Fr.6.1–2). He then develops the consequences. Reality
cannot come into being or pass away, or change or move, or exhibit any variation or
imperfection, since to represent it as subject to any of these processes or conditions
would require explicit or implicit use of “is not” (Fr. 8.1–49). The perceptible world as
represented in ordinary human beliefs about reality fails to satisfy these constraints.
“Mortal opinions,” as Parmenides calls them, are desperately confused because they roll
up “is” and “is not” together—they fail to make the critical decision between the two
(Fr. 6 and 7; cf. Fr. 8.15–18).44
In book V of the Republic, Plato follows Parmenides in his characterization of opinion
as the mental condition of ordinary people who take what their senses tell them for
knowledge (Rep. 5.474b–480a). Nonetheless, it sounds from his formulations as though
he construes the state of the perceptible world primarily in terms of Heraclitean flux
rather than any corresponding Parmenidean category—so that Aristotle may well have
been right in construing reflection on Heracliteanism as a key metaphysical ingredient
in the motivation for positing Forms. Thus in the account of the beautiful in the
Symposium, Diotima begins (211a): “First, it always is and neither comes to be nor passes
away, neither waxes nor wanes”; and there are generic formulations in terms of change
or coming to be and passing away in the Phaedo (Phd. 78c–e) and the Republic (Rep.
6.485b). The reality with which philosophers are concerned in asking “What is X?” is
specified in these texts by means of a contrast—implicit or explicit—with the Heraclitean
flux of the perceptible world.
None of these texts makes the argument that there cannot be knowledge of what is in
flux, so knowledge has to be concerned with entities—the Forms—quite separate from
the coming to be and passing away of perceptible things. For anything like that, we have
to wait until the end of the Cratylus. On the supposition that (as many scholars think
probable) Cratylus postdates the three dialogues just mentioned, we might hypothesize

43 Guthrie, History of Greek Philosophy, Vol. 4, 35.

44 For presentation and discussion of the fragments of Parmenides’s poem, see Kirk et al., Presocratic
Philosophers, ch. 8. A good brief account is in Sedley, “Parmenides,” 7.229–35. A seminal modern
interpretation is G. E. L. Owen, “Eleatic Questions,” Classical Quarterly 10 (1960), 84–102. The major
edition is A. H. Coxon, The Fragments of Parmenides (Assen, 1986). Philosophical studies that
illuminate Parmenides’s influence on Plato are I. Crystal, “Parmenidean Allusion in Republic V,”
Ancient Philosophy 16 (1996), 351–63; P. K. Curd, The Legacy of Parmenides: Eleatic Monism and Later
Presocratic Thought (Princeton, N.J., 1998), ch. 6.2; and J. A. Palmer, Plato’s Reception of Parmenides
(Oxford, 1999) (see ch. 4 on the Parmenidean dimension of Rep. 5.474b–480a).
60   Malcolm Schofield

that Plato here acknowledges the debt the metaphysics of those dialogues owes to the
Heraclitean view of the perceptible world that he had first got to know in Cratylus’s
company all those years ago. Perhaps, indeed, it was only belatedly—after writing
Symposium, Phaedo, and Republic—that he came to appreciate the importance of his
conversations with Cratylus in shaping his approach to questions of metaphysics and
epistemology when he turned eventually to tackle them. In general, Cratylus emerges
from the Cratylus as one of the least impressive thinkers put on stage in the dialogues. It
is at least a pleasant thought that Plato used its last couple of pages to flag up what he
nonetheless now realized he learned from him.

6. The Academy and

the Late Dialogues

The literary and philosophical temper of Plato’s late dialogues, as every reader notices, is
much changed from the writings that precede them. Their hiatus-free prose can be
extraordinarily crabbed and involved; they are comparatively lacking in dramatic life
and color; anonymous and anonymized figures—the Eleatic Visitor (in Sophist and
Statesman), the Athenian Visitor (in the Laws)—conduct most of the relatively wooden
conversation (Socrates has a lead role only in Philebus), while Timaeus (in Timaeus) and
Critias (in Critias) resort to uninterrupted monologue. There is less sense of contextual-
ization of philosophical dialogue within a real world. Dialogues such as the Sophist,
Statesman, and Philebus (and, earlier, the Parmenides) read like texts for the Academy
and, indeed, reflect discussion within the Academy.
In the case of the Sophist and Statesman, the argument for that hypothesis derives
almost wholly from their combination of pedagogical and didactic concern to instill
understanding of correct dialectical method with forbiddingly abstract or technical
content. The Parmenides exhibits the same combination, but its preoccupation with
the critique and proper interpretation of the theory of Forms locates it within a
well-documented debate in the Academy about Forms, to which Eudoxus,
­
Speusippus, Xenocrates, and Aristotle all contributed (neither Aristotle nor Speusippus
accepted Plato’s theory in any version).45 Whether Plato wrote the Parmenides to ini-
tiate debate, or whether he was responding to an incipient or ongoing controversy
already launched (as one might conjecture from the dialogue’s consideration of the

45 For accounts of the metaphysical issues and positions that preoccupied the early Academy, see
W. D. Ross, Plato’s Theory of Ideas (Oxford, 1951), chs. 9–17; Burkert, Lore and Science in Ancient
Pythagoreanism, ch. 1; and J. Annas, Aristotle’s Metaphysics, Books M and N (Oxford, 1976). For
treatments of the contributions of Speusippus and Xenocrates, see also the literature referred to in note
38 of this chapter; for Eudoxus, consult Guthrie, History of Greek Philosophy, Vol. 5, 452–53, and
M. Schofield, “Eudoxus in the Parmenides,” Museum Helveticum 30 (1973), 1–19; for Aristotle, see
G. Fine, On Ideas: Aristotle’s Criticism of Plato’s Theory of Forms (Oxford, 1993).
 Plato in his Time and Place   61

possibility—associated by Aristotle with Eudoxus—that Forms are immanent in

­particulars: Metaph. A9, 991a12–20), we do not know. But he must have intended it
primarily for a readership within the Academy. The star example of such a dialogue,
however, is the Philebus.
Philebus is a heady and often esoteric mixture of ethics, methodology, and metaphysics.
The methodology and the ontology—couched as they are in terms of number, ratio,
limit, the unlimited—reflect the Pythagorizing turn in Academic metaphysical speculation
that is attested particularly in books M and N of Aristotle’s Metaphysics. This is especially
true for Speusippus, but there is also Pythagorizing in Plato’s own “unwritten doctrines,”
with which the ontology of the Philebus has an apparent affinity.46 In making pleasure
and arguments for and against hedonism the focus of the dialogue’s ethical enquiry,
Plato was not merely participating in an Academic debate but acting as adjudicator
(in fact, the dialogue represents itself as awarding prizes—in rank order—to the most
convincing candidates for what determines the goodness of a good life). We know—
again primarily from Aristotle—that Eudoxus argued for hedonism and Speusippus
against it. What Aristotle tells us about Eudoxus’s hedonism (in book 10 of the
Nicomachean Ethics) makes it very probable that it is his account of pleasure as the good
that Plato is reproducing for discussion in the Philebus (Phlb. 20c–21a). Later in the dia-
logue, various other thinkers who provide ammunition for anti-hedonist conclusions
are referred to under designations such as the “difficult” people (44b–e) or the “subtle”
people (53c). It is tempting to try to strip them of their anonymity (could “difficulty” be
what Speusippus was known for?).47 But whether identifications can be secured or not,
there is no mistaking that Plato here introduces contemporary voices—presumably
Academic—into the argument.
How much did Plato’s own thinking as evidenced in the late dialogues owe to the
other leading philosophers who worked with him in the Academy? Everything we have
seen of Plato in this chapter suggests that he was someone whose philosophizing was
invariably nourished by engagement with his immediate intellectual and political
environment. His originality lies in the versatility and fertility of his response to it. So—
to return for a moment to the Philebus—the very idea of writing once again about
­hedonism (already discussed in Protagoras, Gorgias, and book IX of the Republic),
but within a framework shaped by Pythagorizing metaphysics, would surely never

46 The music theorist Aristoxenus reports a famously unintelligible lecture by Plato on the Good
(Harmonics 2.30.16–31.3), which is often connected with Aristotle’s reference to “unwritten doctrines”
of Plato (Phys. 4.2, 209b11–16). It is generally supposed that these must have included an idea he makes
central to Platonic metaphysics in Metaph. A6 and debates at length in Metaph. M and N: that numbers
(identified with Forms) are to be analyzed in terms of a formal principle (the one) and a material
principle (the large and the small). The Tübingen school of Platonic interpretation sees this as the true
centerpiece of Plato’s entire philosophy, only hinted at in the dialogues; at the other extreme is the
thoroughgoing skepticism of Cherniss, Riddle of the Early Academy, chs. 1 and 2. For a balanced and
informative review of the evidence and the controversy (with ample bibliography), see Guthrie, History
of Greek Philosophy, Vol. 5, ch. 8.
47 See M. Schofield, “Who Were oἱ Δυσχερεῖς in Plato, Philebus 44Aff?” Museum Helveticum 28
(1971), 2–20, 181.
62   Malcolm Schofield

have occurred to him without the stimulus of arguments about these issues in the
Academy. The great cosmological project of the Timaeus must owe much “to the
research of and discussion with other members of the Academy in the 350s, especially
mathematicians and astronomers.”48 It is possible to suspect the impact of much more
specific ideas generated there, too. The Statesman makes central to its concept of
statesmanship something that might well take the reader of the Republic by surprise: the
idea that the knowledge required by someone involved in practical activity must be a
capacity for measured judgment of what is appropriate and timely—in short, of what is
“removed to the middle from the extremes” (Statesman 284b–e). Is this an entirely
spontaneous innovation by the elderly Plato? Or is it his appropriation of a theory of
virtue as occupying a mean determined by practical knowledge that had already been
worked out by the young Aristotle?49

7. The Laws in its Time and Place

The Laws was evidently designed for a wider public—although it has sometimes been
suggested that, nonetheless, its existence bears an intimate relationship with the purposes
for which the Academy existed. G. R. Morrow, one of the leading twentieth-century
authorities on the dialogue, is one of many distinguished scholars convinced that the
Academy was a school for statesmen, which prepared its members for the role “by the
study of Greek law and politics,” inter alia.50 T. J. Saunders, another major authority on
the Laws, believes we can infer from it the sorts of policies and procedures that Academic
political “advisers” would have been taught to recommend to those who consulted
them.51 But while some who had associated with Plato in the Academy did become
involved in the politics mostly of their home cities, as might be expected of aristocrats,
the case is flimsy for seeing them as emissaries from the Academy primed for their task
in the way Morrow and Saunders imagine, or for thinking the Academy had its own
political agenda.52 No doubt, its members talked politics during their stay in the
Academy. To judge from the evidence of other late dialogues, however, philosophical
discussion would have been devoted mostly to questions of metaphysics and ethics and
to the dialectical methods appropriate for tackling them.

48 Ostwald and Lynch, “Growth of Schools and Advance of Knowledge,” 609.

49 For example, M. Schofield, “The Disappearance of the Philosopher King,” Proceedings of the
Boston Area Colloquium in Ancient Philosophy 13 (1997), 224–26.
50 G. R. Morrow, Plato’s Cretan City: A Historical Interpretation of the Laws (Princeton, N.J., 1960), 5.
51 T. J. Saunders, “ ‘The RAND Corporation in Antiquity?’ Plato’s Academy and Greek Politics,” in
J. H. Betts, J. T. Hooker, and J. R. Green (eds.), Studies in Honour of T. B. L. Webster (Bristol, 1986),
1.200–10.
52 For example, P. A. Brunt, “Plato’s Academy and Politics,” Studies in Greek History and Thought
(Oxford, 1993), ch. 10; M. Schofield, “Plato and Practical Politics,” in C. J. Rowe and M. Schofield (eds.),
The Cambridge History of Greek and Roman Political Thought (Cambridge, 2000), ch. 13.
 Plato in his Time and Place   63

What seems hard to doubt is that the Laws was written with practical intent, as a guide
to the principles that should inform the communal life of a well-ordered Greek city and
as a blueprint for their detailed implementation on a monumental scale. “No work of
Plato’s,” said Morrow, “is more intimately connected with its time and with the world in
which it was written than the Laws.”53 This huge dialogue is dense with reference explicit
and (more often) implicit to the political and sociocultural institutions and practices of
the Greek city-state, and, above all, of Athens itself. In fact, Plato’s extensive and intricate
legal code is a reworking of contemporary Athenian law, embodying a radical new utili-
tarian penology based on the Socratic view that, since nobody does wrong willingly,
criminality is a disease (Gorg. 466d–480d). In consequence a much more inquisitorial
form of procedure before the courts was in his view required, reducing the scope for the
rhetoric the Gorgias thought so pernicious.
One particularly fascinating dimension of the code is its elaborate and un-Athenian
differentiation of penalties for offenses, according to whether the perpetrator is a citi-
zen, a slave, a temporary visitor, or a long-term resident alien (a metic). As in Athens
(but not Sparta), Plato allows for a class of metics: persons such as Cephalus in the
Republic, needed for occupations regarded in the Laws as harmful to the soul—notably
commerce. But unlike at Athens, their residence is to be subject to a time limit (20 years).
The way their alien status is marked can be illustrated from the highly baroque structure
of laws covering assault (Laws 9.879b–882c). Mostly, the metics are to be subject to more
severe penalties than citizens. Some prominence is given to the rule that if a foreigner
whose assault on a citizen can be proven—after an examination that pays proper respect
to the god who protects foreigners—to have been designed to insult and humiliate, he or
she is to be subjected to as many strokes of the lash as the blows he or she inflicted in
order to put a stop to “foreign uppitiness” (thrasuxenia: a word, as Saunders points out,54
that occurs nowhere else and was evidently coined for this occasion). There is no rule
covering citizen behavior of this sort (in Athens, imprisonment and loss of citizen rights
was probably the penalty).
Morrow went so far as to describe the society Plato was intent on defining through his
legislative template as “an idealized Athens.”55 Certainly, there is an explicit preoccupa-
tion with Athens (as with Sparta and Persia) in book III, where a historical approach is
taken to the task of working out what the ideal social and political system would be like.
It begins with the flood and the emergence of the first simple postdeluvian communities,
and ends with a discussion of Athens and Persia as societies that, in the past, combined
the three prime desiderata of wisdom, freedom, and friendship by balancing in their
system of government a monarchic with a democratic principle. Since the time of Cyrus
the Great (in Persia) and, less explicitly, Solon (in Athens), the balance has become

53 Morrow, Plato’s Cretan City, xxix.

54 T. J. Saunders, Plato’s Penal Code (Oxford, 1991), 275 n.45.
55 Morrow, Plato’s Cretan City, 592. But he did not underestimate the importance of the Spartan and
Cretan model for the ideal city of the Laws—it is, after all, to be a closed and tightly controlled society.
Uniquely, the dialogue is set not in Athens but in Crete, where the Solonian figure of the Athenian
Visitor can find sympathetic, if challenged, recipients for the blueprint he proposes.
64   Malcolm Schofield

fatally disturbed. Persia has degenerated into tyranny, Athens into what Plato calls
“theatocracy”: the self-indulgence of a society under the control of the illusion that
anybody’s judgment is as good as anybody else’s.
Thirty years on Plato here plays new variations on the old analysis of the malaise of
Athenian democracy familiar from the Gorgias. But the contrast between Solonian
Athens and the decline since the days of Marathon has a contemporary flavor. Around
the time the Laws was being composed, Isocrates was vainly appealing for the rein-
troduction of what he called “the democracy bequeathed by our ancestors” in his
Areopagiticus, written (probably in 355 bc) as a wake-up call to Athens in the aftermath
of its second brief attempt to sustain an empire.56 In the time of Solon and Cleisthenes,
Athens enjoyed a balanced, well-ordered constitutional settlement, which did not as
now educate the citizens “to regard licentiousness as democracy, lawlessness as freedom,
outspokenness as equality, and the license to do these things as happiness” (Areop. 20).
The Laws has a different agenda, but it breathes the same air.
An autobiographical dimension to the Laws has often been perceived.57 In the 360s,
Plato had made two further visits to Sicily, to the court of the young Dionysius II: one
probably in 366, very soon after his accession to power after his father’s death; the other
in 361. Both were undertaken to oblige Dion, who had hopes of influencing the new
tyrant and, initially (according to the Seventh Letter), of Plato’s turning him into a
philosopher ruler (Ep. 7.327b–328d). Both were wretched failures, with Dionysius turning
out to be a dilettante in philosophy and interested only in using Plato in regional political
machinations (again according to the Letter).58 The Laws notoriously gives no room to
the aspiration for rule by philosophers articulated in the Republic, and the dialogue
insists that absolute power will almost inevitably bring about the moral corruption of
anyone who wields it. Although the Republic itself has plenty to say about the corruptions
of power, and the corruptibility especially of those naturally endowed for philosophy
(Rep. 6.491a–495b), and although the Laws is designed to work out an approximation to
what the Republic always conceived as a scarcely feasible ideal, readers have diagnosed
personal disillusionment on Plato’s part. It is certainly hard to think that his recent
Sicilian experience did not somehow color his thoughts about tyrants in the Laws.
At any rate, it is interesting that when in book IV he comes to sketch the preconditions
that might favor the creation of a well-ordered polity, he specifies a location well inland,
far from any port (Syracuse had a great harbor), and a young tyrant prepared to work
with a lawgiver—thanks to his “orderly” character (that sounds ironic to the point of
sarcasm).59

56 The political history of the period is a tangled tale: for example, Hornblower, Greek World, chs. 16
and 17. A helpful brief account of the “second Athenian confederacy” is available in S. Hornblower and
A. Spawforth (eds.), The Oxford Classical Dictionary, 4th ed. (Oxford, 2012), 1337.
57 A classic statement of this interpretation is G. Vlastos, “Socratic Knowledge and Platonic
‘Pessimism,’ ” Platonic Studies (Princeton, N.J., 1973), ch. 9.
58 For a historical narrative, see Rhodes, History of the Classical Greek World, ch. 21.
59 As is argued by Schofield, “Disappearance of the Philosopher King,” 230–41.
 Plato in his Time and Place   65

Bibliography
Annas, J. Aristotle’s Metaphysics, Books M and N (Oxford, 1976).
Barney, R. “The Sophistic Movement,” in M. L. Gill and P. Pellegrin (eds.), A Companion to
Ancient Philosophy (Malden, Oxford, and Carlton, 2006), 77–97.
Barton, J. “Hippocratic Explanations,” in P. J. van der Eijk (ed.), Hippocrates in Context (Leiden,
2005), 29–47.
Blank, D. L. “Socratics versus Sophists on Payment for Teaching,” California Studies in Classical
Antiquity 1 (1985), 1–49.
Bouchet, C., and P. Giovanelli (eds.), Isocrate: entre jeu rhétorique et enjeux politiques (Lyon,
2015).
Brunt, P. A. “Plato’s Academy and Politics,” Studies in Greek History and Thought (Oxford,
1993), ch. 10.
Burkert, W. Lore and Science in Ancient Pythagoreanism (Cambridge, Mass., 1972).
Cherniss, H. The Riddle of the Early Academy (Berkeley, Calif., 1945).
Coxon, A. H. The Fragments of Parmenides (Assen, 1986).
Crystal, I. “Parmenidean Allusion in Republic V,” Ancient Philosophy 16 (1996), 351–63.
Curd, P. K. The Legacy of Parmenides: Eleatic Monism and Later Presocratic Thought (Princeton,
N.J., 1998).
Dancy, R. M. Two Studies in the Early Academy (Albany, N.Y., 1991).
Dillon, J. The Heirs of Plato (Oxford, 2003).
Dodds, E. R. (ed.) Plato: Gorgias (Oxford, 1959).
Dunn, F. “On Ancient Medicine and Its Intellectual Context,” in P. J. van der Eijk (ed.),
Hippocrates in Context (Leiden, 2005), 49–67.
Eucken, C. Isokrates: Seine Positionen in der Auseinandersetzung mit den zeitgenössischen
Philosophen (Berlin, 1983).
Field, G. C. Plato and His Contemporaries (London, 1930).
Fine, G. On Ideas: Aristotle’s Criticism of Plato’s Theory of Forms (Oxford, 1993).
Ford, A. “Sophists without Rhetoric: The Arts of Speech in Fifth-Century Athens,” in Y. L. Too
(ed.), Education in Greek and Roman Antiquity (Leiden, 2001), 85–109.
Furley, D. J. Cosmic Problems (Cambridge, 1989).
Furley, W. D. Andokides and the Herms: A Study of Crisis in Fifth-Century Athenian Religion
(London, 1996).
Gifford, M. “Dramatic Dialectic in Republic Book 1,” Oxford Studies in Ancient Philosophy
20 (2001), 35–106.
Griffith, M. The Authenticity of Prometheus Bound (Cambridge, 1977).
Grote, G. History of Greece (London, 1850).
Guthrie, W. K. C. A History of Greek Philosophy, Vols. 3, 4, 5 (Cambridge, 1969, 1975, 1978).
Heidel, W. A. “The Pythagoreans and Greek Mathematics,” American Journal of Philology
61 (1940), 1–33.
Horky, P. S. Plato and Pythagoreanism (Oxford, 2013).
Hornblower, S. The Greek World 479–323 b.c., 3rd ed. (London, 2003).
Hornblower, S., and Spawforth, A. (eds.), The Oxford Classical Dictionary, 4th ed. (Oxford,
2012).
Huffman, C. A. Archytas of Tarentum: Pythagorean, Philosopher and Mathematician King
(Cambridge, 2005).
Huffman, C. A. Philolaus of Croton: Pythagorean and Presocratic (Cambridge, 1993).
66   Malcolm Schofield

Huffman, C. “Archytas,” in The Stanford Encyclopedia of Philosophy (Fall 2016 Edition). https://
plato.stanford.edu/archives/fall2016/entries/archytas/.
Irwin, T. H. “The Theory of Forms,” in G. Fine (ed.), Plato 1: Metaphysics and Epistemology
(Oxford, 1999), 143–70.
Isnardi Parente, M. Frammenti: Senocrate, Ermippo (Naples, 1982).
Johnson, R. “Isocrates’ Methods of Teaching,” American Journal of Philology 80 (1959), 25–36.
Jouanna, J. Hippocrates (Baltimore, 1999).
Kagan, D. The Peloponnesian War: Athens and Sparta in Savage Conflict, 431–404 bc (London,
2003).
Kahn, C. H. The Art and Thought of Heraclitus (Cambridge, 1979).
Kahn, C. H. Plato and the Socratic Dialogue (Cambridge, 1996).
Kahn, C. H. “Plato’s Funeral Oration: The Motive of the Menexenus,” Classical Philology 58
(1963), 220–34.
Kahn, C. H. Pythagoras and the Pythagoreans: A Brief History (Indianapolis, Ind., 2001).
Kennedy, G. The Art of Persuasion in Greece (Princeton, N.J., 1963).
Kerferd, G. B. The Sophistic Movement (Cambridge, 1981).
Kirk, G. S. “The Problem of Cratylus,” American Journal of Philology 72 (1951), 225–53.
Kirk, G. S., Raven, J. E., and Schofield, M. The Presocratic Philosophers, 2nd ed. (Cambridge,
1983).
Knorr, W. R. The Ancient Tradition of Geometric Problems (Boston, 1986).
Knorr, W. R. “On the Early History of Axiomatics: The Interaction of Mathematics and
Philosophy in Greek Antiquity,” in J. Hintikka, D. Gruender, and A. Agazzi (eds.), Theory
Change, Ancient Axiomatics and Galileo’s Methodology (Dordrecht, 1981), 145–86.
Livingstone, N. A Commentary on Isocrates’ Busiris (Leiden, 2001).
Lloyd, G. E. R. Demystifying Mentalities (Cambridge, 1990).
Lloyd, G. E. R. (ed.) Hippocratic Writings (Harmondsworth, 1978).
Lloyd, G. E. R. Magic, Reason and Experience (Cambridge, 1979).
McCabe, M. M. Platonic Conversations (Oxford, 2015).
Mirhady, D., and Too, Y. L. (trans.) Isocrates I (Austin, TX, 2000).
Morrow, G. R. Plato’s Cretan City: A Historical Interpretation of the Laws (Princeton, N.J.,
1960).
Nails, D. The People of Plato (Indianapolis, Ind., 2002).
Norlin, G. (ed.) Isocrates, Vols. 1–2 (London, 1928, 1929).
Ober, J. Mass and Elite in Democratic Athens (Princeton, N.J., 1989).
O’Brien, M. J. The Socratic Paradoxes and the Greek Mind (Chapel Hill, N.C., 1967).
Ostwald, M., and Lynch, J. P. “The Growth of Schools and the Advance of Knowledge,” in
D. M. Lewis, J. Boardman, S. Hornblower, and M. Ostwald (eds.), The Cambridge Ancient
History, Vol. 6, 2nd ed. (Cambridge, 1994), ch. 12a.
Owen, G. E. L. “Eleatic Questions,” Classical Quarterly 10 (1960), 84–102.
Palmer, J. A. Plato’s Reception of Parmenides (Oxford, 1999).
Papillon, T. L. (trans.) Isocrates II (Austin, Tex., 2005).
Price, S. Religions of the Ancient Greeks (Cambridge, 1999).
Rhodes, P. J. A History of the Classical Greek World 478–323 b.c. (Oxford, 2006).
Riginos, A. Platonica: The Anecdotes concerning the Life and Writings of Plato (Leiden, 1976).
Ross, W. D. Plato’s Theory of Ideas (Oxford, 1951).
Saunders, T. J. Plato’s Penal Code (Oxford, 1991).
 Plato in his Time and Place   67

Saunders, T. J. “ ‘The RAND Corporation in Antiquity?’ Plato’s Academy and Greek Politics,”
in J. H. Betts, J. T. Hooker, and J. R. Green (eds.), Studies in Honour of T. B. L. Webster
(Bristol, 1986), 1.200–10.
Schiefsky, M. J. Hippocrates On Ancient Medicine (Leiden, 2005).
Schofield, M. “Antisthenes,” in E. Craig (ed.), Routledge Encyclopedia of Philosophy (London,
1998), 1.314–17.
Schofield, M. “The Disappearance of the Philosopher King,” Proceedings of the Boston Area
Colloquium in Ancient Philosophy 13 (1997), 213–41.
Schofield, M. “Eudoxus in the Parmenides,” Museum Helveticum 30 (1973), 1–19.
Schofield, M. “Plato and Practical Politics,” in C. J. Rowe and M. Schofield (eds.), The
Cambridge History of Greek and Roman Political Thought (Cambridge, 2000), ch. 13.
Schofield, M. “Who Were oἱ Δυσχερεῖς in Plato, Philebus 44Aff?” Museum Helveticum 28
(1971), 2–20, 181.
Schofield, M., and Griffith, T. (ed. and trans.), Plato: Gorgias, Menexenus, Protagoras
(Cambridge, 2010).
Sedley, D. N. “Parmenides,” in E. Craig (ed.), Routledge Encyclopedia of Philosophy (London,
1998), 7.229–35.
Sedley, D. N. Plato’s Cratylus (Cambridge, 2003).
Tarán, L. Speusippus of Athens (Leiden, 1981).
Taylor, C. C. W. and Lee, Mi-Kyoung, “The Sophists,” in The Stanford Encyclopedia of Philosophy
(Winter 2016 Edition) https://plato.stanford.edu/archives/win2016/entries/sophists/.
Thompson, W. H. The Phaedrus of Plato (London, 1868).
Too, Y. L. The Rhetoric of Identity in Isocrates: Text, Power, Pedagogy (Cambridge, 1995).
van Hook, L. (ed.) Isocrates, Vol. 3 (London, 1945).
Vlastos, G. Socrates: Ironist and Moral Philosopher (Cambridge, 1991).
Vlastos, G. “Socratic Knowledge and Platonic ‘Pessimism,’” Platonic Studies (Princeton, N.J.,
1973), ch. 9.
Wardy, R. B. B. The Birth of Rhetoric (London, 1996).
Zeyl, D. (ed.) Encyclopedia of Classical Philosophy (Westport, Conn., 1997).
Zuntz, G. Persephone (Oxford, 1971).
 chapter 3

The Pl aton ic Cor pus

T. H. Irwin

1. Contents of the Corpus

If we open an edition of the Greek text of Plato, such as the five volumes of the Oxford
Classical Texts (OCT) edition,1 we find 36 works,2 in nine “tetralogies” (groups of 4 works),
plus the Definitions and 6 works listed under “Spurious.”3 Each work is supplied with
marginal numbers and letters (beginning with the Euthyphro, at 2a); these numbers and
letters refer to “Stephanus pages.”4 The Platonic corpus is unusual among the works of
Greek authors by being, as far as we know, complete. No reference in any ancient author
attests the existence of any work by Plato that does not appear in our Platonic corpus.5

1 I have benefited from helpful comments by the editor and by Charles Brittain. Platonis Opera, ed.
J. Burnet (5 vols., Oxford: OUP, 1900–07). This edition is in the process of being replaced by a new
OCT. So far vol. 1 (ed. E. A. Dukes et al, 1995) and the Republic (ed. S.R. Slings, 2003) have appeared.
2 The two longest dialogues, the Republic and the Laws, are also divided into books. The divisions
are probably not derived from Plato. See n15 below.
3 The titles of the works correspond to the titles in the catalogue of Plato’s works that appears in
Diogenes Laertius (DL) iii 57–61. Each dialogue in the catalogue has an alternative title and a generic
classification. (Hence the Phaedo, e.g., is listed as “Phaedo, or On the Soul; ethical” (DL iii 58).)
Anonymous, Prolegomena in Platonis Philosophiam (ed. Westerink), 21.11–14, comments that it is not
always easy to see how the dialogue deals with the subject matter mentioned by the second title.
The six works listed as spurious are De Iusto, De Virtute, Demodocus, Sisyphus, Eryxias, and
Axiochus. On the tetralogies see Section 4 below.
4 The edition printed and published by Stephanus (the Latin name of Henri Estienne) in 1578 was
the first edition of Plato to divide its pages into sections marked by letters. Modern editions of Plato
still use these references.
5 According to DL iii 5–6, Plato took up painting and writing poetry, especially tragic drama, before
he met Socrates, but burned his tragedies after he had heard Socrates. This “biographical” information
may simply be invented on the basis of the dialogues. The authenticity of the short poems attributed to
him (see, e.g., DL iii 29–33) is questionable. W. Ludwig, “Plato's Love Epigrams”, Greek, Roman, and
Byzantine Studies 4 (1963), 59–82, argues that they are all spurious, except for the epigram on Dion.
(They are collected at the end of Plato, Collected Dialogues, ed. J. M. Cooper and D. S. Hutchinson
(Indianapolis: Hackett, 1997).)
70   T. H. Irwin

Most, but not all, the works in the Corpus are dialogues. The Apology purports to be
Socrates’ defence at his trial. The Menexenus is a funeral speech. The ninth tetralogy
ends with 13 letters under Plato’s name. The Definitions is a list of definitions of philo-
sophical terms. The remaining works, including the ones listed as “Spurious,” are usually
called “dialogues” because they contain philosophical conversations between more or
less sharply characterized individuals, many of whom are known to us from other sources.6
Often, but not always, the main or the sole speaker is Socrates. Plato never speaks under
his own name.
The Corpus is found in the manuscripts of Plato, produced in the ninth century ad
and later. In the 1,300 years between the lifetime of Plato and the production of our man-
uscripts, a canon of Plato’s works was formed, someone tried to distinguish genuine
from spurious works, and the works were arranged in a fixed tetralogical order. To see
when and how this all happened, we need to consider the history of the Corpus from its
creation to the production of our manuscripts. Many stages in this history are obscure,
and the evidence is often fragmentary.
Plato is not the author of the whole Corpus. The non-Platonic works fall into four
groups:

(1) The manuscripts correctly designate six short dialogues as “Spurious.”

(2) The Definitions probably is not Plato’s work, but may originate in the Academy.
(3) Unlike the works just mentioned, the 13 letters fall within the tetralogies, but
most modern students reject almost all of them as spurious. The most important
letter about which serious dispute remains is Letter VII. If this letter is genuine,
it is important, for two reasons. First, it offers some autobiographical detail about
Plato. Second, it puts forward philosophical claims that have no parallel in the
dialogues. Probably, however, Letter VII is neither Plato’s work nor a reliable
source for his life and philosophy.7
(4) Some further works within the tetralogies are probably spurious. Among those
that modern students have doubted are Hippias Major, Hippias Minor, Menexenus,
Alcibiades I–II, Hipparchus, Amatores, Theages, Clitopho, Minos, and Epinomis.8
The doubts about the first three of these are probably mistaken, but in the case of
the other works doubts are reasonable.

The fact that the Corpus contains some works that are certainly or probably not by
Plato will be relevant later, when we consider the history of the Corpus after Plato’s death.

6 Evidence on the characters in the dialogues is assembled by D. Nails, The People of Plato
(Indianapolis: Hackett, 2002).
7 A recent discussion of Letter VII is M. F. Burnyeat and M. Frede, The Pseudo-Platonic Seventh Letter,
ed. D. Scott (Oxford: OUP, 2015). See esp. Scott’s “Editor’s Guide”, at 85–97. I have discussed it briefly in
“The Inside Story of the Seventh Platonic Letter: A Sceptical Introduction”, Rhizai 6 (2009), 7–40.
8 This list is taken from Frede and Burnyeat, Seventh Letter, 5.
 The Platonic Corpus   71

2. The Order of the Dialogues

We do not know much about the absolute dates at which the different works in the
Corpus were composed. The division into tetralogies is useless for this purpose. Dramatic
dates are provided, more or less precisely, in many of the dialogues; that is to say, the
characters allude to specific historical events that give some idea of the setting of the
dialogues. These dates are mostly useless for fixing the dates of composition, since most
of them are much earlier than any date at which Plato could conceivably have written the
dialogues.9 Some incidental references fix a terminus post quem, but no absolute date,
for some dialogues, but they do not fix absolute dates.10 We do not know when Plato
began to write philosophical dialogues,11 or when the latest dialogue was written.12
Can we, then, find plausible relative dates?13 We have to consider internal evidence.14

9 D. Nails, The People of Plato, Appendix I, gives a useful table of dramatic dates.
10 The Symp. probably refers to events in the mid-380s. See K. J. Dover, ed., Plato’s Symposium
(Cambridge: CUP, 1980), 10. Theaetetus was fatally wounded is the fighting around Corinth in 369
(Tht. 142ab). This is the latest dateable event in the Corpus.
11 Some critics assume, for no good reason, that Plato did not write any dialogues during the
lifetime of Socrates. W. K. C. Guthrie, History of Greek Philosophy (Cambridge: CUP, 1975), iv, 54–56,
surveys the unconvincing arguments that have been offered on each side.
12 On the Laws see Anon., Prolegomena 24.12–15: “They say that the Laws were written last, because
he left them uncorrected (adiorthôtous) and disordered (or ‘confused’; sunkechumenous), because he
lacked the time to arrange them because of his death. And if they now appear appropriately arranged,
that is because they were arranged not by Plato himself, but by a certain Philip of Opous, a successor in
the Platonic school.” Anon. quotes Proclus disputing the authenticity of the Epinomis, Proclus asks “How
could Plato have been unable to correct the Laws because he did not live long enough, and after that
been able to write the Epinomis?” (Anon. Prol. 25.6–8). According to DL iii 37, Philip of Opous “copied
his [Plato’s] Laws, which were written in wax.” Wax tablets allowed easy erasure and revision. On these
stories about the Laws see G. R. Morrow, Plato’s Cretan City (Princeton, NJ: Princeton UP, 1960), 515–18.
Though Proclus tells us that the Laws was unfinished at Plato’s death, it need not have been the last
dialogue to be written. Plato may have put it aside to work on other things.
13 Though ancient critics consider the question of relative date, they offer nothing more helpful than
the suggestion that the Phaedrus must have been the first dialogue, since the subject matter is suitable
for a young man, and the Laws must have been the last, since it was left unfinished at Plato’s death
(Anon. Prol. 24.7–19; cf. DL iii 38). Most of the discussion in Anon., Prol. concerns the appropriate
order for reading the dialogues (26.21–44). Anon. omits most of the “early” (or “Socratic”) dialogues,
and treats the Parmenides as the culmination of Platonic doctrine (26.13–44). He follows Iamblichus,
who picks out 12 dialogues, divided into two groups, “natural” and “theological.” He picks out two
“complete” (or “perfect”, teleioi) dialogues, the Timaeus in the natural group and the Parmenides in the
theological group. He remarks that some instructors also think it appropriate to discuss the Republic
and the Laws (26.45–47). The Alcibiades often comes first because it introduces the central Platonic
theme of the division between body and soul. On the order of reading dialogues see A. J. Festugière,
Etudes de philosophie grecque (Paris: Vrin, 1971), 533–50.
14 Questions about dating would be complicated if there were clear evidence that Plato revised
dialogues and published second editions. (Cf. K. J. Dover, Aristophanes’ Clouds (Oxford: OUP, 1968),
lxxx–xcviii.) According to Dionysius of Halicarnassus, De Comp. Verborum 25, “Plato kept on combing
and curling and in every way braiding his dialogues even when he had turned eighty; for all lovers of
literature (philologoi) are familiar with the stories told about this man’s industry—including the story
72   T. H. Irwin

Some arguments for an ordering of the dialogues have received fairly widespread
support over the past 150 years or so.15 These arguments rely on different sources of
evidence for relative dates:

1. Style and language. If we suppose that the Laws is the latest of the dialogues, we
can order the others by the degree of linguistic and stylistic resemblance to the
Laws.16 Linguistic and stylistic tests pick out a fairly clear group of late dialogues,
and less sharply defined groups of early and middle dialogues.
2. Character. Some dialogues are short, vividly characterized, and ostensibly negative
in their conclusions. Some are more didactic than dramatic, and seem to concentrate
on the exposition of a doctrine rather than the cross-examination of interlocutors.
We can see this contrast if we compare the Charmides and Laches with the Sophist,
Timaeus, and Laws. The first type of dialogue is earlier, by linguistic and stylistic
tests, than the second type.
3. Philosophical content. Some dialogues speak of “forms” or “ideas” as non-sensible,
stable realities that are grasped by intellect rather than sense, in contrast to the
changeable things grasped by sense. Other dialogues speak of forms without this
sharp contrast. In some dialogues Socrates argues for the identification of virtue

about the writing-tablet that is said to have been found after his death, the one that had many different
versions of the beginning of the Republic that reads ‘I went down yesterday . . . ’ ”. DL iii 37 has the story
about the Republic. One might wonder whether Dionysius’ general claim rests on anything more
definite than the story in DL about the Republic, combined with Phdr. 278de; see A. S. Riginos,
Platonica (Leiden: Brill, 1976), 186. Both Dionysius’ hairdressing metaphors and his one example
suggest cosmetic changes rather than radical revisions of dialogues.
Anon. in Tht. 3.28–37 (in Corpus dei Papiri Filosofici Greci e Latini, iii (Florence: Olschki, 1995),
268–69, 486) mentions a spurious alternative prologue to the Tht., which shows nothing about revision
to the body.
The two apparent versions of a passage in the Cra. (437d10 in the new OCT; cf. Sedley, Cratylus 7–10)
may simply indicate that Plato discarded a passage and failed to delete it.
A change of plan in the Republic (R. Hackforth, “The Modification of Plan in Plato's Republic,”
Classical Quarterly 7 (1913), 265–72) does not imply two editions. Similarly the story in Aulus Gellius,
NA xiv 3.3 may refer to the publication of the Rep. in six books. See H. Alline, Histoire du texte de
Platon (Paris: Champion, 1915), 14–15. It does not imply a first edition.
G. R. Morrow, “Aristotle’s Comments on Plato’s Laws” in Aristotle and Plato in the Mid-Fourth Century,
I. Dȕring and G.E.L. Owen eds. (Gothenburg: Almquist and Wiksell, 1960), 145–62, at 160–62, suggests
that when Aristotle wrote his criticism of the Laws in Politics ii, he knew only Laws iii–vii (and some of
the later books). Since the Laws was unfinished at Plato’s death, Aristotle may have had access to some
sections of Plato’s work in progress. There is no reason to suppose that a first edition was published.
We have no good reason, therefore, to believe that revised editions were published. Plato apparently
preferred to move on to new projects rather than complete old ones. He never finished the Laws. Only
a part of the Critias, and none of its anticipated successor, the Hermocrates, is extant. The Philosopher, a
successor to the Politicus, was apparently projected, but never written. See note 57 below.
15 This ordering has become popular enough to be described (often pejoratively, by those who reject it)
as “the standard view.” See Annas and Rowe in New Perspectives on Plato, ed. J. Annas and C. J. Rowe
(Cambridge, Mass.: Harvard UP, 2002), p. ix.
16 The evidence and arguments are summarized by L. Brandwood, “Stylometry and Chronology” in The
Cambridge Companion to Plato, ed. R. Kraut (Cambridge: CUP, 1992), and by C. H. Kahn, “On Platonic
Chronology”, in New Perspectives, ed. Annas and Rowe, ch. 4. On the date of the Laws cf. note 12 above.
 The Platonic Corpus   73

with knowledge and against the possibility of weakness of will. Other dialogues
recognize different and potentially conflicting desires in the soul, apparently
allowing non-rational desires to overcome rational desires, contrary to Socrates’
views on weakness of will.
4. The testimony of Aristotle. According to Aristotle, (1) Socrates did not separate the
forms from sensible things, but Plato did separate them. (2) Socrates treated virtue
as knowledge, but Plato did not. Aristotle calls the historical Socrates simply
“Socrates,” but calls the Platonic character “the Socrates.” He often uses the imperfect
tense when he uses “Socrates” (“Socrates used to say . . .”), but the present tense
(“the Socrates says . . .”) for the views of the character in a dialogue.17

These different tests tend to order the dialogues in the same way. With some excep-
tions, dialogues that appear to be early on the basis of linguistic tests also tend to be
dramatic rather than didactic, and tend to hold the doctrines that Aristotle ascribes to
Socrates. The apparently later dialogues tend to be more didactic and less dramatic, and
tend to hold the views that Aristotle describes as Platonic.
These arguments suggest that Plato began by agreeing with Socrates, but developed
his distinctive, and in some respects non-Socratic, philosophical outlook. He extends
his philosophical range beyond Socrates’ primary concern with ethics, and rejects some
central Socratic claims even within ethics.
If, then, we want to study the dialogues in a fairly probable chronological order, we
might follow this plan:

1. We might want to begin with the Apology, since it defends Socrates’ life and work;
we can test Socrates’ claims about himself in the light of our reading of the shorter
dialogues on ethical topics (Laches, Charmides, Lysis, Euthyphro, Hippias Minor,
Ion, Crito).
2. The Protagoras, Gorgias, and Euthydemus are more elaborate works on similar
themes. They record confrontations between Socrates and professional intellectu-
als who discuss philosophical topics. These interlocutors are sophists, orators, and
eristics, rather than the philosophical nonspecialists who appear in Group 1.
3. The Meno reflects on some of the themes in these previous dialogues, and intro-
duces epistemological and metaphysical claims that are explored further in the
Cratylus, Hippias Major, Phaedo, and Symposium.
4. The Republic is Plato’s most elaborate effort to construct a broad philosophical
theory that embraces the themes of most of the previous dialogues. The Phaedrus
explores questions about rhetoric (raised in the Gorgias), love (raised in the
Symposium), and moral psychology (raised in the Republic).
5. An ostensible version of the metaphysics and epistemology of the previous dia-
logues (Groups 3–4), and especially of their Theory of Forms, is examined in the

17 In (1) I take a disputed position on the legitimacy of “Fitzgerald’s canon.” I give some further
details and references in Plato’s Ethics (Oxford: OUP, 1995), §5.
74   T. H. Irwin

Parmenides. Plato’s more systematic reflexions on epistemology and metaphysics

begin in the Theaetetus.
6. These reflexions are continued in the Sophist and Politicus. The Timaeus places them
in a sketch of cosmology and natural philosophy.
7. The Philebus, Politicus, and Laws are major works in moral and political philosophy
that seem to reflect on, and to modify, some of the doctrines introduced in the
previous dialogues.

Different parts of this arrangement rest on the different arguments mentioned above.
Linguistic and stylistic evidence isolates Groups 6 and 7 (not necessarily in that order) as
the latest dialogues, and places Groups 4 and 5 (not necessarily in that order) before
them. It places Groups 1–3 before the others without fixing any definite order of groups
or within groups. Doctrinal considerations (mentioned in points (3) and (4) above) jus-
tify us in placing Group 3 after Groups 1 and 2.
Many students of Plato agree that something like this is a reasonable account of the
order of the dialogues. In speaking of “early” or (because of Aristotle’s testimony) “Socratic”
dialogues, they normally have in mind Groups 1 and 2. Groups 3 and 4 are usually called
“middle” dialogues, and Groups 5–7 are “late” dialogues.
Some particular dialogues have been subjects of dispute among those who accept this
division of the dialogues:

(a) Some place the Euthydemus and Cratylus in Groups 5 or 6, contrary to stylistic
evidence.
(b) Some place one or both of the Protagoras and Gorgias after the Meno.
(c) Most people take the Parmenides to presuppose the exposition of the Theory of
Forms in Groups 3 and 4. We will take a different view of its significance if we
place it before some or all of these dialogues. Stylistic evidence places it later than
the Phaedo and Symposium, but less obviously after the Republic.
(d) Since the Timaeus maintains central elements in the Theory of Forms, Plato does
not seem to have thought that the Parmenides undermined the theory. Some
have supposed that Plato could not have responded in this way to the Parmenides,
and therefore they place the Timaeus before the Parmenides, and hence before
Groups 5–7. This place for the Timaeus conflicts with the stylistic evidence about
the order of the dialogues.18

If we accept the order of the dialogues that has been defended, we cannot, for instance,
take the Philebus and the Protagoras to express, respectively, Plato’s earlier and later
views on pleasure; nor can we take the political theory of the Politicus to precede that of

18 G. E. L. Owen, “The Place of the Timaeus in Plato’s Dialogues,” in Logic, Science, and Dialectic
(London: Duckworth, 1986), ch. 4 (from Classical Quarterly 3 (1953), 79–95) defends an early date for
the Timaeus. The reply by H. F. Cherniss, “The Relation of the Timaeus to Plato’s Later Dialogues,” in
Studies in Plato’s Metaphysics, ed. R. E. Allen (London: Routledge, 1965), ch. 17 (from American Journal
of Philology 78 (1957), 225–66) is generally (though not in every detail) convincing.
 The Platonic Corpus   75

the Republic; nor again (as we have just seen) can we take Plato to have abandoned the
Theory of Forms in all the dialogues that are later than the Parmenides.

3. Questions about the Order

of the Dialogues

Some critics have questioned the order that was defended in the previous section. They
have relied on the following arguments, among others:

1. Perhaps Plato sometimes chose to write dramatic dialogues, but sometimes chose
to write didactic dialogues. Stylistic differences may result from these choices, and
may not indicate any temporal sequence.
2. Apparent doctrinal divergences indicate a temporal sequence only if (1) it is clear
which of two conflicting doctrines came first, (2) Plato was aware of the diver-
gence; and (3) Plato intended to affirm a particular position, rather than simply to
explore it. This third assumption depends on the assumption (which I will discuss
later) that Plato’s intention in the dialogues is often dogmatic, and not purely
exploratory. But even if we accept the assumption, we need not take apparent
conflicts at face value. Plato may, for instance, have set out a deliberately one-sided
case for treating virtue as knowledge, so that he could answer it in another dialogue;
or he may have outlined his own position in one dialogue, and then explored the
opposite position in a later dialogue.
3. Aristotle may be wrong. Perhaps he noticed some of the differences between the
dialogues that we have mentioned above, and mistakenly inferred that they were to
be explained by Plato’s different attitudes to Socrates.

Some critics have inferred that we ought not to concern ourselves about the order of
the dialogues. Perhaps they are partial expositions of a philosophical system.19 Or
perhaps they are explorations that were not meant to express a coherent doctrine.20
Admittedly, some of the signs of the “late” style seem to result from conscious choice.
For instance, in the Phaedrus and (allegedly) later dialogues, Plato sharply reduced the
incidence of hiatus.21 This was probably a conscious decision, given the sharp difference
between the dialogues that allow it and those that avoid it. Hence, one might argue that
Plato could turn it on and off at will. Other stylistic features are less likely to be consciously

19 See P. Shorey, What Plato Said (Chicago: U. Chicago Press, 1933), 64–73.
20 This view of the dialogues is defended by G. Grote, G., Plato and the Other Companions of
Socrates, 4 vols. (London: Murray, 1888).
21 A word ending with a vowel followed by a word beginning with the same vowel creates a hiatus,
as in “Carla Adams.” Spoken English in England sometimes inserts “r” between the vowels to avoid
such a hiatus.
76   T. H. Irwin

adopted; but one might argue that they are unconscious results of the conscious adoption
of a particular style.
The difference between didactic and dramatic dialogues, however, does not explain
the common stylistic features of the later dialogues. The Philebus, for instance, has some
of the characteristics of earlier dialogues, and so we might expect Plato to have decided to
write it in the style of the Gorgias or Republic i, but it belongs stylistically with the late
dialogues. Similarly, the Phaedrus differs linguistically and stylistically from the Symposium
(despite its similar subject matter), in ways that connect it with the Theaetetus and
Republic.22 The Philebus, the Timaeus, the Sophist, and the Laws differ in their subject
matter, dialogue form, and (in certain respects) style, but they share the features that
mark out the late group. Apparently, then, the features of Plato’s late style are constant
across dialogues that are otherwise quite different. Their constancy suggests that they
probably do not result from a conscious decision to adopt different styles for dialogues
of different types.
It is more difficult to decide whether apparent doctrinal conflicts result from conscious
decisions by Plato to present one side of an issue on which he favours the other side.
Does he, for instance, present a one-sided discussion of incontinence in the Protagoras
even though he already accepts the views that he presents only in Republic iv?23 If we find
(as I believe we do) that the one-sided presentation does not even seem to take seriously
the position that Plato defends in another work, we have good reason to believe that it is
not intentionally one-sided.
Aristotle’s testimony is difficult to reject. Admittedly, he is not always a reliable histo-
rian, and we might wonder whether his claims about Socrates depend on inferences
from the dialogues, rather than on independent and reliable information.24 But why
would the mere study of the dialogues suggest to him that some dialogues, but not oth-
ers, defend the views of the historical Socrates? He believes that the character called
“Socrates” in some dialogues presents the views of Plato rather than Socrates. He can-
not, then, have naively supposed that the character Socrates always puts forward the
views of the historical Socrates. He must have supposed that he had some reason for dis-
tinguishing the position of the character Socrates in some dialogues from his position
in other dialogues, and for treating only one of these positions as the position of the
historical Socrates.
What could his reason be? He might have done what modern readers do, and noticed
the apparent doctrinal conflicts between different dialogues. But why should he infer
some dialogues express Socrates’ views and others do not? Does he perhaps assume that
the dialogues with the fullest biographical remarks about Socrates are intended to pres-
ent the historical Socrates? He does not assume this. The Phaedo has more to say than

22 See L. Brandwood, “Stylometry and Chronology” in The Cambridge Companion to Plato, ed. R. Kraut,
(Cambridge: CUP, 1992), 113–15.
23 I have discussed this question further in “The Parts of the Soul and the Cardinal Virtues,” in Platon:
Politeia, ed. O. Höffe (Berlin: Akademie Verlag, 1997), ch. 6.
24 C. H. Kahn, Plato and the Socratic Dialogue (Cambridge: CUP, 1996), 79–87, discounts Aristotle’s
testimony.
 The Platonic Corpus   77

any other dialogue about Socrates’ life and philosophical development, but Aristotle
takes it to present a Platonic, not a Socratic, theory of forms. The internal evidence of the
dialogues could not have persuaded Aristotle that some, but not others, are about Socrates.
Probably, then, Aristotle ascribes some of the views of the Platonic Socrates to the
historical Socrates because he has external evidence about the views of the historical
Socrates. He had access to such evidence, and had a good reason to look for it and to use
it. He joined Plato’s Academy in 367, just over 30 years after the death of Socrates in 399.
He knew people who had known Socrates, and he could find out what they thought about
Socrates. Since Plato was not the only one who wrote Socratic dialogues or expressed
views about Socrates, Aristotle had a reason to try to find out what Socrates had actually
thought, and whether any of Plato’s dialogues accurately described his views.
For these reasons, we have a plausible cumulative argument for the order of the dia-
logues that was set out in Section 2 above. We can rely on it to show that the Theaetetus,
for instance, is later than the Protagoras, and that the Timaeus is later than the Phaedo.
We cannot rely on it to settle every question that arises about the order of the dialogues.
It does not settle, for instance, the relation of the Gorgias to the Meno. But we have good
grounds for taking some version of the standard view as our starting point for discus-
sion of the dialogues.

4. Performance and Publication

The first audience for Plato’s written compositions may have been a group of people
who came for a private reading. The next stage of dissemination was the delivery of the
work to a bookshop; in Socrates’ time Athenian bookshops sold philosophical works
even by non-Athenian writers (Ap. 26d–e).25 By Plato’s time Athenians were used to
the idea of buying a copy of a play for private reading, and dramatic dialogues probably
also found readers.26
Sale by a bookshop did not replace oral dissemination. The Parmenides begins after
Zeno the Eleatic has finished reading a treatise of his that he has brought to Athens for
the first time (Parm. 127c–e). In the Theaetetus a slave reads Eucleides’ written version
of a conversation he had heard (143ab). Socrates heard someone reading from a book of
Anaxagoras (Phd. 97c). Bookshops were centres for public reading where interested
people might come to hear recent books read, and to decide whether they wanted to buy

25 On books in fifth and fourth century Greece see Oxford Classical Dictionary, ed. S. Hornblower
and A. Spawforth 3rd ed. (Oxford: OUP, 1996), 250–01; L. D. Reynolds and N. G. Wilson, Scribes and
Scholars (3rd ed., Oxford: OUP 1991), 1–5 (an admirably clear and instructive book); B. M. Metzger,
The Text of the New Testament (3rd ed., OUP, 1991), ch. 1.
26 Aristophanes, Frogs 52–54, refers in passing to private reading as something familiar. On the uses of
written texts in classical Athens see H. Yunis, ed., Written Texts and the Rise of Literate Culture in Ancient
Greece (Cambridge: CUP, 2003), especially the chapters by Yunis, “Writing for Reading,” ch.9, 189–212;
C. H. Kahn, “Writing Philosophy”, ch. 7, 139–61; R. Thomas, “Prose Performance Texts,” ch. 8, 162–88.
78   T. H. Irwin

a copy.27 Before they could buy a copy, the original delivered by the author—or an
unauthorized copy (Parm. 128de)—had to be copied by hand.
Plato held discussions in the gymnasium (which was also a public meeting-place) in
the area called Academeia, after the precinct of the hero Hecademus. He bought land
in this area and established a house and garden for himself. Here he began the association
that came to be called the Academy.28 The oral delivery of the dialogues (whether before
or after they were available to booksellers) may have been connected with discussions in
the Academy.29
Some questions about the performance of the dialogues arise from their literary form.
Though we speak of “dialogues,” only some of them are “direct” dialogues in the sense in
which a play is a dialogue between different speakers. In most of the early dialogues the
characters speak (as in a play) without any narrator. The Crito, Ion, Euthyphro, Laches and
the Gorgias, for instance, have sharply defined characters whose interaction with Socrates
has the character of a real conversation; the interlocutors reveal their aims, limitations,
vanity, rashness, and so on. These dialogues show the interactions between Socratic
ethical beliefs and the ordinary ethical assumptions of Socrates’ contemporaries, who
grapple with the implications of Socrates’ questions.
Other “dialogues” are really monologues in which the narrator says he gives a verbatim
report of a dialogue. In the earlier reported dialogues, including the Charmides, the Lysis,
and the Protagoras, Socrates is the reporter; he introduces his own reported remarks with
“I said.” The Euthydemus begins with direct dialogue between Socrates and Crito, and
continues with a report by Socrates, interrupted by a brief direct dialogue (at 290e1).
Monologues containing reported dialogue were familiar to Plato’s audience, who were
used to oral performances of Homer. A long section of the Odyssey is narrated in the first
person by Odysseus; he reports dialogues between himself and others, and often reports
the reactions of others and himself to what is said. Plato intends his readers to recall the
Odyssey; in the Protagoras Socrates illustrates his comments on the various sophists
with a quotation from Odysseus’ comments on the shades in the underworld.30
These monologues ostensibly report conversations. Socrates as narrator comments on
the attitude of the different characters, including himself (e.g., Protagoras 333d, 335c, 339e).
He describes the effect of his questions on other people. He says that Thrasymachus is at
first impatient (Republic 335b), then triumphant (344d), and then embarrassed (350cd).
We might have gathered these changes in Thrasymachus from direct dialogue, as we
gather similar changes in Callicles from the direct dialogue in the Gorgias. But we gather
them more clearly from Socrates’ mention of Thrasymachus’ swaggering and blushing.

27 See, e.g., DL vii 2–3.

28 DL iii 7, 20, 41. On the character of the Academy see J. M. Dillon, The Heirs of Plato (Oxford:
OUP, 2003), ch. 1.
29 Favorinus (c.AD 85–155) reports that when Plato read his work “On the Soul,” everyone got up and
left before he was finished, except for Aristotle (DL iii 37). “On the Soul” is sometimes used as a title for
the Phaedo. Plato is reported to have given a public lecture on the good (not a reading of a dialogue); for
reports see Aristotle, Complete Works, ed. J. Barnes (2 vols., Princeton, NJ: Princeton UP, 1984), vol. 2,
2397–99.
30 Pr. 315c, quoting Homer, Odyssey xi 601.
 The Platonic Corpus   79

These monologues would be more entertaining if a skilful narrator took the part of
Socrates. Dickens exerted and even exhausted himself in the public reading of excerpts
from his works, and moved his audience to laughter and tears.31 The narrator of a
Platonic reported dialogue might also have entertained and moved the audience.32
Plato’s contemporaries attended performances of the Homeric poems by “rhapsodes,”
who supplied appropriate gestures, signs of emotion, and other dramatic features. The
Ion—a direct dialogue—examines the pretensions of Ion the rhapsode to knowledge of
the subject matter of the Homeric poems. For the purposes of the argument it is assumed
that such knowledge is necessary for the appropriate performance of the poems. Socrates’
objection to the rhapsode would be answered if Plato played the part of Socrates in the
monologues that contain reported dialogue.
In other monologues, the narrator does not take part in the reported dialogue, but
reports it in the third person. The third-person reports give the narrator an opportunity
to comment on Socrates as well as on his interlocutors. Sometimes the monologue is
part of an introductory direct dialogue. The Symposium begins with a direct dialogue
between Apollodorus and an unnamed friend. Apollodorus reports a conversation
between himself and Glaucon. In this reported conversation Apollodorus reports to
Glaucon the story told by Aristodemus about a banquet held many years earlier, just
after Agathon had won first prize at a dramatic festival. The reported dialogue helps us
to look back on Socrates’ contemporaries and to reflect on the significance of Socrates’
life and arguments from the perspective of the narrators. The Phaedo begins with a brief
direct dialogue between Echecrates and Phaedo, but most of it is narrated by Phaedo,
who ends with an account of Socrates’ death. In these monologues the third-person
report gives us someone else’s view of Socrates.
In later dialogues, the function of the reported dialogue is not obvious. In the first
part of the Parmenides the only speaker is Cephalus, who reports Antiphon, who reports
a third man, Pythodorus, who reports Parmenides, Zeno, and a third man, Socrates, in
conversation. The Theaetetus begins with a direct dialogue between Eucleides and
Terpsion. Eucleides offers to ask a slave to read Eucleides’ written report of Socrates’
report of a conversation between Socrates, Theodorus, and Theaetetus. Then Eucleides
says he will omit the “I said” and “he said” that would belong to Socrates’ report. He
proceeds to a direct dialogue between the three main characters (Tht. 144bc).
Perhaps, then, Plato got tired of reported dialogues. If he took Socrates’ part in the
monologues of Socrates, he may have been exhausted by the strenuous work of performing
the Protagoras or (even more) the Republic. Similarly, he may have got tired of the multiple
intermediaries in the Parmenides and Theaetetus. The dialogues that are probably later
than the Theaetetus are direct dialogues.
In most of the later dialogues the characters are not sharply defined, long passages
are primarily expository, and the interlocutors often say little more than Yes and No.33

31 On Dickens’s public readings from his work see E. Johnson, Charles Dickens: His Tragedy and
Triumph (London: Allen Lane, 1977), 456–57, 467, 472–75.
32 G. Ryle, Plato’s Progress (Cambridge: CUP, 1966), ch. 2, speculates on oral performances.
33 This is not to say that their literary form is unimportant. See C. J. Gill and M. M. McCabe (eds.),
Form and Argument in Late Plato (Oxford: OUP, 1996) esp. Gill’s ch. 10.
80   T. H. Irwin

In the Sophist Socrates asks the Stranger whether he would prefer to give a continuous
discourse or to use the dialogue form that Parmenides once used in a conversation
with Socrates (an allusion to the Parmenides). The Stranger replies that if he can find an
interlocutor who will not give any trouble, he would prefer a dialogue, since he is reluctant
to impose a long discourse on Socrates and his friends on their first meeting. Here and
in the Politicus, the announced role of the interlocutor is simply to make the exposition
sound less like a public speech (Sph. 217e). The demand for a pliable interlocutor marks a
sharp contrast with the assertive interlocutors in the early dialogues. In the Timaeus a short
direct dialogue that includes Socrates introduces Timaeus, who gives a long continuous
discourse. The unfinished Critias begins in the same way.
The Philebus, however, is a late work that uses direct dialogue as more than an expository
device. The interlocutors are not pliable. Philebus is stubborn, though not voluble, and
Protarchus is assertive and argumentative; he often raises objections where Socrates’
position needs defence. Plato returns to one of the topics of the early dialogues—­pleasure
and the good—and partly returns to the form of the early dialogues as well.
Plato, therefore, may have written direct dialogues, and monologues reporting dia-
logues, for different reasons on different occasions and at different times in his career.
Perhaps he wrote the Politicus or the Laws as a dialogue only because he was in the habit
of writing dialogues, and not because this form now seemed especially suitable for what
he wanted to say. Perhaps little or nothing would have been lost if Plato had given up not
only reported dialogue but also direct dialogue, and had written some of his latest works
as continuous treatises.

5. The Dialogues and

Plato’s Philosophy

Since Plato wrote dialogues, he does not address his reader directly. We assume that
Aristotle’s Metaphysics and Aquinas’ Summa present us with the views of their authors,
except when the authors say (as they often do) that they are presenting the views of
others. But we do not assume that because Oedipus is the leading character in two of
Sophocles’ plays, he presents Sophocles’ views. We may learn something from Sophocles’
plays about Sophocles’ views; we may, for instance, find that he emphasizes some
questions rather than others, that he presents some points of view more fully or more
sympathetically than others, and that he differs from other tragedians on these points
(so that he is not simply following conventions of the genre). What is Plato’s relation to
his dialogues and to the character of Socrates?
Since Aristotle often cites the views of the Platonic Socrates in order to criticize Plato,
he takes them to be Plato’s views. If most of his contemporaries had not assumed that the
Platonic Socrates represented Plato’s views, Aristotle’s criticism would have seemed
 The Platonic Corpus   81

misguided. If we were to criticize Shakespeare for holding the views of Hamlet or Othello,
or if we supposed Shakespeare held inconsistent views because Hamlet and Othello do
not agree, we would misunderstand Shakespeare. But we do not know of any contempo-
rary who accused Aristotle of a similar misunderstanding of Plato.
Later, however, the Sceptical Academy treated the dialogues as critical exercises that
examine the arguments for and against each side in order to leave the reader in a puzzled
condition (aporia). The point of these exercises is to induce suspension of judgment
about dogmatic claims. According to the Sceptical Academy, the dialogues are primarily
“peirastic” (“testing”) rather than dogmatic. This description seems to fit parts of some
earlier dialogues (those usually called “Socratic” by modern students), parts of the
Theaetetus, and most of the Parmenides. Ancient readers who gave generic titles to
the dialogues called some of them “peirastic,” thereby recognizing some plausible
elements in the Sceptical interpretation of Plato.
The Sceptical Academy tried to enlist Plato in a Sceptical project. We have no reason
to believe that this was Plato’s project. The later Academy returned to a doctrinal inter-
pretation of the dialogues.34 We may therefore claim the support of most ancient readers
for a doctrinal interpretation—though readers differed among themselves about what
doctrines should be ascribed to Plato.
Were Aristotle and other ancient readers wrong to treat the views of the Platonic
Socrates as Plato’s views? We would have good reason to disagree with Aristotle if we found
that the views of the Platonic Socrates are so unstable, fragmentary, or ­inconsistent that
they are better understood as starting points for inquiry than as firm conclusions. If,
however, we find that a reasonably coherent philosophical outlook and a reasonably
intelligible line of philosophical development can be ascribed to the Platonic Socrates,
we have some grounds for claiming to have found Plato’s views.
For this reason, the interpreter of Plato cannot do without philosophical reflexion on
the content of the dialogues. We might have preferred to establish a firm framework for
philosophical interpretation by appeal to “extra-philosophical” evidence—the use of
the dialogue form, chronology, characterization, dramatic dates, and so on—in the
hope that it would be less disputable. All of this evidence is indeed useful, even indis-
pensable, for the understanding of the dialogues, but we cannot rely on it independently
of philosophical interpretation. To understand the significance of the dialogue form, or
the character of Socrates, or the use of characters and dramatic dates, we have to make
our minds up about the doctrines—if any—that Plato reaches in different dialogues.

34 A doctrinal interpretation of Plato apparently left its mark even on the ancient manuscripts of
Plato. DL iii 66 mentions that an asterisk in the margin of manuscripts was used to indicate harmony
(sumphônia tôn dogmatôn) in Plato. See Alline, Texte 93. Alline (187–88) found 15 asterisks in the
surviving manuscript T, mainly in the myth in the Phaedrus. F. Solmsen, “The Academic and
Alexandrian Edition of Plato’s Works,” Illinois Classical Studies 6 1981, 102–11, at 107, doubts whether
they serve the purpose mentioned by DL.
82   T. H. Irwin

6. The Early Editions of the Corpus

Wherever35 the oral performances took place—in the Academy, in other invited gatherings,
or in bookshops—Plato’s dialogues were circulated widely enough in Athens to be
the subject of parody and allusion in comedy.36 One comic fragment alludes to discussions
carried on in the Academy, but it offers no help on their subject matter, apart from
the suggestion that definitions were discussed.37 If we could accept the suggestion that
Aristotle’s Topics was designed for the conduct of discussions within the Academy,38 we
would know some of what happened in the Academy, but we would not know whether
the dialogues played any part in it.
To ensure circulation of his works outside Athens, Plato needed someone to travel
abroad with them, as Zeno the Eleatic did on behalf of his own works (according to the
Parmenides). Hermodorus, a Syracusan member of the Academy, introduced Plato’s
works to Sicily, with Plato’s permission.39 It is not clear how widely the dialogues were
circulated,40 but some stories suggest that at least some of them were read or heard
outside Athens.41
After Plato’s death his nephew Speusippus became the head of the Academy. Plato’s
will does not say anything about his books and papers (DL iii 41–3). One might suggest,
then, that his works survived only through the various copies that had been made by buyers
in his lifetime, and the further copies made from these copies. If this suggestion is right,
there was no standard or authorized text.
This suggestion is less likely, however, than the alternative suggestion that the Academy
is the ultimate source of our texts. If Plato left the Laws incomplete,42 it was probably
never published (in the ways described above) in Plato’s lifetime. The same is true of the
incomplete Critias. Since these works are in our present Corpus, someone after Plato’s
death possessed his unpublished works, and allowed them to be disseminated. Most
probably, Speusippus and the Academy had received these works from Plato.

35 A clear and lively account of this question appears in. Grote, Plato, vol. 1, ch. 6 (“The Platonic
Canon”). On the history of the Corpus see Alline, Texte. Much of the relevant evidence is collected by
H. Dörrie, Der Platonismus in der Antike, vols. 1–2 (Stuttgart: Frommann-Holzboog, 1987–1990).
36 DL iii 26 mentions an allusion to the Phaedo by the comic poet Theopompus.
37 See the fragment of Epicrates translated and discussed by Dillon, Heirs 7–8.
38 See Ryle, Plato’s Progress, ch. 4.
39 On Hermodorus see Academicrorum Philosophorum Index Herculanensis ed S. Mekler (Berlin:
Weidmann, 1902), 34. Cicero mentions Hermodorus, ad Att. xiii 21.4 (= Shackleton Bailey 327),
remarking that (unlike Cicero’s publisher) Hermodorus did not act without Plato’s permission. See
D. R. Shackleton Bailey ad loc. in Cicero: Letters to Atticus vol. 5 (Cambridge: CUP, 1966).
40 DL vii 31 suggests a restricted circulation outside Athens.
41 Themistius 23, 295b (Teubner) mentions Axiothea of Arcadia (who, after reading the Republic, put
on men’s clothes and went to Athens to listen to Plato) and a Corinthian farmer (who, after reading the
Gorgias, abandoned his farm and became a disciple of Plato).
42 See note 12 above.
 The Platonic Corpus   83

Did the Academy maintain an authorized edition of Plato’s works? The earliest
papyrus fragments of Plato display considerable textual variation.43 Anyone who had
access to a copy, and could pay a scribe, could make a further copy from the copy. Repeated
copying multiplies errors. But the Academy may have tried to counteract this process of
corruption. According to one curious story, anyone who wanted to read Plato’s works
had to pay a fee to the owners.44 Instead of selling the Corpus to a bookshop, the Academy
apparently required prospective readers to pay for a copy made from the texts in the
Academy. This would not only make money for the Academy, but would also require
prospective owners of the Corpus to make their copies from a (supposedly) accurate
copy. Such a measure would not prevent A from buying a copy, and then lending it to B
to make a further copy from A’s copy, but it would mean that B would at least borrow a
better copy from A than either A or B might have obtained from a bookshop or from
someone who had obtained a copy from a bookshop. Plato’s successors may have been
doing their best to prevent damage that might result from inexpert copying. According
to one story, Aristotle’s works suffered from this sort of damage.45
We do not know how many people read Plato’s works seriously in the generation after
his death. Plato is said to have given Aristotle the nickname of “The Reader” (anagnôstês),
perhaps because he was an unusually assiduous student of written texts, including the
texts of Plato. Aristotle’s works contain many criticisms of points in the dialogues, and
some complaints about their obscurities, but he never says he asked Plato what he meant.46
Perhaps Aristotle encouraged the practice of careful reading of philosophical texts.
Our knowledge of other people’s use of the Corpus is limited, because, apart from
Aristotle, most philosophical writings that survive from the period between the death of
Plato and the lifetime of Cicero are fragmentary. But we have some evidence of interest
in Plato. Zeno the Stoic was a student of Polemon the head of the Academy.47 Both Zeno
and Chrysippus discuss and criticize Platonic dialogues.48 Under Arcesilaus the Sceptical

43 The history of the Platonic corpus in the Hellenistic period is discussed by G. Jachmann, “Der
Platontext” in Textgeschichtliche Studien, ed. C. Gnilka. Beiträge zur Klassischen Philologie 143 (1982),
ch.2. His extreme scepticism about whether there was an Academic and Alexandrian “standard text” at
all is endorsed by J. Barnes, “Ancient Plato”, in Mantissa: Essays in Ancient Philosophy, IV Oxford:
OUP 2015), ch. 15, 204–43. For a less sceptical view see L. Taran, “The Manuscript Tradition of Plato’s
Phaedo,” in Collected Papers (Leiden: Brill, 2001), ch.14, from Gnomon 48 (1976), 760–68.
44 DL iii 66 attributes this information to Antigonus of Carystus (fl. c.240). Since it comes from
Antigonus’s life of Zeno, it may refer to the situation in Zeno’s lifetime (335–263).
45 According to Strabo xiii 1.54, a damaged copy of Aristotle came into the hands of Apellicon, who
filled the gaps he found in the text, but did not know enough to fill the gaps correctly. This dramatic
story about the fate of the (allegedly) sole copy of the Aristotelian Corpus may well be (at the least)
exaggerated, but it suggests what could happen to texts in the process of circulation and recopying.
46 Aristotle “the reader”: Vita Marciana §6. See I. Düring, Aristotle in the Ancient Biographical
Tradition (Gothenburg, 1957), 98. Düring comments on it at 108, 368. See also J. Annas, “Aristotle on
Inefficient Causes,” Phil. Quart. 32 (1982), 311–26, at 326.
47 DL vii 2. On Polemon see Dillon, Heirs, ch. 4.
48 Zeno’s Republic; see DL vii 32. Chrysippus on Plato; see Plutarch, Stoic. Rep. 15–16.
84   T. H. Irwin

Academy studied the dialogues from their sceptical point of view.49 The Epicurean
school also seems to have been familiar with Plato.50 Poseidonius (c.135–51 b.c.) and
Cicero (106–43 b.c.) provide fuller evidence of knowledge of Plato. Poseidonius dis-
cussed Plato’s views on the division of the soul, and examined passages in the Phaedrus
and Timaeus in some detail. Cicero translated the Timaeus into Latin, and partly imi-
tated the Republic and Laws in two of his own works.51 From Cicero onward Plato is
familiar in philosophical circles, both Platonic and Stoic.52 Outside philosophical cir-
cles, Plato was known to orators, poets, literary critics, and many of their readers, from
the fourth century onward.53

7. Early Arrangements of the Corpus

In what form did early readers read the Platonic Corpus? Three stages in its transmission
deserve attention:

1. The Library of Alexandria was organized by Demetrius of Phaleron, a pupil of

Theophrastus, on the model of the library of Aristotle’s Lyceum. It was well stocked
with philosophical works, and so probably included Plato’s works. Demetrius prob-
ably used the edition that was in the Lyceum, and this was probably the Academic
edition. The Alexandrian copy was probably derived from the Academic edition.
2. Aristophanes of Byzantium (?c.257–180? b.c.), working in Alexandria, arranged
some of the Corpus in trilogies (DL iii 61–2). Hence the Alexandrian and the
Academic editions probably had not been arranged in any canonical order. About
this time, scholarly editing of Plato began. Some editions included editorial marks
to indicate spurious intrusions and textual corruptions (DL iii 65–6).
3. The division into tetralogies precedes the mid-first century bc.54 Thrasyllus
(d. ad 36)55 kept it in his edition. He is said to have claimed that Plato himself
had arranged his works in tetralogies, following the arrangement of tragedies

49 DL iv 32 suggests that Arcesilaus’ acquisition of the Platonic corpus was a significant event, as
though he had not always had easy access to it. Cf. the story about Zeno, note 45 above.
50 One of the spurious works in the Platonic corpus, the Axiochus, presents a curious combination
of Socratic and Epicurean views about death.
51 See Kidd’s commentary on F 31, 85, 290–91 (in L. Edelstein and I. G. Kidd, eds., Posidonius, 4 vols.
(Cambridge: CUP 1972–1999)); Dillon, Heirs 216–31; A. E. Taylor, Commentary on Plato’s Timaeus
(Oxford: OUP, 1928), 32–33.
52 A helpful short account of the treatment of Plato in Hellenistic philosophy is given by J. Barnes,
“Ancient Plato.”
53 On knowledge of Plato see R. Hunter, Plato and the Traditions of Ancient Literature (Cambridge
CUP, 2012).
54 The terminus ante quem for the tetralogies is the mid-first century b.c., since Varro (116–27 b.c.)
refers to the Phaedo as “Plato in the Fourth” (Varro LL vii 37, published before Cicero’s death in 43 b.c.).
Cf. F. Solmsen “The Academic and Alexandrian Edition.”
55 On Thrasyllus’ dates see H. Tarrant, Thrasyllan Platonism (Ithaca, NY: Cornell UP, 1993), 215 (T1).
 The Platonic Corpus   85

performed at Athenian festivals (DL iii 56).56 Internal references suggest that the
Republic, Timaeus, Critias, and Hermocrates were intended to form a group.57 Of
these four the Hermocrates was apparently never written, and the Critias was left
unfinished. Similarly, the Theaetetus, Sophist, Politicus, and Philosopher (never
written) may have been intended to form a group.58 Plato’s apparent interest in
tetralogies may have encouraged later readers and editors to arrange all the dia-
logues tetralogically.

Thrasyllus collected not only the nine tetralogies, but also the acknowledged spurious
works (DL iii 62). Since the spurious works fall outside the tetralogies, they were separated
from the genuine works at or before the time when the tetralogies were introduced. Hence
they were introduced into the Corpus before the mid-first century bc. If some or all of
the dubious works in the tetralogies are also spurious, the compilers of the tetralogical
list took some spurious works to be authentic.
It would probably have been difficult to introduce spurious works into the Corpus
after it had been placed in the Alexandrian Library and was regularly studied by scholars
such as Aristophanes. By the time the tetralogies were compiled, six works had been
recognized as spurious. Probably the Academic edition already included spurious works.
To see how this could happen, we should recall that the Academy inherited from Plato not
only his published works, but also his unfinished works, and his finished but unpublished
works (if there were any). Plato and the Academy may also have kept on their shelves
some dialogues by members of the Academy. The First Alcibiades, for example, may have
been composed to do for the Academy what it did for later Platonists (see Anon. Prol. 26.
23–6)—to introduce the main themes of Plato’s philosophy, with a special emphasis on
psychophysical dualism. The De Virtute is a short essay on the theme of the Protagoras
and Meno, “Can virtue be taught?.” The Sisyphus discusses questions that arise from the
Meno and from Aristotle’s Ethics. Other spurious works, including the Epicurean-
influenced Axiochus, were written later.59

56 “But Thrasyllus says that he published his dialogues corresponding to the tragic tetralogy. For, they
contended with four plays . . ., one of which was a satiric drama, and the whole four plays were called a
tetralogy.” Apparently, then, Thrasyllus did not represent himself as introducing the tetralogical
arrangement. Albinus, Eisagoge 4, attributes a tetralogical arrangement to Dercyllidas (of uncertain date)
and to Thrasyllus. See J. Mansfeld, Prolegomena (Leiden: Brill, 1994), ch. 2; Tarrant, Platonism, esp. chs. 1–4.
57 See Tim. 17b–19a (summary of part of the Republic), 27ab; Crit. 108ab (expecting a discourse by
Hermocrates). It is by no means clear, however, that the reference to the Republic implies that Plato
intended this dialogue to be the first member of a tetralogy. He may simply have intended to write a
trilogy beginning with the Timaeus. See F. M. Cornford’s reasonable doubts in Plato’s Cosmology
(London: Routledge, 1937), 4–5. Still the belief that Plato had intended a tetralogy beginning with the
Republic may have encouraged the effort to divide all the dialogues into tetralogies.
58 See Sph. 253e; Pol. 257a, 258a. As Cornford remarks, Plato’s Theory of Knowledge (London:
Routledge, 1934), 168, Socrates implies that he and Young Socrates will be the main speakers in the
Philosopher. M. L. Gill, Philosophos (Oxford: OUP, 2012), argues that Plato never intended to write the
fourth dialogue that is announced. Others (see Gill, Philosophos 1n3) identify the Philosopher with some
extant dialogue. Given the Critias and Hermocrates, I see no reason to doubt that the Philosopher marks
another uncompleted project.
59 The influence of Stoicism is more difficult to decide, partly because Stoic ethics is quite close to
Socratic and Platonic ethics on many points, as we can see in the Eryxias.
86   T. H. Irwin

This expansion of the Corpus to include works that are non-Platonic in doctrine as
well as origin may result from the fact that members of the Academy, from Speusippus
onward, held various philosophical views, some of them opposed to Plato’s. It would not
be surprising if they wrote dialogues from their different points of view. If this is so, the
Corpus contains not only the works of Plato, but also a number of post-Platonic reflex-
ions on Platonic themes.60 These spurious dialogues deserve study in their own right.

8. The Manuscripts

Our main sources for the text are 51 Byzantine manuscripts, copied from the ninth
century ad onward in the Greek-speaking areas ruled from Constantinople.61 Plato’s
complete works survived until the Byzantine period (in contrast to the works of many
other ancient authors) because he was widely studied in late antiquity in the later Platonic
schools (which modern writers refer to as “Middle Platonist” and “Neoplatonist”),62
which also produced commentaries on the dialogues. As we have seen in Diogenes
Laertius’s report, readers attempted to correct the text. No doubt their attempts were
sometimes mistaken, but they may also have removed some errors, and so may have
partly counteracted the progressive corruption resulting from repeated copying.
Our oldest surviving manuscript (normally referred to as B) was copied by John the
Calligrapher, who finished his work (containing the first six tetralogies) in 895, on the
orders of Arethas, a deacon in Patras who later became archbishop of Caesarea.63
Knowledge of Plato spread from the Byzantine Empire to Italy and the rest of Western
Europe during the fourteenth and fifteenth centuries. The first manuscript of Plato to
reach Italy may have been the (probably) eleventh-century manuscript W, which arrived
in the fourteenth century.64 A Latin translation of Plato by Marsilius Ficinus was
­published in 1484 and became a bestseller. The first printed edition of Plato in Greek
appeared in 1513.65

60 A. Carlini, “Alcuni dialoghi pseudoplatonici e l’Accademia di Arcesilao” Annali della Scuola Normale
di Pisa ser. ii 31, 1962, 33–63, argues that some spurious dialogues (he discusses Clt., Thg., IIAlc., Rivals)
reflect the influence of the Sceptical Academy under Arcesilaus, and that they may have been included in
a collected edition that Arcesliaus produced in order to defend his interpretation of Plato from an
authentic text. The arguments are highly speculative, but the possibility is worth considering.
61 For a list of the Plato manuscripts see R. S. Brumbaugh and R. Wells, The Plato Manuscripts (New
Haven, CT: Yale UP, 1968). G. J. Boter, The Textual Tradition of Plato’s Republic (Mnemosyne Supp. 107
(1989)), p. xx, suggests that still more Plato manuscripts may survive in uncatalogued libraries.
According to G. Pasquali, Storia della Tradizione e Critica del Testo (2nd ed., Florence: Le
Monnier, 1952), 247, Plato is the classical author with the richest textual tradition after Homer.
62 On later Platonism see Brittain’s chapter 28 in this volume.
63 On Arethas see Reynolds and Wilson, Scribes 64–65. Their Plate III shows part of Arethas’
manuscript of Plato (now in the Bodleian Library, Oxford).
64 See E. R. Dodds, ed. Plato’s Gorgias (Oxford: OUP, 1959), 39.
65 Omnia Platonis Opera, published and printed by Aldus Manutius, Venice, 1513. On Ficinus see
Reynolds and Wilson, Scribes 155.
 The Platonic Corpus   87

Modern printed editions from the sixteenth century onward choose their text from
different manuscripts, and from conjectural emendations. Editors and textual critics try
to establish the most probable Greek text by comparison of the different manuscripts.
At the foot of each page in the OCT of Plato we find textual variants whose sources
are indicated by letters referring to manuscripts or by the names of modern (sixteenth
century and later) critics who have proposed emendations not found in any manuscript.
In some editions we also find a list of “testimonia”—lists of places in ancient authors
where a given passage of Plato is quoted or closely paraphrased.66 Why is all this appa-
ratus necessary?67
The transmission of the Corpus in antiquity explains why the manuscripts have
different texts, and why one needs to exercise some judgment in choosing among variant
readings. Copyists make mistakes; later copyists multiply mistakes because of ignorance
or inattention or illegible handwriting. Copies were originally written in capital letters
(“uncials”) without punctuation and without spaces between words. Small (“minuscule”)
letters were introduced, probably in the eighth century,68 and eventually punctuation
was also introduced.69 These changes offered opportunities for new errors.
The surviving manuscripts of Plato display variations that show the effects of these
processes of textual corruption. Editors have tried to undo some of these effects. If a
particular sentence in manuscript Z* is so grossly ungrammatical as to be unintelligible, but
the corresponding sentence in Y* is slightly different in ways that make it grammatical
and intelligible, Y* has probably preserved what Plato wrote.70 But once we go beyond
relatively easy corrections, we need to know more about the manuscripts. If Z* and Y*
both offer a reading that makes sense, or if neither makes sense but we want to know
which might be a better guide to what Plato wrote, we need to know something about
the credentials of each manuscript. The study of Greek and Latin texts advanced signifi-
cantly in the nineteenth century, when critics undertook the systematic comparison of
manuscripts and began to trace their affiliations. Recent editions differ from the early
printed editions because their textual proposals rest on a wider acquaintance with Plato,
with Greek, and with the relations among manuscripts.
Should we rely on the earliest manuscript? If Y* is earlier than Z*, Y* should be less
affected by the process of corruption that inevitably follows repeated copying. We have

66 See, e.g., the back of the new OCT i and of the new OCT of the Republic. Testimonia for the
Republic are listed and discussed by Boter, Textual Tradition 285–365.
67 I have derived most of the information on texts and textual criticism from. Dodds, Gorgias,
34–58; R. S. Bluck, ed., Plato’s Meno (Cambridge: CUP 1961), 129–47; Boter, Textual Tradition; Reynolds
and Wilson, Scribes; A. Carlini, Studi sulla tradizione antica e medievale del Fedone (Rome:
Ateneo, 1972); S. Martinelli Tempesta, La tradizione testuale del Liside di Platone (Firenze: la Nuova
Italia, 1997), ch. 7.
68 On the difficulty of reading ancient books see Reynolds and Wilson, Scribes 4. They discuss the
transition from uncial to minuscule at 59f.
69 S. R. Slings ed., Plato’s Clitopho (Cambridge: CUP 1999), 342, maintains that ancient (i.e.,
pre-Byzantine) prose texts had punctuation.
70 I use asterisks to indicate imaginary manuscripts introduced for the sake of examples. Letters
without asterisks are used to name actual manuscripts, following the names in the OCT.
88   T. H. Irwin

enough information about the absolute dates of some manuscripts to form reasonable
estimates of their relative dates. In some cases, the scribe helpfully adds a date at the end
of his manuscript (as John the Calligrapher did with B, mentioned above). Sometimes
we can find relevant information in library catalogues. On this basis we can distinguish
different Greek hands, and can sometimes use this information to estimate the date and
provenance of different manuscripts. This information about dates influences editors in
their decisions about which manuscripts they should follow. A glance at an edition of
Plato (e.g., OCT i p. 2) will show that editors regularly report the readings of only a few
of the surviving manuscripts, listing the other manuscripts as “more recent” (recentiores),
and citing only occasional readings from them. Some editors have based their text on the
readings of the earliest manuscript, regarding the readings in later manuscripts as errors
or (when they are probably correct) as conjectural emendations.
Most recent editors, however, have rejected this rigid rule of following the earliest
manuscript.71 Imagine this situation: (1) We have two surviving manuscripts, Y* and Z*.
(2) Y* is earlier than Z*. (3) Z* was copied from an ancestor manuscript W*, and Y*
from an ancestor X*. (4) W* is earlier than X*. Hence (5) the later manuscript Z* pre-
serves an earlier form of the text (from W*) than the earlier manuscript Y* preserves
(from X*). And so the textual critic should try to find the readings of W* and X*, and
should not simply follow Y* or Z*.72
Sometimes we have reason to believe that the lost manuscripts W* and X* underlie
the surviving Y* and Z*. Scribal errors may help editors, since an error that is common
to one group of manuscripts and absent from another group may show us that the first
group descend from an earlier manuscript. One group of manuscripts, for instance may
contain the same lacuna (a shorter or longer passage omitted from a copy), because the
scribe copying an earlier manuscript skipped a line (e.g.) in copying and transmitted his
error to the manuscripts copied from his copy. If, then, the surviving manuscript Y*
contains lacunae that are absent from Z*, this may suggest that Y* was copied from W*
(which had the lacunae), whereas Z* was copied from X* (which lacked the lacunae). In
that case, we should be more interested in the relative dates—if we can estimate them—of
W* and X* than in those of Y* and Z*.
Some shared errors may indicate the relative dates of manuscripts. Uncials and
minuscules give rise to different errors.73 Hence if Z* seems to contain more errors arising
from uncials than Y* contains, Z* and Y* were probably copied from different manuscripts,
W* and X*, and W* is probably earlier than X*. Hence the text preserved by Z* may be
earlier than the text preserved by Y*.
Fortunately, this actually happens with the manuscripts of Plato. Some of them, notably
the Vienna manuscript F and its affiliates, contain variants arising from uncial errors.74

71 See, e.g., Boter, Tradition 7–10.

72 Boter, Tradition, gives a detailed account of the manuscripts of the Republic, and of their relations.
Slings, Clitopho 340–43, states the conclusions briefly.
73 Since, e.g., C and G in the Latin alphabet look more similar than c and g, we are more likely to
confuse C and G than to confuse c and g.
74 The importance of F was pointed out by J. Burnet, “A Neglected MS. of Plato,” Classical Review 16
(1902), 98–101. See also Dodds, Gorgias 41–47; Boter, Tradition 12, 99–100.
 The Platonic Corpus   89

Though these are not the earliest manuscripts, they may preserve an earlier text, and
they may help us to correct the text offered by earlier manuscripts. This is only one small
example of evidence that justifies attention to one or another manuscript. Progress in
the study of the Plato manuscripts has sometimes shown, for instance, that a manuscript
previously thought to be simply an inferior copy is in fact an independent source.75
The text of Plato is generally sound. If we compare the critical apparatus on a random
page of a modern edition of Plato and of most other Greek authors, we will see that the
text of Plato contains many fewer variants significantly affecting the sense. The text of our
manuscripts is generally confirmed by the abundant testimonia in other ancient writers.76
A comparison of the old and the new OCT confirms the general soundness of the text.77
The critical apparatus in the new edition differs significantly from the old, partly because
of further research (such as that just described) on the manuscripts and their relations.
But the text printed in the two editions generally agrees, and the changes in the new edition
are not radical. Moreover, some of the more significant changes do not result from further
research on the textual tradition. In some cases they result from closer consideration of
the testimonia.78 In other cases they are conjectural (and sometimes questionable) alter-
ations by the editors with no basis in the manuscripts.79
This generally good condition of the text is no accident. The continuous interest in
Plato in antiquity helped to preserve the text and to maintain its general integrity. The
text of Aristotle is far more imperfect; it may have suffered from declining interest in
him in the early Hellenistic period.80

Bibliography
Academicrorum Philosophorum Index Herculanensis ed. S. Mekler. Berlin: Weidmann, 1902.
Alline, H., Histoire du texte de Platon. Paris: Champion, 1915.

75 The treatment of the manuscript tradition of Plato by various editors is discussed by Pasquali,
Storia 247–69. Boter, Tradition 22, on the Republic, argues that the relative value of different
manuscripts is still not completely understood.
76 The quantity and distribution of the ancient testimonia for the Republic may be gathered from the
70-page index in Boter, Tradition 291–365. At 366–76 Boter gives a list of the authors cited. In
considering variations between the testimonia and the manuscripts of Plato, we need to bear in mind:
(1) The testimonia are derived from manuscripts that may be less sound than the Plato manuscripts.
(2) Ancient writers often quoted from memory. (3) Sometimes they emended the text of Plato on
doctrinal grounds; see J. M. Dillon, “Tampering with the Timaeus”, American Journal of Philology 110
(1989), 50–72. Hence the testimonia are not always to be preferred to the Plato manuscripts.
77 For some discussion of the new OCT i see C. J. Rowe, “Plato Re-edited”, Classical Review 47
(1997), 272–74; D. B. Robinson, “Textual Notes on Plato’s Sophist,” Classical Quarterly 49 (1999), 139–60.
78 S.R. Slings, Critical Notes on Plato’s Politeia (Leiden: Brill, 2005), 126, cites Plotinus in defence of a
convincing emendation of 518d10.
79 See, e.g., Slings’s arbitrary deletion of kaiper noêtôn ontôn meta archês in Rep. 511d, defended in
Critical Notes, 114–19. At Cra. 385b3–d1 the editors of the new OCT unwisely follow M. Schofield
(“A displacement in the text of the Cratylus,” Classical Quarterly 22 (1972), 246–53) in bracketing the
passage. They go even further than Schofield in believing it has no place in the dialogue as it now
stands (note on 387c5). See also D. Sedley, Plato’s Cratylus. Cambridge CUP, 2005, 10–13.
80 See n46 above.
90   T. H. Irwin

Annas, J., “Aristotle on Inefficient Causes,” Philosophical Quarterly 32 (1982), 311–26.

Annas, J., and Rowe, C. J., eds., New Perspectives on Plato. Cambridge, MA: Harvard UP, 2002.
Anonymous, Prolegomena in Platonis Philosophiam, ed. L. G. Westerink. Paris: Les Belles
Lettres, 1990.
Aristotle, Complete Works, ed. J. Barnes. 2 vols., Princeton, NJ: Princeton UP, 1984.
Barnes, J., “Ancient Plato,” in Mantissa: Essays in Ancient Philosophy, IV. Oxford: OUP 2015),
ch. 15, 204–43.
Bluck, R. S., ed., Plato’s Meno. Cambridge: CUP, 1961.
Boter, G. J., The Textual Tradition of Plato’s Republic. Mnemosyne Supp. 107. ______, 1989).
Brandwood, L., “Stylometry and Chronology” in The Cambridge Companion to Plato, ed.
R. Kraut, Cambridge: CUP, 1992.
Brumbaugh, R. S., and Wells, R., The Plato Manuscripts. New Haven, CT: Yale UP, 1968.
Burnet, J., “A Neglected MS. of Plato,” Classical Review 16 (1902), 98–101.
Burnyeat, M. F., and M. Frede, The Pseudo-Platonic Seventh Letter, ed. D. Scott. OUP, 2015.
Carlini, A., “Alcuni dialoghi pseudoplatonici e l’Accademia di Arcesilao” Annali della Scuola
Normale di Pisa ser. ii 31, 1962, 33–63.
Carlini, A., Studi sulla tradizione antica e medievale del Fedone. Rome: Ateneo, 1972.
Cherniss, H. F., “The Relation of the Timaeus to Plato’s Later Dialogues,” in Studies in Plato’s
Metaphysics, ed. R. E. Allen. London: Routledge, 1965) ch. 17. From American Journal of
Philology 78 (1957), 225–66.
Cicero, Letters to Atticus, ed. D. R. Shackleton Bailey. Cambridge: CUP, 1966.
Cornford, F. M., Plato’s Cosmology. London: Routledge, 1937.
Cornford, F. M., Plato’s Theory of Knowledge. London: Routledge, 1934.
Corpus dei Papiri Filosofici Greci e Latini, vol. 3. Florence: Olschki, 1995.
Dillon, J. M., The Heirs of Plato. Oxford: OUP, 2003.
Dillon, J. M., “Tampering with the Timaeus,” American Journal of Philology 110 (1989), 50–72.
Dodds, E. R., ed. Plato’s Gorgias. Oxford: OUP, 1959.
Dörrie, H., Der Platonismus in der Antike, vols. 1–2. Stuttgart: Frommann-Holzboog, 1987–90.
Dover, K. J., ed., Aristophanes’ Clouds. Oxford: OUP, 1968.
Düring, I., Aristotle in the Ancient Biographical Tradition. Gothenburg: Almquist and Wiksel
[?], 1957.
Festugière, A. J., Etudes de philosophie grecque. Paris: Vrin, 1971
Gill, C. J., and McCabe, M. M., (eds.), Form and Argument in Late Plato. Oxford: OUP, 1996.
Gill, M. L., Philosophos. Oxford: OUP, 2012.
Grote, G., Plato and the Other Companions of Socrates, 4 vols. London: Murray, 1888.
Guthrie, W. K. C., History of Greek Philosophy, vol. 4. Cambridge: CUP, 1975.
Hackforth, R., “The Modification of Plan in Plato's Republic,” Classical Quarterly 7 (1913),
265–72.
Hunter, R. L., Plato and the Traditions of Ancient Literature. Cambridge: CUP, 2012.
Irwin, T. H., Plato’s Ethics. Oxford: OUP, 1995.
Irwin, T. H., “The Parts of the Soul and the Cardinal Virtues,” in Platon: Politeia, ed. O. Höffe
(Berlin: Akademie Verlag, 1997), ch. 6.
Irwin, T. H., “The Inside Story of the Seventh Platonic Letter: A Sceptical Introduction,” Rhizai
6 (2009), 7–40.
Jachmann, G., “Der Platontext” in Textgeschichtliche Studien, ed. C. Gnilka. Beiträge zur
Klassischen Philologie, 143 (1982).
Johnson, E., Charles Dickens: His Tragedy and Triumph. London: Allen Lane, 1977.
 The Platonic Corpus   91

Kahn, C. H., “On Platonic Chronology,” in New Perspectives, ed. Annas and Rowe, ch. 4.
Kahn, C. H., “Writing Philosophy,” in Yunis, ed., Texts, ch. 7, 139–61.
Kahn, C. H., Plato and the Socratic Dialogue. Cambridge: CUP, 1996.
Ludwig, W., “Plato's Love Epigrams,” Greek, Roman, and Byzantine Studies 4 (1963), 59–82.
Mansfeld, J., Prolegomena. Leiden: Brill, 1994.
Martinelli Tempesta, S., La tradizione testuale del Liside di Platone Florence: la Nuova Italia, 1997.
Metzger, B. M., The Text of the New Testament, 3rd ed. Oxford: OUP, 1991.
Morrow, G. R., “Aristotle’s Comments on Plato’s Laws” in Aristotle and Plato in the Mid-Fourth
Century, ed. I. Dȕring and G. E. L. Owen. Gothenburg: Almquist and Wiksell, 1960, 145–62.
Morrow, G. R., Plato’s Cretan City. Princeton, NJ: Princeton UP, 1960.
Nails, D., The People of Plato. Indianapolis, IN: Hackett, 2002.
Omnia Platonis Opera. Venice: Aldus Manutius, 1513.
Owen, G. E. L., “The Place of the Timaeus in Plato’s Dialogues,” in Logic, Science and Dialectic.
London: Duckworth, 1986, ch. 4, from Classical Quarterly 3 (1953), 79–95.
Oxford Classical Dictionary, 3rd ed., ed. S. Hornblower and A. Spawforth. Oxford: OUP, 1996.
Pasquali, G., Storia della Tradizione e Critica del Testo, 2nd ed. Florence: Le Monnier, 1952.
Plato, Collected Dialogues, ed. J. M. Cooper and D. S. Hutchinson. Indianapolis, IN: Hackett, 1997.
Platonis Opera quæ extant omnia, ed. H. Stephanus. Geneva: H. Stephanus, 1578.
Platonis Opera, ed. J. Burnet. 5 vols., Oxford: OUP, 1900–1907.
Platonis Opera,. vol. 1, ed. E. A. Dukes et al. Oxford: OUP, 1995.
Platonis Respublica, ed. S. R. Slings. Oxford: OUP, 2003.
Poseidonius, ed. L. Edelstein and I. G. Kidd, 4 vols. Cambridge: CUP 1972–1999.
Reynolds, L. D., and Wilson, N. G., Scribes and Scholars. 3rd ed. Oxford: OUP 1991.
Riginos, A. S., Platonica. Leiden: Brill, 1976.
Robinson, D. B., “Textual Notes on Plato’s Sophist,” Classical Quarterly 49 (1999), 139–60.
Rowe, C. J., “Plato Re-edited,” Classical Review 47 (1997), 272–74.
Ryle, G., Plato’s Progress. Cambridge: CUP, 1966.
Schofield, M., “A Displacement in the Text of the Cratylus,” Classical Quarterly 22 (1972), 246–53.
Sedley, D. N., Plato’s Cratylus. Cambridge: CUP, 2005.
Shorey, P., What Plato Said. Chicago: U. Chicago Press, 1933.
Slings S. R., ed., Plato’s Clitopho. Cambridge: CUP 1999.
Slings, S. R., Critical Notes on Plato’s Politeia. Leiden: Brill, 2005.
Solmsen, F. “The Academic and Alexandrian Edition of Plato’s Works,” Illinois Classical Studies
6 1981, 102–11.
Taran, L., “The Manuscript Tradition of Plato’s Phaedo,” in Collected Papers. Leiden: Brill,
2001, ch. 14, from Gnomon 48 (1976), 760–68.
Tarrant, H., Thrasyllan Platonism. Ithaca, NY: Cornell UP, 1993.
Taylor, A. E., A Commentary on Plato’s Timaeus. Oxford: OUP, 1928.
Thomas, R., “Prose Performance Texts,” in Yunis, ed. Texts, ch. 8, 162–88.
Yunis, H., ed., Written Texts and the Rise of Literate Culture in Ancient Greece. Cambridge:
CUP, 2003.
Yunis, H., “Writing for Reading,” in Yunis, ed., Texts, ch.9, 189–212.
 chapter 4

Pl ato’s Ways
Of W r iti ng

Mary Margaret M c cabe

1. The Dialogue Form: A Prospectus

Plato’s writing scintillates. And most—if not all1—of what comes down to us from
Plato’s hand is in the dialogue form, somehow or other. But should we speak of “the”
Platonic dialogue form? After all, the dialogues come in all sorts of different forms:
some are dramatic, others merely formalized discussion (compare the Phaedo and
the Statesman); some are in direct speech, others narrated (compare the Gorgias and the
Symposium); some seem to have a beginning, a middle, and an end, whereas others begin,
or end, in the middle of things (compare the Euthydemus and the Philebus); some have
Socrates in the central role, and others are dominated by less engaging, but more author-
itative figures (compare the Theaetetus and the Sophist).2 We may miss the complexity of
Plato’s ways of writing if we reduce his dialogues to a single and canonical shape. Such
reduction might be hopelessly banal (because vastly general), or else simply false.3
Still, we rightly call them dialogues. Plato’s characters talk to each other: about where
they are going (e.g., Euthyphro 2–4), who they met yesterday (e.g., Euthydemus 271),
where they are planning to have dinner (e.g., Symposium 174); and about virtue (e.g., Meno),
about knowledge (e.g., Theaetetus), about truth and falsehood (e.g., Sophist), about what

1 This depends on whether we take the Letters to be genuine; compare K. Sayre, Plato’s Literary
Garden: How to Read a Platonic Dialogue (Notre Dame, Ind., 1995). I include Apology by virtue of the
passage of dialogue within it, see Section 7 of this chapter.
2 For all the figures who take the leading role, strong claims to authority are made: Parmenides at
Parmenides 127–29, the Eleatic Stranger at Sophist 216, Timaeus at Timaeus 20, the Athenian Stranger
in his authoritative leading of the discussion from the beginning of Laws I. The position of Socrates is
more complicated; see Section 3 of this chapter.
3 See A. Long’s excellent remarks, “Character and Dialectic: The Philosophical Origins of the
Platonic Dialogue,” Ph. D. dissertation (University of Cambridge, 2004); Long, “The Form of Plato’s
Republic,” in R. Osborne, ed., Debating the Athenian Cultural Revolution (Cambridge, 2007).
94   Mary Margaret Mccabe

there really is (e.g., Parmenides). Some of what they say seems utterly trivial, but some
seems universal and of abiding importance. So, for example, in the Protagoras, the short
conversation between Socrates and his friend at 309c–310a has the air of gossip, while
the discussion about the virtues between Socrates and Protagoras at 332a–333b appeals
to general principles and seeks an abstract conclusion. Does the dramatic detail of the
dialogues have any bearing on their philosophical purposes?
I shall say that it does—that, indeed, we cannot properly make sense of what Plato does if
we ignore the effect on the arguments of dramatic context, allusion, characterization—
indeed, of all the aspects of the style and drama of a dialogue. This effect is felt both
particularly (where the dramatic detail alters radically how we understand individual
arguments) and generally (where various strategies render the reader carefully reflective
on what is said). I shall conclude that the philosophical content of a dialogue is to be
found, at least, in the dialogue as a whole. How Plato writes, therefore, is indivisible from
what he is trying to say.

2. A Different View:
Pictures and Frames

But this conclusion does not command universal assent. For many suppose that the
philosophical picture of a dialogue lies just in its arguments, while the fictional a­ pparatus—
the “who said what to whom”—is only the frame in which they are presented.4 So, for
example, the Euthyphro is constituted by sections of argument, clearly demarcated,
about what is the holy (5d–11b, 11e–15e), exhibited in a frame discussion about Socrates’
meeting with Euthyphro as they are both hurrying to court, and Euthyphro’s retreat at
the end (2a–5d; 15e–16a). These argumentative sections of the dialogues can, often, be
formalized. For example, Parmenides 152a–e purports to show that the one both is and
becomes both older and younger than itself, and it can be formalized as a linear sequence
without the answers that young Aristotle gives to Parmenides’s questions.5 If those
answers make little difference, is the rest of the descriptive material of a dialogue irrele-
vant to its arguments save as their frame?

4 The “frame/picture” contrast is an old one; compare, for example, D. Gallop, Plato: Phaedo
[Phaedo] (Oxford, 1975), 74. Gail Fine rightly reminds me that the relation between frame and picture
is a variable one. I retain the terminology in what follows, to mark the difference between what is said
and the setting in which it is said, while noting that my conclusion—that philosophical significance is
to be found in each dialogue as a whole—supposes that the frames are, indeed, a part of the picture.
Consider the complex play on frame in Magritte, René. The Human Condition. 1933, National Gallery of
Art, Washington DC.
5 F. M. Cornford, Plato and Parmenides (London, 1939), misses them out. This is not, surely, the
Aristotle who wrote the Metaphysics—but who is to say that Plato did not choose this name for his
character with deliberation?
 Plato’s Ways Of Writing   95

One answer would be that some parts of a dialogue give us the emotive cast, the tone,
of what will follow.6 Thus the Republic opens with an optimistic discussion of how
Socrates and Glaucon fell in with Polemarchus, Adeimantus, and friends and how their
long conversation about justice ensued. The Symposium describes—in delightful detail—
Socrates’ wayward progress to Agathon’s dinner party and what happened next. In darker
contrast, several dialogues turn on Socrates’ trial and execution: the Phaedo describes
Socrates’ death, surrounded by friends, whose grief Socrates finds inappropriate. This
theme is picked up in other dialogues, too: Meno (Meno 80b) warns Socrates of the
dangers of practicing his puzzling method of examination; in the Hippias Major, Socrates’
alter ego tells him that he will never make a really fine speech in court until he knows what
fine really is (304c–e); and the Theaetetus, presented as a memorial to Socrates’ young
lookalike, Theaetetus, ends with Socrates going off to court (to his meeting, we suppose,
with Euthyphro in the King’s Porch; Theaetetus 210d). All this affects us, makes us care
about what we read. However, it makes no substantial difference to the real philosophi-
cal business in hand—or so this account of the dialogue form would go.7 Perhaps, then,
discussions about issues of abstract generality (virtue, knowledge, truth) count as the
philosophical content of the dialogues, and details, such as where Socrates was sitting in
the gymnasium when his friends arrived, do not. Instead, those details may serve to
seduce and attract us into reading, to please and sometimes anger us—so that we set foot
on the harsh road to understanding the arguments beyond; but they are not—so this
account would say—otherwise part of the dialogue’s philosophical destination.
Is the frame so easily separated from its picture, or the emotional tone of Plato’s
writing from its content?8 We might think Parmenides’s comparison of himself to an
aged racehorse unimportant (Parmenides 136e–137a) and yet be quite sure that there is
some significance to the arguments of the theme that Socrates is a young man (127c, 130e,
135d).9 Or we might suppose that Socrates’ impending death makes no difference to the
Phaedo’s arguments about the immortality of the soul, even if it adds urgency, but still
wonder whether the calmness of Socrates’ death by hemlock is an anomaly that tells us

6 Compare here the dramatic opening of the Theaetetus and the ancient story that there existed
another, “more frigid” introduction that was not used by Plato: Anonymous Commentary on the
Theaetetus, 3.28–37.
7 Emotion, one might say, does not affect truth value; compare Jonathan Barnes’s view of the poverty
of Plato’s arguments, The Cambridge Companion to Aristotle (Cambridge, 1995), xvi.
8 In recent years, many have argued that it is not: see, for example, M. C. Stokes, Plato’s Socratic
Conversations [Conversations] (London, 1980), and R. Blondell, The Play of Character in Plato’s
Dialogues [Play] (Cambridge, 2002). A subtheme of this line of interpretation is the question of Plato’s
attitude to tragedy: see M. C. Nussbaum, The Fragility of Goodness: Luck and Ethics in Greek Tragedy
and Philosophy (Cambridge, 1986), and S. Halliwell, The Aesthetics of Mimesis: Ancient Texts and
Modern Problems [Mimesis] (Princeton, N.J., 2002); compare D. Roochnik, The Tragedy of Reason
[Tragedy] (New York, 1990).
9 This is a commonplace in accounts of how the arguments against forms are to be construed: for
example, R. E. Allen, Plato’s Parmenides, Translation and Analysis (Oxford, 1983), 100.
96   Mary Margaret Mccabe

something important about the nature of a philosopher.10 How are we to decide which
bits of the dialogue to include in a philosophical interpretation, and which we may safely
leave out? I suggest that if we treat the settings as dispensable frames for the arguments,
we miss both the ways in which the frames are involved in particular arguments and
how the frames determine the reader’s reflective stance.

3. The Interlocutors

In the dialogues, people talk to each other. Some are richly characterized and act accord-
ingly. Euthyphro, for example, comes to the discussion on his way to court, to prosecute
his father for the murder of a slave (4a–d); in this extraordinary act11 he has complete
confidence, by virtue of the expertise he claims in matters moral and religious (4e–5a).
This comes across as an arrogance so deep-seated that even when Socrates shows it to be
unfounded, Euthyphro cannot forsake it but quits the scene in haste (15c–e). Charmides
is introduced as a young man as noble of spirit as he is handsome (Charmides 154d),
and his discussion with Socrates reveals that his character is appropriately modest for a
discussion of the virtue of self-control or temperance, sôphrosunê. Ctesippus is said to
be headstrong (Euthydemus 273a); accordingly, he is the first to succumb to the attractions
of sophistic argument (298b ff.).
And there is Socrates: regularly the protagonist, but variously portrayed. Sometimes
he is Socrates the expert in erotics (e.g., Lysis 204b–c, Symposium 177e)—at times appar-
ently inflamed by beautiful young men (Charmides 155c–e, Symposium 216), at times
coldly self-controlled (Symposium 217b ff.). Sometimes he is a solitary (late for dinner
because he is thinking alone, Symposium 174–75; compare Euthydemus 272e); but some-
times he insists on the company of friends as the best way to proceed in philosophy
(e.g., Charmides 166d; Gorgias 486e ff.; Cratylus 391a). He is regularly described as brave
(e.g., Laches 181), and he certainly seems to have the courage to discuss anything with
anyone; but at times he is apparently terrified by the arguments against him (Euthydemus
293a). Sometimes he is modest (e.g., Euthydemus 272c–d), sometimes arrogant (Gorgias
482)—and sometimes the modesty seems to be a ploy, designed to flummox those to whom
he presents himself (compare Euthyphro 5a–c with 15d–16a).12 He is the self-­deprecating
inquirer after knowledge who proposed to the Athenians that as a punishment he should

10 C. Gill, “The Death of Socrates,” Classical Quarterly 23 (1983): 25–28, and E. Bloch, “Hemlock
Poisoning and the Death of Socrates: Did Plato Tell the Truth?” in T. Brickhouse and N. Smith, eds.,
The Trial and Execution of Socrates (Oxford, 2001).
11 The historical context is vital here to understanding just how extraordinary this would be in
ancient Athens; see Schofield, chapter 2 in this volume.
12 Socrates’ apparent irony, however, does not merely entitle us to negate whatever he says (note
Alcibiades’ remarks at Symposium 216): see, for example, G. Vlastos, Socrates, Ironist and Moral
Philosopher [Socrates] (Cambridge, 1991); A. Nehamas, The Art of Living: Socratic Reflections from
Plato to Foucault [Art] (Berkeley, Calif., 1998); and M. M. McCabe, “Irony in the Soul: Should Plato’s
Socrates Be Sincere?” in M. Trapp, ed., Socrates (London, 2007).
 Plato’s Ways Of Writing   97

be given a state pension (Apology 36). And his tone often seems insincere when he sets
up some opponent for a fall—however much he may demand sincerity from others.13 In
this, he seems less of the cooperative friend than a competitor, and often readers of a dia-
logue are incensed by his pride in his own “human wisdom” and his apparent contempt
for the terminal ignorance of everyone he meets (Apology 21–23).
This figure of Socrates is often unattractive and always difficult to interpret;14 it might
properly deter us from seeking a single and uniform account of the dialogues, as much
as from looking for a single Socrates persisting through the dialogues.15 For its Socrates
may be particular to each dialogue and may both remind us of the other Socrateses and
discourage us from supposing that any portrait aims for verisimilitude. Still, Socrates is
usually a vivid figure. Some dialogues, by contrast, present their main figures in an exig-
uous way. The Eleatic Stranger, for example, is introduced as “godlike” (Sophist 216); that
may be why he has little individual character.16 His interlocutors, Young Socrates and a
subdued Theaetetus, are as thinly characterized. The contrast between these dialogues
and those rich dramas of character has tempted commentators to posit a difference in
their dates: to suggest that “the” dialogue form was a literary device that lost its appeal
later in Plato’s life, to be replaced by an inadequate gesture in the direction of style. But
maybe we should pause: How are we to understand the relation between the richly por-
trayed Theaetetus of his eponymous dialogue and his thin counterpart in the Sophist?
Or, within a single dialogue, the benign figure of “father” Parmenides in his eponymous
dialogue, compared with his meager portrayal in the second half of the same dialogue?
Is the line of demarcation between character and argument to be drawn so easily? And is
it so to be drawn for Plato?

4. Identification and Transparency

For those cases where philosophical discussion is directly and richly portrayed—the Meno,
for example, or the Phaedo or the Philebus—we might think we should imagine ourselves
as one of the characters in question: should sympathize with their positions, should take

13 For example, at Euthyphro 9d; Protagoras 331c–d; Republic 346a. See M. M. McCabe, Plato and
His Predecessors: The Dramatisation of Reason [Predecessors] (Cambridge, 2000), ch. 2.
14 Nehamas, Art.
15 The debate on the date of composition of the dialogues and the development of Plato’s thought
has been much at issue in the past 50 years or so; see, for example, G. E. L. Owen, “The Place of the
Timaeus in Plato’s Dialogues,” Logic, Science and Dialectic (London, 1986). Vlastos, Socrates, had the
effect of focusing recent attention on the question of Plato’s attitude to the “historical” Socrates. On
some of this see, for example, R. Kraut, “Introduction to the Study of Plato,” in R. Kraut, ed., The
Cambridge Companion to Plato (Cambridge, 1992), 1–50; C. H. Kahn, Plato and the Socratic Dialogue:
The Philosophical Use of a Literary Form [Dialogue] (Cambridge, 1996); J. M. Cooper, Plato: Complete
Works [Plato] (Indianapolis, 1997), xii–xviii; J. Annas and C. Rowe, eds., New Perspectives on Plato,
Modern and Ancient (Cambridge, Mass., 2002); and the sane observations of D. N. Sedley, The
Midwife of Platonism: Text and Subtext in Plato’s Theaetetus (Oxford, 2004).
16 His austerity is anticipated, perhaps, in the digression of the Theaetetus 172–78: he is a
philosopher.
98   Mary Margaret Mccabe

up the point of view they espouse. So we might share Cebes’s worries about whether the
soul is immortal, while sympathizing with his earnest desire that it be so, indeed (Phaedo
88a ff.). The arguments that follow, then, would engage with that position and show us
where it should be modified and resolved. Or we might imagine ourselves as Meno, or
Laches—and feel for them as their less than rigorous collections of beliefs are subjected
to Socratic argument, feel with them the sense of puzzlement (of aporia), of frustration
and irritation, or just sheer embarrassment (Meno 80a ff.; Laches 194a ff; and compare
Protarchus’s more robust response, Philebus 20).17 We may feel some sympathy, even,
for rebarbative characters such as Protagoras or Critias (Protagoras 333–34; Charmides 166).
And we can empathize with the characters thus just because they are portrayed in vivid
ways, such that we can clothe ourselves, as we read, with their character and attitudes.
But we then find the attitudes and views that we thus adopt subjected to dialectical scrutiny,
and this serves a direct philosophical purpose. So—we could say—the point of having
these characters represented to us is that the representation is somehow transparent,18
available for us to identify with the characters on the dialogue’s stage and to suffer their
philosophical fate.
Well, who has not sympathized with the squirming embarrassment sometimes felt by
Socrates’ interlocutors (e.g., Hippocrates, Protagoras 312a; Nicias, Laches 200)? Who has
not felt Callicles’s irritation (Gorgias 489 ff.) or Thrasymachus’s annoyance (Republic
336–37, 343) that Socrates is merely manipulating the arguments against them, missing
the heart of the problem of justice for the individual? Our emotional sympathy is engaged,
indeed, by the brilliance of Plato’s writing: as Socrates supposes to occur when we hear
recitations of Homer:

When even the best of us hear Homer or some other tragedian imitating one of the
heroes sorrowing and making a long lamenting speech or singing and beating his
breast, you know that we enjoy it, give ourselves up to following it, sympathize with
the hero, take his sufferings seriously, and praise as a good poet the one who affects
us most in this way (Republic 605c–d).19

And then the dialogue’s effect might be somehow therapeutic:20 the examination of
these views we have adopted is beneficial to us (at least it purges us of error). By the end
of the dialogue, we may be brought to identify, instead, with Socrates, peculiar figure
though he may be. So in our imaginative engagement with the dialogue, our views are
transformed. Is that what we should say about these rich dialogues? And if it is, what
should we say about those dialogues that seem obstinately poor?

17 The same seems to happen to Socrates himself: for example, Protagoras 339, Euthydemus 293.
18 A. Nehamas, “Plato and the Mass Media,” The Virtues of Authenticity [Authenticity] (Princeton,
N.J., 2000); Halliwell, Mimesis, 91.
19 Translations throughout are from Cooper, Plato.
20 D. N. Sedley on the Anonymous Commentator on the Theaetetus in “The Theaetetus: Three
Interpretations” [“Interpretations”], in C. Gill and M. M. McCabe, Form and Argument in late Plato
(Oxford, 1996).
 Plato’s Ways Of Writing   99

5. The Trouble with Imitation

Imaginative identification may prove problematic, however, even where it might seem
plausible (identifying with the Eleatic Stranger, or with Young Socrates of the Statesman
is a harder task). In the Republic, Socrates is disapproving of dramatic performance
(605e ff.), since it takes over our emotions under reason’s inadequate guard.21 Socrates
denies that this kind of sympathy is intellectually healthy—especially sympathy with those
who are not exemplary characters (as many of Socrates’ interlocutors are not)—for in
imitating poor or uncontrolled characters, we shall ourselves practice their inadequacy,
and so much the worse for us.
But even the imitation of good characters will be problematic as a means of learning
wisdom. For wisdom—or so, at least, Socrates seems to say on several occasions—is not
something that is transmitted by our passive absorption of what we learn (for example,
Protagoras 313–34; Symposium, 175; compare Euthydemus 285, Republic 345b). Instead,
the search for wisdom is a hard road, whose traveling we cannot delegate to anyone else
(see Apology 23a–b; Euthydemus 281). So neither imaginative identification with Socrates’
interlocutors nor even with Socrates himself seems to be the right way of going about
philosophical inquiry. If we read, in the dialogues, representations of philosophy being
done, that tells us nothing about how we should do philosophy with the dialogues.
How far, then, does either the criticism of poetry and drama that Socrates puts forward
in the Republic, or the repudiation of a passive model of learning, target Plato’s own
writing, too?

6. Doctrines and Impasses

To answer this question, we may ask another. In the recent explosion of discussions
about Plato’s literary skills, the question “Who speaks for Plato?”22 has been posed,
asking what we may suppose to be the views of Plato, transmitted through his represen-
tations of others in dialogue. Is the point that we should end up switching our allegiance
away from the interlocutor toward either Socrates or the Eleatic Stranger? Are we to
think, then, that Socrates (or the Eleatic Stranger, or Timaeus, or the Athenian Stranger),
by occupying the protagonist’s role, is the mouthpiece of the author?23 What Socrates

21 On this complex issue, see, notably, M. F. Burnyeat, Culture and Society in Plato’s Republic:
The Tanner Lectures on Human Values (Cambridge, Mass., 1999); Nehamas, Authenticity;
Halliwell, Mimesis.
22 G. A. Press, ed., Who Speaks for Plato? Studies in Platonic Anonymity (Lanham, M.D., 2000),
and D. Clay, Platonic Questions: Dialogues with the Silent Philosopher [Questions] (University Park,
Penn., 2002).
23 Of course, we should not say so for a drama—is King Lear Shakespeare?
100   Mary Margaret Mccabe

says, on this account, would be what Plato himself believes.24 And in that case, the
importance of everything else in the dialogue may diminish, turning, at worst, into the
“merely” literary curlicues to make the doctrines put forward by the master palatable.25
But this explanation cannot be universal. After all, in many dialogues Socrates ends
up without a doctrine—or, at least, without a doctrine that has withstood serious critical
attack26—but only a state of impasse, aporia. In some dialogues, further, it is not clear
just what has happened to some thesis employed along the way (a good example is the
theory of perception from Theaetetus 15627). And in some dialogues, it appears that
progress genuinely is made, but it is circumscribed by caution (for example, the account
of false statement in Sophist, see 261; the account of the mixed life in the Philebus, see
67b; the “likely story” of the Timaeus).
Instead, the dialogue form may be meant to offer us something that is not Platonic
doctrine. Some have seen there a thoroughgoing, nondogmatist, skepticism: any view,
even the view of a Socrates, is liable to be overturned, and so the best thing to do is to
suspend judgment.28 Some have found the dialogues to be in some other, vaguer way,
“open-ended”—noncommittal, indeterminate about what we should say about these big
questions: committed, merely, to the project of going on looking.29 But neither of these
broad views takes seriously enough the determinate differences that do occur within
the arguments: between a view that is denied—for example, that justice is the interest of
the stronger; Republic 343a—and one that is not—for example, that justice is where each
does his own; Republic 433.
Perhaps, instead, the different attitudes to different arguments imbue the positive
conclusions with a kind of (variable) provisional status, so the frequently puzzling
nature of the endings, the tentative way in which Socrates declares his commitments en
route would be meant to disavow authority for them.30 “This,” the protagonist may
sometimes be understood to be saying, “is a likely story; but still open to critical attack.”
Socrates’ arguments (and even the procedures of the Eleatic Stranger) are presented
to make clear this disavowal of authority, to advance the conclusions as tentative, but

24 But see, e.g., J. Beversluis, Cross-Examining Socrates: A Defense of the Interlocutors in Plato’s
Early Dialogues (Cambridge, 2000), and C. Gill, “Speaking up for Plato’s Interlocutors,” Oxford Studies
in Ancient Philosophy 20 (2001): 297–321.
25 Some ready-made contrast between the “literary” and the “philosophical” begs the very question
I am asking.
26 This happens not only in the dialogues conventionally labeled “Socratic” (such as Laches,
Charmides, Lysis, and Protagoras) but also in Theaetetus and Parmenides.
27 See M. F. Burnyeat, “Plato on the Grammar of Perceiving,” Classical Quarterly 26 (1976): 29–51;
Burnyeat’s magisterial The Theaetetus of Plato, trans. M. J. Levett (Indianapolis, 1990); and J. M. Cooper,
“Plato on Sense-Perception and Knowledge (Theaetetus 184–6),” Phronesis 15 (1970): 123–65.
28 See Sedley on the Theaetetus as interpreted by Academic skepticism, in “Interpretations.”
29 See, e.g., Roochnik, Tragedy, and Clay, Questions,; compare H. Gadamer, Dialogue and Dialectic:
Eight Hermeneutical Studies of Plato, trans. P. C. Smith (New Haven, Conn., 1980), with the comments
by C. Gill in. “Critical Response to the Hermeneutical Approach from an Analytic Perspective” in
G. Reale and S. Scolnicov, eds., New Images of Plato: The Idea of the Good [Good] (Sankt Augustin,
Germany, 2002).
30 Stokes, Conversations; M. Frede “The Literary Form of the Sophist,” in C. Gill and
M. M. McCabe, eds., Form and Argument in Late Plato (Oxford, 1996).
 Plato’s Ways Of Writing   101

as conclusions nonetheless.31 To identify, by the dialogue’s end, with Socrates is to

agree with him, with reservation, and to take this view of the nature of philosophical
progress seriously.
This would not make the dialogues weapons of skepticism or leave them “open-ended.”
It explains the differences between what the characters say in different dialogues,
while at the same time allowing the reader an active role in the process of reading: the
reservation. So it does, I suggest, tell some of the story. It grants to Plato genuine
commitment to—and responsibility for32—the views about, for example, politics and
psychology presented by Socrates in the Republic, while allowing them to be subject to
revision. And it tells us something about the appropriate general response to the explicit
arguments when we read. But it still does not account for the multifarious forms of the
dialogues, nor for how we should respond to that multifariousness when we read.

7. Philosophical Fiction

It would be an obvious mistake to take the Symposium, or the Gorgias, or even the
Parmenides, to describe some historical event, just as it happened. For each dialogue is a
work of fiction, somehow artfully composed in such a way that we notice the artistry
itself.33 Consider, for example, the pastoral tone of the Phaedrus—a work set ostenta-
tiously outside Athens (whither Socrates allegedly only ventured twice34); or the high
tragedy of the Phaedo (apparently recorded by Plato, even though he was away sick (59b);
or the logically low comedy of the Euthydemus (Socrates’ encounter with a pair of
sophists with poor historical credentials35). Even those works that make vigorous claims
to historicity at the same time bear the marks of fiction. The Apology claims to be
Socrates’ speech in his own defense, but it includes an improbable philosophical discussion
with his accusers (24d–28a).36 The Parmenides describes a visit to Athens paid by the
two great Eleatics, Parmenides and Zeno, and their discussion with the young Socrates:
bien trouvé, but are we meant to think it actually happened?37 The dramatic dates and
places of the dialogues, therefore, may be trickier than they at first appear.38 And what of

31 The change of protagonist might be designed to produce this tentativeness in itself; that would
account, perhaps, for the apparently dogmatic conclusions of the Timaeus.
32 See M. F. Burnyeat, “Sphinx without a Secret?” [“Sphinx”], New York Review of Books (May 30, 1985).
33 See, e.g., A. Nightingale, Genres in Dialogue: Plato and the Construct of Philosophy (Cambridge,
1995); R. Rutherford, The Art of Plato (London, 1995); and Blondell, Play.
34 On campaign at Potidaea and Delium, see Symposium 219e ff and Crito 52b. We should be
attentive, therefore, to how the Phaedrus makes play with his being out of place, 230d; see G. R. F
Ferrari’s classic Listening to the Cicadas: A Study of Plato’s Phaedrus [Cicadas] (Cambridge, 1987).
35 See D. Nails, The People of Plato (Indianapolis, Ind., 2002), 152. At Euthydemus 271–72, it is
suggested that the brothers are “new-fangled.”
36 On the genre of Socratic logoi, see Kahn, Dialogue.
37 The dialogue takes care with the relative ages of the interlocutors (127a–b), echoing the elaborate
timing of the transmission of the story itself, but there is no independent evidence that the meeting
actually took place.
38 R. Hunter, Plato: Symposium (Oxford, 2004), 3 n.1.
102   Mary Margaret Mccabe

those dialogues that somehow lack location in particular time and place? For example,
the Sophist and the Statesman—led by the anonymous Eleatic Stranger39—appear to be
dramatically tied to the Theaetetus; but they are oddly bereft of particular detail. I sug-
gest that there are explanations—both general and particular—for a dialogue’s elaborate
fictionality and that those explanations are philosophical ones.40
So the dialogues are not merely verbatim reports of actual events, but artful presenta-
tions of conversation. We may distinguish, therefore, between:

• The philosophy that is represented in the fictional encounters of the dialogues:

Do the discussions between Socrates (often) and various interlocutors constitute a
dialogue’s philosophical content?
• The representing of those encounters: Or is Plato doing more—from a philosophical
point of view—than merely recording these discussions when he writes?

If there is some kind of contrast thus to be drawn between what is represented and its
representation, what happens when we read that representation? Is it that we take it in,
envisage it, so that the representation is somehow directly transmitted to the reader? Or
is there a more complex relation between what Plato writes and what happens when we
read? If the dialogues are fiction, and we are brought to notice it, then the business of
reading may be a more active process than merely absorbing whatever it is that the rep-
resentation is meant to “say.” So there is a third point of contrast:

• Reading these representations: When we read, are we passive to what is represented

there, or actively and critically engaged?

That Plato saw these three things—the represented, the representation, and the business
of reading—to be distinct may be attested by the Phaedrus’ story, in which Thamus, king
of Egypt, castigates Theuth for rejoicing at his discovery of writing. For:

Writing shares a strange feature with painting. The offsprings of painting stand there
as if they are alive, but if anyone asks them anything they remain most solemnly
silent. The same is true of written words. You’d think they were speaking as if they
had some understanding, but if you question anything that has just been said,
because you want to learn more, it continues to signify just that very same thing
forever. When it has once been written down, every discourse rolls about everywhere,
reaching indiscriminately those with understanding no less than those who have no
business with it and it doesn’t know to whom it should speak and to whom it should
not. And when it is faulted and attacked unfairly, it always needs its father’s support;
alone, it can neither defend itself nor come to its own support. (275d–e)

39 Xenos may mean “stranger” or “visitor”; in antiquity, visitors were often strangers.
40 It is impossible, of course, to demonstrate this for every word in every dialogue, and it is a fortiori
impossible to do so in a short essay such as this. My hope, merely, is to shift the burden of proof onto
those who suppose that some bits of the Platonic dialogues are easily dispensable by the philosophical
reader. I discuss various strategies in M.M.McCabe, Platonic Conversations [Conversations] (Oxford 2015).
 Plato’s Ways Of Writing   103

The story is told by Socrates, who wrote no philosophy (this is the direct representation).
His telling of the story, however, is represented in such a way that we notice as we read
that the story somehow undermines its own mode of representation (it is told in writing).
This is not, then, mere reportage. Instead, the writing of a tale told against writing shat-
ters the confidence of his readers that what they read here can be taken on trust.41 But
now this gets worrying. If writing is somehow unreliable, why—if we are searching for
wisdom—should we read the dialogues?
Some would say that the dialogues are somehow inherently contradictory, testimony
to the deep down slipperiness of the way we write—or even talk.42 Others would say that
they are somehow second best to the oral tradition of philosophy within the Academy.43
Maybe their surface meaning is even the disguise for a coded message underneath.44 Each
of these suggestions, however, may underplay the overt self-consciousness of the paradox
about writing. For in challenging its own mode of presentation, it asks how the search
for wisdom should proceed, and this question, itself a philosophical one, is provoked by,
and so reflective on, the written dialogue itself. In what follows, I suggest that the dia-
logues have this philosophical quality through and through—not only the represented
dialogues, but also Plato’s representing of them—and that this should determine our
reading of them. Even the most unlikely aspects of their composition45 may be best
understood from the default position that Plato writes nothing in vain.

8. Question and Answer

In almost all the dialogues, people talk about big abstract questions: the nature of reality,
truth, virtue, knowledge. These conversations come in all sorts of shapes and sizes,
but they are often conducted by question and answer, by one person asking questions
of another.46
Why does Plato use conversation like this? Is it the compulsion of culture, the regular
practice of classical antiquity for theoretical discussion? Compare, for example, the
debates in Thucydides,47 or the formalized conversations of drama,48 or the dissoi logoi of
the sophistic tradition.49 Or is it—more strongly—a philosophical claim: that theoretical

41 Compare Protagoras 329a–b, and Epistulae VII 341 ff.

42 J. Derrida, “Plato’s Pharmacy,” in Dissemination, trans. Barbara Johnson (Chicago, 1981), and
Ferrari, Cicadas.
43 For example, H. J. Krämer, Plato and the Foundations of Metaphysics, trans. J. R. Caton (Albany,
N.Y., 1990), and T. A. Szlezak, Reading Plato (London, 1999); see also Reale and Scolnicov, Good.
44 For example, L. Strauss, The City and the Man (Chicago, 1964), but see Burnyeat, “Sphinx.”
45 Aristotle’s answers in the Parmenides are still a challenge, though . . .
46 The three-sided conversation of Laches 194 ff. is an exception.
47 For example, the Mytilenean debate, 3.37 ff., and the Melian debate, 5.87 ff.
48 For example, the debate between the right and the wrong arguments at Aristophanes’ Clouds, 900
ff., or between Antigone and Creon at Sophocles, Antigone 435 ff.
49 Diels/Kranz 90.
104   Mary Margaret Mccabe

discussion can only be carried out within a particular culture?50 (Tough luck for us,
reading Plato from a distance.) This might condemn in advance any attempt by us to
abstract arguments from the dialogues or to find some kind of disengaged philosophical
viewpoint therein. Can there be a view from outside the culture upon what happens within,
whether an Archimedean point from which to make judgment about the arguments
represented there, or a way for us here and now to read the dialogues?51
Well, it seems that the business of question and answer is an explicit method of pro-
ceeding in philosophical inquiry. Socrates sometimes suggests that question and answer
is either the right way to proceed or the only way he is able to proceed:52 consider the
extraordinary moment in the Protagoras (334c–d) where he complains that he cannot
remember long speeches—even though he is the narrator of the whole dialogue. Other
protagonists have the same commitment to question and answer: for example, the Eleatic
Stranger at Statesman 285 and Parmenides at Parmenides 137. And many interlocutors
explicitly accept this way of going about the business in hand (for example, Gorgias
at Gorgias 457 ff.; Protarchus at Philebus 19; and Euthydemus and Dionysodorus at
Euthydemus 275–7653), even if others refuse, either in practice, or directly (for example,
Protagoras, at Protagoras 335c, 348b–c; compare Hippias Major 291a). As a consequence
of this self-consciousness, proceeding by question and answer is itself subject to scrutiny:
What recommends it?
First, the process is sequential: one answer provokes another question, and so on
(compare Symposium 204d). So the salience of answer to question, and of question to the
preceding answer, brings order to the discussion.54 This shows up in the limiting cases,
where what the interlocutor says precludes his answering any further question: for
example, the monists of Sophist 244c are unable to sustain a conversation at all, and this
is taken to be so successful a rebuttal that it amounts to murder.
Second, this order is connected to a constraint on both answer and question: the views
put forward by any discussant should be internally consistent (Charmides 164c–d; Gorgias
491b–c; Euthydemus 287a–b; Euthyphro 15c). This seems, indeed, to be a constraint on
what is said or believed that is somehow fundamental to the interlocutors themselves
(Gorgias 482b–c). Inconsistency, that is to say, is somehow a fault or a danger, so much
so that its discovery provokes all sorts of emotional anguish, evidenced by physiological
effects: blushing (Republic 350d), gaping (Charmides 169c); and psychological disturbance,
including irritation (Gorgias 489) and confusion (Meno 80).

50 See R. Rorty, Philosophy and the Mirror of Nature (Princeton, N.J., 1979).
51 Compare B. A. O. Williams, Ethics and the Limits of Philosophy (London, 1985), with Williams,
Shame and Necessity (Berkeley, Calif., 1993).
52 On the “Socratic method,” see, among many others, Vlastos, Socrates; T. Brickhouse and
N. Smith, Socrates on Trial (Princeton, N.J., 1989); Brickhouse and Smith, The Philosophy of Socrates
(Boulder, Colo., 2000); H. H. Benson, Essays on the Philosophy of Socrates (New York, 1992); and
Benson, Socratic Wisdom: The Model of Knowledge in Plato’s Early Dialogues (New York, 2000).
53 Here the process is subverted and made even more self-conscious: the brothers claim, 275e, that
whatever their interlocutor says he will be refuted.
54 For example, the connectives at Charmides 159c–160b, and the explicit conclusions at 160b3 and
7; or the way in which the method of collection and division is allied to questioning to produce order at
Statesman 281.
 Plato’s Ways Of Writing   105

Third, as a consequence both of the ordered nature of question and answer and of its
demand for consistency, the process is also reflective. If a later answer is inconsistent
with an earlier one, or out of order, this causes trouble, which is then explicitly discussed
(for example, Phaedo 92c; Charmides 164d; Laches 193e). Contrariwise, some sophists
notoriously reject this condition on conversation (Euthydemus and Dionysodorus at
Euthydemus 283–88, Protagoras at Theaetetus 151–71), but Socrates supposes that rejection
to be self-defeating (Euthydemus 288a; Theaetetus 178–79). As these conversations are
presented, then, they are presented under reflective scrutiny, not only for the content of
what is said but also for its integrity and good order.
But, fourth, there is not merely a demand for consistency at all costs—mere consistency
might just give us a coherent collection of someone’s beliefs, with no claims on the truth,
or it could be secured at the price of refusing to engage in philosophical conversation at
all. On the contrary, the burden of the questioning is repeatedly to elicit the reasons
that some earlier claim might be true: the dominant question is a demand for some kind
of explanation (for example, Euthyphro 10–11; compare Phaedo 96 ff.). Explanation is
fundamental, indeed, to the structure of question and answer, or at least as it is practiced
in these Platonic conversations.

9. Conversation and Dialectic

Conversations, then, could be analyzed formally in terms of sequence, order, ­consistency,

and explanatory structure. Just as the discussions between Parmenides and Aristotle
could be restated as linear arguments, the same might be done for more complex passages:
for example, the last argument for the immortality of the soul at Phaedo 102–755 or the
argument to deny the possibility of weakness of will at Protagoras 353–58.56 In that case,
could we edit out a great deal of the individual dialogues, forsaking the conversational
details altogether? This would be a mistake. For it misses, not only the variety of the
dialogue form but also the way in which the argumentative material extends—I say
indefinitely—beyond the passages that seem to be the best candidates for formalization.
And it misses the way in which this breadth brings the conditions for argument them-
selves into philosophical view.
To repeat the point with which I began, these dialogues are not uniform. And they
are not committed to conversations of a particular sort, or even to conversation at all.
Conversation predominates when Socrates is engaged on the inquisition of some
unfortunate acquaintance—encountered going about his business, and falling into the
trap of explaining what he is doing to the indefatigable Socrates (Euthyphro or Ion,

55 This argument is demarcated by the interruption of the frame at 102a–b and terminated with
Simmias’s qualified agreement at 107a. Compare the formal analysis of Gallop, Phaedo.
56 This argument is demarcated by the imagined participation of “the many,” expressly begun at 353a
and terminated at 358a. Compare the formal analysis of G. Santas, Socrates: Philosophy in Plato’s Early
Dialogues (London, 1979), ch. 7.
106   Mary Margaret Mccabe

for examples). In such cases, one party seems to have—to begin with, at least—some
definitive and determinate point of view, which comes under scrutiny in the questions and
answers that follow.57 Here, then, the person who answers is being examined. Sometimes,
both interlocutors seem to have a view: for example, in the Gorgias, Callicles maintains that
natural justice is the exercise of power without the control of reason, while Socrates
denies it (482 ff.); in the Philebus, Socrates starts out as the advocate of the life of pure
reason, Protarchus the advocate of some kind of hedonism (12–22). On other occasions,
it is the questioner who seems to occupy the philosophical position, but still his inter-
locutors perform a pivotal role. For example, the two objections made by Simmias and
Cebes at Phaedo 84d ff. dictate the course of the rest of the dialogue; Glaucon’s challenge
at Republic 357a ff. informs the nine books that follow. In other dialogues, where the role
of the answerer seems exiguous (e.g., Sophist or Statesman),58 the significance of question
and answer is brought into sharp relief at the moments when the interlocutor does actually
say something that matters (for example, Statesman 299e, Euthydemus 290).59
Sometimes, instead, there is a striking contrast between lengthy exegesis and conver-
sation. The most notable example may be the paratactic structure of the Symposium, which
presents a series of separate speeches about love60 but whose final speech contains a
dialogue between Socrates and Diotima, the role of protagonist sharply reversed.61 On
other occasions, one question and answer session is oddly embedded in another; this
occurs when the interlocutors imagine some third party responding to questions (e.g.,
Apology 26b ff; Republic 476–80; Protagoras 351 ff.; Theaetetus 170). Differently again, in
the Timaeus, the long speech of Timaeus is preceded by a prologue between Socrates
and the others present.62 Yet again, some dialogues make play with other forms of
discourse: the Phaedrus offers the reading of some speeches—of which at least one is
ostentatiously spurious; the Gorgias, Phaedo, and Republic close with elaborate myths.63
These differences—both within and between dialogues—matter. For the display of
different modes of argument goes along with a discussion of how arguments should
work, and this discussion is provoked even by those cases where conversation either
ostentatiously fails or just runs out (e.g., Euthydemus 303; Protagoras 334–35; Gorgias
523a; Phaedo 90–91). In these cases, there is a critical relation between what is said (about
the serious significance of conversation for philosophical discussion) and how it is said

57 Sometimes, as in the cases of Euthyphro or Ion, this is a claim to knowledge—the same goes for
the sophists Socrates talks to: Gorgias, Protagoras, Hippias, Euthydemus, and Dionysodorus.
Sometimes the view to be examined has to be teased out of him: as in Charmides, at Charmides 158, or
Lysis and Menexenus at Lysis 207.
58 Aristotle is introduced as someone “who will give the least trouble” (Parmenides 137b); likewise,
Theaetetus at Sophist 217c–d.
59 At Sophist 252d, Theaetetus is positively prolix; this has the effect of emphasizing his claim there
about contradiction; and compare 256b (although the text there is suspect).
60 The arrangement is emphasized by Aristophanes’s famous attack of the hiccups that displaces the
planned order of the speeches.
61 The Euthydemus has a similar paratactic construction.
62 See M. F. Burnyeat, “First Words” [“Words”], Proceedings of the Cambridge Philological Society
43 (1998): 1–20. The Menexenus has the same form.
63 On Platonic myths, see, e.g., J. Annas, “Plato’s Myths of Judgement,” Phronesis 27 (1982): 119–43,
and K. Morgan, Myth and Philosophy: From the Presocratics to Plato (Cambridge, 2000).
 Plato’s Ways Of Writing   107

(in conversation, or failing conversation), and this relation is itself the focus of philo-
sophical attention. As Socrates suggests at Republic 534d8–535a1, conversation is the way
that philosophy should proceed:

‘‘So you would legislate, would you, that they should most of all receive that educa-
tion through which they would be able to ask and answer questions in the most
knowledgeable way?’’
‘‘Yes, I would so legislate—and you with me, too.’’
‘‘So do you suppose,’’ I said, ‘‘that dialectic lies at the top for us, like a coping-stone
on our studies, and that there is no other subject that should rightly be put higher
than it, but that it provides now the end to our inquiries into education?’’

It should proceed that way, it seems, both by dialogue between persons and by dialogue
within a person—for that is what thinking is: a “silent dialogue within the soul”:

Socrates: A talk which the soul has with itself about the objects under its considera-
tion. Of course, I am only telling you my idea in all ignorance; but this is the kind of
picture I have of it. It seems to me that the soul when it thinks is simply carrying on
a discussion in which it asks itself questions and answers them itself, affirms and
denies. And when it arrives at something definite, either by a gradual process or a
sudden leap, when it affirms one thing consistently and without divided counsel, we
call this its judgment. (Theaetetus 189e–190a)64

The paradigm of philosophical activity, then, is dialectical discussion;65 this is portrayed

and reflected on by the dialogues. Philosophical discussion thus takes place at both levels
of discourse within the dialogues—it takes place, therefore, across frame and picture, so
it involves both what is represented and its representation. What is more, the contrast
between different modes of discourse becomes evident when we read, just because the
dialogues are composed in ways that make us notice their composition—such as, for
example, their fictionality. The conditions for philosophical discourse, as a consequence,
are themselves subject to the reader’s active and reflective scrutiny.

10. The Psychology of Dialectic

Should we still say that this allows us to select those parts of the dialogues that constitute
argument, of the general, abstract sort, and to dispense with the rest? Plato’s conversa-
tions, in their variety, cannot be reduced to an impersonal form, however. For built in to
these models of conversation are (at least) four psychological conditions.

64 Theaetetus 189–90; Philebus 38c ff.; and Sophist 263e. On Republic 523–25, see M. M. McCabe
“Dialectic Is as Dialectic Does,” in The Virtuous Life in Greek Ethics, ed. B. Reis (Cambridge, 2006).
65 The technical expression dialektikê is derived from dialegesthai, “to converse”; see especially
Republic 454a; 511b–c; 532–33. Aristotle follows in the same tradition: Metaphysics B1.
108   Mary Margaret Mccabe

The first is the way in which views are held by the interlocutors. Sometimes someone
offers a thesis that they hold dear (for example, Critias at Charmides 164d). This leads the
characters to difficulties in the sequel: when their claim is attacked, they themselves
mind about it.66 At other times, responsibility for the views in question is delegated to
someone not present (for example, at Republic 476e ff.), but these absent characters
are themselves characterized as committed to the view (Republic 476d). How far is this
feature of the dialogues ineluctably ad hominem? How far does it fail a demand that
philosophical discussion should be somehow general? And what does it have to do with
us at a distance, as we read?
The second psychological condition invites the same objection. Sometimes an inter-
locutor is reduced by Socrates to puzzlement, to aporia. Meno, for example, complains
that Socrates has reduced him to numbness by showing that the views he thought he
held consistently cannot be sustained together (Meno 80a ff.). This puzzlement, Socrates
protests, has its own dynamic: for puzzlement is part of what compels one to continue to
search for the answers to the questions with which one began. It seems, then, a part of the
psychology of the drama we see: a feature attributed to one (or both) of the interlocutors
and particular to them. Formal arguments don’t feel puzzled.
Third, more happens in the dialogues than simply the reduction of someone to aporia.
For the interlocutor is sometimes able to rethink his position, to alter his allegiance, to
change his mind. This reflects a feature of some conversations: that they represent two
different points of view, in tension with each other, and that it is possible, engaged in
such a conversation, to take up different positions, to shift perspective (e.g., Protarchus
at Philebus 23a). This, indeed, seems to be what the “silent dialogue” claims: that we are
able, when we think, to occupy different viewpoints (or to have different parts of our
souls do so), to conduct an internal conversation in which one side asserts and the other
denies the same thing—without risking contradiction or psychological damage. This
suggests, further, that we are able to entertain a view, instead of committing ourselves to
it.67 Likewise, engaged in a conversation, we may find ourselves stepping outside the
position we originally occupied and understanding a different point of view.
Fourth, this “stepping outside” may occur at a remove. For not only do the characters
of a dialogue engage in a conversation and reflect on their own and their interlocutor’s
positions from outside. They are also able to notice, inspect, think about the whole
­process of a conversation. Consider occasions where the interlocutors imagine a conver-
sation with someone else—imaginary or otherwise absent: for example, the conversation
with the “other Socrates” at Hippias Major 286c ff.68 Here the frame interlocutors imagine
the two different points of view and their resolution, from the outside. Their position

66 Of course, since these are fictional characters, the sense of “themselves” minding about it is odd.
67 Compare the occupation of a position for the sake of argument, as by Glaucon at Republic 358,
and by Simmias and Cebes at Phaedo 77e.
68 Compare Republic 476–80, or the framing of the Parmenides in an outer narrative.
 Plato’s Ways Of Writing   109

need not be committed to either point of view; instead, it is broadly reflective on both.
This is a psychological phenomenon, not merely a logical one: call it detachment.69

11. Detachment

Detachment, this reflection at a remove, is neither limited to the characters within the
dialogues nor just exemplified therein. For it is also something we do when we read: we
look at the conversations represented in the dialogue, and at the reflections on them
provided by the dialogue, from outside. This, I suggest, is the role of the elaborate play
between history and fiction in which some dialogues engage: in puzzling about whether
this or that story is true, we bring into focus the relation between ourselves (reading here
and now) and whatever did or did not happen then. The dialogues’ tension between his-
toricity and fiction thus renders our role as readers self-conscious. Then we are not pas-
sive spectators, not mere recipients of some knowledge conveyed by one character or
another. On the contrary, we are actively engaged in reflection, both on the represented
dialogue and on the way the dialogue is written for us to read.
Consider some generic differences in the way conversation is presented from dialogue
to dialogue. Some are directly presented: the Laches, for example, offers us a play-like
conversation among Lysimachus, Laches, Nicias, and Socrates (compare, for exam-
ple, Meno, Euthyphro, Cratylus, and Philebus). Indeed, this formula may encourage us to
identify with one or another character, as well as to change our identification to Socrates,
as the dialogue proceeds. Three dialogues reinforce this thought, for they are narrated
by Socrates, told from his perspective: Charmides, Lysis, and Republic. Here we might
imagine that Socrates is speaking to an imaginary audience—or even that he speaks
directly to us as we read. Should we think, then, that his privileged voice is the only one
we should hear or obey, or should we notice the contrivance itself?
The Sophist and the Statesman, too, proffer the philosophical discussion directly in
dialogue, but now the privileged voice is that of the Eleatic Stranger. These two dialogues
are explicitly tied to the Theaetetus, which has an elaborate opening of a quite different
sort. For the story is told—years later—within a direct dialogue between Euclides
and Terpsion. Euclides has a written record (with good credentials: it was checked by
Socrates, 143a) of the encounter among Socrates, Theodorus, and Theaetetus, and he
proposes to read it out, but as if in direct speech, missing all the “and he said . . . ” bits.
This frame dialogue should frame not only the discussion that follows but also the two
sequels, even though Euclides makes no promise to tell us about the meeting with the
Eleatic Stranger. This could be mere carelessness, of course: the connection among
the three dialogues is trivial, a surface matter, and of no significance to how we understand
the dialogue. But the opening of the Theaetetus calls our attention to how the dialogue is

69 On detachment, in the context of two independent discussions, see McCabe, Predecessors, e.g. 125;
Halliwell, Mimesis, 79 ff.
110   Mary Margaret Mccabe

presented, especially to the difference between direct and indirect reportage. At the very
least, this makes us notice the contrivances of the fiction, as well as the differences of
style, and reflect on our distance from the story we are told.
In the Phaedo, too, the story is told by two people (Echecrates and Phaedo) and
provoked by a death and its circumstances. But later in the dialogue the frame interrupts
the narration, twice. First, at 88c, the two objections of Simmias and Cebes against the
arguments for the immortality of the soul, have been voiced, and Phaedo and Echecrates
discuss the devastating effect of those objections, both on Socrates’ friends as they sat in
prison and on Echecrates and Phaedo as they consider it from their later viewpoint.
Second, at 102a, when Socrates has just pronounced some of the most vexed and impen-
etrable arguments of the dialogue, Echecrates interrupts again, saying that “he made these
things wonderfully clear to anyone of even small intelligence.”70 Both interruptions
reflect on the state of mind of the people hearing the arguments: both the audience
within the narrated story and those who hear it later. But they also serve to remind the
reader of her own position. For while the first interruption may underline our sympathy
for the characters of the dialogue, the second emphasizes our alienation from them. In so
doing, the pair of interruptions provokes us to think more directly about our own
perspective on what we are being told.
The same device of the interruption of the frame dialogue is found at the central point
of the Euthydemus. Socrates, narrating the previous day’s events to Crito, says that
Cleinias has made a certain objection to Socrates’ argument and (disingenuously,
perhaps) that the objection is a good one. Crito interrupts with a vigorous disagreement
(Euthydemus 290), incredulous that it could have been Cleinias who made the objection.
As a consequence, the framed argument continues in the frame, between Socrates and
Crito, and discussing not only who could have said what but also what constraints on
argument and agreement there should be (291b, 292e). The shift to the frame both adds
this methodological reflection and brings to our attention the framing procedure itself.
We are shaken from merely identifying with the characters (since we are no longer sure
who said what) into something far more detached and reflective—and the reflection is
focused not only on the original subject but also on the nature of argument itself.71
In cases such as these, the reader may have two different views of what she reads. The
rich portrayal of character and situation—Socrates, his friends, his opponents, at dinner,
in prison, at the gymnasium—encourages her to identify with one or another of the
characters (sometimes alternating, as they speak) and to feel their theoretical commit-
ments, either as if they were her own or at least with sympathetic understanding. In this
way, we come to understand different arguments and philosophical positions, to con-
sider, at firsthand, the arguments put forward in the dialogue. By contrast, when we are
shocked, by devices in the frame, into looking at what is going on, we do so from this

70 The existence of a vast literature on the interpretation of the method of hypothesis bears witness
to the impenetrability of Socrates’ words: this interruption, then, should surprise us.
71 Compare the elaborate frame of the Parmenides, which vanishes by the end of the dialogue and
leaves us with an extreme, and unresolved, contradiction. See McCabe, Conversations, chs. 1, 6, 8.
 Plato’s Ways Of Writing   111

position of detachment, contemplating both points of view from the outside. We both
consider the sorts of conditions that will come to bear on either point of view, including
the conditions for argument as such, and think about the subject in question from a
position of detached reflection. So, I suggest, the active role of the reader is crucial to
understanding why the dialogue is multiform.

12. Ethical Rationalism

Still, does this reflectiveness account for the dialogue form in its particularity, or of the
local purposes of Plato’s varied use of dialogue? Suppose that there are some central
philosophical issues addressed in each dialogue—the nature of virtue, the possibility of
knowledge, the structure of reality. Allow, further, that in some dialogues Plato focuses
attention on just how we may address those issues—by thinking, for example, about
either the possibility of discovery (Meno 80 ff), or the nature of truth and the possibility
of falsehood (Theaetetus and Sophist), or the constraints on contradiction (Euthydemus),
or the nature and significance of philosophy itself (for example, Phaedo, Republic,
Theaetetus 172–77, and in the Sophist/Statesman/[Philosopher] trilogy). How are these
aspects of the dialogues connected? We might think that the dialogue form gives us access
to the general, methodological questions but that this is quite separate from the issues of
virtue and reality, which are tackled by Plato’s characters. How are we to make the
connection between detached reflection, and what we should reflect about?
I have suggested that the questions about how to do philosophy, the general method-
ological questions raised in the frame dialogues, are not to be construed as generating
merely formal constraints on argument, for, I argued, these formal constraints include
psychological conditions. The methodology, therefore, already has considerable content.
But there is more.
Return to those dialogues where we see Socrates investigating the views of his
interlocutor—the dialogues where Socrates seems to advance little that is positive of
his own—apart, of course, from the vital discussions we find in each case of how to go
about answering the questions. In the Crito, for example, the discussion between
Socrates and Crito about the nature of moral justification affects the action they are
considering: Should Socrates escape from prison? Or Euthyphro’s false claims to expertise
underpin, and undermine, what he does: If he has no expertise, how can he justify
prosecuting his father? As the frame dialogue articulates the long sections of argument
by asking just what the conditions for knowledge are (at 6d, 8a, 9c, 11b, 14c, 15c), it
addresses what both Socrates and Euthyphro should do. Here claims to knowledge are
connected to what should be done; if they fail, this should have an immediate effect on
what is done. But the claims about knowledge themselves are subject to the constraints
about explanation and consistency governing the formality of dialectic. The formal
conditions for argument and for explanation, therefore, are not separable from the
ethical questions that are the content of the arguments.
112   Mary Margaret Mccabe

This is not merely an issue for those who lay direct claim to knowledge about moral
matters. After all, in many passages, Socrates suggests that all knowledge is integrated
(for example, Euthydemus 281; Republic 508d ff.). This, at least, asks how we should
understand the relations between knowledge of one sort (of shoemaking, for example)
and of another (of the good, for example), or the relation between knowledge and the
virtue of wisdom. And it is a question that seeks an ethical answer, as well as an episte-
mological one. But the relation between moral matters and formality goes deeper.
Consider the attack by the sophists of the Euthydemus on the possibility of contra-
diction (285d–288a). They take the view (if a view it could be called) that—since everything
they say is true—no one can contradict anyone else, or even themselves. If this is correct,
consistency does not matter and they themselves cannot be refuted. The consequence of
this, of course, is that the rules for argument break down altogether, including the rules
for the argument that would rebut it. How, in the face of this, are we to reclaim the possi-
bility of argument? Socrates offers metaphysical suggestions—suggestions about the
persistence and integrity of the person72 to whom the rules of argument apply—to
support his claim that the sophists are wrong. But those metaphysical suggestions are
compelling because they are based on ethical considerations: without persistence and
integrity, there is no such thing as a life such that we might live it. Since a life—the best
life—is what we seek, we are bound to suppose that we persist and then to agree that
consistency matters. The constraints—if this is right—on argument come from a complex
network of theory: from assumptions about knowledge, about reality and identity, and
about value, which provide the content of the individual discussions of individual
dialogues. I call this “ethical rationalism,” and it runs through the dialogues, in frame
and picture alike. As a consequence, Plato commits himself—in his ways of writing—to
a substantial philosophical position: that there is no line of demarcation between the
constraints of logic and those of ethics, of psychology or metaphysics, or epistemology.
The dialogues, in different ways, with different interests, with varying assumptions and
starting points, suppose that the relation between what is said, how it is said, and who
said it is intimate.

13. Representation and

Reading the Dialogues

We can still draw a line between where the characters represented stand and where Plato
stands as he represents them: the line between the frame and the picture is not a contrast
between what does and what does not matter to philosophy but between the orders of

72 At Protagoras 311b, Socrates asks of Hippocrates: “Who would you become [if you learn from the
sophist]?” At Euthydemus 291a, Crito is incredulous that Ctesippus could have said what Socrates reports:
“What sort of Ctesippus [said that]?” The opening word of the Phaedo is “yourself ”; see Burnyeat, “Words.”
 Plato’s Ways Of Writing   113

reflection on what is said. Equally, therefore, we can draw a line between what Plato
writes and what stance we may occupy in reading. For as we read, we identify with
different positions, and then—also as we read—we are shocked out of that identification
into a position of detachment, of reflection on the positions with which identification is
possible. What happens to us then is exactly parallel to—but not identical with—what
happens to any interlocutor: we take a position, then look at it from without. This
detachment allows us to see just how complex a matter it is to occupy a philosophical
position at all and how doing so—if ethical rationalism is true—involves all our various
commitments.
This, we may now see, gives us an account of why the dialogues represent people in
conversation with one another, and in quite different ways. There may be no single story
to be told of how, for example, personal identity, or psychological unity, is maintained for
all the dialogues. The questions each dialogue asks range over a wide range of abstract
questions, and they do so by virtue of setting direct conversations in a dramatic context,
in all the many ways in which that is done.
This means, in the first instance, that we should read the dialogues one by one, alert to
the internal nuances of context and reference, and read them whole. This is not, I have
been arguing, merely a matter of tone or of emotional engagement but, rather, a matter
of the constraints on argument and, connectedly, on what arguments can reasonably
maintain. Consider how the first part of the Parmenides (127–136)—where, notoriously,
the venerable figure of Parmenides puts apparently lethal objections to Socrates’ theory
of forms—portrays Socrates as a young man. Frequently, this is construed as a coded
message, that we should not take Parmenides’s arguments to be decisive. But we should
avoid reading such coding piecemeal: What, in the same context, are we to say of the
figure of Parmenides, apparently a committed monist (and so opposed to Socrates’
theory), who announces that without the theory of Forms we lose the power of dialectic—
the power that he himself wields throughout the dialogue (when he loses it in the Sophist,
he dies)? The setting of this dialogue makes our response to the arguments indeterminate
because both sides are loaded. The framing of the dialogue, striking as it is, makes this
indeterminacy itself the focus of critical attention and reflects, in turn, on the standing
of the arguments against the forms.
Or the Theaetetus finds itself unable to say just how someone could come to believe
falsely, at least in the absence of perception (200b–c). The puzzle about falsehood, in this
dialogue, is partly a problem in the philosophy of mind—a problem of explaining just how
we could come to be wrong about things. But it follows the denial of Protagoras’s claim
that everything is true—a denial that rests on a dialectical refutation of Protagoras’s
view. That refutation, in turn, relies on the suggestion that if dialectic is possible, then we
can be wrong about things and can disagree. If the interlocutors engage in conversation,
then disagreement is possible; if we disagree, then the disagreement is realized as we read.
The Philebus opens (and closes) in the middle of a discussion of the best life, which
should be perfect and self-sufficient (20d–21a). Discussion should have the same
characteristics, and be determinate, complete, and sufficient (19e–20a), but discussion
114   Mary Margaret Mccabe

with Philebus can never be so since his extreme hedonism admits of no limitation at all.
The way the discussion with Philebus fails73 is designed to show that the preferred life,
the life rationally chosen, must be itself susceptible to rationality and that extreme
­hedonism is intolerable. It does so once we reflect on the conditions for this dialogue:
it does so when we read, reflectively and detached.

14. Intertextuality

But there is something else again: for the dialogues invite us not only to be self-conscious
readers of any one dialogue but also to attend to others, in an equally self-conscious way.
The interconnections among Theaetetus, Sophist, Statesman, and the missing Philosopher,
for example, are peculiar. The Theaetetus is a story told years later about the discussion
among Socrates, Theodorus, and Theaetetus (along with some of Theaetetus’ friends).
The Sophist continues the same conversation (now in direct speech; the recall of the
Theaetetus has vanished) the following day, with a new character, the Eleatic Stranger.
Then the Statesman continues the conversation (at some unspecified time) but fails to
complete the discussion: the dialogue of the Philosopher does not exist.74 There is, there-
fore, a formal continuity between the three extant dialogues, so that we are asked to
notice that they are related. But the relation is very much more complicated than just
that the Sophist and the Statesman provide us with more of the same. Instead, the Sophist
reflects vividly back on places where the Theaetetus failed (e.g., at 236e) and also, perhaps,
on places where the Theaetetus asks questions the Sophist does not answer (there is nothing
in the Sophist that explains how we could come to make a mistake about an unrealized
state of affairs, even if it tells us what it is to have made the mistake: false beliefs are false
statements in our heads75). And the Statesman wonders, in its turn, just how—or whether—
the reflections on dialectic to be found in the Sophist are to be integrated into life lived in
the state.76
We might say the same of the relation between the Republic and the Timaeus,77
dialogues that at first seem to be describing the same event—a two-day conversation
Socrates had after the festival of a goddess (Republic 327a, Timaeus 21a). But then it
becomes clear that the Republic’s goddess is the Thracian Bendis, while the Timaeus’s is
Athena, the patron of Athens.78 Is this mere carelessness? Well, once this forces the

73 See McCabe, Predecessors, ch. 4.

74 Of course, Plato might have died before he wrote it, but the complex interconnections among
Theaetetus, Sophist, and Statesman suggest that its omission was deliberate.
75 See Section 9 of this chapter on the silent dialogue.
76 Notice the unclarity of the myth about just what constitutes the best life, or even about what
constitutes a life at all: 269–74.
77 The Republic does not specify Socrates’ audience; this is not inconsistent with the Timaeus’s
representation that Socrates was, on the occasion to which he alludes at Timaeus 17a, talking to
Timaeus, Critias, and Hermogenes.
78 See Republic 354a10 and Timaeus 26e.
 Plato’s Ways Of Writing   115

reader to think harder about the relation between the Republic and the Timaeus, she
should notice, too, that what seems like a summary of the Republic at the opening of the
Timaeus in fact leaves out all the central metaphysical and epistemological material of
books V–VII. The “compare and contrast” that this invites is critical: What is the impor-
tance of the central books to the theory advanced by Socrates in the Republic? How is
that altered by the account of cosmic teleology offered by Timaeus? How are the two
accounts of the way things really consistent? The setting of the Timaeus prevents us from
restricting our reading of it to the Timaeus alone, and it brings the Timaeus’s relation to
the Republic into philosophical view.
We may say the same of the extraordinarily large intertextuality of other dialogues.
Sometimes the trick is turned by the same sort of language (e.g., the motif of sticks and
stones at Phaedo 74b ff, Parmenides 129d); sometimes by the allusion to some particular
person (e.g., Euthydemus, whose theory of total assimilation is described at Cratylus 386d);
and sometimes in ways that are rather more specific (e.g., the reference to an argument
for recollection at Phaedo 72e79). Sometimes one passage seems to offer a critique of
another (e.g., Euthydemus 293b–296d on Meno 81 on recollection); sometimes the point,
or the direction, of an allusion remains obscure.80 The one characteristic of these passages
is that they are allusive, inexact, puzzling. They run true to what I have described as the
detaching effect of the dialogues—for they urge not only a similarity between the two
passages in question but also a difference. In so doing, they militate against the thought
that what we have here are somehow references to a single body of fixed doctrine, under-
pinning all the dialogues. They equally, and for the same reasons, militate against the
thought that we are being offered here coded references to some esoteric thinking,
private to the initiated. For, on the contrary, the critical reflection to which we are invited by
these cross-references is symmetrical with the reflection invited in us by other passages
and with the reflection portrayed in the characters at the many levels of the narration.
Nothing here requires us to suppose, that is to say, that we have here merely a vehicle for
Platonic doctrine or, contrariwise, to suppose that the dialogues repudiate positive
views altogether. Instead, views are indeed put forward in the dialogues, and for some of
those views, the author must take responsibility. But in writing the way he does, he
engages his readers, too, in active scrutiny of what is said: a large part of the philosophical
work, therefore, is done by us, the readers of what Plato writes.81

79 This cross-reference invites the reader to remember, in a context where what it is to remember is
at issue.
80 What is the relation between the sudden mention of dialecticians at Euthydemus 290c and the
account of the philosopher-kings of the central books of the Republic? On these passages see McCabe,
Conversations, chs. 10 and 13.
81 I have, as always, benefited from discussion with many people about these topics: my particular
thanks to Peter Adamson, Hugh Benson, Sarah Broadie, Gail Fine, Christopher Gill, Verity Harte,
Dominic Scott, Nick Smith, Raphael Woolf, and especially to Owen Gower, Jonas Green, Alex Long,
and Imogen Smith. My thanks, also, to the Leverhulme Trust for the Major Research Fellowship during
the tenure of which I wrote this essay.
116   Mary Margaret Mccabe

Bibliography
Allen, R. E. Plato’s Parmenides, Translation and Analysis (Oxford, 1983).
Annas, J. “Plato’s Myths of Judgement,” Phronesis 27 (1982): 119–43.
Annas, J., and Rowe, C., eds. New Perspectives on Plato, Modern and Ancient (Cambridge,
Mass., 2002).
Barnes, J. The Cambridge Companion to Aristotle (Cambridge, 1995).
Benson, H. H. Essays on the Philosophy of Socrates (New York, 1992).
Benson, H. H. Socratic Wisdom: The Model of Knowledge in Plato’s Early Dialogues (New York,
2000).
Beversluis, J. Cross-Examining Socrates: A Defense of the Interlocutors in Plato’s Early Dialogues
(Cambridge, 2000).
Bloch, E. “Hemlock Poisoning and the Death of Socrates: Did Plato Tell the Truth?” in
T. Brickhouse and N. Smith, eds., The Trial and Execution of Socrates (Oxford, 2001).
Blondell, R. The Play of Character in Plato’s Dialogues (Cambridge, 2002).
Brickhouse, T., and Smith, N. The Philosophy of Socrates (Boulder, Colo., 2000).
Brickhouse, T., and Smith, N. Socrates on Trial (Princeton, 1989).
Burnyeat, M. F. Culture and Society in Plato’s Republic. The Tanner Lectures on Human Values
(Cambridge, Mass., 1999).
Burnyeat, M. F. “First Words,” Proceedings of the Cambridge Philological Society 43 (1998): 1–20.
Burnyeat, M. F. “Plato on the Grammar of Perceiving,” Classical Quarterly 26 (1976): 29–51.
Burnyeat, M. F. “Sphinx without a Secret?” New York Review of Books, May 30, 1985.
Burnyeat, M. F. The Theaetetus of Plato, trans. M. J. Levett (Indianapolis, Ind., 1990).
Clay, D. Platonic Questions: Dialogues with the Silent Philosopher (University Park, Penn., 2000).
Cooper, J. M. Plato: Complete Works (Indianapolis, Ind., 1997).
Cooper, J. M. “Plato on Sense-Perception and Knowledge (Theaetetus 184–6),” Phronesis 15 (1970):
123–65.
Cornford, F. M. Plato and Parmenides (London, 1939).
Derrida, J. “Plato’s Pharmacy,” in Dissemination, trans. Barbara Johnson (Chicago, 1981).
Ferrari, G. R. F. Listening to the Cicadas: A Study of Plato’s Phaedrus (Cambridge, 1987).
Frede, M. “The Literary Form of the Sophist” in C. Gill and M. M. McCabe, eds., Form and
Argument in Late Plato (Oxford, 1996).
Gadamer, H. Dialogue and Dialectic: Eight Hermeneutical Studies of Plato, trans P. C. Smith
(New Haven, Conn., 1980).
Gallop, D. Plato: Phaedo (Oxford, 1975).
Gill, C. “The Death of Socrates,” Classical Quarterly 23 (1973): 25–28.
Gill, C. “Speaking Up for Plato’s Interlocutors,” Oxford Studies in Ancient Philosophy 20 (2001):
297–321.
Gill, C. “Critical Response to the Hermeneutical Approach from an Analytic Perspective” in
G. Reale and S. Scolnicov, eds., New Images of Plato: The Idea of the Good (Sankt Augustin,
Germany, 2002).
Gill, C., and McCabe, M. M., eds. Form and Argument in Late Plato [Form] (Oxford, 1996).
Halliwell, S. The Aesthetics of Mimesis: Ancient Texts and Modern Problems (Princeton, N.J. 2002).
Hunter, R. Plato: Symposium (Oxford, 2004).
Kahn, C. H. Plato and the Socratic Dialogue: The Philosophical Use of a Literary Form
(Cambridge, 1996).
Krämer, H. J. Plato and the Foundations of Metaphysics, trans. J. R. Caton (Albany, N.Y., 1990).
 Plato’s Ways Of Writing   117

Kraut, R. “Introduction to the Study of Plato,” in R. Kraut, ed., The Cambridge Companion to
Plato, 1–50 (Cambridge, 1992).
Long, A. G. “Character and Dialectic: The Philosophical Origins of the Platonic Dialogue,”
Ph.D. dissertation, University of Cambridge (2004).
Long, A. G.“The Form of Plato’s Republic,” in R. Osborne, ed., Debating the Athenian Cultural
Revolution (Cambridge, 2007).
McCabe, M. M. “Dialectic Is as Dialectic Does,” in B. Reis, ed., The Virtuous Life in Greek
Ethics (Cambridge, 2006).
McCabe, M. M. “Irony in the Soul: Should Plato’s Socrates Be Sincere?” in M. Trapp, ed.,
Socrates, (London, 2007).
McCabe, M. M. Plato and His Predecessors: The Dramatisation of Reason (Cambridge, 2000).
McCabe, M. M. Platonic Conversations (Oxford 2015).
Morgan, K. Myth and Philosophy: From the Presocratics to Plato (Cambridge, 2000).
Nails, D. Agora, Academy and the Conduct of Philosophy (Dordrecht, 1995).
Nails, D. The People of Plato (Indianapolis, Ind., 2002).
Nehamas, A. The Art of Living: Socratic Reflections from Plato to Foucault. (Berkeley, Calif., 1998).
Nehamas, A. The Virtues of Authenticity (Princeton, N.J., 2000).
Nightingale, A. Genres in Dialogue: Plato and the Construct of Philosophy (Cambridge, 1995).
Nussbaum, M. C. The Fragility of Goodness: Luck and Ethics in Greek Tragedy and Philosophy
(Cambridge, 1986).
Owen, G. E. L. Logic, Science and Dialectic (London, 1986).
Press, G. A., ed. “Who Speaks for Plato?”: Studies in Platonic Anonymity (Lanham, M.D., 2000).
Reale, G., and Scolnicov, S., eds. New Images of Plato: The Idea of the Good (Sankt Augustin, 2002).
Roochnik, D. The Tragedy of Reason (New York, 1990).
Rorty, R. Philosophy and the Mirror of Nature (Princeton, N.J., 1979).
Rutherford, R. The Art of Plato (London, 1995).
Santas, G. Socrates: Philosophy in Plato’s Early Dialogues (London, 1979).
Sayre, K. Plato’s Literary Garden: How to Read a Platonic Dialogue (Notre Dame, Ind., 1995).
Sedley, D. N. The Midwife of Platonism: Text and Subtext in Plato’s Theaetetus (Oxford, 2004).
Sedley, D. N. “The Theaetetus: Three Interpretations,” in C. Gill and M. M. McCabe, eds., Form
and Argument in Late Plato (Oxford, 1996).
Strauss, L. The City and the Man (Chicago, 1964).
Stokes, M. C. Plato’s Socratic Conversations (London, 1980).
Szlezak, T. A. Reading Plato (London, 1999).
Vlastos, G. Socrates, Ironist and Moral Philosopher (Cambridge, 1991).
Williams, B. A. O. Ethics and the Limits of Philosophy (London, 1985).
Williams, B. A. O. Shame and Necessity (Berkeley, Calif., 1993).
 chapter 5

Eu th y phro, A pology,
a n d Cr ito
The Examined and Virtuous Life

Hugh H. Benson

1. Introduction

Plato’s Euthyphro takes place outside the hall of the King-Archon before Socrates’
attendance at a preliminary hearing associated with his prosecution for corrupting the
youth and impiety. In the Apology Plato presents Socrates’ unsuccessful defense (apologia)
against these charges. About a month later the Crito depicts the attempt to persuade
Socrates to escape from his cell and thereby avoid his pending execution in the next day
or two. The Phaedo depicts the day of Socrates’ execution with Socrates trying to relieve
his friends’ fears and grief by persuading them that the soul is immortal and that genuine
philosophers want to die. The dramatic context suggests reading these dialogues together.
Nevertheless, I focus primarily on Socrates’ answer to the question how one should live
one’s life in the first three dialogues.

2. The Apology

The Examined Life

At Apology 38a5–6 Socrates maintains that the unexamined life is not worth living. He
says this in the penalty portion of his trial, after having been found guilty of the charges
mentioned previously. He considers the possibility of exile in silence and offers a two-
part explanation for why he cannot propose this alternative penalty. First, being silent
120   Hugh H. Benson

would be disobeying the god (37e5–38a1), and second, making logoi1 concerning virtue
and other things the jurors have heard him discussing, and examining himself and others
is the greatest good for an individual (38a1–5).2 Making such logoi, the greatest good for
an individual, makes life worth living.
The recommendation to live this examined life has its origins in the defense portion
of the Apology. At Apology 18a7–19a7, Socrates distinguishes between two sets of charges
leveled against him—the official ones brought by Meletus, Anytus, and Lycon, and earlier
unofficial ones imagined as follows:

Socrates is guilty of wrongdoing in that he busies himself studying things in the sky
and below the earth; he makes the worse into the stronger argument, and he teaches
these same things to others. (Apology 19b4–c1)3

Socrates maintains that it is more difficult to respond to these earlier accusations.

Nevertheless, he repudiates them outright. He disavows knowledge of any of the things
alleged, and so the ability to teach them, though he “would pride and preen himself if
[he] had this knowledge” (20c1–2). He explains, however, that the cause of his slander-
ous reputation is that he possesses a certain kind of wisdom—human wisdom, but not
the wisdom he is thought to possess.
As a witness to his human wisdom he appeals to the god of Delphi. As Socrates tells it,
Chaerephon asked the oracle at Delphi whether anyone was wiser than Socrates, and the
oracle responded that no one was. Socrates, however, was at a loss because he was aware
of (sunoida) not being wise about anything large or small (21b4–5), and yet the god could
not speak falsely (21b6). Consequently, Socrates began to investigate as follows. He
examined one of those reputed to be wise—a politician—and discovered that though he
seemed wise to others and to himself, in fact he was not, and Socrates tried to show him
this. Socrates concluded that:

I am wiser than this man; it is likely that neither of us knows anything fine and good
(kalon kagathon), but he thinks he knows something when he does not, whereas
when I do not know, neither do I think I know; so I am likely to be wiser than he to this
small extent, that I do not think I know what I do not know. (Apology 21d1–7)

Socrates next examined other politicians, and then poets, and craftsmen, all to similar
effect.4 The result of these examinations was that Socrates became disliked and acquired
a reputation for wisdom because he was believed to be wise concerning things knowledge

1 The Greek word logos and it cognates (logoi is the plural of logos) can be translated in a variety of
ways: word, statement, story, discussion, speech, thought, reason, account, calculation. I leave it
transliterated in what follows.
2 Plato’s use of anthrôpôi at most restricts the greatest good of the examined life to free Athenian
males. All such individuals have the ability to live such a life. (See (Kraut 1984:201 n.18)).
3 Translations are from (Cooper and Hutchinson 1997) subject to minor revisions.
4 In examining the craftsmen Socrates discovers that they did know some of the things they thought
they knew—the many fine things associated with their crafts—but they also thought they knew other
 Euthyphro, Apology, and Crito   121

of which he had shown the others lacked (23a3–5). But in fact, according to Socrates, the
god is wise and the puzzling response to Chaerephon meant that:

human wisdom is worth little or nothing, and that when he says this man, Socrates,
he is using my name as an example, as if he said: “This man among you, mortals, is
wisest who, like Socrates, understands (egnôken) that his wisdom is worthless.”
(Apology 23a6–b4)

No one is wiser than Socrates because no one else recognizes that he does not know the
things he does not know. Consequently, Socrates continues to seek and to examine
­anyone whom he thinks is wise and when he discovers that he is not tries to show him5
that he is not.

The Disavowal of Knowledge

Socrates’ description of the examined life highlights his disavowal of knowledge, or, as
Socrates puts it, his awareness of being wise about nothing large or small (21b4–5)—his
human wisdom. It is sometimes thought that Socrates is professing to know that he
knows nothing and then it is immediately pointed out that such a profession is at best
paradoxical,6 at worst self-refuting.7 But Socrates professes no such thing.8 Socrates
does assert at 21b4–5 that he is aware of being wise about nothing great or small. But, to
get a paradox or self-refutation Socrates also needs to be committed to both:

A1 If A is aware of p, then A knows p, and

A2 If A is wise about nothing great or small, then A knows nothing.

So, according to 21b4–5, Socrates knows (by A1) that he knows nothing (by A2). Of
course, one reason to doubt Socrates’ commitment to A1 and A2 is that such a commit-
ment would lead to paradox or self-refutation given 21b4–5. But we have independent
reason to doubt Socrates’ commitment to A2. At least twice later in the Apology Socrates
professes to know something. At 29b6–7 Socrates professes to “know . . . that it is wicked
and shameful to do wrong, to disobey one’s superior, be he god or man” and again at

fine things which they did not. And so, like the politicians and the poets, they too failed to be as wise as
Socrates. Only he failed to think that he knew fine and good things, which he did not.
5 (Doyle 2004:28) points out that 23b7 does not say “show him that he is not [wise]” (emphasis
added) as Grube translates, but “show that he is not [wise].” (Woolf 2008:8) connects 23b7 with 21c8–d1
and points out that the latter passage is qualified by “I tried to show him that he was not [wise].” The
point is that part of Socrates’ aim in pursuing the examinations of others is to show them their lack of
wisdom, when they lack it, whether or not he is ever successful at this. In doing so he encourages them
to care for wisdom and virtue. For possible successes see Ion at Ion 542b1–4, Hippocrates at Protagoras
313c3–4, Charmides at Charmides 162b9–10, and 176a6–b4, and Crito at Crito 50a4–5.
6 See, e.g., (Austin 1987). 7 See, e.g., (Kraut 1984:272 n. 44).
8 For a persuasive argument to this effect see (Fine 2008).
122   Hugh H. Benson

37b7–8 that he knows that imprisonment or exile would be evil. So, whatever else
Socrates believes, he evidently does not believe that in being wise about nothing great or
small, he knows nothing. Appealing to Socrates’ professions of knowledge at 29b6–7
and 37b7–8, however, avoids one textual tension only to raise another. At 21d4–5,
Socrates rephrases his disavowal of knowledge as failing to know (eidenai) anything fine
and good (kalon kagathon), and whatever else we think about the knowledge professions at
29b6–7 and 37b7-8 they are professions to know something fine and good. Indeed, on
the counterfactual (as I think) assumption that Socrates would claim to know that he
knows nothing fine and good, the paradox or self-refutation of Socrates’ disavowal of
knowledge returns though in a different way. Knowing that he knows nothing fine and
good is presumably to know something fine and good.
Textual tensions like this are typically met by attempts to disambiguate some key
concept, and the current tensions are no exception. Vlastos distinguishes between elenctic
knowledge, which Socrates professes to have, and certain knowledge, which Socrates
disavows.9 The former, according to Vlastos, is something like justified true belief and is
fairly common and easily acquired, at least on Socrates’ part. The latter is something
like Cartesian certainty and belongs only to the gods. Others distinguish between
high-level knowledge and low-level knowledge.10 The nature of high-level knowledge
differs according to these accounts but the general idea is that it amounts to a robustly
systematic “truth-entailing cognitive condition”11 requiring the ability to explain or define
key features of a field of inquiry, analogous (if not identical) to the cognitive condition of
an expert cobbler or medical practitioner, and extremely difficult (if not impossible) to
acquire. Low-level knowledge is, on the other hand, an isolated (or nearly isolated) truth
entailing cognitive condition analogous (if not identical) to justified true belief, fairly easy
to acquire, and fairly ubiquitous among Socrates’ peers. It is the former that Socrates
disavows, and the latter that he sometimes professes. An important difference between
these latter accounts and Vlastos’s account is that though Socrates disavows the posses-
sion of high-level knowledge he remains committed to acquiring it. This is in part what
makes his human wisdom valuable, but more on this in a moment.
Much can be said in favor of these latter accounts,12 though I doubt that they adequately
explain Socrates’ profession of knowledge at Apology 29b6–7.13 Nevertheless, despite this
problematic passage, Socrates’ general disavowal of knowledge is neither paradoxical
nor self-refuting. Either Socrates fails to be committed to A1 and so fails to assert that he

9 (Vlastos 1985).
10 For this terminology see (Fine 2008) and Taylor, Chapter 18 of this volume. Others who defend
significantly different versions of this sort of account include (Woodruff 1990), (Brickhouse and
Smith 1994:30–45), and (Taylor, Hare and Barnes 1999:45–51).
11 (Fine 2008:53). 12 For my sympathy with these accounts see (Benson 2000:233–38).
13 The point of 29b6–7 is to contrast something Socrates knows with something he does not know.
On the one hand, he disavows knowledge (eidôs) of anything in the underworld explicitly referring
back to what makes no one wiser than him (viz. his awareness that he lacks high-level knowledge) and,
on the other hand, he professes to know (oida) that it is wicked and shameful to disobey a superior. The
contrast between something he knows and something he does not know loses its force if the levels of
knowledge disavowed and avowed are different. The knowledge disavowed is high-level and so is the
knowledge avowed.
 Euthyphro, Apology, and Crito   123

knows that he knows nothing great or small, or he is merely asserting that he knows at a
low-level that he knows nothing great or small at a high-level. If a tension remains
among Socrates’ avowals and disavowals that cannot be resolved by recognizing different
levels of knowledge, it does not concern his general disavowal. Rather, it concerns his
apparent professions of knowledge of something great or small or fine and good. But,
the recognition that Socrates is disavowing high-level knowledge (whether or not he
recognizes any low-level knowledge) reduces the number of such conflicting professions.
Only those texts in which Socrates avows high-level knowledge of something fine and
good are problematic, and those are few and far between. Indeed, only Socrates’ knowledge
profession at 29b6–7 seems best understood as high-level knowledge of something fine
and good.14 Perhaps it is an unfortunate, though pointed, overstatement made in the
heat of the moment.15
However we treat Socrates’ profession at 29b6–7, Socrates’ focus is on the knowledge
he lacks, not on the knowledge he has. Socrates does not understand the oracle’s puz-
zling response to indicate that no one is wiser than he because no one has more low-level
knowledge than he does or that no one has low-level knowledge concerning some subject
matter that he has. Rather he takes the oracular response to indicate that his cognitive
state, whatever it is (low-level knowledge of some sort or no knowledge of any sort at all
concerning anything fine and good), is worthless (23b3–4). What makes no one wiser
than Socrates is not the knowledge he has (if he has any), but the awareness of the
knowledge he lacks. It is his reflective awareness that he lacks any high-level knowledge
of anything fine and good at all16 (if there is more than one level of knowledge), and so
his failure to believe that he has such knowledge when he does not, unlike everyone else
he has examined.

The Virtuous Life

After Socrates has, as he thinks, quickly refuted the official charges put forward by Meletus,
he considers whether he is ashamed to have lived a life that has put him in jeopardy. In
response Socrates connects the examined life with the other life he recommends—the
good and virtuous life.17 Once Socrates has come to understand the puzzling oracular
response, he does not cease his examination of himself and others—his philosophizing—
but continues it according to the god. He does so in order to encourage others to care

14 Apology 23b3–4, 37b2-8, Republic 351a5–6, Euthydemus 296e3–297a2, Gorgias 521c7–d3, and
Hippias Major 304e6–9 are more amenable to low-level knowledge professions. Matthews, Ch. 16 of
this volume, offers an alternative reading of this tension. Against his reading see my brief defense later
on in this chapter of Socrates’ commitment to the priority of definitional knowledge.
15 See, e.g., (Forster 2006, 14–16) and (Irwin 1977:58).
16 Does Socrates ever reflectively profess to have any high-level knowledge? Perhaps, but if so it is
outside the so-called Socratic dialogues. At Symposium 177d7–8 Socrates indicates that the only thing
he knows is the art of love and at Theaetetus 149a4 he professes to have the expertise (technê) of
cognitive midwifery.
17 See also 36c2–d1 and 41e1–42a1.
124   Hugh H. Benson

about virtue, wisdom, and truth, as opposed to wealth, reputation, and honor (29e1–3,
see also 30a9–b2). Socrates does this by examining anyone he meets who professes to
care about the former things in order to determine whether they have the wisdom they
think they do (see 41b5–7). If they do not, he reproaches them for failing to care about
the most important things (29e5–30a3).
Socrates here indicates the value of his human wisdom. Examining whether one has
the high-level knowledge one thinks one has evidently is a way of examining whether
one cares about what one should. The idea seems to be that if one cared about wisdom,
truth, and virtue, one would either have the knowledge one professed to have, or one
would recognize that one did not and seek to acquire it. The latter is Socrates’ situation.
Socrates’ human wisdom promotes his care for virtue.18 Recognizing his lack of high-
level knowledge concerning anything fine and good leads him to try to acquire it—to
acquire the wisdom and virtue he lacks. Caring about wisdom and virtue, does not
require being wise and virtuous,19 but it does require recognizing one is not, when one is
not, as Socrates does, and so attempting to become wise and virtuous, again as Socrates
does. In lacking knowledge of anything fine and good one lacks virtue. Socrates does
not explicitly assert the equivalence of virtue and knowledge of anything fine and
good here.20 His description of the examined life, however, appears to presuppose it.
Nevertheless, the knowledge that Socrates takes to be equivalent to virtue is not the low-
level knowledge that Socrates and his compatriots may have (perhaps in abundance—if
they have any at all), but the high-level knowledge that he and everyone else he has met
fail to have—at least concerning anything fine and good. This is what Socrates cares
about and why he examines himself and others. The examined life encourages the virtu-
ous life. To see this in practice we turn to the Euthyphro.

3. The Euthyphro

Euthyphro is prosecuting his father for killing his dependent who had killed one of their
household slaves. Euthyphro’s father had had the dependent bound and thrown into a
ditch, and then sent off to the priest in Athens to find out what ought to be done. In the
meantime the dependent, predictably, died. Euthyphro explains that his father and other
relatives are angry at him, for they think “it is impious for a son to prosecute his father
for murder” (4d9–e1). Socrates proposes that Euthyphro would never have pursued
such a prosecution unless he thought that he knew what piety was (4e4–8 and 15c11–d8),
fearing lest his relatives are correct that in prosecuting his father he is acting impiously.
Consequently, Socrates challenges Euthyphro to teach him the nature of piety, so that he

18 It does not, however, make him virtuous. For that the possession of the wisdom he lacks is required.
19 Although being wise and virtuous is sufficient for caring about wisdom and virtue.
20 See (Brickhouse and Smith 2010:153–92). Devereux, Ch. 17 of this volume, maintains that while
Socrates treats knowledge and virtue as equivalent in the Euthydemus, he rejects this view in the Gorgias.
 Euthyphro, Apology, and Crito   125

can defend himself against the charge of impiety brought by Meletus. When Euthyphro
agrees that he knows all such things (4e9–5a2), Socrates asks Euthyphro to tell him what
he claims to know—what piety and impiety are (5c9). What follows is an example of liv-
ing the examined life. Socrates and Euthyphro make logoi concerning piety and Socrates
examines Euthyphro in order to determine whether he has the wisdom he professes to
have, that is, the knowledge of what piety is, or whether he thinks he is wise (concerning
piety) but is not.

Proposed Answers to “What Is Piety?”

Much of the remainder of the Euthyphro consists of a series of answers to the question
“What is piety?,” none of which withstands Socratic scrutiny. Euthyphro first answers
that:

the pious is to do what I am doing now, to prosecute the wrongdoer, be it about

murder or temple robbery or anything else, whether the wrongdoer is your father or
your mother or anyone else; not to prosecute is impious. (5d8–e2)

When Socrates points out, and Euthyphro agrees, that many other actions besides
­prosecutions are pious, he explains that Euthyphro has failed to answer his question. He
did not ask for one or two of the many pious things, but for the “form itself (auto to
eidos) by which all pious things are pious” (6d10–11). As a result Euthyphro proposes his
second answer: piety is what is dear to the gods. Socrates follows this answer by reducing
it to absurdity. He maintains that given Euthyphro’s commitment to disagreement
among the gods (5e5–6c7), it follows from Euthyphro’s second answer that the same
things are pious and impious. But this is absurd since the pious and the impious are
opposites (7a9–b1). This leads to Euthyphro’s third, revised, answer: piety is what all the
gods love and impiety what they all hate (9e1–3).
The structure of the argument against Euthyphro’s third answer is relatively clear.
According to Euthyphro all three of the following are true:

E1 [a] x is loved (phileitai) by the gods because x is pious, and [b] it is not the case
that x is pious because x is loved by the gods (10d6–8 and 10e2–5).
E2 [a] x is god-loved21 because x is loved by the gods, and [b] it is not the case that
x is loved by the gods because x is god-loved (10c10–14, 10d9–11, and 10e6–9).
E3 [a] if piety is the same (tauton)22 as the god-loved, then [b] if x is loved by the
gods because x is pious, then x is loved by the gods because x is god-loved, and [c] if

21 I take philomenon and theophiles to be used equivalently in the argument and to theophiles to be
how Socrates understands Euthyphro’s third answer. Only in this way can the argument avoid being a
non sequitur.
22 The nature of this relation, that is, what it is for the F to be the same as the G where the G answers
the “What is F-ness?” question, is controversial, but see my brief comments about real definitions later
on in the chapter.
126   Hugh H. Benson

x is ­god-loved because x is loved by the gods, then x is pious because x is loved by

the gods (10e10–11a3).23

But, E1a and E2b entail that E3b is false and so that piety is not the same as the god-loved,
and E1b and E2a entail that E3c is false and so, again that piety is not the same as the
­god-loved. So, Euthyphro’s third attempt to answer the “What is piety?” question is
incompatible with E1, E2, and E3.
Despite the relative clarity of the structure of the argument, the meaning of E1 through
E3 and why Euthyphro should accept them are obscure. Socrates offers a partial explana-
tion at 10a5–c13. He first briefly distinguishes between a carrying thing (pheron) and a
being carried thing (pheromenon),24 and then between something being carried (pheretai)
and something being a carried thing (pheromenon). The latter distinction is obscure even in
the Greek, but Euthyphro readily agrees that something is the latter (a carried thing)
because it is the former (being carried) and not the other way round. That Euthyphro
should readily agree to this depends on how we understand the because-of relation.
If the because-of relation is analogous to the causal relation, Euthyphro should not agree
that a thing’s being carried causes it to be a carried thing. Whatever else is involved in a
causal relation the cause temporally precedes its effect. But a thing’s being carried is
simultaneous with its being a carried thing. The former does not precede the latter. Again,
if the because-of relation is analogous to a reason an agent might have for doing some-
thing, it is difficult to see why Euthyphro should agree that a thing’s being carried is the
reason the thing is a carried thing. Indeed, it is difficult to even make sense of such
relation in the context of a thing’s being carried and being a carried thing. But, if the
because-of relation is analogous to the grounding relation,25 then the idea is that some-
thing’s being carried serves as the ground of a thing’s being a carried thing. And this,
perhaps, has some intuitive force. The same force, for example, which would lead us to
agree that the disjunctive proposition that Socrates is wise or Plato is an Athenian is
true in virtue of the truth of one of its disjuncts, for example, that Plato is an Athenian.
The truth of the latter grounds the truth of the former.26 Of course, the precise nature
of the grounding relation is unclear, though Euthyphro expresses his familiarity with the
idea earlier when he agrees to Socrates’ request to tell him the form itself by which pious
things are pious. Socrates is asking for what grounds something’s being pious. He is
asking what makes pious things pious. Despite being imprecise, then, the idea that a thing’s

23 To theophiles and to hosion are ambiguous at 10e10–11a3. Their first uses refer to the properties of
being god-loved and being pious (or piety); thereafter they are used to refer to a god-loved thing and a
pious thing. See (Judson 2010:52 ff.).
24 For simplicity I will only use the carrying and being carried example, though Socrates includes others.
25 See, e.g., (Schaffer 2009), (Rosen 2010), and (Correia and Schnieder 2012). I follow
(Evans 2012:11–12) in restricting the relation to metaphysical as opposed to conceptual ground.
Henceforth, I call this relation the grounding relation. My reading is indebted to Evans, although
differences remain.
26 I owe this example to (Rosen 2010).
 Euthyphro, Apology, and Crito   127

being a carried thing is grounded in its being carried has some intuitive appeal.27 In any
case Euthyphro readily agrees without any argument. Indeed, in light of this, Euthyphro
agrees to E2—a thing is god-loved (or a loved thing) because of (i.e., is grounded in) the
thing’s being loved by the gods, and not the other way round (10c10–13).28
The example of something’s being a carried thing because it is carried is offered in
order to explain the question that Socrates had asked Euthyphro at 10a4, viz. “Is some-
thing loved by the gods because it is pious or is it pious because it is loved by the gods?.”
Though the discussion of this example leads Euthyphro to accept E2, it was aimed at
helping him understand Socrates’ question. When Socrates asks the question again at
10d4, though in an altered form, Euthyphro answers immediately and without argument
that (E1) x is loved by the gods because it is pious and it is not the case that it is pious
because it is loved by the gods (10d6–8). But no argument has been offered on behalf of
E1. Some have thought that Socrates here is appealing to the fact that the gods have
reasons for loving the things they do,29 and the only reason on offer is that it is pious. But
this supplies a defense of E1 at the cost of making Socrates’ explanation of his question
irrelevant. Having explained his question in terms of the grounding relation, Socrates
accepts an answer in terms of what we might call the reasons relation. Moreover, if the
relations in E1 and E2 are not the same—the reasons relation in E1 and the grounding
relation in E2—equivocation threatens the argument. Thus, charity recommends reading
E1 as follows: x’s being loved by the gods is grounded in x’s being pious and not the other
way round. Euthyphro accepts E1 either because now that he (and the reader) under-
stands the question it is in some way intuitive or because he thinks that if the gods’ love is
not grounded in something (and being pious is the only thing on offer) their love would
be unreasonable, but the gods’ love is not unreasonable.30
E3 is not included in the argument proper (10d1–e1); rather it shows up in Socrates’
explanation of the argument at 10e2–11a6. After reminding Euthyphro that he has agreed
to E1 and E2, Socrates asserts E3 from which the conclusion that the pious and the god-
loved are different follows (11a3–6). E3 might be justified as a substitution instance of the
following more general principle.

E3* [a] if the F is the same as the G, then [b] if x is H because x is F, then x is H
because x is G, and [c] if x is F because x is H, then x is G because x is H.

27 Since the grounding relation is metaphysical its relata are distinct facts (pragmata) or things
(onta), requiring a fine-grained ontology.
28 The grounding relation is asymmetrical; if x grounds y, then not (y grounds x).
29 (Judson 2010:41–46) offers a plausible argument for why Socrates and Euthyphro may believe this;
see also (Evans 2012:23–29).
30 That something is loved by the gods is grounded in its being pious is plausibly intuitive, at least to
anyone inclined toward theistic ethics. The plausibility of the second defense increases proportionally
to the wisdom and perfection of the gods. Something like this is the intuition of those who take the
Euthyphro problem to be a serious challenge to theological voluntarism. If God’s reason for willing
some activity is not grounded in something “deeper” ontologically or morally than the willing itself, the
respect, reverence, or veneration of God is thereby diminished.
128   Hugh H. Benson

But such a principle looks fallacious because it permits substitution into intensional
contexts introduced by the because-of relation.31 The because-of relation understood as
the grounding relation, however, can be read extensionally. Suppose that this glass’s
value is grounded in its fragility. If we were to discover that molecular arrangement
answers the question “What is fragility?,” it would not be fallacious to conclude that this
glass’s value is grounded in its molecular arrangement. Indeed, if the glass’s value were
not so grounded we would doubt the discovery.
Thus, once Euthyphro understands that Socrates has in mind the grounding relation
he consents to E1, E2, and E3. Though Euthyphro is given no argument for these agree-
ments, they are not implausible or fallacious. Having secured Euthyphro’s consent
Socrates rightly concludes that they are at odds with his answer that the pious is the
god-loved. Socrates thereupon encourages him to try yet again to answer his “What
is piety?” question.
After expressing his frustration in being unable to say what he thinks (11b6–8) and
after considerable Socratic prodding (11e7–12d11), Euthyphro proposes that piety is the
part of justice concerned with care of the gods (12e6–8). Socrates, however, points out
that according to Euthyphro to care for something is to seek to benefit the thing cared
for, but Euthyphro denies that piety benefits the gods. Consequently, Euthyphro responds
with a revised, fifth answer: piety is the part of justice concerned with service to the gods
(13d8–9).32 Socrates presses Euthyphro to say what work service to the gods aims to
complete and Euthyphro abandons this answer and offers his final answer that piety is
knowledge of how to sacrifice and pray (14c6). When Euthyphro agrees that knowledge
of how to sacrifice to the gods amounts to knowledge of what to give to the gods and that
what we ought to give to the gods is what pleases them, Socrates concludes that piety is
what is pleasing and so loved by the gods (15b1–6).33 But we have already seen that piety
is not what is loved by the gods (15c1–3). So, Socrates maintains

Either we were wrong when we agreed before, or, if we were right then, we are wrong
now. (Euthyphro 15c8–9)

The Priority of Definition

The depiction of the examined life here in the Euthyphro highlights Socrates’ interest in
definition.34 One way to make the logoi described in the Apology is to ask what things
such as piety, courage (Laches), temperance (Charmides), virtue (Meno), and justice

31 See, e.g., (Judson 2010:48). (Evans 2012:15–22) defends the extensional reading.
32 On the presumed significance of this answer see (Brickhouse and Smith 1983) and
(McPherran 1985).
33 The last three answers seem to characterize pious persons, while the first characterizes pious
actions, and the two intermediate ones can characterize both.
34 In what follows “definition” is shorthand for an answer to a “What is F-ness?” question.
 Euthyphro, Apology, and Crito   129

(Republic I) are. It is to ask the “What is F-ness?” question, though the examined life
need not proceed in this way.35
The Euthyphro indicates that in looking for an answer to the “What is piety?”
­question Socrates is looking for the thing (the form)36 by which something is pious
(6d10–11). Socrates is looking for the answer to the question “What grounds a thing’s
being pious?.” “Piety” is its name, but what is it that “piety” names? Whatever it is it will be
extensionally equivalent with a thing’s being pious. A prosecution of a religious wrongdoer
(Euthyphro’s first answer) may be a pious action, but there are pious things (actions or
persons) that are not such prosecutions. Perhaps, all pious persons serve the gods
(Euthyphro’s fifth answer), but only some of those who serve the gods (those whose
­service accomplishes the gods’ work) are pious.37 These answers fail to be extensionally
equivalent to piety, and so they fail to adequately answer the “What is piety?” question. But
extensional equivalence alone does not suffice for successful definition. As Socrates’
discussion of Euthyphro’s third answer indicates Socrates is also looking for what grounds
something’s being pious. All and only pious things may be loved by all the gods, but
being loved by all the gods fails to ground something’s being pious. Something like this
is indicated by Socrates’ summary following his discussion of Euthyphro’s third answer.

I’m afraid, Euthyphro, that when you were asked what piety is, you did not wish to
make its nature (ousia) clear to me, but you told me . . . [a] quality (pathos) of it, that
the pious has the quality of being loved by all the gods, but you have not yet told me
what the pious is. (Euthyphro 11a5–b1)

This sort of answer is sometimes referred to as a real definition, as opposed to a nominal

definition.38 They are the sort of answers that H2O provides in response to the question
“What is water?” or that a general seeks when asking what courage is.39 They are not
questions about the meaning of the word “piety”. They are questions about the thing that
“piety” names.
The Euthyphro also indicates why Socrates is interested in such questions: knowledge
of what piety is is prior to knowledge that something is pious.40 For example, one does
not know that Euthyphro’s prosecution of his father is pious if one fails to know what

35 See the Ion, Hippias Minor, and Crito.

36 Socrates’ use of eidos at 6d11 and cognates of idea at 6e1 and 6e4 (and even paradeigmati at 6e6)
has suggested to some an early version of a Platonic theory of Forms. See, e.g., (Allen 1970) and
(Prior 2004). I do not pursue this issue here except to say that a substantive answer to “what is piety?”
according to which piety is the form Piety and so x’s being pious is grounded in x’s participating in (or
whatever the relation is) Piety (see Phaedo 100c9–d8) would indicate the beginnings of a theory of
Forms. But if Socrates (or Plato) endorses such a substantive answer in the Euthyphro it is exceedingly
underdeveloped.
37 This is not Socrates’ objection to Euthyphro’s fifth answer. Unfortunately, Plato does not depict
Euthyphro offering an answer that fails to be a sufficient condition. In other dialogues he does. See
Laches 192b9–c1.
38 See, e.g., (Vlastos 1965, 156 n. 26) and (Fine 1992:202). 39 (Penner 1973:40).
40 He also believes that knowledge of what piety is is prior to knowledge that piety is a virtue or that
piety is good, but I do not pursue that here. Evidence for such a Socratic commitment is absent in our
three dialogues.
130   Hugh H. Benson

piety is. Indeed, it is Socrates’ commitment to this priority that leads to the definitional
journey undertaken in the Euthyphro. When Socrates points out that Euthyphro must
think that he knows that prosecuting his father is pious (4e4–8), Euthyphro does not
demur. Rather, he professes to know all such things (5a2), referring back to Socrates’
remark that he must think that he knows how things hold concerning the gods and
pious and impious things (4e4–6). Socrates investigates whether Euthyphro has the
knowledge he professes to have. Socrates’ own disavowal of this knowledge is indicated
by his seeking to become Euthyphro’s pupil. But when Socrates turns to his examination
of Euthyphro, he does so by asking him to say what piety is, which Socrates claims
Euthyphro just now maintained he clearly knew (5c8–d1). This, however, is not what
Euthyphro professed to know. Rather he professed to know “all such things” “how things
hold concerning the gods and pious and impious things.” It is the commitment to the
priority of definitional knowledge that permits Socrates’ gloss. Euthyphro does not
know “all such things,” if he fails to know what piety is.
A commitment to the priority of definitional knowledge, however, is not necessary to
explain Socrates’ gloss. Socrates instead may be committed to the view that definitional
knowledge is required only to know in general which things are pious (“all such things”).
But, if Socrates is to conclude, as surely he does, that Euthyphro fails to know that pros-
ecuting his father for murder is pious rather than that Euthyphro fails to know in general
which things are pious, something stronger is required. I do not mean that Socrates
concludes that Euthyphro recognizes his failure. He does not, as we will see in a moment.
Rather Socrates comes to see (as do Plato’s readers) that Euthyphro fails to know that his
prosecution is pious whether Euthyphro recognizes it or not. But all that Socrates has
shown is Euthyphro’s inability to say what piety is and thereby presumably his failure to
know what it is. Socrates must at least be committed to thinking that failure to know
what piety is suffices for the failure to know whether Euthyphro’s prosecution is pious.
Perhaps this is because Euthyphro’s prosecution is particularly controversial. In this
case, a different commitment would suffice; for example, knowledge of what piety is is
necessary to know controversial instances of piety. But this commitment is at odds with
Socrates’ disparagement of the beliefs of the many, which he discusses explicitly in the
Crito. Recall that Euthyphro’s relatives believe that it is impious for Euthyphro to prose-
cute his father in this case. Euthyphro’s care for virtue and wisdom motivates his need to
know whether the prosecution is pious or impious. Relying on the beliefs of his relatives
will not suffice for this knowledge. Rather, this knowledge requires knowledge of what
piety is, which Euthyphro fails to have.41
Socrates’ commitment to the priority of definitional knowledge fits well with his
disavowal of knowledge in the Apology. It is often objected to the priority of definitional
knowledge that it is obviously false. As Peter Geach put it “We know heaps of things

41 For a longer defense of Socrates’ commitment to the priority of definitional knowledge, see
(Prior 1998) and (Benson 2000: chap 6). Matthews, Ch. 16 of this volume, provides an aporetic reading
of some of the passages offered as evidence for Socrates’ commitment to the priority of definitional
knowledge.
 Euthyphro, Apology, and Crito   131

without being able to define the terms in which we express our knowledge.”42 Besides
misconstruing Socratic definitions as nominal rather than real, Geach’s suggestion that
we know heaps of things is not something that the Socrates of the Apology would
endorse—at least if the knowledge supposed here is high-level knowledge of fine and
good things. But this is the only kind of knowledge Socrates cares about. Euthyphro
cares about doing the pious thing—so much so that he risks the anger of his relatives.
If so, then, according to Socrates, Euthyphro cares (or should care) about high-level
knowledge concerning whether what he is doing is pious. Socrates tests this knowledge
by asking him to say what piety is, that is, to say what grounds a thing’s being pious at
least in a way consistent with the other things Euthyphro (and perhaps Socrates)
believes.43 Euthyphro’s inability to do so displays his failure to know what piety is. The
ability to provide such an answer may not suffice for definitional knowledge, but if
Socrates is to conclude that Euthyphro lacks such knowledge, it must be required. Thus,
Euthyphro lacks (high-level) knowledge of what the pious thing to do in this circum-
stance is, without realizing it—evidently like everyone else Socrates has met. As a result
Euthyphro, despite his profession, fails to care about wisdom and truth in the way that
he should.

What Should Euthyphro Do?

Following Socrates’ objection to Euthyphro’s sixth attempt to answer the “What is piety?”
question, he once again encourages Euthyphro to teach him what piety is. Euthyphro,
however, has had enough. He offers to teach Socrates another time. But now he has an
appointment and rushes off. The dialogue concludes with Socrates expressing his disap-
pointment in not having learned what piety is from Euthyphro.
Euthyphro fails to recognize his lack of knowledge. Earlier, following the discussion
of the third answer, he professes only to be unable to say what he thinks (or knows)
(11b6–7). At the conclusion of the dialogue he does not maintain that he is unable to
teach Socrates because he fails to have the relevant knowledge himself. He suggests
instead that he will teach Socrates another time. But suppose, contrary to fact, that
Euthyphro did recognize his ignorance of the nature of piety as a result of the examina-
tion in the Euthyphro. What then should Euthyphro do? To a certain extent the answer is
clear. He should make logoi concerning piety and the other things Socrates discusses,
and examine himself and others to see who is wise and who merely thinks he is wise but
is not, in order to acquire the knowledge of piety he now recognizes he lacks so that he
can do the pious thing with respect to prosecuting his father and not risk acting badly.
He should, that is, live the examined and virtuous life. But what, we might ask, should

42 (Geach 1966:371). See also (Brickhouse and Smith 1994:51 n.34).

43 I hear allude to the debate between constructivist and non-constructivist interpretations of
Socrates’ arguments (his elenchoi). See, for example, Matthews, Ch. 16 of this volume, for a fair
discussion of this debate.
132   Hugh H. Benson

he do in the meantime—prior to acquiring the requisite knowledge of piety—about

prosecuting his father?44 If Socrates is any indication, the knowledge Euthyphro lacks is
a long way off, and yet Euthyphro needs to act now. He must either continue his prose-
cution or drop it. Because of his lack of knowledge, he risks acting impiously in either
case. If his family is right, then he will be acting impiously by prosecuting his father. If he
is right, he will be acting impiously if he does not. His counterfactually recognized lack
of knowledge makes him unable to know (in a high-level way) what to do. He must decide
without such knowledge. How is such a decision to be made? The Crito suggests an answer.

4. The Crito

The Crito takes place in jail where Socrates awaits his execution following the guilty
verdict of the Apology. Crito arrives, having heard that the ship whose arrival betokens
the pending execution has been observed off the coast. Socrates has just awoken from a
pleasant dream, which he understands as indicating that his execution draws near. Crito
explains that he has arranged things to facilitate Socrates’ escape and he attempts to per-
suade Socrates to do so.

Crito’s Advice
Crito offers primarily two considerations on behalf of escape. First, if Socrates does
not escape, he and his friends will appear cowardly and evil to others, especially to the
many (44b5–c5 and 45d8–46a2). Second, in helping Socrates to escape, Crito and his
friends will be acting rightly (45a1–3), and in refusing to escape Socrates will be acting
wrongly (45c6–7). Crito implores Socrates to escape based on these considerations.
Time is running out.
Socrates’ response to Crito is immediate:

We must . . . examine whether we should act in this way or not, as not only now but
at all times I am the kind of man who listens only to the logos that on reflection
seems best to me. (Crito 46b3–6)

He explains that he cannot abandon the logoi he has made in the past just because of
what has happened to him now, and that unless better logoi can be provided he will be
persuaded only by his former logoi no matter the consequences. Socrates is referring
here to the making of logoi and the examination of himself and others that he takes to be
the greatest good, which he practiced in attempting to understand the Delphic oracle,

44 For importantly different answers to this question see (McPherran 2002), (Benson 2013), and
(Smith 2013).
 Euthyphro, Apology, and Crito   133

and which Plato displays in the Euthyphro. These logoi and the examinations of himself
and others—the examined life—will determine whether he will escape from jail or remain.

Socrates’ Examination of Previous Logoi

The examination that follows falls into two parts: an examination of the logos that some
beliefs ought to be respected and others not (46d1–2), and an examination of the logos
that it is not living but living well that matters most (48b4–5). In the first examination
Socrates considers the case of the athlete and maintains that the athlete does not respect
everyone’s beliefs, but rather only the beliefs of the one who knows—the doctor or the
physical trainer. She does this because respecting the beliefs of the many who have no
relevant knowledge will harm (kakon)45 and destroy her body. Socrates generalizes to
other cases, especially those concerning right and wrong, shame and honor, good and
bad. We ought to especially respect the beliefs of those who know about these latter
things. This is because while life is not worth living when the body is destroyed by the
bad advice of the ignorant-many concerning health and disease, life is even less worth
living when the soul is destroyed by the bad advice of the ignorant-many concerning right,
and the rest.46 Socrates concludes that Crito did not begin correctly when he maintained
that we ought to respect the beliefs of the many concerning right and wrong, etc. Rather,
the logos made in the past, that we ought to respect the beliefs of some (the one or ones
who know) and not the beliefs of others (the ignorant-many) remains (48b2–4).
The second examination begins with Crito agreeing that the following two logoi
also remain: that it is living well, not living, that is most important (48b4–5) and that living
well and living rightly are the same (48b7–8). Socrates offers no argument here for these
logoi. He simply gets Crito’s agreement that they remain. Nevertheless, he concludes
from this agreement that they must only examine whether it is right or not to escape
(48b10–d6), since it follows that we must live rightly. After highlighting that there are no
exceptions to this third logos that one must live rightly, Socrates expands it to include
never returning wrong for wrong (49b9–c11).47 Next, Socrates confirms the centrality of
these logoi, or as he now calls them “agreements,” and points out that few hold them in
common and that for those who do not common deliberation is absent (49c11–d5).
The three questions that follow look like the beginnings of a practical deliberation.
First, Socrates asks “ought a man to do what he has agreed to do, provided it is right, or

45 I here render kakon as “harm” as it almost certainly must be in this passage. Nevertheless, a bit
later Socrates identifies kakôs poiein or kakourgein with adikein (49c7–8). If this is rendered as “doing
harm is the same as doing wrong” Socrates may be committed to thinking that doing physical damage
of any sort (even in time of war for example) is wrong. See, e.g., (Kraut 1984, 26). The appropriate
rendering of kakon and its cognates in the Crito is important but I cannot pursue it here.
46 The argument here assumes that the body stands to health and disease as the soul stands to right
and wrong—the former benefitting and preserving, the latter harming and destroying. See note 45 of
this chapter.
47 After identifying doing harm and doing wrong, Socrates includes never returning harm for harm
in this passage and thereafter. I set aside this complication for brevity. See note 45 of this chapter.
134   Hugh H. Benson

may he violate his agreements?” (49e6–7). Crito responds that one must do what one
has agreed to do. Socrates has just been reminding Crito that they have agreed in the
past, and the current circumstances do not alter that agreement, that one must never do
wrong even in return for wrong. Such an agreement can never require one to do some-
thing wrong. So, they must keep their agreement:

(C1) Never act wrongly even in return for wrong.48

This is, as it were, the major premise of a practical deliberation. Next, Socrates asks two
questions: “If we go away from here without the consent of the state, are we doing harm
to the very ones to whom we least ought to do harm, or not?” (49e9–50a2). And, “Are we
abiding by what we agreed was right, or not?” (50a2–3). A positive answer to the second
question will determine Socrates’ decision, in light of the positive answer to the first
question, and the third question makes this clear. But at this crucial juncture Crito
demurs. He responds that he cannot answer because he does not know (50a4–5).
Notice the similarity of this point in the Crito with the imagined counterfactual situa-
tion at the end of the Euthyphro. Euthyphro needs to decide about prosecuting his father
in light of his counterfactually recognized lack of knowledge of the nature of piety on the
basis of which he had originally decided to act. Here in the Crito Socrates needs to
decide about escaping from jail while recognizing his lack of knowledge of the nature
of virtue.49 He listens to Crito’s advice to escape and then he proposes to examine his
previously made logoi, because he is the sort of person who always listens to the logos
that appears best on reflection. He examines these logoi to determine whether, having
been made in the past, they remain in the present circumstances of his pending execu-
tion. He is less concerned to argue on behalf of these logoi, than he is to establish that
they remain for both himself and for Crito. They have been made before. Socrates does
not maintain that the stability of these logoi is sufficient for knowledge concerning how
to decide,50 but he does suggest it helps in decision-making.
The examined life—the previous making of logoi and the examination of them in
the present circumstances—provides the general principles of practical deliberation.
Nevertheless, the problem of completing the deliberation remains. Examining previous
logoi to see if they remain in the specific circumstances may provide stable general
principles on the basis of which to make such a decision, but it leaves unaddressed how
those general principles apply to the specific decision at hand. Does prosecuting one’s
father for murder in the specific circumstances described in the Euthyphro fall under a
general prohibition to never act impiously even in return for impiety? Does escaping
from jail in Socrates’ circumstances fall under the general prohibition to never act

48 For this reading of these questions, especially the first see (Lane 1998:315 & 321–22).
49 See note 51 of this chapter.
50 See the conclusion of the Crito where Socrates appears open to further counterarguments from
Crito. For the necessity of stability for knowledge see the analogy of the statues of Daedalus at
Euthyphro 11b6–e1 and Meno 97d6–98b6.
 Euthyphro, Apology, and Crito   135

wrongly, even in return for wrong? Answers to these questions supply the minor
premises of the practical deliberations, but Crito’s response to Socrates’ questions at
50a4–5 indicates that Crito, at least now, does not profess to know the answers. Socrates
cannot take Crito’s advice in these circumstances, as he might take the advice of the
doctor or physical trainer in circumstances concerning care for the body. Crito’s exper-
tise in the present circumstances has been found wanting. But Socrates lacks the relevant
knowledge as well.51 How, then, is Socrates to answer these questions and decide whether
to escape?

The Speech of the Laws

The Euthyphro suggests that one way to begin is by asking what rightness and wrongness
are. An answer to this question will help in, if not suffice for, recognizing which of the
actions available to one is right and which wrong.52 But this is not how the Crito proceeds.
Rather, Socrates follows Crito’s disavowal of knowledge by terminating his dialogue
with Crito. He offers instead a prolonged speech in the name of the Laws of Athens.
The Speech of the Laws falls into three waves.53 The first two waves provide the following
argument for the minor premise of the practical deliberation initiated at 49e6–50a3.

C2 It is wrong to disobey the laws of Athens, because to do so either would be to

attempt to destroy them (first wave—50a6–51c5) or to break a voluntary agreement
to obey them (second wave—51c6–53a8).
C3 Socrates’ escaping is disobeying the laws of Athens.
C4 So, Socrates’ escaping is acting wrongly, even though the laws may have
wronged Socrates.

C3 looks trivial. The Laws aim to establish C2. We will return to it in a moment. In the
third wave, the Laws maintain that escaping would wrong Socrates, his friends, and his
children and so again he should not escape (53a9–54d2). The last wave avoids the poten-
tial of conflicting deliberations. Even if the first two waves successfully established that
in escaping Socrates acts wrongly, if Crito is right that not escaping wrongs Socrates’
friends and his children, Socrates would face a moral dilemma. Whatever he does will be
acting wrongly. The third wave eliminates that potential.
But, do the first two waves successfully establish that in escaping Socrates acts wrongly?
Only if C2 and C3 are true. But C2 looks problematic. Not only does it look morally
unattractive—maintaining, when combined with the major premise (C1), that one should

51 See Socrates’ general disavowal of knowledge as well as his specific disavowal of knowledge of the
right (to dikaion) at Republic 354c1. Socrates’ disavowal of knowledge, however, is not prominent in the
Crito. He does not explicitly disavow knowledge anywhere in the dialogue, nor, however, does he profess it.
52 That Socrates may be committed to the sufficiency of definitional knowledge is suggested at
Euthyphro 6e3–6; see (Benson 2000:ch. 7).
53 See (Harte 1999:137 n. 35).
136   Hugh H. Benson

never disobey the law no matter what it demands,—but it also looks incompatible with
Socrates’ other commitments. First, it appears incompatible with C1 itself. According to
C1 one should never act wrongly, and yet it would seem possible for a law to require that
one act wrongly. Consider, for example, the Fugitive Slave Act enacted in 1850, which
required that U.S. citizens return escaped slaves to their slaveholders. But C2 also looks
incompatible with various descriptions of Socrates’ own behavior in the Apology. In at
least three instances Socrates describes his virtuous behavior in terms that suggest a
willingness to disobey the Laws of Athens. In testifying to his unwillingness to yield to
anyone contrary to what is right for fear of death, Socrates reminds the jurors of his
unwillingness to obey their attempt to try together the 10 generals at the battle of
Arginusae and again his unwillingness to obey the Thirty when they ordered him to get
Leon of Salamis (32a4–e1). A bit earlier he explains to the jurors that if they offer to set
him free on the condition that he cease philosophizing, he will obey the god rather than
them and continue to philosophize (29d3–5). None of these cases are explicitly incom-
patible with C2. Socrates points out that to try the 10 generals together was illegal, as
the Athenians came to recognize later, and disobeying the Thirty (perhaps an illegally
imposed authority as a result of defeat by the Spartans) is not obviously disobeying
the Laws of Athens. Finally, the jurors have no legal authority to order him to cease
philosophizing54 and so his refusal to obey such a demand would again not be disobedience
to the Laws of Athens. Consequently, these cases may not be explicitly incompatible
with C2. Nevertheless, Socrates’ refusal to obey the jurors’ hypothetical demand may
indicate that he would have disobeyed them had the demand been legal. Whether that is
so or not, the first incompatibility remains.
Various solutions have been proposed to attempt to mitigate Socrates’ commitment
to such a strict adherence to the rule of law.55 For example, some have maintained that
Socrates’ claim that one should either persuade or obey the laws at, for example, 51e7–52a3,
commits him only to obeying those laws that one has not attempted to change through
persuasion.56 Others have defended Socrates’ commitment to a strict adherence to the
law by distinguishing between blameworthy and non-blameworthy wrongdoing. Thus,
according to Socrates, the mid-nineteenth century New Yorker who returns a runaway
slave to her owner would be acting wrongly, but not in a blameworthy way given
the enactment of the Fugitive Slave Act.57 Still others have found the difficulties with the
Speech of the Laws too difficult to overcome. Socrates has attempted to persuade the
Laws (or their representatives on the jury) that he is innocent of the charges and that

54 The jurors have no legal authority to impose a penalty on Socrates if they do not find him guilty
of Meletus’s charges, and according to Athenian law, if they do find him guilty, the jurors can only
impose a penalty suggested either by the prosecution or the defendant. But neither Meletus nor
Socrates propose the penalty of ceasing to philosophize.
55 For a nice discussion of many of these solutions see (Bostock 1990).
56 See (Kraut 1984:54–90).
57 See (Brickhouse and Smith 2004:212–41). Their view is much more complex (as is Kraut’s) than
I here have space to present, and they have responses to my rushed objections. In fact, I believe that
(Brickhouse and Smith 2004) offer the most thorough and plausible defense of Socrates’ commitment
to the Speech of the Laws
 Euthyphro, Apology, and Crito   137

he should not be executed, and yet he is obliged to obey their verdict. And, the distinction
between blameworthy and non-blameworthy wrongdoing appears to lack textual
support and is at odds with the apparently blanket refusal to do wrong in C1. Moreover,
Socrates’ allusion at the end of the dialogue to the noise associated with the Corybantic
rites interfering with his ability to hear anything other than the Laws suggests that
Socrates does not endorse their argument for C2.58 But then we are left without an
argument for C2. Harte, for example, suggests that Socrates’ decision to remain in jail
derives instead from his dream at the beginning of the dialogue and indicates that
Socrates’ decision is based not on the practical deliberation suggested at 49e6–50a3, but
on obedience to the god.59 But this is difficult to square with Socrates’ statement that his
decision will be based only on the logos that seems best to him (46b3–6), as well as on his
commitment to the examined life. Deciding on the basis of a dream hardly looks like
deciding on the basis of the logos that seems best on reflection.
One might wonder whether taking Socrates as committed to the argument of the
Laws offers any comfort to this worry. The Speech is hardly representative of the sort of
logoi-making Socrates usually engages in. It does not depict Socrates making logoi and
examining himself and others to see who is wise and who thinks he is but is not. Indeed,
if Socrates ever was examining Crito to determine whether Crito was wise or thought
he was but was not, that examination has been terminated after Crito disavows his
knowledge at 50a4–5. Crito responds to a Socratic question only three times in the next
four-and-a-half pages. Nor is Socrates examining the reputed wisdom of the Laws. The
Laws address their questions to Socrates, not the other way round, and even so, Socrates
only responds to their questions two or three times (50d5, 50e1, and maybe 52a6). Nor
does the Speech depict an examination of logoi made in the past. This is as it should be.
Logoi on behalf of minor practical premises such as C4, at least, will not be the sorts of
things one would or should examine previously. Each case will be specific to its circum-
stances and will require a new logos.
In the Crito, then, Socrates is depicted as deciding what to do when no one with the
relevant knowledge is available whose advice can be relied on—neither himself nor the
individual he is examining. Crito explicitly professes his lack of such knowledge, and
Socrates nowhere in the Crito professes to have it. Nevertheless, Socrates must decide
what it is right to do—escape or remain, and if he gets it wrong he will harm that part
of him he most values. In such circumstances, he relies on his own defeasible rational
reflection. He lives the examined life—making logoi and examining himself and others
to determine who is wise and who merely thinks he is but is not. In this way he identifies
stable, though defeasible, general principles on the basis of which he should act. When
it comes to applying these general principles to the specific circumstances at hand,
Socrates’ decision procedure is less clear. The Crito does not depict Socrates engaging in
his customary method of question and answer to examine the professed wisdom of

58 See (Weiss 1998:134–45) and (Harte 1999).

59 See (Harte 1999:144). (Weiss 1998:72–81) maintains that C2 has been established prior to the
Speech, but I am skeptical.
138   Hugh H. Benson

others in order to decide what to do. Rather, he relates a speech of the Laws. This speech
may depict a different method of defeasible rational reflection, one that does not rely on
the examination of others. Alternatively, we may need to look elsewhere for how one
should apply defeasible general principles to the circumstances at hand. In either case,
defeasible rational reflection looks to be what Socrates recommends.

5. Conclusion

The three dialogues in which Plato depicts the weeks leading up to Socrates’ last day
are replete with various philosophical explorations. Among those explorations is the
question of how to live our lives. On the one hand, Socrates is clear and straightforward.
We should live the examined life—making logoi and examining ourselves and others in
order to determine whether we are as wise as we think we are, and we should live the
virtuous life. This is how Socrates lives his life. He makes logoi examining himself and
others discovering that everyone else thinks he is wise in ways he is not. In disavowing
knowledge of anything fine and good, only Socrates fails to think he knows things he
does not. He also lives, or at least seeks to live, the virtuous life. His life reflects an
­unwavering sense of right and wrong, at one point going so far as to assert that he knows
it is wrong to disobey a superior while at the same time reminding us of his disavowal of
knowledge. On the other hand, the examined life undercuts, or at least should undercut,
the confidence with which he seeks to live the virtuous life. It may help bring some sta-
bility to the general principles by which he lives his life, but it can do so only defeasibly
and without certainty. Moreover, the application of these general principles to the
specific circumstances at hand is even more uncertain given their specificity. Lacking
knowledge, we risk harming our souls, in the same way that in lacking knowledge of
health and disease we risk harming our bodies. The risk is real, yet perhaps unavoidable.
This risk is depicted in these dialogues as well. It is confronted, if at all, in the character
of Socrates. It is Socrates’ attempt to successfully integrate both the examined life and the
virtuous life, however he manages it, that so attracted, and yet puzzled, Plato and others.60

Bibliography
Allen, R. E. (1970) Plato’s Euthyphro and the Earlier Theory of Forms (New York: Humanities
Press).
Austin, Scott (1987) “The Paradox of Socratic Ignorance (How to Know That You Don’t Know),”
Philosophical Topics 15, 23–34.

60 Thanks to Ray Elugardo and Gail Fine for helpful comments and/or discussions. I also wish to
thank the National Endowment for the Humanities for a fellowship during a portion of which this
essay was composed.
 Euthyphro, Apology, and Crito   139

Benson, Hugh H. (2000) Socratic Wisdom: The Model of Knowledge in Plato’s Early Dialogues
(New York: Oxford University Press).
Benson, Hugh H. (2013) “What Should Euthyphro Do?” History of Philosophy Quarterly 30,
115–46.
Bostock, David (1990) “The Interpretation of Plato’s Crito,” Phronesis 35, 1–20.
Brickhouse, Thomas C., and Smith, Nicholas D. (1983) “The Origin of Socrates’ Mission.”
Journal of the History of Ideas 44, 657–66.
Brickhouse, Thomas C., and Smith, Nicholas D. (1994) Plato’s Socrates. New York: Oxford
University Press.
Brickhouse, Thomas C., and Smith, Nicholas D. (2004) Routledge Philosophy Guidebook to
Plato and the Trial of Socrates (New York and London: Routledge).
Brickhouse, Thomas C., and Smith, Nicholas D. (2010) Socratic Moral Psychology (Cambridge:
Cambridge University Press).
Chalmers, David J., Manley, David, and Wasserman, Ryan, eds. (2009) Metametaphysics: New
Essays on the Foundations of Ontology (Oxford: Oxford University Press).
Charles, David, ed. (2010) Definition in Greek Philosophy (Oxford: Oxford University Press).
Cooper, John M., and Hutchinson, D. S., eds. (1997) Complete Works of Plato (Indianapolis,
IN: Hackett).
Correia, Fabrice, and Schnieder, Benjamin, eds. (2012) Metaphysical Grounding: Understanding
the Structure of Reality (Cambridge: Cambridge University Press).
Doyle, James (2004) “Socrates and the Oracle,” Ancient Philosophy 24, 19–36.
Evans, Matthew (2012) “Lessons from Euthyphro 10A–11B,” Oxford Studies in Ancient Philosophy
42, 1–38.
Everson, Stephen, ed. (1990) Ancient Greek Epistemology (Cambridge: Cambridge University
Press).
Fine, Gail (1992) “Inquiry in the Meno,” in Kraut 1992, 200–26.
Fine, Gail (2008) “Does Socrates Claim to Know That He Knows Nothing?,” Oxford Studies in
Ancient Philosophy 35, 49–88.
Forster, Michael N. (2006) “Socrates’ Demand for Definitions,” Oxford Studies in Ancient
Philosophy 31, 1–47.
Geach, P. T. (1966) “Plato’s Euthyphro: An Analysis and Commentary,” Monist 50, 369–82.
Hale, Bob, and Hoffmann, Aviv, eds. (2010) Modality: Metaphysics, Logic, and Epistemology
(Oxford: Oxford University Press).
Harte, Verity (1999) “Conflicting Values in Plato’s Crito,” Archiv Für Geschichte der Philosophie
81, 117–47.
Irwin, T. H. (1977) Plato’s Moral Theory (Oxford: Oxford University Press).
Judson, Lindsay (2010) “Carried Away in the Euthyphro,” in Charles 2010, 31–61.
Kraut, Richard (1984) Socrates and the State (Princeton, NJ: Princeton University Press).
Kraut, Richard, ed. (1992) The Cambridge Companion to Plato (Cambridge: Cambridge
University Press).
Lane, Melissa (1998) “Argument and Agreement in Plato’s Crito,” History of Political Thought
19, 313–30.
McPherran, M. L. (1985) “Socratic Piety in the Euthyphro,” Journal of the History of Philosophy
23, 283–309.
McPherran, M. L. (2002) “Justice and Pollution in the Euthyphro,” Apeiron 35, 105–29.
Penner, T. (1973) “The Unity of Virtue,” Philosophical Review 82, 35–68.
Prior, William J. (1998) “Plato and the ‘Socratic Fallacy,’” Phronesis 43, 97–113.
140   Hugh H. Benson

Prior, William J. (2004) “Socrates Metaphysician,” Oxford Studies in Ancient Philosophy 27, 1–14.
Rosen, Gideon. (2010) “Metaphysical Dependence: Grounding and Reduction,” in Hale and
Hoffmann 2010, 109–36.
Schaffer, Jonathan (2009) “On What Grounds What,” in Chalmers, Manley, and Wasserman
2009, 347–83.
Smith, Nicholas D. (2013) “Socrates on Practical Deliberation,” History of Philosophy Quarterly
30, 93–113.
Taylor, C. C. W., Hare, R. M., and Barnes, Jonathan, eds. (1999) Greek Philosophers: Socrates,
Plato, Aristotle (Oxford: Oxford University Press).
Vlastos, G. (1965) “Anamnesis in the Meno,” Dialogue 4, 143–67.
Vlastos, G. (1985) “Socrates’ Disavowal of Knowledge,” Philosophical Quarterly 35, 1–31.
Weiss, Roslyn (1998) Socrates Dissatisfied: An Analysis of Plato’s Crito (Oxford: Oxford
University Press).
Woodruff, P. (1990) “Plato’s Early Theory of Knowledge,” in Everson 1990, 60–84.
Woolf, Raphael (2008) “Socratic Authority,” Archiv Für Geschichte der Philosophie 90, 1–38.
 chapter 6

The Protag or as
A n d G orgi as

James Warren

Anyway, when we knocked at the door, he opened it, took one look at us.
“Ugh! Sophists!” he said.
Protagoras 314d (trans. T. Griffith)

1. Reading the Protagoras

and the Gorgias as a Pair

It is not difficult to see why the Protagoras and Gorgias are often considered as a pair.1 In
both, Socrates takes on a famous intellectual visiting Athens from out of town and in
both Socrates and this famous visitor clash on important issues of philosophical method
as well as philosophical substance. In both dialogues the interlocutors are unlike the
group of relatively compliant friends who participate in dialogues such as the Phaedo
and the bulk of the Republic. Instead, here Socrates locks horns with some of the most
famous of his intellectual contemporaries who have arrived in Athens in part to advertise
their intellectual skills.2 And Socrates’ method, especially his preference for the brief

1 The two dialogues appear together in the sixth tetralogy of Thrasyllus’s ordering of Plato’s works.
They certainly differ in various ways. For example: the Gorgias is presented as a direct dialogue while
the Protagoras has a brief frame dialogue (309a–310a) before the remainder is taken up by Socrates
recounting to his friend the story of his discussion with the great sophist earlier that day.
2 See the roll call of famous names at Protagoras 314e–316b and note the list of well-connected
Athenians who have come to see these visitors: Protagoras 314e–315a. At Protagoras 317c Socrates
suspects that Protagoras is keen to show-off to the assembled crowd. At the beginning of Gorgias it is
clear that Gorgias has just finished giving some kind of public demonstration (epideixis) of his skill that
has also drawn quite a crowd: Gorgias 447a.
142   James Warren

question-and-answer style of discussion, is itself subjected to disagreement by those

who prefer to be allowed to make extended speeches in favor of a particular view and as
a way of displaying intellectual and rhetorical skill. Protagoras offers the following
contrast: “Your strength, Socrates, is asking the right question. Mine is that I enjoy giving an
answer to people who ask the right question” (Protagoras 314d), and Socrates too is often
keen to distinguish his method from that of the orators and sophists (e.g., Gorg.
471e–472c) But that contrast is not always evident: for example, Socrates is no less adept
than Protagoras at providing a demonstration of skill in creative poetic exegesis
(Protagoras 338e–348e) and at the end of the Gorgias (523a–527a) he offers an extended
eschatological myth that is certainly the rival of Protagoras’s story of the origins of
justice and teachability of virtue (Protagoras 320c–328d).3
There are other important themes that appear in both works. In both, Socrates is
interested in exploring what makes a good human life and what counts as human virtue
(aretē). Consideration of different views of a good human life prompts thoughts about
the ways in which we may be mistaken about what is good for us and may therefore fail
to desire what we ought, in the sense of failing to desire what would in fact be best for us.
Closely connected with this danger is the associated risk of being persuaded by clever
speakers or by our fellow citizens to choose or desire things it would be better for us not to
desire; this is why the participants in both dialogues think about power, in particular the
power of rhetoric and the ability to persuade, and the general danger of misleading
appearances of goodness. Rhetorical persuasion and the mere appearance of goodness
are then contrasted with genuine understanding of and an ability to discriminate what is
in fact good. This, in turn, leads the discussion back to the initial questions about how such
an understanding may be acquired and whether (and by whom) virtue can be taught.4
These similarities encourage speculations about the position of these two dialogues in
the development of Plato’s thought. And, more specifically, there is a long-running set of
questions concerning the relative chronology of the Protagoras and Gorgias. This latter
dispute turns on a set of particular interpretative problems in making sense of each
dialogue, particularly the moral psychology outlined in each dialogue, and whether it
is possible to determine in either case a set of commitments on Socrates’ part such that
differences emerge between the works. In particular, a lot of effort has been expended
on two related points of interpretation. First, the Protagoras ends with a discussion in
which Socrates and Protagoras settle on using pleasure and pain to calculate the value
of actions or objects of pursuit in such a way that some commentators are convinced
that Socrates approves of this hedonist axiology. In the Gorgias, however, Socrates offers
a number of arguments against Callicles’s view that a life is good insofar as it involves

3 See Long 2013, 26–45 on the complicated nature of Socrates’ conversations with the sophists and,
in particular, the choice between long speeches and short question-and-answer discussion. Gorgias
boasts that he is equally skilled at giving long epideixeis and speaking concisely (Gorgias 449b–c; cf.
Protagoras 335a).
4 Many of these themes are also discussed in Daniel Devereux’s contribution to this volume,
chapter 17, esp. Sections 2–4.
 The Protagoras And Gorgias   143

satisfying great desires: a claim Socrates takes to be tantamount to a profession of

­hedonism (495d).
Second, that same argument in the Protagoras is aimed at discrediting the idea that it
is possible to think some course of action better than a rival but choose the apparently
worse option nevertheless. And it does so by casting all such choices as evaluative calcu-
lations and insisting that we always pursue what appears to us to be the better option.
This is generally taken to be an intellectualist account of human motivation. The Gorgias,
however, contains remarks that some have taken to suggest that, according to Socrates, a
human soul is complex and we can act not only on the basis of what we reason to be best
but also on the basis of non-rational good-independent desires. So there is an apparent
disagreement between the two dialogues on two counts: on the relationship of pleasure
to the good life and on the complexity or simplicity of our motivations. It is then, of
course, a further question whether any such differences, should there be any, do indeed
point to changes in Plato’s views and therefore point to a change over time in his position.
And such interpretative decisions themselves turn on more general views about the
relationship between the character of Socrates in the dialogues, Plato’s own views, and
the nature of the use of the dialogue form as a means of writing philosophical works.5
I will turn to look at both of these important passages later on in this chapter. But it is
worth noting at the outset the way in which the treatment of these passages has affected
how these two dialogues have been viewed, both individually and as a pair. It is certainly
worth asking, for example, whether Socrates is committed to a kind of hedonism in
Protagoras or whether in Gorgias he is committed to a view that posits more than one part
or aspect of an individual’s soul. But these questions are hard to disentangle from
whatever prior thoughts we have about the other dialogue in the pair. If we think that
Socrates does adopt hedonism in the Protagoras, what should we conclude from his
fierce criticism of Callicles’s life of pleasure and desire-satisfaction in the Gorgias? Is it
more likely that Plato became more and more dissatisfied with hedonism and therefore
the Gorgias must be the later dialogue? If we think that in the Gorgias Socrates has a view
that posits more than one part or aspect of an individual’s soul, including non-rational
and good-independent desires, then how should we account for his argument in the
Protagoras against the possibility of thinking one possible object of choice better but
choosing one that is thought worse? But if, on other grounds, we think the Protagoras is
the later of the two, should that affect our interpretation of Socrates’ commitments in
the argument against akrasia? There is a danger that such discussions become circular,
with preconceptions about the likelihood of changes of mind across Plato’s career deter-
mining the view we take on the chronological order of his works.6

5 The majority opinion is that the Protagoras predates the Gorgias. But the reverse order has been
championed by Kahn 1981 and is followed by Schofield in Griffith and Schofield 2010, viii: “I think the
Gorgias is the work of an angry young man, the Protagoras the product of more detached middle age.”
6 Some examples: Dorion 2012, 48 n. 69 sees a complex moral psychology in the Gorgias, which
recognizes non-rational and good-independent desires as evidence for it being later than the Protagoras
and closer to the Republic. Kahn thinks similarly about the psychology of the Gorgias but has it predate
Protagoras; he therefore argues that Plato does not endorse the “omnipotent rationalism” of the
144   James Warren

In part, these interpretative difficulties arise from a characteristic of both dialogues,

which I noted previously. Socrates is engaged in serious and substantial discussions of
important ethical and psychological questions with people who begin those discussions
with views that are often radically at odds with what he himself finds plausible and are
often recalcitrant if not genuinely hostile to him and to his philosophical method.
Callicles, for example, threatens not to respect the usual rule of answering Socrates’
questions sincerely when he foresees how the exchange is designed to expose an incon-
sistency in his position (Gorgias 495a–c). And at Gorgias 506c–507c, Socrates famously
takes on both the roles of questioner and answerer, constructing a dialogue within the
dialogue, because Callicles has for the moment decided not to participate.7
Unlike the “definitional” dialogues that consist principally in the examination of some
interlocutor’s opinions, in the Gorgias and Protagoras Socrates is perhaps more prepared
to express his preference for a certain view but, unlike in dialogues such as the Phaedo
or Republic, he does not have the opportunity or the audience that would allow him to
secure his interlocutor’s agreement with the psychological and metaphysical foundations
necessary for a full defense of his position. Instead, we might say that in both dialogues
in terms of his metaphysical and psychological commitments he is traveling as light as
possible, as he often does when he can rely only on what he thinks his interlocutor might
accept. It is therefore hardly unexpected that Socrates’ own views are not set out in as
much clarity as some would like. In fact, as the ongoing disagreements over fundamental
questions of interpretation show, it is extremely difficult to pin down precisely what
Socrates’ own commitments are in either of these dialogues.

2. Rhetoric and Teaching Virtue

Socrates clashes with Protagoras and Gorgias in various ways but they are all three
engaged in the project of assisting people to live excellent lives. That shared aim gener-
ates friction between them because of their varied understandings of what an excellent
life is and of how people are best taught to live it. They differ from one another in other
important ways too; in particular, Protagoras charges for his services (Protagoras
310d–e, 328b–c).
Protagoras says that he can teach someone good judgment (euboulia) in both private
and public affairs. He can teach people how to manage their own households well and,
when it comes to politics, how to be “most powerful” (dunatōtatos) in action and in

Protagoras (Kahn 1996, 232; cf. Kahn 2003). Compare Russell 2005, 239–40, who offers as a
consideration against the interpretation that Socrates endorses hedonism in the Protagoras the fact that
we would then have to posit a substantial change of mind on Plato’s part when he comes to write the
Gorgias. Irwin 1995 argues that in the Protagoras Socrates does endorse “epistemological hedonism”
(85–92) and that the Gorgias therefore “expresses Plato’s further thoughts about hedonism” (113–14).
7 Doyle 2006, 87–88 notes that the first words in the Gorgias are “war and conflict” (polemou kai
makhēs), setting the tone for the discussion to come.
 The Protagoras And Gorgias   145

speech (Protagoras 318e–319a). It is not clear whether these are two aspects of one and
the same skill but it is likely that they are: the city is regularly thought of as a household
writ large, and the same skill at management and ruling will be applicable in both arenas;
in any case the political aspect is what drives the remainder of the discussion. As his long
mythological narrative makes clear, Protagoras thinks that all adult humans have learned
justice since otherwise there could be no political communities at all. He thinks that
justice is indeed teachable and is generally learned in various more or less informal ways
from a person’s parents and teachers, poets and musicians, fellow citizens, and the city’s
laws and modes of punishment (324d–328c).8 Nevertheless, there is still room for a
sophist such as him since he offers to hone and improve this general and common skill—
for a price—and to teach some to excel at it.9
Gorgias too offers lessons to his clients designed to ensure that they are able to wield
political influence through the power of rhetoric. He calls himself an orator and prom-
ises to teach his clients also to excel in the art of rhetoric (Gorgias 449a–d). Polus takes
up his cause by claiming that orators wield the greatest power in the city since through
their powers of persuasion they are able to exile or kill any rival they want (466b–d). This
ability to acquire influence and manipulate a populace through the power of persuasion
also seems to lie behind Polus’s and Callicles’s admiration for the life of a tyrant who is
able to allow his desires to be unchecked because of his unassailable power and position
in a city. It also explains why these dialogues turn quickly from the discussion of rhetoric
to psychological questions, since they are interested both in the psychology of the pow-
erful figures the interlocutors admire and also in the mechanism by which rhetoric
can manipulate its audience. Gorgias himself, however, soon becomes entangled by
Socrates’ questioning when pressed to say something more about what the art of rheto-
ric is and, in particular, what its specific sphere of expertise might be. Socrates argues
that it cannot be the same as the art of medicine or physical training or indeed many
other arts even though they will all involve speaking and communication to some
degree. Eventually, Gorgias admits that the art of rhetoric is principally the art of
persuasion, particularly political persuasion (452e). But the persuasion it produces is
not the same as is generated by expertise in a certain field as, for example, an expert
mathematician can teach a pupil by using his knowledge of the subject. Instead, Gorgias
agrees that his form of persuasion does not teach but instead produces “conviction”
(454e–455a). The very general applicability of this art seems to disqualify it from claiming
any particular sphere of expertise beyond a certain expertise in generating conviction in
an audience.
Perhaps the promise of acquiring this ability might still appeal, particularly in a cul-
ture such as Athens in which a premium was placed on political prestige and democratic
acclaim. But Socrates next turns to undermine even the claims to being a skill (teknē) of

8 At 324a–c Protagoras maintains that the function of punishment is educational and forward-
looking. In this respect, he and the Socrates of the Gorgias are in agreement.
9 See Denyer 2013.
146   James Warren

this promised ability to generate conviction.10 Socrates reserves the term tekhnē for
those arts that aim at some good, as medicine and gymnastics aim at the good of the
body and legislation and justice aim at the good of the soul. They are contrasted with
mere “knacks” (empeiriai) which instead aim only to please. Here cookery is the coun-
terpart to medicine since it aims only at bodily pleasure and not at the body’s good, and
fashion is the counterpart to gymnastics, offering the mere pleasing semblance of
physical beauty (464b–465d). It is no surprise that Socrates categorizes rhetoric also as
an empeiria, aimed at pleasing an audience through sycophancy rather than truly
benefiting it; “it relates to the soul as cookery does to the body” (465d). Polus takes this
to be an outrageous claim and takes over as the next principal interlocutor.
The participants in these two dialogues share the assumption that we should all strive
to make our lives as good as possible. They disagree, however, on what a good life
consists in and therefore how it might be achieved. For Gorgias and Protagoras, a good
life is one that is best attained by acquiring certain skills that will allow a person to wield
and maintain political power and prestige through oratory. And these are precisely the
skills the professional sophists offer to teach, for a price.
In both dialogues it becomes clear that in order to understand properly how it might
be possible to teach people to live a good life it is necessary to understand more about
what a good life is, what people desire, and how people may be mistaken in desiring
things that will not in fact make their lives good. Socrates’ distinction in the Gorgias
between merely pandering to what is pleasant for the body or soul and producing genu-
ine benefit suggests that there may well be things we like and enjoy—and in that sense
may think to be good—but which are not genuinely beneficial. This allows Socrates to
claim that an orator works only with what the audience thinks is good since he is trying
to win its favor, and it is not part of the orator’s skill (or “knack”) to know what is genu-
inely good. Both dialogues then go on to explore the relationships between virtue, the
good life, and the psychology of desire.

3. Desire and the Good 1:

The Denial of Akrasia in
the Protagoras

At Protagoras 329b–d Socrates asks Protagoras to give him a detailed and clear account
of virtue (aretē): Is it just one thing and its parts are wisdom, temperance, courage, justice
and piety? Or are these all just names for one and the same thing? Protagoras prefers the
first alternative. Then Socrates asks about these “parts.” Are they like the various compo-
nents that make a human face or are they like bits of gold, not differing from one another

10 At this point Polus steps in on Gorgias’s behalf, accusing Socrates of being overzealous in his
search for an inconsistency in Gorgias’ position (461b–462e).
 The Protagoras And Gorgias   147

or from the whole? This matters because Socrates is interested in whether it might be
possible to acquire one virtue without the others or whether they all come as a package.
Protagoras then finds it difficult to articulate how the various virtues are different
from but nevertheless like one another and also how their opposites may be related,
and it becomes clear to Socrates that Protagoras himself is dissatisfied with the
answers he has been giving and also with the general method of their inquiry
(334c–335d).11 At this point the dialogue almost comes to a halt but the intervention of
some of the bystanders suggests a new approach and the discussion resumes based
around the critical exegesis of the view of human goodness expressed in a passage of
Simonides (338e–349a).12 When the discussion eventually turns back to the question of
the various virtues, Protagoras has changed his position and now asserts that all of the
virtues but courage are names of one thing. At 349d he claims that courage is different
from the rest: there can be courageous but unjust people; indeed some people are
extremely courageous but impious and altogether wicked in other ways.13
Socrates tries first to trap Protagoras in a brief conversation that attempts to make
courage and confidence equivalent (349e–351a). But Protagoras notes that he wants to
say only that courageous people are confident, not that confident people are courageous.
Nevertheless, by the end of this section Protagoras has stated that he takes courage to be
attained by nature and the proper nurture of the soul. Confidence, on the other hand,
comes from skill or emotion or madness. At this point (351a) the discussion takes an
abrupt turn. Socrates begins to ask about the characterization of a good life. Protagoras
agrees that a painful life is not good and that someone who has completed his life
pleasantly has lived a good life, and by 351c they have asserted that to live pleasantly is
good and unpleasantly is bad.14
Protagoras then adds a caveat: someone who lives pleasantly, he says, lives a good life
provided that “the things which give him pleasure in his life are admirable (kala)” (351c).
Socrates tries to force Protagoras to agree that even things he thinks are not admirable
are good if they are pleasant, but Protagoras insists that they should look into it further
before he will accept that position (351e). The use of pleasure as a marker of goodness in
the subsequent discussion is principally justified by a reference to it being what “the
many” take to be good, and it can therefore be used as a placeholder for any simple crite-
rion of value and as a useful concession in a dialectical argument against what the many
think about our desires. Neither Socrates nor Protagoras wholeheartedly claims the

11 On the unity of virtue in the Protagoras see: Vlastos 1972, Penner 1973, Vlastos 1981, O’Brien 2003,
and Manuwald 2005; and for the fortunes of the thesis in other dialogues, including the Gorgias, see
Sedley 2014.
12 At 325e–326a Protagoras had included poetic works as sources of important ethical education and
advice. On the discussion of Simonides’s poem see Trivigno 2013, Woodruff 2017.
13 See Laches 198d–199e for an argument that discards a proposed definition of courage on the
grounds that it picks out not one part (morion) of virtue but rather virtue as a whole (sumpasa aretē).
14 Note that the discussion is at the level of a completed life, not of an episodic evaluation of one’s state.
This is important because the dimension of time and its role in pleasure will be crucial in what follows.
148   James Warren

view for their own and they both voice doubts about it early on.15 Indeed, 352a–b shows
that what Socrates and Protagoras are really interested in is not whether pleasure is the
good but, rather, the role of knowledge in securing a good life. They are for the moment
allies in thinking that knowledge is the most powerful thing in a person’s life: a position
on which it is reasonable to think they will agree given that each of them is committed to
the possibility of improving a person’s life through some kind of intellectual education.
But “the many” disagree and say that although knowledge is often in charge it can on
occasion be overridden especially by pleasure or pain (352b). This would be a problem
for Socrates and Protagoras because it holds out the possibility of there being someone
who is perfectly well educated and able to determine what is good but who nevertheless
is not living a good life because of obstructing psychological forces beyond his or her
control. Socrates and Protagoras agree instead that knowledge is the most powerful and
can never be “dragged about” like a slave (352c–d).
From 353b Socrates and Protagoras aim to remove this objection by revealing an
inconsistency in the views of the many. They claim that the many cannot hold both that
what makes a life good is the pleasure it contains and that it is possible to act contrary to
what one thinks at the time is the better course of action because of being “overcome by
pleasure.” This second phenomenon is what is often termed akrasia: a lack of rule, in the
sense that some choice to pursue a particular course of action fails to exercise its normal
authority because the agent is overwhelmed by some other motivation or desire. “The
many” had earlier cited various possible causes of such phenomena: lust, anger, and so
on, but these have now been narrowed down to pleasure alone (352b, 352e; the reference
to factors other than pleasure is dropped by 353a). The first objection to the many turns
simply on this admission: if pleasure is the good then it is absurd to say that someone
can fail to secure what is good because of being overcome by pleasure since that would
amount to saying that someone has been overcome by what is good (355c–e).
This prompts the many to clarify that they mean that a person may opt for a lesser but
temporally closer pleasure over a greater but remote one, or may avoid a pain but in
doing so miss out on a greater later pleasure. Now their proposal is trying to explain
failures to maximize the good over time. This allows Socrates to offer his alternative
account of what in fact happens in those cases the many think should be identified as
instances of akrasia. Socrates’ explanation begins by building on the distinction between
immediate and more remote pleasures and pains and then between pursuing something
because of its immediate character and because of its later consequences. He establishes

15 Similarly, Callicles in the Gorgias does not begin the relevant discussions by clearly endorsing
hedonism. In both cases Socrates directs the conversation so that it narrows the terms by which things
are considered good or bad until pleasure and pain are the relevant bearers of value. Callicles says that
there are shameful pleasures that should be avoided (Gorgias 499b–e, cf. 495a–b). Later in the
Protagoras at 358a–b Socrates turns to ask Protagoras directly whether he accepts the claim that what is
pleasant is good and that actions aimed at living pleasantly are fine. It remains unclear, however,
whether Socrates is merely insisting that he and Protagoras must agree that this is a necessary
consequence of the views of the many they have been investigating. See Taylor 1991, 201–02, 208–10
and compare Taylor 2003.
 The Protagoras And Gorgias   149

that sometimes nearer pleasures are forgone and nearer pains are endured in order to
secure more remote but greater pleasures; in these cases the decision is taken in favor of
what is thought to be the action that produces more pleasure and less pain overall.
On this model, pleasures and pains are commensurable; that is the import of the
“weighing” metaphor at 356a–c.: we weigh pleasures against pleasures and pains against
pains insofar as we evaluate comparatively the pleasures and pains involved in various
possible courses of actions.16 We should note, however, that there is very little interest
shown here in precisely what pleasure and pain are. Instead, we are left to infer what
notion of the nature of pleasure the discussion relies upon from various clues here and
there in the text. Socrates needs no such elaboration since he is entitled to rely on what-
ever “the many” take pleasure to be; after all, they are the ones who have suggested that it
is possible to be overcome by pleasure, and it is that proposal which is being scrutinized.
Similarly, Socrates has no need to be committed to the idea that pleasure and pain are
quantifiable, only that we can compare different pleasures and see overall which course
of action is better in hedonic terms. He does appear to support, however, the possibility
of making reasonable assessments of the likely consequences of our immediate and
longer-term objects of pursuit such that we can make relatively confident predictions
about which course of action is the better by comparing all available choices against a
single scale. He does not assume, however, that we are good at making those predictions
and that we are skilled at assessing and comparing the relative values of possible courses of
action. Indeed, he will diagnose what the many take to be cases of akrasia as instances in
which we fail to assess those things properly. But he does set out a general framework for
thinking about our decision-making, casting it in two phases: first, an estimation of the
values of different outcomes according to some criterion of value (here: pleasure and
pain) and, second, a comparison of those outcomes so as to produce a winner. He thinks
this framework is sufficiently plausible that we can assume the many will also accept it.
Clearly, the cases of akrasia offered by the many now make sense only if we take into
account the immediate and longer-term consequences; they are bad because overall the
action generates less good even if it does involve some short-time pleasant experience.
But in that case, these people are not really being “overcome by pleasure” when they act
in this way. What is happening instead? Here Socrates uses an analogy. Just as some-
times we mistakenly think a nearer but smaller object is larger than a more distant but
larger object, so too we sometimes mistakenly think a temporally nearer but smaller
pleasure is larger than a more distant but larger pleasure. The different distances—one
spatial, the other temporal—sometimes make it difficult successfully to assess compara-
tively two objects or two pleasures. Nearer ones sometimes seem larger than they should
in comparison with more distant ones. But just as we can learn to compensate for the
difference caused by spatial distance so too we can become more skilled at making the
correct evaluations of pleasures and pains, even if some are closer than others. We need
to be more skilled at judging the comparative value of goods, and in principle there is no
obstacle to us getting it right. Socrates terms the necessary skill the “art of measurement”

16 Compare: Richardson 1990; Warren 2014, 105–19.

150   James Warren

(tekhnē metrētikē) and this is something we can acquire and practise. Armed with this
skill we can pursue with maximal success the goal of living the most pleasant life possible.17
This account of how we do and how we should make prudential choices and how we
are sometimes misled in our calculations has the important consequence that what the
many call akrasia is nothing of the sort. The people who are said to be “overcome by
pleasure” are simply making an incorrect judgement about the better course of action.
They are all trying to maximize the good in their lives but do not perform the correct
comparative evaluation and fail to take into account the distorting effects of temporal
proximity. The further upshot is that people’s lives are improved by improving their
beliefs and intellect, and this is something congenial to both Protagoras and Socrates
and their shared commitment to the supreme importance of knowledge (357e). In order
to maintain the claim that knowledge is sufficient for virtue (352c), they need to show
that it is impossible for someone to know what is good but fail to act accordingly. And in
order to show that akrasia is not possible, Socrates and Protagoras have adopted an
account of human motivation that has borrowed hedonism as a plausible axiological
position and has diagnosed cases of apparent akrasia as instances of mistaken compara-
tive evaluation.
The overall importance of this section of the dialogue is to show that Protagoras
cannot maintain that courage stands apart from the other virtues unless he is prepared
to weaken his own claims to be able to teach people in a way that will ensure that they
live good lives. Protagoras’s popular audience—his clientele—are both attracted to
­hedonism and to the possibility of akrasia; Protagoras can maintain his professional
claim to be a teacher of virtue by accepting the first and denying the second.18
The other thing to remember is that Socrates has replaced the hypothesis of akrasia
with the claim that there is often a distinction to be drawn between what seems to us to
be better and worse—what course of action appears to be choiceworthy—and what is
in fact better and worse for us. We all are trying to make sure our lives are as good as
possible—we all want to live a good life—but we are often misled by what seems better
but in fact is not, and end up choosing a worse option thinking it the better.

4. Desire and the Good 2:

Socrates and Polus in the Gorgias

On this view, Socrates in the Protagoras ends the dialogue by outlining an intellectualist
view of the soul and of human desire and motivation. That is, while he allows that various
things may “appear” to us to be good that are not in fact so, our desires are guided by
what we take to be good and there are no good-independent non-rational desires that

17 Cf. Moss 2009. 18 Cf. Russell 2005, 244–48; compare Shaw 2015.
 The Protagoras And Gorgias   151

may conflict with our evaluations and cause cases of akrasia. In the case of the Gorgias,
however, things are not so clear.
Earlier in the dialogue, in the discussion with Polus, Socrates offers an argument the
conclusion of which is consonant with the upshot of the last section of the Protagoras.
Polus takes over the discussion at 461b, noting that Gorgias is embarrassed to agree with
Socrates that rhetoric is not a skill (tekhnē).19 He and Socrates begin by disagreeing over
the role of orators in a city; Polus insists—and insists that this is the common view of the
matter—that orators wield the greatest power in the city since they are able to “put to
death anyone they will, confiscate property, and banish from their cities anyone they please”
(466b–c), while Socrates insists that in fact orators have the least power in the city.
Note that Polus uses two different expressions here. He says both that orators have the
most power (megiston dunantai 466b4) and that orators can persuade people to do
whatever they, the orators, see fit (ho an boulōntai 466c1). Socrates instead insists that
“orators and tyrants have the least power in cities” (smikrotaton dunantai). “They do
virtually nothing of what they will (boulontai)—though they do as they please (ho an
doxēi beltiston)” (466d8–e2). Socrates evidently wants to make a distinction between
true power that comes from doing what you want (for which he reserves the verb “to
will/want” boulesthai, the verb which Polus himself used in his original formulation)
and merely doing what seems best (dokei beltiston) to you. Orators are able to do only
what seems best.
In what follows Socrates tries to explain his view, basing it on an analysis of a general
method of reasoning. Sometimes when a person wants (and Socrates again tends to use
boulesthai throughout this section) X they want X not for X itself but because of Y; in
that case, Y is that for the sake of which we want X. Socrates adds to this model a distinc-
tion between things that are good, bad, or neither good nor bad (467e). He and Polus
agree that health, wealth, and wisdom are good and their opposites are bad. Some things
are neither good nor bad. Walking is neither good nor bad but when we walk we do so
“in pursuit of the good” and “because we think it is better” (oiomenoi beltion einai 468b2),
and we will think it is better if we think that it promotes or is in some other way a means
to something that is good, such as health.
Now Socrates turns to their disagreement over the power of an orator. If the preced-
ing general model for understanding intentional action is correct then when an orator
puts someone to death, this must also be because he thinks that this is for some good.
And this is where Socrates thinks he can show that Polus must be mistaken (468c). The
argument here turns on Socrates reminding Polus that they have agreed that we want
(boulometha) only what is good (468c5–7).
Next, Socrates asks about the case of an orator or tyrant who is successful in having an
enemy put to death. The orator takes that action supposing that to be better (oiomenos
beltion) for him, but in fact it turns out to be worse for him. In that case, Socrates agrees
that the orator does what seems (dokei) best but, given that the outcome is worse and we

19 When Callicles takes over as the principal interlocutor he notes that the same thing has happened
to Polus as Polus said had happened to Gorgias: 482c–e.
152   James Warren

want only what is good, then the orator cannot want the outcome and also cannot in fact
want to put the person to death given that it promotes this detrimental outcome.
There are various concerns we might have about this argument.20 Evidently, Socrates
at the end of the discussion is using the terms good and bad, better and worse to qualify
the actual outcome for the subject—putting the rival to death was in fact worse for the
orator—in a way that allows them to differ from how the subject originally conceived of
the plan. The outcome seems better to the orator but is in fact worse. And this restricts
what we can actually want (boulesthai) to those things that are in fact good. On Socrates’
model, merely thinking or believing that something is good is not enough for us to want
it. But this is not likely to have been what Polus understood by the important claim on
which Socrates relies, namely that we want only what is good; that seemed to be justified
by a very general thought about the structure of our choices and desires that, roughly
speaking, amounted to the idea that a necessary condition of wanting something is
merely thinking it good, regardless of whether it is in fact good. Socrates has restricted
“being powerful” to “getting what one wants” and, in turn, has restricted “wanting” to
thinking good something that is in fact good.
Socrates is successful in refuting Polus only to the extent that Polus is also attracted to
the idea that there is a distinction to be made between what we think good and what is in
fact good, and is also inclined to restrict what we want to the latter rather than the
­former. After all, there is something amiss with the orator who puts a rival to death
thinking this will be beneficial but who is in fact harmed as a result; it matters whether
our conceptions of what is good match reality.
One important thing to note is that there is nothing in this stretch of argument to
suggest that either Socrates or Polus is working with a model of desire that includes
non-rational good-independent desires. It seems instead to be very close to the discussion
at the end of the Protagoras and the distinction between what appears good and what is
good with the accompanying claim that we fail to live good lives because we make
mistaken rational choices and not because we are assailed by non-rational good-
independent desires. Throughout the discussion, even those desires that are for things
that turn out to be in fact bad or detrimental are described in terms according to which
the object seems (dokei) good or preferable or the subject thinks (oietai) that the object
is good or preferable. That is to say, some reference to apparent goodness is part of the
account of why the object is desired. The crucial point once again is that it is possible to
be mistaken and to desire something that seems but is not in fact good.21
Socrates then returns to the claim that it is an ability to kill or exile rivals that is the
marker of the excellence of an orator’s life and insists that in fact acting unjustly is harm-
ful and it is much preferable to suffer than to commit an unjust act (469b–c; cf. 470e).
This completes a trio of criticisms of these great orators: they do not have any genuine

20 See Vlastos 1991, 148–54.

21 This is also a cousin of an argument at Meno 77b–78b according to which Meno is wrong to state
that virtue is “wanting good things and being able to secure them” in part because “wanting good
things” does not distinguish the virtuous from the non-virtuous; everyone wants what they want
thinking it good. They may nevertheless be wrong about what is in fact good.
 The Protagoras And Gorgias   153

tekhnē, they are unable to do what they will, and if they do what seems best to them and
commit an injustice then this contributes to their being wretched. Polus now says he is
astonished. He imagines someone who could commit injustice with impunity: a perfect
tyrant—someone who is beyond punishment and reproach, who is sufficiently powerful
to be confident that he will not be overthrown. They consider the example of Archelaus,
tyrant of Macedonia, who usurped his brother’s rightful throne (471a). Polus agrees that
this person is unjust but finds it hard to believe that he is therefore miserable, especially
on the assumption that he is never held to account or punished for his transgressions.
Surely it is better to be this person than someone subject to his injustices.
Socrates offers the following argument (474c–475e). Polus thinks that suffering an
injustice is worse than committing an injustice and also that committing an injustice is
more shameful (aiskhion) than suffering injustice. He admits that what is admirable
(kalon) is not the same as what is good (agathon), and what is shameful (aiskhron) is not
the same as what is bad (kakon). Socrates then gets him to accept that things are admi-
rable (kala) because they are useful or produce pleasure and therefore if something is
more shameful then it must be because it is less useful or produces less pleasure. But
Polus agreed both that committing an injustice is more shameful than suffering an injus-
tice and also that it is certainly not more painful.22 So it seems he must concede that
unjust actions are less useful than just actions. Perhaps Polus ought to object that he has
conceded only that committing injustice is less useful than suffering it and that does not
contradict his starting claim that it is better to commit an injustice than to suffer it. But
then he would have to spell out some sense in which something might be more shameful
insofar as it is less useful but nevertheless not worse. So Archelaus’s unjust actions are
shameful and less useful but somehow still better than just actions. But this is clearly an
unstable position.
Now Polus has accepted that it is better to suffer than to commit an injustice, Socrates
asks whether it is better to be punished for one’s unjust acts. If the punishment is just and
admirable then it is good, indeed good for the recipient of the punishment. So Polus
should agree also that people who commit an injustice should recognize that it will be
better for them to be punished for it than to get away scot-free (476a–477a). And now
his picture of the admirable life of an unrestrained tyrant subject to no law or sanction
has been thoroughly undermined. Socrates returns to the analogy between the health of
the body and the health of the soul that he first used when discussing the “knack” of
rhetoric with Gorgias. The health of the body is something good and worth choosing,
and Polus also agrees that there is something analogous in the case of the soul and,
crucially, at 477b that the soul, like the body, can be corrupted. Corruptions of the body
include disease, ugliness, and weakness; corruptions of the soul include injustice, igno-
rance, and cowardice. It is more shameful to have a corrupt soul than a corrupt body.

22 According to Vlastos 1967, Polus should not have conceded this point since he has agreed only to
the idea that things that are admirable or beautiful (kala) are so to those who behold them (475d8–9)
either because of pleasure or usefulness. Polus can agree that acting unjustly is not more painful than
suffering injustice for the unjust person (475b8–c3) although it may be more painful for others to see
the unjust prosper than the just suffer. For a critical appraisal of Vlastos’s view and a review of some
other reactions to it see Berman 1991.
154   James Warren

So injustice is particularly shameful, and it is all the more important for any such
­corruption in the soul to be cured.23 Anyone who wishes to promote the health of his own
soul, should he commit an injustice, will willingly submit to punishment.24 (Socrates
only later draws out as a further implication the famous “Socratic paradox” that, since
no one wishes to live a miserable life, no one will do wrong willingly (509e).) Polus has
very rapidly moved away from an estimation of the value of a life based on wealth,
power, influence, and the like and is now brought to accept Socrates’ alternative radical
new estimation of what is of value, which depends entirely on the health of the soul.
Evidently, Polus’s initial concession that an unjust action is shameful causes him some
difficulty. The acceptance that acting unjustly might be shameful, even if it is ultimately
preferable to suffering an injustice, shows that Polus has not entirely abandoned various
evaluative attitudes that are critical of these tyrannical strong men. He has a certain
admiration for their power and perhaps envies their lifestyle but he remains of the view
that their unjust actions are nevertheless shameful and certainly more shameful that
suffering an unjust act at another’s hands. In fact, shame plays an important role more
than once in the dialogue. Both Polus and Callicles appeal to a vision of a strong and
powerful ruler as someone with various admirable qualities. Callicles’s ideal person is
powerful, brave, and wise (492a2: he has andreia—literally, “manliness”—and phronēsis)
and he thinks that such a person is criticized by the many only because they are ashamed
of their weakness. Callicles may think that he is remedying Polus’s mistake by not
subscribing to the views of the many, who brand the tyrant’s life as one of shameful
wantonness (akolasia). But Callicles too is eventually forced to see that his picture of a life
dedicated to fulfilling unrestrained desires may also conflict with what he is prepared to
accept as seemly when Socrates presents him with the life of someone dedicated to ful-
filling great desires to scratch an itch or to being a passive partner in male homosexual
sex (494c–d).25 Callicles fancies himself as a brave and manly person (Socrates calls him
andreios at 494d4) and evidently wants his picture of an ideal person to be similarly
manly. And he finds Socrates’ examples shameful (494e7).26 So both Polus and Callicles,
for all their anti-conventionalist posturing, are still wedded to conventional concerns
about shame and seemliness for the lives of their ideal unrestrained agents, and Socrates
is quick to exploit the argumentative potential in uncovering this inconsistency.27

23 Compare Crito 47d–48a.

24 See Mackenzie 1981, 179–91. Socrates elaborates on the benefits of punishment for the soul in the
myth that ends the dialogue. See Sedley 2009.
25 The exchange at 495a–c shows that in order to escape an inconsistency Callicles is prepared to
agree that even shameful pleasures are good. His own pride overcomes his initial distaste for this
consequence of his position.
26 Callicles also finds it difficult to accept that cowards may experience just as much pleasure as
brave people and therefore, on a hedonist criterion, live equally good lives (498a–499b).
27 On the appeals to notions of shame in the Gorgias see Moss 2005, esp. 139–53, 160–69;
Tarnopolsky 2010, 56–88.
 The Protagoras And Gorgias   155

5. Desire and the Good 3:

Socrates and Callicles
in the Gorgias

In the conversation with Polus, at least, there is no clear sign of significant differences
between the treatments of desire in the Gorgias and the Protagoras. However, there are
also various passages in the Gorgias that have been taken to suggest that Socrates here
has a more complex conception of the soul. On this view, Socrates envisages multiple
parts of the soul and, in particular, a distinction between non-rational good-independent
appetites or desires and our beliefs and desires about what is good. Later in the dialogue,
in the discussion with Callicles, Socrates refers to psychic order (taxis) and arrangement
(kosmos 504d) and he advocates self-control and restraining appetites. Perhaps most
interesting is his reference at 493a3–4 to: “this part of the soul where desires are” (tēs
psukhēs touto en hōi epithumiai eisi). These have been taken as indications that Socrates
has a view of a complex soul, perhaps similar to the picture familiar from the psychology
of dialogues such as the Republic.28
We should first remember the context of these remarks. At this stage of the dialogue
Callicles has begun to outline a position that approves of strong agents acting in an unre-
strained manner in order to satisfy their desires. It takes some time for Callicles to offer a
clear account of the sort of person he has in mind but it becomes clear eventually that he
is thinking of a political leader who understands the affairs of the city and is also coura-
geous, effective, and strong (491a–b); this is the sort of person who should be in charge
and is the sort of person we should all aspire to be. It is Socrates who introduces the
question whether such a person ought also to rule himself and, when asked to explain
what that might mean, reaches for what he takes to be the popular understanding of that
notion (491d–e): “I mean what most people mean, being moderate, his own master,
ruling the pleasures and desires within himself.” Callicles rejects that and instead
expands on his preference: “The person who is going to live in the right way should allow
his own desires to be as great as possible, without restraining them. And when they are
as great as can be, he should be capable of using his courage and understanding in their
service and giving them full measure of whatever it is, on any particular occasion, his
desire is for” (492a).

28 The translation is by Griffith. The Greek does not express explicitly a word for “part”; Irwin 1979,
67 offers the more literal “that of our souls with appetites in it.” See Dorion 2012, 41, 43: “Although in
the Gorgias Plato never explicitly asserts a bipartition of the soul into reason and desire, one can
conclude nevertheless (in the light of 491d and 493a–b) that Plato envisages a bipartition of this sort.”
This is a “genuine innovation” (47). Irwin 1979, 195 ad loc.: “[T]he Greek does not show whether
Socrates thinks of parts, or, more generally of aspects of the soul. . . . Though Socrates does not say so,
these desires seem to be liable to conflict with other desires. . . . This recognition of ‘good independent’
desires is incompatible with the Socratic paradox.” Cf. Irwin 1995, 116–17.
156   James Warren

This is the background of the discussion that begins at 492d with Socrates wondering
whether it is better to be constantly filling leaking jars with a sieve—like the mythologi-
cal punishment of the daughters of Danaus—or to have a soul that is more like a small
and watertight jar that is easy to satisfy and retains whatever fills it. We have now reached
the point at which Socrates refers to “the part of the soul where the desires are” (493a),
within the account he ascribes to a clever or ingenious (kompsos) storyteller. This is the
part that is imagined to be either leaking and insatiable or watertight and easily satisfied.
The ascription of this account to a storyteller need not itself suggest that Socrates is not
attracted to this conception of the soul but it certainly shows that this is not a detailed
and robust account that Socrates is prepared on this occasion personally to defend in a
way that will secure his interlocutor’s reasoned commitment. All that is needed is an
account that Callicles is prepared to consider as plausible for the purposes of further
scrutinizing his claim about the nature of the best life.29 It does suggest that the soul can
be thought of as a complex item, since there is an implicit contrast between the desires
that are filled and the means of filling them. But it does not require that the soul must
include non-rational good-independent desires any more than did the earlier conversa-
tion with Polus.
It is certainly true that Socrates himself later asserts that the self-controlled life is the
one we should choose to live, and we should pursue that end by not allowing our desires
to become undisciplined (akolastous 507d); punishment and discipline are useful means
to that end. But this too shows relatively little about Socrates’ own preferred under-
standing of the nature of an individual soul. Just a little earlier he and Callicles had been
discussing the need for design (taxis) and order (kosmos): a good house that shows taxis
and kosmos and the establishment of those qualities is the aim of a builder. Similarly
physical health arises from bodily order, and psychic health and excellence (virtue) will
arise from psychic taxis and kosmos. For the body, taxis and kosmos may just mean a
proper arrangement of the various elements that make up a body, whatever they may be.
(Socrates reprises this argument in discussion with himself at 506cff.) A house is certainly
a complex thing insofar as it is made of various parts in a certain arrangement. But here
taxis need mean no more than that: it is well-put-together and orderly.
An orderly soul may similarly be a well-put-together soul, not in the sense that it has
set in order various competing and discrete motivational sources but, for example, in the
sense of having a consistent and harmonious set of desires: a position that is perfectly
compatible with the intellectualist idea that desire are, at base, all judgments of a certain
kind. In other words, Socrates’ promotion of psychic taxis and kosmos is still neutral
between various different ideas about the nature of the soul (and he and Callicles may
well have differing assumptions about that topic).30 The argument here merely requires

29 Cf. Cooper 1999, 63: “This device [the citing of an unnamed person’s view] serves clearly to
distance Socrates from at least the details of what he brings into the discussion by its means.” Socrates
often uses such devices (e.g. Meno 81a–b) and they are best understood on a case-by-case basis
dependent on the particular dialectical situation and the characters present in the dialogue in question.
30 Cf. Rowe 2007, 153–56; Cooper 1999, 65. Compare Carone 2004, who resists both strongly
intellectualist and strongly “irrationalist” readings of the dialogue.
 The Protagoras And Gorgias   157

that there be some kind of psychological complexity such that it is appropriate to talk
about psychological order, and that requirement may be satisfied in variety of ways. For
example, since Socrates and Callicles go on to talk about how best to arrange one’s desires,
the taxis and kosmos of a soul may just be an arrangement and ordering of desires, a
claim that is also compatible with a wide range of views about what desires are and whether
the human soul contains desires of different kinds or desires from different parts.31
There is a temptation to think there must be the possibility of conflict between
psychic parts, and therefore a commitment to good-independent desires, since Socrates
connects psychic order with being sōphrōn, an adjective standardly translated as being
“self-controlled” and that therefore may imply some kind of relationship between an
ordering and restraining part and an unruly but restrained part of the soul. But the
adjective sōphrōn and the associated noun sōphrosunē need not carry that implication.
Here Socrates simply describes such a person as one who neither pursues nor avoids
things when it is not appropriate but rather avoids and pursues what he should and
stands his ground when he must (507b).
To be sure, Socrates does say that he endorses this view (507c) but what he endorses is
compatible with a range of claims about the precise nature of the soul and the details of
moral psychology. Socrates does not need to go any further into those topics and, given
the potentially hostile interlocutor with whom he is dealing, it makes good sense for his
commitments to remain as economical as possible.

6. Conclusions

The Protagoras and Gorgias, whatever the relative chronology, both present conversa-
tions marked by skilful characterization and show Plato’s ability to encourage serious
philosophical reflection through the interplay of vividly drawn participants. The two
dialogues also share a feature that is surely the origin of many of the frustrations felt by
various interpreters. In both cases, Socrates is determined to dismiss certain views of
human virtue and the value of sophistic education and also to insist on the importance
of rational order in a person’s life. But, unlike in dialogues such as the Phaedo and Republic,
he does so as far as possible without setting out and then relying on detailed and substan-
tial claims of his own about the nature and structure of a human soul and the nature of
ethical knowledge. There are evident signs of what his own preferences are, of course,
and he is not averse to letting his interlocutors know which views he wants to carry
the day. Indeed, he makes plain what he thinks is true and takes it as confirmation of the
truth of these views that even Callicles, for example, can come to agree with his conclu-
sions (e.g., Gorgias 507c–509c).32 But neither Gorgias nor Protagoras, neither Polus nor

31 For a discussion of the role of Socrates’ elenctic method in attempting to generate various kinds of
psychic harmony and order, see Woolf 2000.
32 See Long 2013, 48–51.
158   James Warren

Callicles, are prepared to grant for the sake of these arguments the robust psychological
and metaphysical premises Socrates elaborates elsewhere. They are much less friendly to
his outlook than Simmias and Cebes or Glaucon and Adeimantus. Socrates nevertheless
tries to tackle his interlocutors’ positions by relying as far as possible only on psychological
assumptions they too are willing to accept, and such an approach has its limitations.33

Bibliography
Berman, S. (1991) “How Polus Was Refuted: Reconsidering Plato’s Gorgias 474c–475c,” Ancient
Philosophy 11: 265–84
Carone, G. R. (2004) “Calculating Machines or Leaky Jars? The Moral Psychology of Plato’s
Gorgias” Oxford Studies in Ancient Philosophy 26: 55–96
Cooper, J. (1999) “Socrates and Plato in Plato’s Gorgias” in his Reason and Emotion: Essays in
Ancient Moral Psychology and Ethical Theory, Princeton, NJ: 29–75
Denyer, N. (2013) “The Political Skill of Protagoras” in V. Harte and M. Lane eds. Politeia in
Greek and Roman Philosophy, Cambridge: 155–67
Dorion, L.–A. (2012) “Enkrateia and the Partition of the Soul in the Gorgias” in R. Barney,
T. Brennan, and C. Brittain eds. Plato and the Divided Self, Cambridge: 33–52
Doyle, J., (2006) “The Fundamental Conflict in Plato’s Gorgias,” Oxford Studies in Ancient
Philosophy 29: 87–100
Doyle, J. (2010) “Socrates and Gorgias” Phronesis 55: 1–25
Gosling, J. C. B. and Taylor, C. C. W. (1992) The Greeks on Pleasure, Oxford: Oxford University
Press
Griffith, T. (trans.) and Schofield, M. (ed.) (2010) Plato: Gorgias, Menexenus, Protagoras,
Cambridge
Irwin, T. (1979) Plato: Gorgias, Oxford
Irwin, T. (1995) Plato’s Ethics Oxford
Kahn, C. (1981) “Did Plato Write Socratic Dialogues?” The Classical Quarterly 31: 305–20
Kahn, C. (1996) Plato and the Socratic Dialogue, Cambridge
Kahn, C. (2003) “Socrates and Hedonism” in A. Havlíček and F. Karfík eds. Plato’s Protagoras:
Proceeding of the Third Symposium Platonicum Pragense, Prague: 165–74
Long, A. G. (2013) Conversation and Self-sufficiency in Plato, Oxford
Mackenzie, M. (1981) Plato on Punishment, Berkeley, CA
Manuwald, B. (2005) “The Unity of Virtue in Plato’s Protagoras,” Oxford Studies in Ancient
Philosophy 29: 115–35
Moss, J. (2009) “Akrasia and Perceptual Illusion” Archiv für Geschichte der Philosophie 91: 119–56
Moss, J. (2005) “Shame, Pleasure, and the Divided Soul,” Oxford Studies in Ancient Philosophy
29: 137–70
O’Brien, D. (2003) “Socrates and Protagoras on Virtue,” Oxford Studies in Ancient Philosophy
24: 59–131
Penner, T. (1973) “The Unity of Virtue,” Philosophical Review 82: 35–68, reprinted in G. Fine ed.
(1999) Plato II: Ethics, Politics, Religion, and the Soul, Oxford: 78–104

33 I would like to thank Gail Fine, Terry Irwin, and Frisbee Sheffield for comments on earlier drafts
of this chapter.
 The Protagoras And Gorgias   159

Richardson, H. S. (1990) “Measurement, Pleasure, and Practical Science in Plato’s Protagoras,”

Journal of the History of Philosophy 28: 7–32
Rowe, C. (2007) Plato and the Art of Philosophical Writing, Cambridge
Russell, D. (2005) Plato on Pleasure and the Good Life, Oxford
Sedley, D. N. (2009) “Myth, Punishment, and Politics in the Gorgias” in C. Partenie ed. Plato’s
Myths, Cambridge: 51–76
Sedley, D. (2014) “The Unity of Virtue after the Protagoras” in B. Collette-Dučić and
S. Delcomminette eds. Unité et origine des vertus dans la philosophie de l’Antiquité, Brussels:
65–90
Shaw, J. Clerk (2015) Plato’s Anti-hedonism and the Protagoras, Cambridge
Tarnopolsky, C. H. (2010) Prudes, Perverts and Tyrants. Plato’s Gorgias and the Politics of
Shame, Princeton, NJ
Taylor, C. C. W. (1991) Plato: Protagoras (2nd edition), Oxford
Taylor, C. C. W. (2003) “The Hedonism of the Protagoras Reconsidered” in A. Havlíček and
F. Karfík eds. Plato’s Protagoras: Proceeding of the Third Symposium Platonicum Pragense,
Prague: 148–64
Trivigno, F. (2013) “Childish Nonsense? The Value of Interpretation in Plato’s Protagoras,”
Journal of the History of Philosophy 51: 509–43
Vlastos, G. (1967) “Was Polus Refuted?” The American Journal of Philology 67: 454–60, reprinted
in his (1995) Studies in Greek Philosophy Volume 2: Socrates, Plato, and Their Tradition,
Princeton, NJ: 60–63
Vlastos, G. (1972) “The Unity of the Virtues in the Protagoras,” Review of Metaphysics 25: 415–58,
reprinted in his (1981) Platonic Studies (2nd edition), Princeton, NJ: 221–69
Vlastos, G. (1981) “Socrates on ‘the Parts of Virtue’” in his Platonic Studies (2nd edition), Princeton,
NJ: 418–23
Vlastos, G. (1991) Socrates: Ironist and Moral Philosopher, Cambridge
Warren, J. (2014) The Pleasures of Reason in Plato, Aristotle, and the Hellenistic Hedonists,
Cambridge
Woodruff, P. (2017) “Why Did Protagoras Use Poetry in Education?” in Olof Pettersson and
Vigdis Songe-Møller eds., Plato’s Protagoras: Essays on the Confrontation of Philosophy and
Sophistry (Philosophical Studies Series 125), Cham, Switzerland: 213–27
Woolf, R. (2000) “Callicles and Socrates: Psychic (Dis)harmony in the Gorgias,” Oxford Studies
in Ancient Philosophy 18: 1–40
 chapter 7

The M eno

Lindsay Judson

1. Summary

Socrates’ main interlocutor in the Meno is a young man, about 18 or 19 years old,
­well-educated, aristocratic, rich, handsome, and possibly rather unpleasant.1 There are
cameo roles for one of Meno’s slaves and for Anytus, another unappealing person who
was to be one of the accusers at Socrates’ trial. The dialogue opens rather abruptly with
a question from Meno:

Are you able to tell me, Socrates–is aretē teachable? Or is it not teachable, but some-
thing acquired by practice? Or does it come to people neither as something acquired
by practice nor as something learnt, but by nature or in some other way? (70a1–4)

This reopens the question which formed the main subject of the Protagoras. The word
aretē is usually translated as “virtue” or “excellence.” In that earlier dialogue, Protagoras
claims to teach “good judgement about one’s own affairs, so that one can run one’s
household in the best way, and about political affairs, so that one can both act and
speak in the strongest way in relation to political affairs” (318e5–a2). This suggests that
aretē is something like “how to make a success of your life”; Protagoras construes what
this involves in a relatively conventional way. Meno’s initial account of aretē for an
adult male citizen is very similar–“to be up to the task of conducting political affairs
and, in so doing, to help his friends, harm his enemies, and take good care not to be
harmed himself ” (71e2–5). On the other hand, Meno finds it hard to disagree that
aretē involves achieving these things justly (78d2–e6; cf. 73a7-b1), and accepts a close
connection between having aretē and having virtues such as justice and temperance
(73d9–10 74a4–6); at 79a3–6 he agrees that these virtues are parts of aretē. I shall with
some hesitation translate aretē as “virtue.”2

1 See Xenophon, Anabasis II.6.21–29; Meno 78c4–d3 and 91a1–6.

2 There is a helpful discussion of “excellence” and “virtue” as translations of aretē in Aristotle’s
Nicomachean Ethics in C.C.W. Taylor, Aristotle, Nicomachean Ethics Books II–IV (Clarendon Aristotle
Series: Oxford, 2006), 1–2.
162   Lindsay Judson

In the Protagoras Socrates seems happy to discuss this question, but in the Meno he
says that he cannot answer it without first discovering what virtue is—and that,
more generally, one cannot know what something is like without first knowing what it
is (71b1–8; cf. 100b4–7). This is an explicit expression of a view that Plato seems to
present Socrates as believing in some earlier dialogues.3 It is controversial both why
Socrates is supposed to hold this view and what restrictions (if any) he is supposed to
place on its scope.4
At this point we might expect the dialogue to follow the same pattern as the so-called
Socratic dialogues, in which Socrates’ interlocutors offer a series of definitions of cour-
age, friendship, etc., each of which is refuted using Socrates’ elenctic method. A later
section of the Meno conforms to this pattern, but before that things take a more theoret-
ical turn. Meno begins by giving a list of different forms or types of virtue: virtue for a
man, virtue for a woman, etc. (71e1–72a5). Socrates’ demand for a unified definition of
what makes these all types of virtue leads to a set piece, if somewhat enigmatic, discus-
sion of definition (74b2–77a2).5 Meno then offers a new definition of virtue as “desiring
fine things and being able to get them.”6 At this point we do get an elenctic refutation,
though it involves Socrates’ deployment of rather more of (what are presented as) his
own philosophical views than we tend to get in the Socratic dialogues—in particular the
intellectualist view that no one ever really desires what is bad.7
With Meno’s definition refuted, Socrates expects the discussion to continue with a
fresh definition; but Meno moves in a quite different direction with a famous set of ques-
tions, usually called “Meno’s paradox” or “the paradox of inquiry”:

And in what way are you going to inquire into this [i.e. what virtue is], Socrates,
when you do not know at all what it is? Which of the things which you do not know
will you set up in your inquiry? Or even if you actually do happen upon it, how will
you know that this is the thing which you did not know? (80d5–8)

3 The idea lies behind Socrates’ inference toward the end of the Euthyphro that Euthyphro must
think he knows what piety is if he thinks that he knows that his action is pious (15d4–8); see also Lysis
223b3–8. Laches 190a6–c2 makes a closely related claim.
4 For discussion see “The Meno’s Characterisation of Knowledge” later in this chapter.
5 For discussion see A. Nehamas, “Confusing Universals and Particulars in Plato’s Early Dialogues,”
Review of Metaphysics 29 (1975–1976), 287–306; I. M. Crombie, “Socratic Definition,” [“Socratic
Definition”], Paideia 5 (1976) Special Plato Issue, 80–102 (reprinted in J. M. Day (ed.), Plato’s Meno in
Focus [Plato’s Meno] [London, 1994], 187–92); D. Charles, “Types of Definition in the Meno,” in
L. Judson and V. Karasmanis (eds.), Remembering Socrates: Philosophical Essays [Remembering
Socrates] (Oxford, 2006), 110–28; V. Karasmanis, “Definition in the Meno,” in L. Judson and
V. Karasmanis (eds.), Remembering Socrates, 129–41; G. Fine, “Signification, Essence, and Meno’s
Paradox: A Reply to David Charles’s ‘Types of Definition in the Meno,’ ” Phronesis 55 (2010), 125–52.
6 77b3–5; cf. 78c6–7.
7 For discussion, see G. X. Santas, “The Socratic Paradoxes,” Philosophical Review 73 (1964), 147–64
(reprinted with revisions as ch. 6 of Santas, Socrates: Philosophy in Plato’s Early Dialogues (London,
1979); T. Irwin, Plato’s Moral Theory [Moral Theory] (Oxford, 1977), 78–82, and Plato’s Ethics (Oxford,
1995), sections 97–98; A. Nehamas, “Socratic Intellectualism,” Proceedings of the Boston Area
Colloquium in Ancient Philosophy 2 (1986), 275–316; H. Segvic, “No One Errs Willingly: The Meaning of
Socratic Intellectualism,” Oxford Studies in Ancient Philosophy 19 (2000), 1–45.
 THE MENO   163

Socrates responds to this by helping one of Meno’s slaves to come to see a geometrical
truth for himself—or so Socrates is made to claim (81e4–85d5). This episode is
­interwoven with an account of the “theory of recollection,” according to which our souls
are immortal and have all along known “all things”: it is this that makes learning (of the
appropriate kind) possible, since it is really the recollection of things we already know
(or have known before).8 Meno somewhat hesitantly goes along with this, but rather
than resume their search for the definition of virtue, he wants to return to his initial
question—is it teachable? Socrates relents, and says that they can pursue this without
knowing what virtue is, if they argue from a hypothesis, an idea introduced by an obscure
parallel with a geometrical procedure.9 The rest of the dialogue from this point onward—
a little under half of the whole work—is one of the strangest passages in Plato’s entire
œuvre. Socrates first argues that virtue is teachable because it is knowledge (87b6–89c4);
then at 89c5–100b4 he argues that it is not teachable, and so it cannot be knowledge, but
rather true belief, which (contradicting what he said about the slave’s progress earlier)
comes not by teaching but “by divine dispensation and without thought” (99e6–100a1).
In the course of his concluding remarks he seems tacitly to change views back again, and
commit himself to the idea that virtue can be taught (99e3–100b1).
In this chapter I shall examine various aspects of a central theme of the Meno, knowl-
edge (epistēmē) and its relation to true belief (alēthēs/orthē doxa10): Socrates’ account of
knowledge as “true belief tied down with a reasoning out of the cause”; the paradox of
inquiry and Plato’s response to it; and the final section of the dialogue (86d3–100c2).

2. The Meno’s Characterization

of Knowledge

At 97e6–98a8 Socrates says:

For in the case of true beliefs as well, as long as they remain they are a fine thing and
they achieve everything good. But they are not willing to remain for a long time, and
instead run away from the person’s soul, so that they are not worth much until one
ties them down with a reasoning out of the cause [aitias logismōi]. And this, Meno

8 81a10–e3 and 85d6–86c4. Almost everything about Socrates’ account is controversial: I discuss
some of the issues in “Meno’s Paradox” and “Plato’s Response” later in the chapter.
9 86d3–87b5. For discussion of the geometrical parallel see, e.g., G.E.R. Lloyd, “The Meno and the
Mysteries of Mathematics,” Phronesis 37 (1992), 166–83; S. Menn, “Plato and the Method of Analysis,”
Phronesis 47 (2002), 193–223; D. Scott, Plato’s Meno [Meno] (Cambridge, 2006), 133–37; V. Karasmanis,
“̓Απαγωγή: Hippocrates of Chios and Plato’s Hypothetical Method in the Meno,” in A. Longo and D. del
Forno (eds), Argument from Hypothesis in Ancient Philosophy (Napoli, 2011), 21–41; L. Judson,
“Hypotheses in Plato’s Meno” [“Hypotheses”], Philosophical Inquiry 41 (2017), 29–39, at 36.
10 Plato uses “true” (alēthēs) and “correct” (orthē) belief interchangeably (see 85c7, 86a7, 97b9, 97e6,
98b7, 98d1, 99a1–2, 99a5; 97b5, 97c4, 97c9, 97d2, 98a8, 98b2, 98c1–2, 98c10). Except in translations,
I shall use “true belief ” throughout.
164   Lindsay Judson

my friend, is recollection, as we previously agreed. When they have been tied down,
they first become pieces of knowledge [epistēmai], and then they are such as to
remain. And it is because of these things that knowledge is more valuable than
correct belief, and knowledge differs from correct belief in being tied down.

This view of knowledge is, presumably, what Socrates presupposes earlier when he says
of the slave:

And although at present these same beliefs have been stirred up in him as if in a
dream, if someone questions him many times and in many ways about the same
things, you may be sure that in the end he will know them no less exactly than
anyone (85c9–d1).

Socrates’ notion of an aitias logismos has been the subject of intense debate; there is now
a reasonably wide consensus that he means the working out of the cause, or the canonical
explanation, of the belief in question being true.11 Knowledge is thus a true belief made
secure by mastery of this explanation: “[Plato’s] view is that the person with knowledge
can explain why what she knows is true. This ability involves a deep and synoptic under-
standing of, and insight into, what she knows.”12
This point deserves stressing all the more since it is not uncommon for Socrates’
claim here to be taken to be the claim that knowledge is justified true belief.13 The

11 For discussion (and some dissent) see M. Burnyeat, “Socrates and the Jury: Paradoxes in Plato’s
Distinction between Knowledge and True Belief, Part I” [“Jury I”], Proceedings of the Aristotelian
Society Supplementary Volume 54 (1980), 173–91; J. Barnes, “Socrates and the Jury: Paradoxes in Plato’s
Distinction between Knowledge and True Belief, Part II” [“Jury II”], Proceedings of the Aristotelian
Society Supplementary Volume 54 (1980), 193–206; G. Fine, “Knowledge and True Belief in the Meno”
[“True Belief ”], Oxford Studies in Ancient Philosophy 27 (2004), 41–81; W. Schwab, “Explanation in the
Epistemology of the Meno” [“Explanation”], Oxford Studies in Ancient Philosophy 48 (2015), 1–36. This
understanding of aitias logismos is also accepted in Scott, Meno, 179, and C. Perin, “Knowledge,
Stability, and Virtue in the Meno” [“Knowledge”], Ancient Philosophy 32 (2012), 15–34, at 17. I shall not
enter into the debate about the precise relationship between (Platonic) cause and “explanation,” and
shall in general speak in terms of “explanation.” For the purposes of this chapter, nothing hinges on the
view we take of this relationship, providing (1) that “explanation” is understood neither as the linguistic
act of explaining nor as what is done in such an act, but rather as the thing(s) that make(s) these
correct; (2) that at least for the cases that are Plato’s primary concern in the Meno (see later in the chapter),
there is a single, canonical cause or explanation of X, which is what matters for knowledge of X.
12 Fine, “True Belief,” 73 (cf. 61). See also Scott, Meno, 179; Burnyeat, “Jury I,” 186–87.
13 This idea was influentially floated in E. Gettier, “Is Justified True Belief Knowledge?”, Analysis 23
(1966), 121–23, at 121, n.1; see also D. Bostock, Plato’s Theaetetus [Theaetetus] (Oxford, 1988), 203 (but
contrast p. 16), T. Irwin, “Recollection and Plato’s Moral Theory” [“Recollection”], Review of
Metaphysics 27 (1974), 752–72, at 754, and G. Fine, “Inquiry in the Meno” [“Inquiry”], in R. Kraut (ed.),
The Cambridge Companion to Plato (Cambridge, 1992), 200–26; reprinted in Fine, Plato on Knowledge
and Forms (Oxford, 2003), 44–65: “At Meno 98a, Plato offers just one definition of knowledge—as
justified true belief (true belief coupled with an aitias logismos)” (p. 204 of the original article). In “True
Belief ” Fine advances the more nuanced claim that, although what Plato says is not that knowledge is
justified true belief, this is nonetheless his view in the Meno (cf. “Inquiry,” 218–19, n.19): I discuss this
claim later in the chapter.
 THE MENO   165

assumption is, perhaps, that since Socrates characterizes knowledge as improved true
belief, he is in the same business as contemporary analyses that start with the same idea
and deploy justification as their opening move. This reading would require Socrates to
mean that the belief is tied down by a working out of the explanation for one’s believing it.
This is an implausible reading of his remark,14 and would fit badly with a number of
other things that Socrates says about knowledge in the dialogue: the remark about the
slave quoted previously, the requirement that one know what X is before one can know
what it is like, and the account of “tying beliefs down” itself. By the same token, these
things make very good sense if we understand Socratic knowledge as involving a grasp
of the explanation of the belief ’s truth. Take, for example, the famous claim at 71b2–9:

Socrates: . . . I blame myself that I do not know about virtue at all. And if I do not
know what something is, how could I know what it is like in some respect? Or do
you think that someone who does not know at all who Meno is could know whether
he is good looking or rich or well-born, or the opposites of these? Do you think he
could? Meno: I do not.

If we take Socratic knowledge to be justified true belief in the everyday sense,15 then
Socrates is vulnerable to Geach’s attack: “we know heaps of things without being able to
define the terms in which we express our knowledge.”16 Since this difficulty is so obvi-
ous, perhaps we should instead infer that Socrates has a different form of knowledge in
mind. If he thinks that to know that virtue is F requires grasping the explanation of why
it is F, it is very plausible that it requires a grasp of the nature of virtue; similarly, it is
plausible that, at least in some cases, understanding why examples of X are indeed
examples of X requires a grasp or understanding of the nature of X itself, of what X is.
Understanding thus construed is also more than having a justified (in the “common or
garden” sense) true belief that the explanation for p is that q: the subject must also see or
grasp the explanation for herself, and not merely accept that this is so.17
It is hardly surprising, then, that some commentators take Plato to be offering an
account of knowledge as understanding.18 If this is right, it is plausible to see the aitias
logismos as invoking not just any explanation of why the belief in question is true, but

14 Cf. Burnyeat, “Jury I,” 187.

15 Later in the chapter I discuss Fine’s very different account of Socratic justification, on which
Geach’s problem does not arise.
16 P. Geach, “Plato’s Euthyphro: An Analysis and Commentary,” Monist 50 (1966), 369–82 (reprinted
in R. Kamtekar (ed.), Plato’s Euthyphro, Apology, and Crito: Critical Essays [Lanham, MD, 2004],
23–34), at p. 371 of the original article.
17 For this idea that understanding involves insight on the part of the person who understands, see
Burnyeat, “Jury I,” 186, and Schwab, “Explanation,” 29–31, contra Barnes, “Jury II,” 202.
18 Burnyeat, “Jury I,” 183; J. Annas, An Introduction to Plato’s Republic (Oxford, 1981), ch. 8, esp.
190–93, 200, and 212–15; A. Nehamas, “Meno’s Paradox and Socrates as a Teacher” [“Socrates as a
Teacher”], Oxford Studies in Ancient Philosophy 3 (1985), 1–30 (reprinted in H. H. Benson, Essays on the
Philosophy of Socrates [Oxford, 1992], 298–316, and in Day [ed.], Meno in Focus, 221–48), at p. 241 of the
reprint in Day (ed.); Schwab, “Explanation.”
166   Lindsay Judson

the canonical explanation, the grasp of which constitutes proper understanding.19 As an

account of the ordinary Greek usage of epistēmē this would be no more plausible than as
an account of contemporary English usage of “knowledge.” This characterization, then,
is best seen as an account of the highest form of knowledge and/or of the form of knowl-
edge at which Socrates is aiming. This formulation is meant to leave it open—as I think
the Meno leaves it open—whether Plato takes understanding to be the only form of
knowledge (see further later in this chapter).20
This view has come under fire, especially from Gail Fine, who is the leading exponent
of the view that knowledge in the Meno is indeed justified true belief. On her view,
“understanding why p is true” specifies the conditions under which the subject’s belief
that p is justified in the special way required for knowledge as Socrates conceives of it.21
I shall consider three objections that Fine makes to the view that Socrates is primarily
concerned with understanding. The first is that Socrates seems to acknowledge a range
of things that can be known which extends well beyond the range of possible objects of
understanding based on a grasp of a thing’s nature: this range includes propositions about
Meno (that he is rich or good-looking) and, famously, about the way to Larisa.22 As Fine
acknowledges, however, the cases of Meno and the road to Larisa may not be meant as
genuine examples of knowledge, but only as analogies to illustrate the point at issue:23
this is an important interpretative choice facing the reader.24 There are some independent

19 Cf. M. F. Burnyeat, “Examples in Epistemology,” Philosophy 52 [1977], 381–98, at 387, and Schwab,
“Explanation,” 7–10 (though note that Schwab thinks that in giving an account of understanding Plato
is not attempting to give an account of knowledge, or at least of ‘what we typically call “knowledge”’: see
“Explanation,” 25–29). Most commentators take this sort of understanding to be systematic (see, e.g.,
N. Gulley, Plato’s Theory of Knowledge [Plato’s Theory] [London and New York, 1962], 14–16; Nehamas,
“Socrates as a Teacher,” 237; Fine, “True Belief,” 68; Scott, Meno, 179; Schwab, “Explanation,” 12–18); but,
as Schwab says, it need not therefore be universal in scope.
20 Some care is needed here over the potential for a vicious regress. If understanding that virtue is F
requires one to grasp why it is F in terms of the nature of virtue, doesn’t this very grasp of what virtue is
have to be a piece of understanding in terms of some more fundamental nature? (See Crombie,
“Socratic Definition,” 173–74, Schwab, “Explanation,” 18–20.) Something like this problem may arise
however one construes knowledge in the Meno (Fine discusses an analogous difficulty in “True Belief,”
74–78). Plato hints at a related problem in Meno’s objection to Socrates’ definition of shape as “that
which always accompanies colour” that this definition will not be satisfactory if one is talking to
someone who “does not know colour” (75c5–7); but if so, he says nothing about how to meet the
difficulty—unless the Theory of Recollection is meant to do this. In any case, one can easily imagine
coherentist (see Schwab previously cited) or foundationalist moves that could be made to try to meet it.
21 See Fine, “Inquiry,” and especially “True Belief.” Fine’s view is shared by Scott, Meno, 184–85.
22 71b2–9, quoted previously, and 97a9–b3. See Fine, “Inquiry,” 218–19, n.19, “True Belief,” 43–45,
60–61, and 69–70; cf. Irwin, Moral Theory, 316, n.17.
23 In the first case, the need to know what X is before one can know what it is like; in the second
case that (the relevant form of) knowledge requires firsthand understanding rather than acceptance of
second-hand beliefs. See G. Fine, The Possibility of Inquiry: Meno’s Paradox from Socrates to Sextus
[Possibility] (Oxford, 2014), 35–38; cf. Sharples, Meno, 13, 124–25, and 182–83, and Scott, Meno, 21–22
and 182.
24 That there is such a choice seems to be ignored by Panagiotis Dimas (“True Belief in the Meno”
[“True Belief ”], Oxford Studies in Ancient Philosophy 14 [1996], 1–32, at 4–9 and 30–32), and by
 THE MENO   167

reasons to think that these cases are merely analogies. (1) Knowing who Meno is is not
obviously the same as grasping a definition—indeed, it seems implausible that Meno has
a (Socratic) definition or essence. (2) If Meno does have such a thing, we cannot expect
that grasping it will enable us to understand why he is F for most values of F (why he is
rich, why he is on the losing side against Artaxerxes). Likewise, if we suppose a weaker
condition—for example that the point is one about securing reference to Meno25—
then while grasping who Meno is might be a necessary part of grasping the aitias logis-
mos of his being rich or good-looking, it will manifestly be a small part of it—and so it
would be a poor illustration of Socrates’ point about the role of the definition of virtue.
(3) A similar point holds for the case of the road to Larisa. If, as Fine suggests, knowledge
of this involves a grasp of the explanation of why this is the (best) way to Larisa, it is far
from obvious that this requires having traveled the road oneself—one could just be
told.26 These responses show that Fine’s objection is something of a two-edged sword,
since knowing who Meno is and traveling the road oneself both seem to come apart
from grasping the (supposed) aitias logismos in these cases. So it is difficult to suppose
that Socrates’ account is meant to apply to a wider range of more everyday cases as well
as to high-level knowledge.
Fine’s second objection is that Socrates talks about knowledge (epistēmē and its
cognates) without further specification—so that we ought not to read him as focused on
a special form of knowledge.27 But most interpreters, including Fine, see Socrates as
implicitly limiting the range of one or more key terms in the Meno—for example by taking
the paradox to be concerned with only one sort of knowledge,28 or taking “the soul
knows all things” (81c9–d3) to exclude contingent truths.29 Analogously, Fine is
committed to the view that, for Socrates, the justification that is required for knowl-
edge varies for different domains;30 but Socrates never says this, nor does he explicitly

Hugh Benson (“Meno’s Paradox and the Theory of Recollection” [Meno’s Paradox”], in H. H. Benson,
Clitophon’s Challenge: Dialectic in Plato’s Meno, Phaedo, and Republic [Clitophon’s Challenge]
(Oxford, 2015), 48–91, at 54, n.20).
25 Cf. Nehamas, “Socrates as a Teacher,” 224; N. P. White, “Inquiry,” Review of Metaphysics 28 (1974),
289–310 (reprinted in Day [ed.], Plato’s Meno, 152–71), at pp. 153–55 of the reprint; Sharples, Meno,
124–25.
26 For the opposite view, see Fine, “True Belief,” 69 (cf. Fine, “Enquiry and Discovery: A Discussion
of Dominic Scott, Plato’s Meno,” Oxford Studies in Ancient Philosophy 32 (2007), 331-67, at 365). See also
Bostock, Theaetetus, 17. I discuss the idea of learning by being told in “Meno’s paradox” below. While
I have argued that traveling the road oneself is not necessary for understanding, Taylor rightly argues
that traveling it even attentively is not sufficient either (Taylor, “Plato’s Epistemology” in this volume).
27 Fine, “True Belief,” 70 (but see “Inquiry,” 218–19, n.19); Barnes, “Jury II,” 204.
28 E.g., J. Moravscik, “Learning as Recollection” [“Recollection”], in G. Vlastos (ed.), Plato I:
Metaphysics and Epistemology (Garden City, NY, 1971), 53–69; reprinted in Day (ed.), Plato’s Meno,
112–28, at p.115 of the reprint.
29 See, e.g., Moravscik, “Recollection,” 118–19; Taylor, “Plato’s Epistemology” in this volume. Fine
takes the claim to be restricted to “general truths about such things as virtue and geometry” (“Inquiry,”
221, n.26; cf. Fine, Possibility, 111–12).
30 Fine, “True Belief,” 45–46, and “Discovery,” 365; cf. Nehamas, “Socrates as a Teacher,” 226 and 241–42.
168   Lindsay Judson

distinguish these domains. It is in any case quite reasonable to suppose that Socrates is
concerned with the form of knowledge that matters to him—and that this is high-level
understanding.31
Fine’s third objection is that the understanding requirement is compatible with the
view that knowledge is justified true belief: Socrates could hold that understanding is
necessary for the subject’s belief to be justified in the way required for knowledge (or for
the relevant form of knowledge).32 This is true, but in the face of Socrates’ characterisa-
tion of the difference between knowledge and true belief, the subject’s being justified in
this way seems at most a necessary feature of knowledge, not part of its nature.33 It is also
unclear what would motivate the view that knowledge-bestowing justification for true
beliefs requires understanding except the prior view that knowledge (or the sort of
knowledge in question) just is understanding.

3. Meno’s Paradox

There is a vast literature on Meno’s paradox and Plato’s response to it.34 I shall focus on a
few key issues and shall ignore many points of controversy. One key issue concerns the
significance—or lack of it—of the differences between Meno’s initial statement of the
paradox (80d5–8) and Socrates’ restatement (80e2–5). (1) Whereas Meno asks, “how
will you inquire into this, Socrates, when you do not know at all [to parapan] what it is?,”
Socrates speaks merely in terms of not knowing. This seems unlikely to be significant,35
since Socrates himself used to parapan in denying that he knew what virtue is (71a6 and
b3), as did Meno in conceding that he does not know what it is (80b4): so it is common
ground that neither of them knows it at all, and Socrates can reasonably be understood

31 So there is no need, on this interpretation, to see him as confusing knowledge and

understanding—an objection made by Barnes (“Jury II,” 204–05), or to see him as not talking about
knowledge (Fine, “True Belief,” 70–71; Schwab, “Explanation,” 25–29).
32 Fine, “True Belief,” 61–67.
33 Cf. Euthyphro 11a6–b1: “though you were asked, Euthyphro, what the pious is, you are in danger of
not wanting to make clear to me its essence [ousia], but only to tell me of some feature [pathos] that it
has, when you say that the pious has this feature, namely being loved by all the gods; but what it is, you
have not yet said.” See also Taylor, “Plato’s Epistemology” in this volume. Schwab makes a similar point
to mine when he argues that “we need to distinguish the question whether Socrates shows any interest
in justification from the question whether he thinks that working out the explanation upgrades true
doxa to epistēmē because it plays a justificatory role” (“Explanations,” 23).
34 Two of the best discussions are those in Fine, “Inquiry,” and M. M. McCabe, “Escaping One’s Own
Notice Knowing: Meno’s Paradox Again” [“Escaping”], Proceedings of the Aristotelian Society 109
(2009), 233–56; reprinted in McCabe, Platonic Conversations (Oxford, 2015), 190–207 (subsequent
references are to the reprint).
35 Cf. Fine, Possibility, 86–87; for the opposing view, see Moravcsik, “Recollection,” 115–16.
 THE MENO   169

as accepting the qualification here.36 (2) Meno apparently identifies two problems facing
inquiry into what one does not know, while Socrates’ restatement does not distinguish
two problems.37 Yet if Socrates were to be understood as deliberately omitting one of
Meno’s problems, it would be the very one on which his solution seems to focus; so it is
plausible that this difference is not significant either—and that Plato is inviting the reader
to work out for herself how Socrates’ solution actually addresses both problems.38
As with Meno’s objection to Socrates’ definition of shape,39 Socrates both describes
the paradox as merely a “disputatious argument” (eristikon logon: 80e1–240) and takes it
very seriously. Why should anyone suppose that the paradox is merely disputatious? The
most plausible answer is that the dilemma Meno sets up simply does not get a grip on
every method of acquiring knowledge—and in particular it does not seem to address
the point that one can acquire knowledge by being told. So if we take the paradox to be
concerned with the acquisition of any sort of knowledge by any sort of method, then it
certainly seems disputatious, since it ignores obvious ways of acquiring knowledge.41 To
be worth taking seriously, the paradox must be understood as restricted in some way.
There are four main possibilities for such a restriction. (1) The very idea of inquiry
rules out learning by being told.42 This seems to me to be plainly false, of the Greek term
zētein as much as of the English term “inquire.” Moravcsik is right to say that inquiry
“must be given direction by the learner himself ” (a point we shall come back to), but
while this means that simply happening to be told something does not count as inquir-
ing, it does not exclude my seeking out an expert to tell me the answer to the question
I am inquiring into.43 (2) In response, one might suggest that the paradox concerns

36 So Nehamas, “Socrates as a Teacher,” 224–28. I will suggest later that Socrates resolves the
paradox, not by showing how one can inquire while not knowing at all, but by arguing that in order to
inquire you need to know in a way and not know in another way.
37 Meno: “which of the things which you do not know will you set up in your inquiry? Or even if
you actually do happen upon it, how will you know that this is the thing which you did not know?”
(80d5–8). Socrates: “nor could he inquire into what he does not know, for he does not know what he
will be inquiring into” (80e4–5).
38 For discussion see Dimas, “True Belief,” 19–20; Scott, Meno, 76–84; Fine, “Discovery,” 340–47;
McCabe, “Escaping,” 190–96; note 70 of this chapter.
39 See note 20 of this chapter.
40 It is controversial that eristikon means “disputatious.” Benson defends the view that it refers to the
aim of the argument—victory rather than truth (“Meno’s Paradox,” 72–73; cf. White, “Inquiry,” 167, n.1).
Fine suggests that it means an argument that is apparently—but only apparently—sound (Possibility,
83–84). For more general discussion of sophism and eristic in Plato, see Nehamas, “Eristic, Antilogic,
Sophistic, Dialectic: Plato’s Demarcation of Philosophy from Sophistry,” History of Philosophy Quarterly
7 (1990), 3–16, and C. C. W. Taylor, “Socrates the Sophist,” in Judson and Karasmanis (eds.),
Remembering Socrates, 157–68.
41 Notice also that if the paradox were supposed to apply to learning by being told, then it would be
much easier to refute than Socrates seems to find it, and it would be very odd indeed that the one thing
that Socrates insists on in the episode with Meno’s slave is that he is not teaching the slave (82e, 84c–d).
I shall not pursue the question of Plato’s attitude to the idea that one can acquire knowledge of things
by perceiving them: but this might count as another type of case that the paradox just seems to ignore.
42 So Moravcsik, “Recollection,” 113; Nehamas, “Socrates as a Teacher,” 228–29; Sharples, Meno, 143.
43 Similarly, Fine is right to say that being told is not a form of inquiry (“Inquiry,” 215–16, n.2); but it
does not follow that it cannot be a part of an inquiry.
170   Lindsay Judson

inquiry when no knowledgeable informant is available. This turns the paradox into
a rather flat-footed problem, however: perhaps we simply have to wait for such an
­informant before we can learn!44 On this reading, moreover, Socrates’ conclusion that
“inquiring and learning are wholly recollection” (81d4–6)45 seems fallacious: it would be
very hard to understand him as meaning only that “all learning in the absence of an
­informant is recollection.” (3) “Inquiry” is to be understood as Socratic inquiry. This is
much more plausible. Socratic inquiry standardly proceeds on the assumption that
Socrates does not know what X is, and it is not halted—indeed on Socrates’ account it is
a sign that further progress can be made—when his interlocutor comes to realize that he
does not know either,46 and no one’s view is accepted without being subjected to Socratic
examination.47 This view imposes a significant restriction on the sort of inquiry envis-
aged in Meno’s paradox, but it is one that makes good sense in the context, since it is
Socratic inquiry that Socrates and Meno are engaged in, and Meno has just conceded
that he does not know what virtue is.48 As with (2), however, Socrates’ conclusion that
“inquiring and learning are wholly recollection,” and Meno’s gloss that “what we call
learning is recollection,” presents a difficulty. For this conclusion to make sense, we
must be being invited to suppose that all learning involves Socratic inquiry, and this
seems quite unmotivated.49 (4) Inquiry here is taken to be the search for understanding.
This would be entirely natural if (but only if) this is the form of knowledge at which
Socrates is aiming. Being told will simply not do for this: one has to master the aitias
logismos for oneself.50

44 A more interesting challenge could be constructed if we asked how knowledge came about in the
first place; but there is no hint that this is what Meno and Socrates have in mind.
45 See also Meno’s reply: “Yes, Socrates, but what do you mean by saying that we do not learn, but
that what we call learning is recollection?” (81e3–4). Cf. 87b6–c1, discussed in Section 5 of this chapter,
“Knowledge and True Belief in the Final Part of the Meno.”
46 84b9–d1.
47 See, e.g., Euthyphro 9e4–7: “should we, therefore, investigate this further, Euthyphro, to see if it is
well said, or should we let it pass and in this way accept things, from ourselves and others, agreeing that
a thing is so if someone merely says that it is so?”
48 79e7–80b4. It will on this reading be nonetheless irrelevant whether Socrates knows the answer
to the geometrical question that he puts to the slave or not: even if he does know it, the point will be
to illustrate how the slave can reach the answer by means of a Socratic elenchus (see note 52 of
this chapter).
49 It is true that a central ingredient of Socrates’ resolution of the paradox, his conversation with the
slave, is clearly signposted as a piece of Socratic elenchus. (Note the repetition of the image of the
“numbing” effect of Socrates’ elenchus, which Meno had applied to his own case at 80a4–7; Socrates’
picking up of Meno’s terms aporia and aporein [79e7–80a4; 84a2–b1; 84b6–7]; Socrates’ reference to the
idea that in his original state of confidence the slave would have felt himself able to speak well “in front
of many people and on many occasions” [84b9–c1: compare Meno’s remark about himself at 80b1–2].
See Irwin, “Recollection,” 754 and n.4; Nehamas, “Socrates as a Teacher,” 233; H. H. Benson, “Meno, the
Slave Boy and the Elenchos” [Elenchos], Phronesis 35 (1990), 128–58, at 137–39; Fine “Inquiry,” 208–09.)
But this just reflects the fact that Socrates’ answer to the question “how is the acquisition of knowledge
possible” involves appeal to the elenchus as a way to help the process of recollection.
50 See Burnyeat, “Jury I,” 184, and pp. 164–66 above.
 THE MENO   171

4. Plato’s Response

What Socrates does to respond to Meno’s paradox is in some ways opaque, for at least
two reasons. (1) He begins by appealing to the pronouncements of “priests and priest-
esses,” rather than to reasoned argument, about the immortality of the soul and its prior
knowledge of “all things,” and he expresses a lack of confidence in the Theory of
Recollection at the end of the exposition—yet he goes on insisting on that inquiry is
possible.51 (2) When he helps the slave to arrive at the geometrical theorem, he is appar-
ently engaged in a demonstrative experiment to show that there can be successful
inquiry, but the rules of the experiment are unclear: What would count as cheating on
his part, and what would count as success? In any case, what really matters is what Plato
invites the reader to think the proper response should be.52 There are two leading inter-
pretations: on one (which I prefer), the Theory of Recollection is central, and the key
idea is recognition; on the other, true belief is the key to the solution, and the doctrine of
recollection is quite secondary.

(1) One of Meno’s difficulties was “even if you actually do happen upon [the object
of your inquiry], how will you know that this is the thing which you did not know?”
(80d7–8): this seems to assume that grasping that X is the right answer requires compar-
ing it to some already known item to see if they match.53 Recognition, by contrast, does
not involve any matching of this sort, though it does require prior knowledge of some
form or other. Suppose that I cannot now say what the famous bust of Aristotle in
Vienna looks like: I cannot describe it or visualize it. I may, nonetheless—if I have seen it

51 81a11ff.; 86b6–c2.
52 In other words, we should treat the episode with the slave as a thought experiment, not an
experiment. Plato is trying to characterize, however indirectly, a process that could in principle lead to
knowledge and that does not involve the mere imparting of it, or learning by being told—but that
rather involves someone coming to a realization of the truth in some other way, with only, in this case,
the Socratic elenchus (see note 49 of this chapter) to help him. For this reason, worries that Socrates’
questions are too leading, that the slave simply says yes out of deference, and that Socrates has not
demonstrated the acquisition of knowledge since the slave only reaches the point of true belief, are not
quite the right worries to have.
53 Fine argues for a different understanding of the problem, on which what is crucial is the idea of
being in a “cognitive blank” about the object of inquiry; Meno’s mistake is then to suppose that knowing
nothing at all about something entails being in a cognitive blank about it, thus ignoring the possibility of
having true beliefs about it (Possibility, 74–90; I discuss Fine’s “true belief ” interpretation below). If
Meno’s view is simply based on a failure to see that one might lack knowledge of X but have true beliefs,
it is hard to see why he holds it: he ought to be prepared to consider the possibility that one could be in a
position to know that a candidate answer about X is the right one on the basis of a knowledge (or beliefs)
about something else, Y. The fact that he appears to ignore this possibility seems to support my reading: it
is easier to see how someone might hold Meno’s view if she assumed that the only way to know that a
candidate answer is right is by seeing if matches a known item (compare the “Augustinian” view attacked
by Wittgenstein in The Blue and Brown Books, cited in note 54 of this chapter).
172   Lindsay Judson

before—be able to recognize it.54 Recognition is a cognitively basic capacity whose

exercise typically (though fallibly) constitutes some form of knowledge (as we ordinarily
conceive of knowledge), and although one can, of course, sometimes conclude that X is
the sought-for item on the basis of an inference from a sense of familiarity, paradigmatic
recognition is non-inferential.55 The first interpretation, therefore, takes the idea of
recognition to show how one can come to grasp that X is the right answer to a question
without already knowing the answer in a way that would make inquiry superfluous.56
On this reading Socrates resolves the paradox by showing that in order to inquire you
need to know in a way (since recognition presupposes causation by former knowledge
that is in some sense or other retained)57 and not know in another way, since inquiry is
needed if one cannot access or articulate this knowledge at will.
This interpretation has had an undeservedly poor press, at least in part because it has
not been understood in its most plausible form. (a) One factor in this may be the idea
that recognition is essentially based on prior perceptual knowledge. This idea is mis-
placed: I can recognize the name of the museum in Vienna that houses the Aristotle
bust, or a train of thought I have had before, or the answer to a problem in logic or men-
tal arithmetic.58 (b) Moravcsik takes the idea of recognition to be merely an analogy for
what Plato really has in mind;59 but then it is unclear what Plato does have in mind or
why we should accept the analogy. In a similar vein Bostock conjectures that “Perhaps
[Socrates] is also relying on a parallel between the feeling one has when one grasps a
solution and the feeling one has when one succeeds in remembering: in each case it
“clicks into place” as something evidently right, and carries conviction with it.”60 Here too
the implication is that Socrates has no business relying on such a parallel, and I would
agree—but I see no reason to suppose that he is. Defenders of this interpretation should
take recognition to be exactly what Plato has in mind. (c) There is a tendency among
commentators (including Bostock) to think that what really matters, epistemologically,

54 Manifestly, recognition of this kind does not occur by matching the sought-for item with a mental
item (see L. Wittgenstein, The Blue and Brown Books [Oxford, 1958], 3).
55 That is, the recognitional belief (like, for instance, paradigm cases of the belief that I am seeing a
tree, or that I am in pain) is simply caused by the relevant psychological states, and is not inferred from
other beliefs. This view is controversial, but I cannot defend it here. On the difficult issue of the
epistemology of non-inferential beliefs see, e.g., J. McDowell, Mind and World (Cambridge, MA, 1994),
especially Lecture VI; M. F. Martin, “Epistemic Openness and Perceptual Defeasibility,” Philosophy and
Phenomenological Research 63 (2001), 441–48; M. McGinn, “Non-inferential Knowledge,” Proceedings of
the Aristotelian Society 112 (2012), 1–27.
56 See Gulley, Plato’s Theory, 11–12 (note that he adds the idea of the answer’s being “indubitably
true” without any warrant from the text: see later in this chapter); Moravcsik, “Recollection,” 116–18;
Dimas, “True Belief,” 27–29.
57 What I say here is compatible with the minimal claim aired by Fine (Possibility, 151–52) that what
is retained might be a disposition that is not itself, strictly speaking, knowledge. If that were the case it
would still be appropriate to say of the possessor of the relevant disposition that she knew in a
way—namely in that she had the right sort of disposition to regain her knowledge.
58 In a similar vein, another factor may be an unfounded assumption that the “recognition” answer
would commit Plato to viewing knowledge as essentially a matter of acquaintance.
59 Moravcsik, “Recollection,” 116–18. 60 Bostock, Theaetetus, 16.
 THE MENO   173

is a feeling of recognition or “inner conviction,” regarded as the basis of an inference that

this is the sought-for item, or, at least, “the ground” for the belief that it is.61 This view is
sometimes then exposed to the criticisms that such a feeling is a poor ground, or, though
a good ground, one that cannot lead to certainty.62 The latter criticism, at any rate, is
clearly misplaced: there is no sign that Plato, or the speakers in the dialogue, are con-
cerned with scepticism or with certainty. But I think that the epistemological value of
recognition does not stem from a feeling: it is, as I have said, a cognitively basic capacity—
like perception, certain forms of memory, and our capacity to form basic beliefs about
our psychological states. From the internal perspective, moreover, while the subject may
take feelings of recognition or familiarity as a sign that she is recognizing, such feelings
are not necessary for her belief that she is (and of course they need not occur at all): the
subject’s belief that she is recognizing can itself be non-inferentially caused by the act of
recognition.63
(2) The second main interpretation sees Plato’s solution as focused on true beliefs.
One can have true beliefs about a subject of inquiry without any knowledge, and reli-
ance on these beliefs enables the inquiry to get started and provides a basis for assessing
candidate answers. Thus inquiry into what one does not know is possible, if one has
appropriate true beliefs.64 As with the recognition interpretation, some clarification is
called for. While at a relatively abstract level the appeal to mere true beliefs might seem
to meet the challenge posed by Meno’s paradox, this seems less plausible when one con-
siders that the challenge is concerned with inquiry thought of as an intentional search
given direction by the learner herself.65 The inquirer must think not only that (some of)
her beliefs about X happen to be true, but must regard them as to some degree justified
(and must be correct in so regarding them), if she is to suppose that she can make genu-
ine progress: she needs either to have reason to suppose that most of her beliefs are true,
or some prima facie justification for some individual beliefs. Fine argues against an anal-
ogous point posed in terms of the inquirer’s knowledge that her beliefs are true;66 while I
agree with her that the inquirer need not know this, I do not see that the point carries
over to the inquirer’s need to have some justification.67 It is implicit in what I have said

61 See N. P. White, Plato on Knowledge and Reality (Indianapolis, IN, 1976), 51–53, and “Inquiry,”
166–68; Dimas, “True Belief,” 27–28; Benson, Clitophon’s Challenge, 85–86.
62 See Gulley, Plato’s Theory, 11–12; J. Gentzler, “Recollection and ‘The Problem of the Socratic
Elenchus’ ” [“Recollection”], Proceedings of the Boston Area Colloquium in Ancient Philosophy 10 (1996),
257–95, at 274, n.42.
63 For a more developed and subtle account along these lines, see McCabe, “Escaping,” especially
196–200 and 205–07.
64 See especially Irwin, Moral Theory, 138–40, and Fine, “Inquiry.”
65 See Moravcsik quoted on p. 169 above; compare Meno’s question: “which of the things which you
do not know will you set up in your inquiry?” (80d6–7), and McCabe, “Escaping,” especially 190–98.
66 “Inquiry,” 212.
67 Fine argues that from an external standpoint the problem does not even arise (“Inquiry,” 212). If
we took the problem simply to be “how is it possible to get to the correct answer?,” this would be right:
Sherlock Holmes might happen to rely on true beliefs in constructing his grand explanation of how the
crime was committed and by whom, and so might happen to get to the correct answer. But first, this
reinforces the point that the internal viewpoint is crucial to the paradox, as Meno’s second question
174   Lindsay Judson

that justified true belief need not amount to knowledge. This is obviously not a problem
for those who think that the sort of knowledge in question in the Meno is understand-
ing; but those who think that knowledge is justified true belief can also accept it, since
they can either think that the justification required for knowledge is something different
from common or garden justification,68 or that the justificatory standard for knowing a
given proposition is typically higher than for merely having a justified belief. This point
is important for at least two reasons. First, it makes for greater continuity with Socrates’
use of hypotheses in the final part of the dialogue—an advantage for the “true belief ”
interpretation, as we shall see. Second, it makes for a sharp discontinuity with the notion
of “true belief ” deployed when Socrates claims that virtue is true belief: this will be
discussed in the next section.
The “recognition” and “true belief ” interpretations have complementary advantages
and disadvantages. The “recognition” interpretation obviously sits very well with the
theory of recollection: the prominence given to the recollection theory is just what we
would expect if Plato’s response to the paradox is that successful inquiry is the recovery
of forgotten or latent knowledge.69 By the same token, however, it has to work harder
than the “true belief ” interpretation to provide a solution to the problem of how one sets
up the inquiry at the outset,70 and unlike that interpretation it provides no obvious link
to the introduction of hypotheses in the next part of the dialogue, which on this account
must be seen as a fresh start. It has some further exegetical advantages over the “true
belief ” interpretation (discussed later in this chapter), but labors under philosophical
disadvantages because it is so closely wedded to the theory of recollection.71
Correspondingly, the “true belief ” interpretation ties up well with the hypothetical
method: what Socrates claims goes on to do (at least in theory) is precisely to use justi-
fied beliefs that fall short of knowledge72 to try to make progress with the inquiry into

(“even if you actually do happen upon [the object of your inquiry], how will you know that this is the
thing which you did not know?”) will not be happily answered by saying that Holmes now has a true
belief about the culprit, which is supported by inferences from other true beliefs. Second, even from
the external standpoint, it is unsatisfactory as an account of rational inquiry: “start with your beliefs,
and hope that the ones you rely on crucially are true.”
68 See Fine, “True Belief,” and Section 2 of this chapter.
69 I shall not pursue the question of which of these Plato thinks we have (but on latent belief see
later in the chapter): for discussion see Fine “Inquiry,” 213; “Discovery,” 353–62; Possibility, chs 4–5;
Taylor, “Plato’s Epistemology” in this volume; D. Charles, “The Paradox in the Meno and Aristotle’s
Attempts to Resolve It,” in Charles [ed.], Definition in Greek Philosophy [Oxford, 2010], 115–50, at 129,
n.18).
70 On this interpretation we must suppose, I think, that the soul’s latent and/or prior knowledge
gives it the wherewithal for starting inquiries. This might be a matter of a low-level grasp of concepts,
and/or capacities to recognize instances, and/or true beliefs. Some commentators find such a view in
the (probably later) Phaedo.
71 On the other hand again, Socrates’ disclaimers in the Meno (see p. 171 above) notwithstanding,
Plato himself does seem wedded both to the immortality of the soul (Phaedo, Republic, Phaedrus), and
to the theory of recollection (Phaedo, Phaedrus).
72 Or so I believe: see Judson, “Hypotheses.” I explain “at least in theory” in “Knowledge and True
Belief in the Final Part of the Meno” in Section 5 of this chapter.
 THE MENO   175

whether virtue is teachable; but the interpretation obviously fits less well than the
“recognition” interpretation with the theory of recollection, which seems redundant if
the “true belief ” interpretation is correct. Perhaps the theory of recollection is Plato’s
explanation of why we have a stock of true beliefs about the objects of inquiry in the first
place, or of why the beliefs we form are more likely to be true than false;73 but if so, Plato
omits mention of this crucial move. Again, while the “true belief ” interpretation may
seem more philosophically pleasing to us, it faces some exegetical problems not faced by
its rival. (a) If Plato’s answer to the paradox is that one gets to knowledge from true
beliefs, it is (not inconsistent but) very odd that the stage of the process he focuses on in
the episode with the slave is that of arriving at a true belief (85c6–d1). (b) If the episode
with the slave is meant to reveal Plato’s response to the paradox, then we must be meant
to see that the slave has some key true beliefs at the very outset of the inquiry, and that it is
these beliefs that enable him to dodge the paradox; it is again (not inconsistent but) very
odd that our attention is never drawn to this—if anything it is drawn to the falsity of
some of the slave’s beliefs (84a3–b1).74 (c) Part of what his discussion with the slave does,
according to Socrates, is to “stir up” true beliefs in him.75 So it looks as if in any case
Socrates is presented as committed to the idea that cognitive items can be present in the
soul in a latent way, and can be brought to the surface by inquiry: if Plato holds the more
philosophically wholesome theory envisaged by the “true belief ” interpretation, which
rejects the need for this machinery in the case of knowledge, it is odd that he commits
Socrates to it in the case of true beliefs.

5. Knowledge and True Belief in the

Final Part of the Meno

Commentators rightly find the final part of the Meno (86c6–100c2) bewildering if not
downright baffling. There are controversies about the hypothetical method,76 but the
serious trouble starts to emerge at 89c5, when Socrates begins his lengthy argument that

73 See Fine, Possibility, chs 4–5.

74 Cf. Nehamas, “Socrates as a Teacher,” 233–34. At 82b64ff. Socrates does ask Meno whether the
slave “is Greek and speaks Greek,” and then asks the slave some basic questions (“is a square a figure
like this [presumably pointing to a diagram]?,” “does a square have four equal sides?,” etc.), which the
slave answers correctly. So it is plausible that the slave does have some true beliefs that are relevant to
the inquiry (cf. Fine, Possibility, 128–34); but my point is that if this is the crux of Plato’s response to the
paradox, it is strange that he makes so little of it. (For example, the only term of epistemic appraisal
that Plato uses of the slave’s views in this passage is gignōskei, which, while it can indicate something
weaker than knowing, more usually does signify knowing; so it is not a good term to use to indicate the
key state between knowing and being in a compete cognitive blank.
75 85c9–10; cf. Gentzler, “Recollection,” 280–81; Nehamas, “Socrates as a Teacher,” 239. Fine disputes
this reading and take Plato’s view to be only that prenatal knowledge explains why we are good at
forming true beliefs (Possibility, 160–61).
76 For a discussion of the issues, and some bibliography, see Judson, “Hypotheses.”
176   Lindsay Judson

virtue is not, after all, teachable, and so is not knowledge but true belief, which comes
“by divine dispensation and without thought” (99e6).77 How can Socrates so readily
contradict the view for which he has just argued—that virtue is knowledge and thus
teachable—and which he is presented as firmly committed to in a number of earlier dia-
logues (most notably the Protagoras)? And how can he then concede, right at the end of
the dialogue, that virtue is teachable—and hence, by his premises, it is knowledge and
not true belief?78 It is very plausible that we are to understand Socrates as being disin-
genuous when he argues that virtue is true belief; but whatever one thinks about that, it
is hard to avoid the conclusion that Plato does not intend the reader to take this section
at face value.79 I shall consider this issue here only insofar as it relates to knowledge and
true belief. One consideration that seems decisive in itself is that Socrates’ claims that
true belief cannot be taught or learned, and that it comes without thought, simply con-
tradict what he claims to have happened at the end of the episode with the slave, who, he
says, has now arrived at true belief about the answer to the geometrical problem.80 There
is simply no satisfactory way to reconcile these claims.81 This means that the trouble
actually began earlier, with Socrates’ premise at 87b5–c9 that X is teachable if and only if

77 Or “without knowledge”: see note 83 of this chapter.

78 “If in this whole discussion we have done well in our inquiry and in what we have said, then
virtue could not come either by nature or from teaching, but would come to people to whom it does
come by divine dispensation and without thought—unless there were to be a statesman of a kind that
could make someone else a statesman too. If there were, he could almost be described as being, among
the living, as Homer says Teiresias is among the dead when he says of him, ‘He alone’—of those in
Hades—‘had life and mind; the rest were shadows.’ That is just what such a man would be like in this
world—in relation to virtue something real among shadows” (99e4–100a7). This is a claim about the
future, but it implies one about the (present) nature of virtue. Socrates’ retraction suggests that he does
not really accept that they have done well in their inquiry. (Note that elsewhere Socrates represents
himself meeting Hippias, Prodicus, and other sophists as Odysseus in Hades—alive, but meeting the
dead (Protagoras 315b9 and c8: see Taylor, “Socrates the Sophist,” 167.)
79 See K. Wilkes, “Conclusions in the Meno” [“Conclusions”], Archiv für Geschichte der Philosophie
61 (1979), 143–53; reprinted in Day (ed.), Plato’s Meno, 208–20 (subsequent page references are to the
reprint); J. M. Day, Introduction,” in Day (ed.), Plato’s Meno, 1–34, at 27-33; Judson, “Hypotheses,” 29,
33–34, and 36–37.
80 85b8–d1. There is some debate, partly fueled by a separate debate about the nature of Socratic
elenchus, as to whether or not Socrates calls the slave’s arrival at true belief about the answer to the
problem “recollection” (see Irwin, “Recollection,” 753–55; Nehamas, “Socrates as a Teacher,” 237–39;
Benson, “Elenchos,” 139–40; Dimas, “True Belief,” 1 and n.1; Gentzler, “Recollection,” 277–81). But it is
plain that the slave’s arrival at the correct answer is an intermediate result of a single elenctic process
(see note 49 of this chapter), which, Socrates claims, can lead the slave to exact knowledge. Socrates
also describes this arrival in terms that imply that the process is bringing latent beliefs to the surface
(85c6–7; see note 75 of this chapter), which is the hallmark of recollection; and the slave certainly does
not arrive at this belief by “divine dispensation”: see note 83 of this chapter.
81 Fine offers an interpretation of teaching as the process of helping someone (in the appropriate
way) to recollect by someone who has knowledge of the answer (Possibility, 115–17): if Socrates does not
know the answer to the geometrical problem, then he does not, after all, teach the slave. This
interpretation does not solve the problem: if Socrates (not knowing the answer) can help the slave in an
appropriate way to recollect a true belief, then a fortiori someone who did know the answer could do
the same (and could use exactly the same elenctic methods)—and that would be teaching.
 THE MENO   177

it is knowledge.82 At 98b1–5 Socrates somewhat surprisingly says that he knows that true
belief and knowledge are “different things” (presumably because he thinks he knows the
definition of knowledge). He does not, however, stop to examine the ambiguity in this
claim. It might mean only that there is some difference between them. This is uncontro-
versial, and of course it follows from the account he has just given of knowledge as true
belief tied down by an aitias logismos; but it does not follow from this that if knowledge
is teachable, true belief is not. Or it might mean that true belief and knowledge are not
really connected, so that it might well be that one is teachable and the other is not. This is
plainly incompatible with Socrates’ account of knowledge, but it seems to be the view to
which he is committed in his argument that virtue is true belief. The epistemological
upshot is that true belief as characterized in this argument is not to be simply equated
with the sort of true belief that the slave arrived at, or that Socrates invokes in his account
of knowledge: there may or may not be some other well-defined epistemological cate-
gory that it represents instead. It is no coincidence that this sort of true belief is treated as
what we might call bare true belief—that is, beliefs held on little or no basis at all. This,
I take it, is the force of “by divine dispensation”: it is a matter of luck (or divine grace or
whim) that the subject is right.83 If these beliefs were ones held with a significant degree
of what I have called common or garden justification84—as the slave’s belief in the right
answer is, for example—then not only would the phrase “by divine dispensation”
be inapposite, but Socrates’ contrast between stable knowledge and unstable belief

82 Benson thinks that Socrates simply makes a mistake best overlooked when he accepts this
biconditional (“The Method of Hypothesis in the Meno,” Proceedings of the Boston Area Colloquium in
Ancient Philosophy 18 (2002), 95–126, at 111, n.52). He thinks that Socrates makes this mistake because
he relies too much on the parallel with the geometer’s biconditional. It is true that in his first argument
Socrates only uses the first half of the biconditional—“if something is knowledge it is teachable” —and
that it is the second half of the biconditional—“if something is teachable it is knowledge” —which
contradicts what he said about the slave. But Socrates does use the second half of the biconditional
(“only knowledge is teachable”) as a key premise in the later argument for the claim that virtue, now
thought of as true belief, comes by divine dispensation (because true belief does not come by teaching).
The incompatibility with what Socrates said about the slave arriving at true belief is blatant.
83 So Wilkes, “Conclusions,” 217. Note that Protagoras is presented as thinking that the options for
how virtue comes to be are teaching, nature, and chance: Protagoras 323c5–d7. (At Meno 99a1–5 Socrates
contrasts true belief with luck: but this is the contrast between something turning out well because of
the subject’s belief and it turning out well in the absence of any belief at all.) At 99e6 Socrates says “by
divine dispensation and without nous.” The latter phrase could mean “thoughtlessly, heedlessly” (so,
e.g., J. M. Day, “Translation,” in Day [ed.], Plato’s Meno, 35–72, at 72), or “without knowledge, without
wisdom” (so, e.g., Sharples, Meno, 121); I prefer the former reading, but do not wish to rely on it here
(my thanks to Gail Fine for discussion of this point). In any case, what Socrates stresses is the idea of
“by divine dispensation,” which he introduced at 99c1–5 and repeated at 99c7–9, c11–d5, and d7–9 as
well as here. (Note that if we sought to unify Socrates’ account by holding that he thinks even the slave
arrives at his true belief by divine dispensation (and presumably without knowledge rather than
without thought), Socrates’ remark at 85c9–d1 that repeated application of the same method would lead
the slave to knowledge would appear to commit him to the view that one can arrive at recollected
knowledge by divine dispensation: this would be hard to accept.)
84 See p. 165 above.
178   Lindsay Judson

(however it is to be understood85) would need to be much more muted. Thus Socrates’

comparison of the two ways of being right about the road to Larisa should be thought of
as between knowing the way (or being in a state analogous to knowing86) and happening
to guess correctly at each junction in the road.
The series of very obvious defects exhibited by Socrates’ argument that virtue is not
teachable confirms the view that Plato does not intend us to take things at face value. (1)
From his claim that there are no teachers of virtue it clearly does not follow that it cannot
be taught.87 (2) The argument that there are no such teachers is itself defective: (a) the
claim that the “great and good” statesmen of the day cannot teach their sons falls to the
objection that Plato himself already made to the same argument in the Protagoras—
that successful teaching requires apt pupils as well as able teachers88–and the argument
that no sophist can teach virtue is woeful. (3) Socrates has no grounds for supposing that
these groups exhaust the candidates for teachers of virtue (96b6–8), and in fact even the
idea that they are good candidates simply ignores Socrates’ insistence at the outset that
by teaching and learning he now means recollection:89 teachers of virtue will have to be
people who can help others to recollect their own forgotten knowledge.90 This makes
these candidates unlikely ones, as does the fact that Socrates has no reason to accept that
the “great and good” statesmen are really virtuous (or that if they are, no one else is).

85 For discussion see Fine, “True Belief,” 72; Scott, Meno, 180–82; A. Millar, “Why Knowledge
Matters,” Proceedings of the Aristotelian Society Supplementary Volume 85 (2011), 63–81, at 65–66; Perin,
“Knowledge,” 15–25.
86 See pp. 166–67 above.
87 This fallacy is apparently encapsulated in the agreement that Socrates secures at 89d6–e3, that X is
teachable if and only if there are teachers and learners of it. See also note 82 of this chapter.
88 A point very clear in the Meno in relation to Meno himself. Protagoras makes the point at
Protagoras 326e6–327c3. It is true that at 93d9–10 Socrates appears to try to forestall this objection: “so
no one would have blamed [Themistocles’] son’s nature, at any rate, as being bad”; but his argument for
this is that the son in question was good at horsemanship, so must be an apt pupil for learning virtue.
89 87b6–c1; cf. 81d4–5, 98a4-5. Daniel Devereux thinks that Socrates’ argument that virtue is not
teachable is actually a good one, if understood as directed ad hominem at Meno who, Devereux thinks,
understands teaching in the non-Socratic sense of the transmission of information (“Nature and
Teaching in Plato’s Meno,” Phronesis 23 [1978], 118–26, at 122-24). But although it is true that Socrates
did talk with the vulgar in something like this sense earlier on in the dialogue, before completing his
response to Meno’s paradox (see, e.g., 81e6–82a3, 82e4–6), he simply does not make this move here: he
introduces the argument with the identification of teaching and recollection, and does not retract this
identification. If he is deploying the “transmission” sense of teaching, moreover, he ought not to accept
the biconditional on which he relies so heavily, that X is teachable if and only if it is knowledge. In a
similar spirit, Michael Forster sees Socrates as accepting both that virtue is non-teachable, divinely
inspired, true belief and that it is teachable knowledge, by supposing that Socrates distinguishes merely
human virtue from “real virtue, which no man has (only god . . . .),” which consists in knowledge
(“Socrates’ Professions of Ignorance,” Oxford Studies in Ancient Philosophy 32 [2007], 1–35, at 11; see also
17–19). While Socrates could have made this distinction, and argued that understood in one way virtue
is teachable (and so is knowledge) and understood in another it is not (and instead comes from the
gods), he does not.
90 Wilkes, “Conclusions,” 217, stresses the idea that Socrates himself is the prime candidate teacher of
virtue (citing 99e4–100a7 [quoted in note 78 of this chapter] and Socrates’ remark in the Gorgias that
he is “one of a few Athenians—not to say the only one—who attempt the true political craft and engage
in politics—the only one among people now” [521d6–8]).
 THE MENO   179

These points suggest that Plato’s main point in making things go so wrong in this part of
the dialogue is to stress that the new method of hypotheses requires no less rigor and
self-scrutiny than the elenctic method.91 There is also an implication for the Meno’s
epistemology, in the contrast between the slave’s true belief and the true belief that
comes by divine dispensation, and in the almost wilful neglect, in the argument that
virtue is true belief, of teaching as helping someone to recollect. The Meno offers two
strands of thought that invoke true beliefs, but they are disjoint rather than continuous.
In developing the idea of true belief that comes by divine dispensation, Plato is not
exploring the contrast between knowledge and true belief as such—for that includes the
state reached by the slave at the end of his discussion with Socrates. Instead Plato is invit-
ing us to consider what sort of cognitive state is available—at best—in the absence of any
learning (i.e., recollection) inspired by philosophical prompting. This is what I have
called bare true belief. Such beliefs are, of course, true; but they constitute a very unap-
pealing goal for our cognitive ambitions, which would be better aimed at things we can
learn, philosophically, and come to understand. While Plato’s depiction of the slave is an
image of someone making progress toward true philosophical understanding, perhaps
these bare true beliefs are a forerunner of the sort of beliefs about beauty that (on some
interpretations) the sight-lovers have in Republic V.92

Bibliography
Annas, J. An Introduction to Plato’s Republic (Oxford, 1981).
Barnes, J. “Socrates and the Jury: Paradoxes in Plato’s Distinction between Knowledge and
True Belief, Part II,” Proceedings of the Aristotelian Society Supplementary Volume 54 (1980),
193–206.
Benson, H. H. “Meno, the Slave Boy and the Elenchos,” Phronesis 35 (1990), 128–58.
Benson, H. H. “The Method of Hypothesis in the Meno,” Proceedings of the Boston Area
Colloquium in Ancient Philosophy 18 (2002), 95–126.
Benson, H. H., “Meno’s Paradox and the Theory of Recollection,” in Benson, Clitophon’s
Challenge: Dialectic in Plato’s Meno, Phaedo, and Republic (Oxford, 2015), 48–91.
Bluck, R. S. Plato’s Meno (Cambridge, 1964).
Bostock, D. Plato’s Theaetetus (Oxford, 1988).
Burnyeat, M. “Examples in Epistemology,” Philosophy 52 (1977), 381–98.
Burnyeat, M. “Socrates and the Jury: Paradoxes in Plato’s Distinction between Knowledge and
True Belief, Part I,” Proceedings of the Aristotelian Society Supplementary Volume 54 (1980),
173–91.
Charles, D. “Types of Definition in Plato’s Meno,” in L. Judson and V. Karasmanis (eds.),
Remembering Socrates: Philosophical Essays (Oxford, 2006), 110–28.
Charles, D. “The Paradox in the Meno and Aristotle’s Attempts to Resolve It,” in Charles (ed.),
Definition in Greek Philosophy (Oxford, 2010), 115–50.

91 See Judson, “Hypotheses,” 36–37.

92 I am grateful to Gail Fine and Ellen Judson for their helpful comments on earlier versions of this
chapter.
180   Lindsay Judson

Crombie, I. M. “Socratic Definition,” Paideia 5 (1976) Special Plato Issue, 80–102; reprinted in
J. M. Day (ed.), Plato’s Meno in Focus (London, 1994), 187–92.
Day, J. M. (ed.), Plato’s Meno in Focus (London, 1994).
Day, J. M. “Introduction,” in Day (ed.), Plato’s Meno in Focus (London, 1994), 1–34.
Day, J. M. “Translation,” in Day (ed.), Plato’s Meno in Focus (London, 1994), 35–72.
Devereux, D. “Nature and Teaching in Plato’s Meno,” Phronesis 23 (1978), 118–26.
Dimas, P. “True Belief in the Meno,” Oxford Studies in Ancient Philosophy 14 (1996), 1–32.
Fine, G. “Inquiry in the Meno,” in R. Kraut (ed.), The Cambridge Companion to Plato
(Cambridge, 1992), 200–26; reprinted in Fine, Plato on Knowledge and Forms (Oxford,
2003), 44–65.
Fine, G. “Knowledge and True Belief in the Meno,” Oxford Studies in Ancient Philosophy 27
(2004), 41–81.
Fine, G. “Enquiry and Discovery: A Discussion of Dominic Scott, Plato’s Meno,” Oxford
Studies in Ancient Philosophy 32 (2007), 331–67.
Fine, G. “Signification, Essence, and Meno’s Paradox: A Reply to David Charles’s ‘Types of
Definition in the Meno,’ ” Phronesis 55 (2010), 125–52.
Fine, G. The Possibility of Inquiry: Meno’s Paradox from Socrates to Sextus (Oxford, 2014).
Forster, M. N. “Socrates’ Professions of Ignorance,” Oxford Studies in Ancient Philosophy 32
(2007), 1–35.
Geach, P. “Plato’s Euthyphro: An Analysis and Commentary,” Monist 50 (1966), 369–82;
reprinted in R. Kamtekar (ed.), Plato’s Euthyphro, Apology, and Crito: Critical Essays
(Lanham, MD, 2004), 23–34.
Gentzler, J. “Recollection and ‘The Problem of the Socratic Elenchus,’ ” Proceedings of the
Boston Area Colloquium in Ancient Philosophy 10 (1996), 257–95.
Gettier, E. “Is Justified True Belief Knowledge?,” Analysis 23 (1966), 121–23.
Gulley, N. Plato’s Theory of Knowledge (London and New York, 1962).
Irwin, T. “Recollection and Plato’s Moral Theory,” Review of Metaphysics 27 (1974), 752–72.
Irwin, T. Plato’s Moral Theory (Oxford, 1977).
Irwin, T. Plato’s Ethics (Oxford, 1995).
Judson, L. “Hypotheses in Plato’s Meno,” Philosophical Inquiry 41 (2017), 29–39.
Judson, L., and Karasmanis, V. (eds), Remembering Socrates: Philosophical Essays (Oxford,
2006).
Karasmanis, V. “Definition in Plato’s Meno,” in L. Judson and V. Karasmanis (eds.),
Remembering Socrates: Philosophical Essays (Oxford, 2006), 129–41.
Karasmanis, V. “̓Απαγωγή: Hippocrates of Chios and Plato’s Hypothetical Method in the
Meno,” in A. Longo and D. del Forno (eds.), Argument from Hypothesis in Ancient Philosophy
(Napoli, 2011), 21–41.
Lloyd, G. E. R. “The Meno and the Mysteries of Mathematics,” Phronesis 37 (1992), 166–83.
McCabe, M. M. “Escaping One’s Own Notice Knowing: Meno’s Paradox Again,” Proceedings of
the Aristotelian Society 109 (2009), 233–56; reprinted in McCabe, Platonic Conversations
(Oxford, 2015), 190–207.
McDowell, J. Mind and World (Cambridge, MA, 1994).
McGinn, M. “Non-inferential Knowledge,” Proceedings of the Aristotelian Society 112 (2012),
1–27.
Martin, M. F. “Epistemic Openness and Perceptual Defeasibility,” Philosophy and
Phenomenological Research 63 (2001), 441–48.
Menn, S. “Plato and the Method of Analysis,” Phronesis 47 (2002), 193–223.
 THE MENO   181

Millar, A. “Why Knowledge Matters,” Proceedings of the Aristotelian Society Supplementary

Volume 85 (2011), 63–81.
Moravscik, J. M. E. “Learning as Recollection,” in G. Vlastos (ed.), Plato I: Metaphysics and
Epistemology (Garden City, NY, 1971), 53–69; reprinted in J. M. Day (ed.), Plato’s Meno in
Focus (London, 1994), 112–28.
Nehamas, A. “Confusing Universals and Particulars in Plato’s Early Dialogues,” Review of
Metaphysics 29 (1975–1976), 287–306.
Nehamas, A. “Meno’s Paradox and Socrates as a Teacher,” Oxford Studies in Ancient Philosophy
3 (1985), 1–30; reprinted in H. H. Benson, Essays on the Philosophy of Socrates (Oxford,
1992), 298–316, and in J. M. Day (ed.), Plato’s Meno in Focus (London, 1994), 221–48.
Nehamas, A. “Socratic Intellectualism,” Proceedings of the Boston Area Colloquium in Ancient
Philosophy 2 (1986), 275–316.
Nehamas, A. “Eristic, Antilogic, Sophistic, Dialectic: Plato’s Demarcation of Philosophy from
Sophistry,” History of Philosophy Quarterly 7 (1990), 3–16.
Perin, C. “Knowledge, Stability, and Virtue in the Meno,” Ancient Philosophy 32 (2012), 15–34.
Santas, G. X. “The Socratic Paradoxes,” Philosophical Review 73 (1964), 147–64; reprinted with
revisions as ch. 6 of Santas, Socrates: Philosophy in Plato’s Early Dialogues (London, 1979),
183–94.
Schwab, W. “Explanation in the Epistemology of the Meno,” Oxford Studies in Ancient
Philosophy 48 (2015), 1–36.
Scott, D. Plato’s Meno (Cambridge, 2006).
Segvic. H. “No One Errs Willingly: The Meaning of Socratic Intellectualism,” Oxford Studies
in Ancient Philosophy 19 (2000), 1–45.
Sharples, R. W. Plato: Meno (Warminster, 1985).
Taylor, C. C. W. Aristotle, Nicomachean Ethics Books II–IV (Clarendon Aristotle Series:
Oxford, 2006).
Taylor, C. C. W. “Socrates the Sophist,” in L. Judson and V. Karasmanis (eds.), Remembering
Socrates (Oxford, 2006), 157–68.
White, N. P. “Inquiry,” Review of Metaphysics 28 (1974), 289–310; reprinted in J. M. Day (ed.),
Plato’s Meno in Focus (London, 1994), 152–71.
White, N. P. Plato on Knowledge and Reality (Indianapolis, IN, 1976).
Wilkes, K. “Conclusions in the Meno,” Archiv für Geschichte der Philosophie 61 (1970), 143–53;
reprinted in J. M. Day (ed.), Plato’s Meno in Focus (London, 1994), 208–20.
Wittgenstein, L. The Blue and Brown Books (Oxford, 1958).
 chapter 8

The Ph a edo
on Phil osoph y
a n d th e Sou l

Luca Castagnoli

Although its ancient subtitle, On the Soul, captures the focal concern of the dialogue,
the Phaedo is more than an inquiry into the nature and destiny of the soul (psuchē). It
also offers a “second defense” (apologia) of Socrates,1 his life and intellectual activity,
and a philosophical tour de force that spans ethics, moral psychology, epistemology,
metaphysics, philosophical method, and protreptic. This chapter cannot do justice to
this complexity and breadth. Section 1 offers a synopsis of the dialogue, outlining the
interconnection of its main sections and themes.2 Section 2 engages with some central
aspects of the Phaedo’s theory of the soul, in particular the nature of the soul-body
“separation.” Section 3 examines one of Socrates’ arguments for the immortality of the
soul, the “Recollection Argument.” Section 4 reconstructs the Phaedo’s subtle and
pervasive reflection upon philosophy, its methods and goals, and especially upon philo-
sophical argument.

1. Synopsis

The Phaedo dramatizes Socrates’ final hours before his execution in an Athenian prison
in 399 b.c. The framework is the Pythagorean Echecrates’ request to his friend Phaedo
to report what Socrates said before his death, and how he died (57a). The question of the

1 For Plato’s presentation of Socrates’ trial defense speech see the Apology.
2 For analyses of the whole dialogue, and reconstructions of its main arguments, see especially
Hackforth 1955, Gallop 1975, Dorter 1982, Bostock 1986, Rowe 1993. For a helpful annotated
bibliography see Ebrey in Oxford Bibliographies Online. The English translation used here, with
occasional changes, is Sedley and Long 2010.
184   Luca Castagnoli

historicity of the ensuing account is immediately foregrounded: while Phaedo, the

narrator, was present,3 Plato’s absence is (uniquely for the Corpus) emphasised: “Plato
was ill” (59b). Should we interpret this as Plato’s attempt to defend the credibility of his
account, via a trustworthy eyewitness, Phaedo, or as a reminder of Plato’s authorship of
(at least some of) the views and arguments presented by Socrates in the dialogue?4
Phaedo agrees to offer as accurate a report as he can, since “remembering Socrates is
always the greatest pleasure of all” (58d). The theme of pleasure and pain is a fil rouge of
the following pages: witnessing Socrates’ final hours was a peculiar mixture of pleasure
and pain and, despite Socrates’ carefree attitude (58e), the philosophical conversation
was not as purely pleasant as it normally was (59a). Socrates’ first words to his friends
also reflect upon pleasure and pain: they never come at the same time, but they always
follow each other as enemies reluctantly joined at their heads (60c). When Socrates
mentions that they would be a good subject for a fable (60c), Cebes asks why Socrates
had started to compose poetry in jail, putting Aesop’s tales and (his own?) proem to
Apollo into verses (60c–d). Socrates’ choice is an ultimate act of piety: Socrates had
often had a dream telling him to “compose mousikē,” and had interpreted it as encour-
aging him to do philosophy, “the greatest music” or “art of the Muses” (60e–61a); but he
wanted to ensure he would not die without honoring his obligation (61b).

1.1 Death and Philosophy (61b–69e)

Socrates’ invitation to tell the poet Evenus that he should die as soon as possible, if he is a
“philosopher” (61c–d), prompts a discussion of the admissibility of suicide. After men-
tioning the obscurity of the “secret accounts” underlying the Pythagorean prohibition—
“we human beings are in a sort of prison” (62b)—Socrates suggests that, since we are
“the gods’ possessions”, “perhaps in this way it isn’t unreasonable (alogon) that one
should not kill oneself until god imposes some necessity” (62c).5 While Cebes finds this
“likely” (eikos, 62c9) and “reasonable” (eulogōs, 62d2), he suggests there is a contradic-
tion: “it is likely (eikos) that the truth is the opposite . . .: it’s fitting that the wise should
resent dying,” since this is leaving the ministration of good masters, the gods (62d–e).
Socrates welcomes Cebes’ attitude of always scrutinizing arguments (cf. also 77a8–9);
he will need to defend himself “more persuasively” (pithanōteron, 63b4) than he did at
his trial. His attitude springs from the conviction that after death he will enter the

3 Besides Phaedo, the following friends of Socrates are named: Apollodorus, Crito, Critobulus,
Hermogenes, Epigenes, Aeschines, Antisthenes, Ctesippus, and Menexenus, all from Athens; foreigners
included Simmias, Cebes and Phaedondes from Thebes, and Euclides and Terpsion from Megara. For a
prosopography of these and other characters mentioned in the Phaedo see Nails 2002. For Plato’s
choice of characters, especially the Pythagoreans Simmias and Cebes, see Sedley 1995.
4 The Phaedo is often considered a “middle period” dialogue, in which Plato introduces views and
arguments that go beyond the earlier “Socratic” dialogues (and, possibly, beyond the views of the
historical Socrates). For the relative chronology of Plato’s dialogues see Irwin’s chapter 3 in this volume.
5 On Socrates’ argument and its context see Warren 2001.
 The Phaedo on Philosophy and the Soul   185

c­ ompany of wise and good gods, and perhaps of better men (62b–c). At this stage
Socrates’ conviction is qualified: “if there is anything of this sort that I would insist on it
is on that.” Asked to convince his friends, Socrates agrees to offer an account (logos) of
how it seems to him “reasonable (eikotōs) for a man who has genuinely spent his life in
philosophy to be confident about his imminent death” (63e). Socrates’ suggestion that
the “sole pursuit of those who correctly engage in philosophy is dying and being dead,”
and therefore it would be absurd for philosophers to resent death (63e), is met by
Simmias’ playful claim that most people (tous pollous) would agree that philosophers
are “near death,” and deserve to die (64b)!
Socrates now invites his friends to ignore these people, and to “speak among them-
selves” (64c), stating what they believe. Death is agreed to be the separation of the soul
from the body (64c). The true philosopher has no concern, beyond what is absolutely
necessary, for the body and its pleasures, and “releases his soul, as much as possible, from
its association with the body” (64d–65a). Also in the acquisition of wisdom the body is
an impediment—the senses are not accurate and deceive the soul (65b); it is “in reasoning
(en tōi logizesthai)—if anywhere—that the soul discovers something real” (65c). Only
the person who uses his thought, so far as possible, by itself, will come closest to knowing
the Just itself, the Beautiful, the Good, Largeness, Health and Strength, and “will hit
upon reality if anyone will” (65d–66a). Therefore, “either knowledge cannot be acquired
anywhere, or it can be acquired when we are dead” (66e). But it is those who remain
“pure” (katharoi) from the “body’s folly” who “in all likelihood” (hos to eikos) will grasp
the truth (67a–b). According to Socrates, “if all this is true” a philosopher should not feel
irrational fear or resentment, since death is the release of the soul from the body, and the
philosophers’ life-long practice is just that (67b–68d). Those who established rites of
purification for a good afterlife (presumably the Orphics)6 were setting a riddle: “wisdom
itself is the kind of rite to purify us” and the few who are inspired are “those who have
pursued philosophy correctly” (69d). Socrates hopes that his friends will find him more
persuasive (pithanōteros) than his jury did (69e).

1.2 Three Arguments about the Soul (70a–84b)

Cebes’ objection that people often worry that the soul is destroyed once separated from
the body indicates that this requires “no little reassurance and proof (pisteōs).” Socrates
agrees to consider thoroughly “whether or not it’s likely (eikos) to be so” (70b).

(OA) The Opposites Argument (70c–72d). He first appeals to the “ancient doctrine
(logos)” of metempsychosis. To support the logos, Socrates argues that, when things
come to be and have opposites, the opposites (ta enantia) must come from their
opposites. For example, if something becomes larger, it must become larger from
being smaller before, worse from better, asleep from awake, and vice versa. But since

6 On Orphism, see, e.g., West 1983, Betegh 2004, and Edmonds 2011.
186   Luca Castagnoli

being dead and being alive are opposites, not only does being dead come from being
alive, but also being alive must come from being dead. Therefore, Socrates concludes,
the souls that eventually will reincarnate must exist somewhere after death. This con-
troversial argument7 meets Cebes’ strong approval.
(RA) The Recollection Argument (72e–78a). Cebes adds that the soul is immortal
also according to the theory (logos) that Socrates “often propounds,” according to
which “for us learning (mathēsis) is actually nothing but recollection (anamnēsis)”
(72e). Simmias’ request to be reminded of the “proofs” (apodeixeis) of this is
answered by an allusion to the geometrical examination of the slave in the Meno
(81a–86c). Socrates then launches into a complex new argument that infers from the
human capacity to recollect Forms beyond the grasp of the senses that we must have
known them before birth; but if this so, our souls must have preexisted this life
(73c–76d; see Section 3 of this chapter).

What RA has shown is that the existence of the Forms and the preexistence of the soul
are “equally necessary” (76d–e). To Simmias’ objection that prenatal existence does not
imply post-death survival (77b), Socrates replies that the latter has already been proved
(apodedeiktai), provided we combine RA with OA (77d).
Although they endorse RA, Simmias and Cebes still fear death; Socrates is asked to
persuade (anapeithein, metapeithein) “the child actually inside us who is afraid of things
like that” (77e). Socrates invites Cebes to keep chanting spells to this inner child until the
fear is “chanted away”; Cebes’ concern that they will soon lose such an enchanter,
Socrates himself, indicates that philosophical arguments are meant to be, after all,
suitable spells.

(AA) The Affinity Argument (78b–84b). AA addresses directly the fear that the
soul may be dissipated after death (78b). Since the soul is most similar to what is
divine, immortal, intelligible, uniform, indissoluble, and unchanging, namely the
Forms, while the body is similar (homoion) and akin (sungenēs) to what is human,
mortal, multiform, dissoluble, and never constant, namely the sensibles, it is proper
(prosēchei)8 for the soul to be either incapable of being disintegrated after death, or
nearly so (78c–80b).9

From the affinity of the soul with the objects of its intellection, Socrates draws
additional consequences for its destiny. Only if during life a soul has cultivated philoso-
phy correctly, regarding as real only what it “sees” alone by itself, will it go to the place

7 For critical assessment of OA and its shortcomings, see, e.g., Gallop 1975: 103–13 (according to
which OA is “better construed as an opening dialectical move than as an argument to which Plato was
seriously committed” [104]); Barnes 1978; Gallop 1982; Bostock 1986: 52–69. See also my brief
comments in Section 2.1 of this chapter.
8 For the suggestion that prosēchei indicates the mere likelihood of the conclusion cf. Rowe 1993:
188–89. Cf. also the use of eikos at 78c7.
9 For critical assessment of AA and its shortcomings, see, e.g., Gallop 1975: 137–46; Bostock 1986:
116–21; Apolloni 1996.
 The Phaedo on Philosophy and the Soul   187

akin to it, noble, pure, and invisible, “Hades as it truly is,”10 where it will be happy
(80c–81a; 82e–83d; cf. 63b–c).

1.3 Reflections upon, and Objections to,

Socrates’ Arguments (84c–91e)
After Socrates acknowledges that what he said “still contains many grounds for suspicion
and counter-attack” (84c), Simmias objects that although the attunement (harmonia) of
a lyre is “something invisible, incorporeal, and utterly beautiful and divine,” clearly it
will be destroyed as soon as the lyre is smashed. Since Simmias, and some unnamed
Pythagoreans, think that the soul is an attunement or proportionate blend of bodily
elements,11 the soul must perish “when our body is loosened or tautened beyond pro-
portion by illness or other evils” (85e–86d).
Cebes argues that although the preexistence of the soul has been “quite sufficiently
proved” by RA, its post-death survival has not (87a). It seems reasonable, on the basis of
AA, that the soul is longer-lasting than the body; nonetheless, even granting that some
souls survive for a number of reincarnations, they need not do so indefinitely, as already
allowed by the disjunctive conclusion of AA (88a–b).
Simmias’ and Cebes’ objections prompt a long reflection on the force of philosophical
argument (88c–91c), to be examined in Section 4 of this chapter.

1.4 Socrates’ Counterarguments, Intellectual

Autobiography, and Final Argument (92a–107b)
Socrates first tackles Simmias’ objection. While Simmias still claims to be strongly
convinced by RA (92a), his suggestion that the soul is harmonia does not “work in con-
cert” with that conviction: how can an attunement of bodily elements exist before the
body (92b–c)? Simmias must choose: and he chooses RA, because unlike the harmonia
theory, which has come “with no proof but with a sort of plausibility (meta eikotos
tinos)” (92d), RA “has been provided by means of a hypothesis (di’hupotheseōs) worth of
­acceptance,” the existence of the Forms (92d–e). Two additional arguments follow, con-
cluding that on the harmonia hypothesis it would be inconsistent to believe that the soul
could oppose the body, or rule over it, and that it could have different degrees of virtue
and vice (93a–95a).
After a painstaking recap of Cebes’ argument (95a–e), Socrates claims that to tackle it
“we must study thoroughly and as a whole the cause of coming-to-be and ceasing-to-be”

10 The Greek Haidēs is etymologized here from aidēs (“invisible”); the etymology is rejected at
Cratylus 404b.
11 For the view that Philolaus was the originator of the doctrine cf. Huffmann 1993 and Sedley 1995.
188   Luca Castagnoli

(95e–96a). He approaches the issue by recounting a sort of intellectual autobiography

that started with the interest in “the sort of wisdom that they call ‘research into nature’
(peri phuseōs historia),” consisting in the search for “the causes (tas aitias) of each thing”
(96a). After the bitter disappointment with Anaxagoras’ book, in which Socrates
expected to find the teleological explanations he was hoping for (97c–99c),12 he
embarked upon a “second voyage in search of the cause” (99d). By an anticipation and
partial reversal of the imagery of the Sun in the Republic (507d–509c), Socrates explains
how he realized that “looking into the truth of things” in logoi, rather than through the
senses, was the correct method. As a result of this, he says, “on every occasion I hypoth-
esise whatever theory I deem more robust, and then I set down as true whatever I think
harmonises (sumphonein) with it—both about cause and about everything—and as
false whatever doesn’t” (100a). Socrates’ “first hypothesis,” the existence of the Forms
(100b), leads to the “safest” way to offer causal explanations: if anything is F, it is F not on
account of some material cause, but by having a share of the F itself—Socrates is non-
committal as to the mode of Form-particular causation (100c–d).13
After a short discussion of how the hypotheses themselves should be tested and
accounted for14 (101d–e), Socrates launches into yet another argument for the immor-
tality of the soul.

(FA) The Final Argument (102d–106e). The kernel of the argument is the suggestion
that a “more ingenious” way to explain what makes a certain thing F (e.g., hot) is to
say that the cause is not the F itself (Hotness), but some X (e.g., fire) that “imports”
F-ness (heat) wherever it comes to be present (106b–c). Even if such Xs do not them-
selves have opposites, like opposites they do not admit whatever form (idean) is
opposed to the “form in them”; for example, when fire approaches, snow must either
retreat or perish, because heat is opposite to the cold in the snow (104b–105b). But
since the soul is what makes a body alive whenever it comes to be present in it (it
“imports” the form of life) (105c–d), and death and life are opposites (105d), the soul
will never admit death (105d, 106b). Therefore, Socrates concludes, the soul is
immortal (athanaton), imperishable (anolethrōn) and indestructible (adiaphthoron)
(106e), and retreats intact when the body dies (106e).15

Although Cebes and Simmias admit that they cannot dispute FA (107a), Simmias retains
some doubt, because of the magnitude of the issue and his awareness of human weak-
ness (107b). This caution meets, once again, Socrates’ approval, with the invitation to all
those present to examine the “first hypotheses” more clearly, and follow the argument
“as far as a human being can follow it up” (107b).

12 Cf. Sedley 1995: 8 for the suggestion that the Timaeus will fulfill Socrates’ expectations.
13 On causation in the Phaedo, see, e.g., Vlastos 1969, Sedley 1998.
14 For the “method of hypothesis” in the Phaedo see, e.g., Robinson 1953, Kanayama 2000,
Bailey 2005, Benson 2015.
15 For the reconstruction and critical assessment of FA, see, e.g., Gallop 1975: 192–222; Frede 1978;
Bostock 1986: 178–93; Denyer 2007; Sedley 2009.
 The Phaedo on Philosophy and the Soul   189

1.5 The Eschatological Myth and Socrates’ Death (107c–118a)

Socrates exhorts his friends to care for their souls during life, since their destinies in
Hades will depend on the degree of purity, wisdom, and virtue they had during life
(107e–108c). A description of the shape of the earth follows, reminiscent of the analogy
of the Cave in the Republic (514a–517a); we human beings actually “dwell in a certain
hollow of the earth,” but on the surface everything is purer and more beautiful
(108d–111c). After the judgment, some souls are sent to Tartarus (forever, if “incurable,”
or temporarily, if guilty of grave offenses), some to the Acheron (if “found to have lived
average lives”), and some to the “pure dwellings” on the earth’s surface, if they have lived
“exceptionally pious lives,” for which they are rewarded by being “freed and separated
from these regions inside the earth, as if from prisons” (113d–114c). Among the latter
group, the souls “who purified themselves sufficiently with philosophy” seem to escape
the cycle of reincarnations awaiting the other souls (107e).16
While Socrates acknowledges that “it does not befit a man of intelligence to insist”
that things are exactly as described in the myth, he claims “it is both fitting and worth the
risk” to insist that at least something like this is true, and “chant such things to oneself ”
(114d)—interestingly, the persuading chant here is a muthos, not a logos. The implication
of the exhortation to “take a risk” (kindunos) is that, pragmatically, it would be worth
believing the myth, even if it were false, because this will encourage us to be better people
(compare “Pascal’s Wager”).
The final pages of the dialogue recount Socrates’ last moments. As Socrates drinks the
poison fearlessly, many of his friends cannot hold back their tears (117b–d), until they
are shamed by Socrates (117e). Socrates’ cryptic and much debated last words are: “Crito,
we owe a cock to Asclepius. All of you must pay the debt and not be careless” (118a).17

2. The Soul: Separation

That human beings have something called a “soul” that is somehow distinct from the
body is taken for granted throughout the Phaedo, starting from the introduction of the
term psuchē, when death is defined as the separation of the soul from the body (64c).
This starting point might appear problematic. To begin with, it is introduced after
Socrates’ invitation to ignore what hoi polloi think, but to say what “we believe,” where
“we” seems to refer to Socrates and his circle of friends. Socrates might be charged with

16 For the significance of the myth within the Phaedo see Sedley 1989 and Kamtekar 2016; for its
relation with Orphic and Pythagorean ideas see Kingsley 1995 and Betegh 2006; for other Platonic
eschatological myths see Annas 1982.
17 Does this suggest that Socrates is thanking the god for healing him from the disease of the
embodied life, or liberating his soul from its bodily prison? Or that he is praying for the conversion of
his friends to the philosophical life? Or does this allude to the healing of Plato, Socrates’ worthy
successor? For alternative interpretations see Most 1993, Crooks 1998, Madison 2002 and Kamen 2013.
190   Luca Castagnoli

preaching to the converted, assuming without argument a conception of human nature

in need of demonstration. Moreover, it has been objected that this account of death begs
the question in support of the soul’s immortality, by presupposing that the soul can be
separated from the body.18
To the first objection one can reply that the existence of the soul is not a controversial
philosophical tenet: the Pythagorean and Orphic doctrines with which Socrates engages
in the dialogue endorse it, and the popular Homeric conception of Hades equally pre-
supposes the soul’s post-death existence. One way of conceiving of Plato’s project in the
Phaedo is that it aims to provide a better account of, and suitable foundations for, those
popular and philosophical ideas. While popular myths describe the survival of the
psuchē after death, they offer a distorted picture of what it amounts to, or could amount
to if one lived a philosophical life: they depict Hades as a bleak place, in which our souls
are wholly deprived of intelligence and identity.19 It is to address this conception that
Cebes asks Socrates to prove that after death the soul not only exists, but “has some power
and wisdom” (70b3–4). The Pythagorean insight that the soul is immortal lacks solid
epistemological and metaphysical foundations, and clashes with another Pythagorean
idea, that the soul is a sort of bodily harmonia20 (compare Philolaus’ lack of clarity as to
why we should not commit suicide (61d), or the mysterious Pythagorean saying that our
body is a prison). Finally, the Orphics’ myths of original sin and metempsychosis,21 and
their rituals, fail to explain why the soul is condemned to a never-ending cycle of rein-
carnations, and to identify the correct (philosophical) means of purification.22 However,
the question remains of whether Plato’s arguments could satisfy critics who reject any
talk about the soul. As for the second objection, the separability of the soul is never
treated as a sufficient condition for its immortality: as we have seen, what worries many
is that the soul might be destroyed immediately or shortly after it separates from the
body, “like breath or smoke” (70a5).
What is more interesting is the variety of ways in which the distinction and separation
of soul and body are described and supported in the Phaedo. In this area Plato’s seam-
less interweaving of metaphysics, epistemology, ethics, moral psychology, and meta-
philosophy is especially apparent. The ontological difference between the soul and the
body is not the focus, unlike in some forms of strong dualism (e.g., Cartesian), accord-
ing to which soul (or mind) and body are separate substances with completely distinct
sets of properties. While AA points toward some kind of ontological distinction, the
kinship between the soul and the Forms is described in terms of similarity: not only does
the soul not have all the properties of the Forms (the soul, for example, is not unchang-
ing), but also it shares some properties with sensible objects, and therefore with the body
(for example, spatial extension, or at least location [cf. 79e: “whenever soul and body are
in the same place”]). Moreover, Socrates first introduces the soul-body distinction by
reflecting upon different kinds of life, the philosophical and the non-philosophical.

18 See, e.g., Gallop 1975: 86–87. 19 Cf. e.g., Meno 100a.

20 Cf. Sedley 1995. 21 Cf. e,g, Meno 81a–d.
22 On the controversial relationship between Orphism and early Pythagoreanism see Betegh 2014.
 The Phaedo on Philosophy and the Soul   191

The philosopher’s constant attempt to turn away from bodily pleasures, pains, and
­concerns, and to use reasoning in the pursuit of truth, becomes an illustration of the very
possibility, and desirability, of attaining already in this life a certain degree of separation
of the soul from the body, understood as intentional and functional independence.23
The separation also has an epistemological dimension: the body is not only a distraction,
but an actual impediment if we misguidedly rely on it to pursue knowledge.
The normative paradigm of the philosophical life suggests that the degree of the
body-soul separation is somewhat fluid: during our lives, our souls can, and ought to,
become more and more separate from our bodies. While the adoption of the metaphor
of the body as a prison of the soul might suggest both strong ontological distinction and
separation (a prison and a prisoner are completely different types of things, and each
can exist without the other),24 the picture is more complex: the soul itself can become
more and more “infected” by bodily elements, and remain so even after death, if it exces-
sively attaches itself to the body and its concerns throughout life (81c–d).25 In fact the
dualism suggested by the prison and release metaphor not only does not match the
metaphor of the “infection” and “purification” of the soul, but is itself qualified later on
by the claim that the body is a clever prison that “works through desire, the best way to
make the prisoner himself assist in his imprisonment” (82e). While certain desires are
sometimes described in such a way as to appear not only to originate from the body, but
also to be of the body, the imprisoned soul ultimately comes to desire and enjoy (and,
strikingly, believe) the same things as the body, thus imprisoning itself (83d). Similarly,
the soul being “weighed down” by the corporeal after death amounts to, or results in, it
fearing something (Hades) and desiring something (corporeal things) (81d–e). Although
there is no suggestion in the Phaedo that the soul is tripartite, unlike in the Republic, the
Phaedrus, and the Timaeus,26 then, the soul itself, as a whole, is (or can become) the
subject of the lowest appetitive desires.
The soul is sometimes said to be imprisoned or enslaved to the body, sometimes to
rule over it. The suggestion that nature instructs the soul to rule, and the body to play the
slave (80a), brings home a normative notion not only of the separation of the soul from
the body, but also of its superiority. As we have seen, the degree of separation that the
soul attains in life also has fundamental ethical and eschatological implications. Only a
purified philosophical soul can be virtuous in this life, and will go on to reap the rewards
of enhanced separation in the afterlife (cf. Section 1.5 of this chapter); on the other hand,
the soul’s self-imprisonment and self-enslavement are the source of vice and are pun-
ished accordingly after death.
According to Socrates, the soul is clearly the best part of us. Does this mean that we
are, ultimately, our souls? The demonstration of the immortality of the soul is expected

23 I borrow these concepts from Johansen 2017. For a different debate on the stance involved in
philosophical separation (ascetic versus evaluative) see Woolf 2004.
24 Although, of course, once separated from the prison the prisoner will no longer be a prisoner
(unless/until transferred to another prison).
25 For Cebes’ different metaphor of the body as a cloak worn (and weaved) by the soul cf. 88a–b.
26 Cf. Lorenz’s chapter 21 in this volume.
192   Luca Castagnoli

to relieve us of our fear of death; this might suggest an affirmative answer, just as Socrates’
reprimand of Crito for his request to leave instructions on what to do with him after
death: “you should cheerfully say that you’re burying my body” (115); “I will not stay
behind when I die, but will depart” (115d). However, Socrates’ own claim is problematic:
if Socrates is his soul, Socrates will not die; and if the death of X is the separation of X’s
soul from X’s body, then X cannot be the same as his soul. Throughout the dialogue the
relationship among an individual, his soul, and his body is described in unstable terms
(cf. e.g., 64e4–6: the philosopher’s “concern is not for the body . . . as far as he can, he
stands apart from it and is turned towards his soul”—which suggests that an individual
is different from both his body and his soul). This is not just a question about the identity
of the human “self ”: there are implications for the soundness of Socrates’ arguments.
For example, the logic of OA seems to require that the subject that undergoes, cyclically,
the opposites “being alive” and “being dead” is a certain individual, and not a soul;27 but
what comes to be alive again is neither the soul (the souls of the dead are not dead souls,
but they are reborn “from the dead”) nor the individual human being, for example,
Socrates, who died, but an individual who happens to have the soul that was formerly
Socrates’ (and who, thereby, does not come to life again). The idea that we ought to
identify, as far as possible, with our soul and with its correct, natural exercise, shows that
the answer to the question “Is X his soul?” might ultimately depend on the degree of
philosophical purification that X has reached.

3. Immortality and Recollection

Socrates’ arguments for the preexistence and immortality of the soul are best studied in
the context of his account of philosophy as a preparation for death, of the soul-body
separation, and of the eschatology of the final myth. It is equally important to examine
their interrelations: Do they all argue for the same conclusion? Are they presented as
carrying different force? Do they build upon each other? Do they rely on different but
compatible sets of premises? In the remainder of this section I will focus on RA (for
short summaries of OA, AA, and FA see Section 1 of this chapter).
As we have seen, RA is introduced by Cebes immediately after OA (72e–73a):

(1) For us learning (mathēsis) is actually nothing but recollection (anamnēsis);

(2) We must have learned at some previous time what we recollect now;
(3) Therefore our souls must have existed before being born in this human form;
(4) Therefore the soul is immortal.

Since OA had received Cebes’ endorsement, RA is intended to offer additional

grounds for the immortality of the soul, and not to build or improve upon OA. But

27 The idea that after its departure from the body the soul exists, but is dead, would also be
incompatible with the later FA.
 The Phaedo on Philosophy and the Soul   193

Cebes’ outline does not explain why, or in what sense, learning is anamnēsis, and does
not offer any insight into how the soul’s preexistence and immortality are supposed to
follow from that tenet in what is, at best, a very elliptical argument (prima facie (3) does
not follow (1) and (2), and (4) does not follow (3)). Cebes answers Simmias’ request to
“refresh his memory” about the proofs for recollection (hupomnēson me) with an allu-
sion to the geometrical examination of the slave in the Meno (81a–86c): “when people
are questioned, provided someone questions well, they themselves come up with true
statements about everything. And yet they wouldn’t be able to do so, if knowledge and a
correct account were not actually inside them. For example, if one confronts them with
diagrams or something else of the kind, that is the situation in which one shows most
clearly that this is the case” (73a–73b). In what follows are we invited by Plato to see the
Phaedo as offering new arguments for, and perhaps a better, more precise account of, the
same “theory” as the Meno’s, or a different version?28
Simmias is convinced by the Meno’s proof of recollection; he now “remembers”
(memnēmai), but still needs to “undergo the very thing that the theory is about, recol-
lecting (anamnēsthēnai)” (73d). The distinction between having one’s memory “jogged”
or “refreshed” (hupo-mnēsis) by someone else, and as a result remembering something
(mnēmē), on the one hand, and recollecting (ana-mnēsis), on the other, is emphasised
here—this might indicate that only recollection guarantees proper learning and
understanding.29
Recollecting the truth of recollection turns out to require a much more complex argu-
ment in the Phaedo than in the Meno. First Socrates clarifies some conditions under
which we ordinarily say that someone “is reminded of,” or “recollects” something:

(1) If X is reminded of (recollects: anamnēsthēsetai) Y, X must have known Y previ-

ously. (73c1–2)
(2) If X, by perceiving something Z, not only recognizes Z, but also thinks of some-
thing else Y, and Z and Y are objects of different knowledge,30 X has been
reminded of (has recollected) Y. (73c5–d1)
Examples: seeing a musical instrument (Z) and being reminded of its owner (Y); seeing
Simmias and being reminded of his friend Cebes; seeing a portrait of Simmias and being
reminded of Simmias himself.
(3) X recollects Y in this way especially when X had forgotten Y, because of pro-
tracted inattention. (73e1–3)
(4) X can be reminded of Y not only by some similar (homoion) Z, but also by some
dissimilar Z (74a) (provided that Z is somehow “associated” to X, 76a3)

28 For discussions of recollection and its context in Plato see e.g. Gulley 1954; Irwin 1974; Scott 1995;
Kahn 2006; McCoy 2011.
29 For the distinction see Phaedrus 274e–275b.
30 For discussion of the clause that the knowledge of Z and Y must be different see e.g. Dancy 2004:
256–64 and Fine forthcoming.
194   Luca Castagnoli

Examples: the lyre is not similar to its owner, but associated with it by habit; Simmias’
portrait is similar to Simmias himself.
(5) When X is reminded of Y by some similar Z, necessarily X thinks whether or
not Z is lacking in similarity in relation to Y. (74a5–8)

Once these conditions are agreed upon, Socrates’ argument proceeds as follows:

(6) We say (or “should say,” phōmen) that there is something equal, besides the
various equal things (“sticks and stones”) we experience, “the Equal itself ”
(auto to ison), and we know (epistametha) what it is. (74a–b)
(7) We thought and got knowledge of the Equal itself from perceiving equal things.
(74b)
(8) The Equal itself is different from these things, since they, unlike it, sometimes
appear equal to something, but not to something else (or appear equal to some-
one, but not equal to someone else; or equal in one respect, but not equal in
another; or equal at one time, but not at another).31 (74b–c)
(9) So we recollect the Equal itself, “whether it is similar or dissimilar to the equal
things.” (74c–d) [From (2), (7) and (8)]

From our recollection of the Equal itself, Socrates infers the soul’s preexistence:

(10) The equal things strive to be like the Equal itself, but fall far short and are
­inferior. (74d)
(11) If X, on seeing Z, thinks that Z strives to be like some other real thing Y, but falls
short of it and is inferior, X must have known Y previously. (74d–e)
(12) Therefore, we must have known the Equal itself, before the time when we first
thought that equal things strive to be like the Equal, but fall short of it. (74e–
75a) [From (10) and (11)]
(13) It must have been before we began to use our senses that we had knowledge of
the Equal itself. (75b)
(14) We began to use our senses just as soon as we were born. (75b)
(15) Therefore, we must have acquired knowledge of the Equal itself before we were
born (75c), and equally of all the other Forms, like the Beautiful itself, the Good
itself, the Just itself. (75d) [From (13) and (14)]
(16) We lost this knowledge when we were born (we forgot), but later we are
reminded by the senses and we learn, that is, we regain knowledge we once had
(analambanomen tas epistēmas has pote kai prin eichomen), we recollect.
(75e–76c)
(17) Therefore, our souls existed before birth, separate from bodies, and had knowl-
edge. (76c) [From (14) and (15)]

31 For discussion of the different Greek texts and translations see Sedley 2007.
 The Phaedo on Philosophy and the Soul   195

Most of these steps have attracted intense scholarly scrutiny, either on account of
some alleged unsoundness or invalidity, or due to some lack of clarity.32 For example, if
(1) expresses a necessary condition for recollection, foreknowledge, it would seem that
(2) cannot express a sufficient condition: the perception of something can prompt the
thought of something new that I never thought (let alone known) before.33 And why
should it be the case that whenever I am reminded of something Z by something similar
to it, Y, I must also consider whether Z is lacking in similarity (5)? Must I always (rather
than often, or typically) reflect upon the degree of similarity between a person and his
portrait, for example, when upon seeing the portrait I am reminded of the person? If
recognizing the lack of similarity is merely realizing that Y and Z are different things,
that is, not mistaking Y for Z, then the implicature of (5), that Y might not lack in simi-
larity, becomes problematic.34 And how can inanimate objects, such as equal sticks and
stones, “strive” to be like what they resemble, the Equal itself (10)?35 Finally, why does
RA conclude that the soul must have preexisted this life “separate from the body”?
Could each incarnation have ensued immediately after the death of the previous “soul-
bearer,” without discarnate “intermissions”?36
The deficiency of the sensibles in their resemblance to the Forms is a central meta-
physical question raised by RA. If we read the Phaedo in the light of other accounts of
the distinction between sensibles and Forms, such as the Republic (e.g., 479a–b) and the
Symposium (210e–211b), it appears natural to assume that the compresence of the oppo-
sites F and not-F in the sensibles, and their being F (and not-F) only qualifiedly, is what
makes them different from, and inferior to, the F itself.37 This reading has been chal-
lenged in favor of the suggestion that conflicting appearances, or diachronic change, are
possible in the case of the sensibles but not of the Forms.38
What is clear is that the Forms become the focus as objects of recollection from (6)
onward.39 In this respect the Phaedo advances upon the Meno’s presentation, which
failed to clarify the nature of the objects of anamnēsis.40 But the reference to the Forms

32 For the reconstruction and critical assessment of RA see e.g. Ackrill 1973; Gallop 1975: 113–36;
Bostock 1986: 60–115; Bedu-Addo 1991; Scott 1995: 53–73; Dimas 2003; Dancy 2004: 253–83;
Franklin 2005; Fine forthcoming.
33 See e.g. Ackrill 1973; Bostock 1986: 63–64. 34 Sedley 2006: 313–16. 35 Sedley 2006.
36 Perhaps, as Gallop 1975: 133 suggests, this envisages the necessity that the soul first learned what it
can now recollect when enjoying some discarnate existence (to avoid the risk of infinite regress: “how
did you learn it then?”). Cf. Phaedrus 245c–257b, in which discarnate vision of the Forms is described
in the myth of the charioteer. For the suggestion that the soul never learned, but always was in a state of
knowing, or “having learnt,” cf. Meno 86a.
37 Cf. e.g. Bostock 1986: 72–94; Irwin 1999. For the questions whether in Plato the sensibles opposed
to the Forms must be understood as sensible particulars, see Harte’s chapter 19 in this volume.
38 The reading of step (8) is central to the debate. Cf. e.g. Gosling 1965; Sedley 2007.
39 The examples in (2) and (4) suggest that they are not the only objects of recollection. But the
recollection that guarantees the preexistence of the soul (and possibly that constitutes learning) must be
recollection of the Forms.
40 The focus on the Forms of Equal, Greater and Smaller in RA can be seen as another link to the
geometry lesson of the Meno. Alternatively (or additionally) these Forms might be singled out because
they are easier to know even for people who are not yet at an advanced stage of philosophical
development (see note 42 of this chapter).
196   Luca Castagnoli

brings to the fore a twofold problem. To begin with, the acceptance of the theory of
recollection, and RA as a whole, now depend upon accepting the existence of the Forms.
According to (6), “we” claim that the Equal itself exists—and later on, in (15), that the
other Forms exist. It is clear that “we” refers, once again, to Socrates himself and his
interlocutors. But this agreement, although already foreshadowed earlier in the dia-
logue (65d–66a), would not be accepted, or possibly even understood, by many outside
the Socratic circle (cf. e.g., 81b4–5). RA therefore rests upon a controversial metaphysi-
cal tenet. Socrates gladly acknowledges this, when he concludes that the existence of the
Forms and the preexistence of the soul are “equally necessary”:41 “if the things which are
our constant refrain really exist, I mean a Beautiful and a Good and all that sort of being,
and if we refer to this being everything derived from our senses . . . then just as these
things exist, so too must our soul exist even before we are born. But if they don’t exist,
then wouldn’t this argument turn out to have been propounded to no effect?” (76d–e).
As we have seen, the role of the Forms as “hypotheses” is confirmed later at 92d–e, and
conceptualized more extensively in FA.
The second problem is that RA relies not only on the fact that “we” believe in the
­existence of the Forms, but also that “we” recollect and know (some of) the Forms ((6),
(7) and (9)).42 On the interpretation of the role of anamnēsis according to which it
accounts for basic concept formation since early childhood, we (with the exception of
small children) all recollect (at least some of) the Forms, however partially and dimly.43
Recollection explains the shared human capacity for conceptualization. On a different
interpretation that has been especially promoted by Scott 1987 and 1995,44 in the Phaedo
(but also in the Meno and the Phaedrus) recollection is an arduous process of philosoph-
ical inquiry and understanding that only few undertake and complete: the mundane
recognition and classification of objects of perception as, for example, equal is not the
result or manifestation of our recollection of the Equal itself (in fact it might be a neces-
sary precondition for recollection). What makes it difficult to adjudicate between these
two interpretive lines45 is that different steps of RA would make one lean toward either
side. Whereas the reflective activity implied by (11) appears difficult to attribute to non-
philosophers, (12), (13), and (14) seem to suggest, cumulatively, that the process of recol-
lection starts very early on in life—although it may end with full grasp of (some of) the
Forms only much later (or, in fact, not end at all). But the use of “we” to refer to philoso-
phers who engage in dialectical inquiry into the Forms in (15) shifts the focus back to
recollection as a higher-level activity. And Socrates’ argument to establish that we are

41 Cf. Gallop 1975: 135 for discussion of this puzzling claim.

42 There is an apparent inconsistency between the claim that “we” know the Equal itself and the later
claim that no one, with the exception of Socrates, can give an account of the Forms, and therefore
know them (76b). The inconsistency has been explained by referring to different scopes of “we”, to
different types or levels of knowledge (e.g. Bostock 1986: 67–69), or to different groups of Forms (e.g.
Scott 1995: 67–68).
43 Cf. e.g. Gulley 1954; Ackrill 1973; Kelsey 2000. Bostock 1986: 68 suggests different levels of
knowledge, so presumably different degrees of recollection.
44 Cf. also Fine 1983: 137–38 and Dancy 2004: 253–83.
45 For critical discussion of Scott’s proposals see Williams 2002.
 The Phaedo on Philosophy and the Soul   197

not born knowing, but we forget at birth and recollect later (16), relies on the premise
that very few people know the Forms, rather than on the observation that young chil-
dren are unable to conceptualize (76a–c).
Another key question raised by RA is what form of innatism, if any, the argument is
trying to establish. The orthodox reading of Platonic recollection as a theory of innate
knowledge has been challenged in recent years. Fine 2014 argues extensively that the
Meno assumes our soul’s prenatal knowledge, but not current latent knowledge, as the
grounds and explanation for our ability to inquire and recollect, and makes a similar
case for the Phaedo.46 Here again Plato’s own words tantalizingly appear to pull in different
directions. On a straightforward reading of Cebes’ paraphrase of recollection, it seems
that at the moment of recollection we do have knowledge in us: “And yet they wouldn’t
be able to do so, if knowledge and a correct account were not actually inside them.”47 On
the other hand, the idea that forgetting is “loss of knowledge” (75d10–11: epistēmēs
apobolēs) and that (16) we forget and “lose” our knowledge of the Forms at birth (75e2–3:
apōlesamen) suggests that there is no knowledge in us at birth, whether explicit or latent.
One could explain the discrepancy by imputing the idea that knowledge is innate to
Cebes’ own poor recollection of Socrates’ lesson. Alternatively, it could be argued that
Cebes does faithfully represent the theory of the Meno, which is liable to be interpreted
as a form of knowledge innatism, but the Phaedo clarifies that anamnēsis is not of innate
knowledge.
I believe that Socrates’ formulation of RA might suggest that human beings “possess”
knowledge, in some sense, even after “losing” it at birth.48 Suppose I lose all my money
in an ill-advised investment; the money is no longer mine, nor is it recoverable in the
future (I could make more money in the future, but that money is lost forever). But if I
lose my wallet in the street, the wallet is still mine; I can hope to recover it. Imagine now
I lost my keys in my house; the keys are still mine in an even fuller sense, even if I am
temporarily unable to find and use them. Finally, I might have lost an important docu-
ment in my office, and even forgotten about its existence; but the document is still there,
and is in my possession (I could come across it in the future while looking for something
else, or I could be somehow reminded of its existence, and start to look for it again). In
other words, there are a number of ways in which I can lose something but still possess
it, even if I am temporarily unable to access and use it—in fact even if I will never be able
to access it again. Knowing (to eidenai) is described as “having got (labonta) knowledge
of something, to hold it (echein) and not have lost it (mē apolōlekenai)” (75d9–10).49

46 Fine 2014: 171–73. The argument is expanded in Fine forthcoming.

47 For the ingenious suggestion that enousa could be taken to refer to past, prenatal presence see
Fine 2014: 172.
48 On this reading, the idea that forgetting is the “loss” of knowledge is different from the idea that it
is “the departure of knowledge” (from the soul) at Symp 208a4: (lēthē gar epistēmēs exodos). Cf. also
Phlb. 33e3, where forgetting is defined as the departure of memory (lēthē gar mnēmēs exodos).
49 Cf. also the use of echontes at 75c7. Contrast this with the Theaetetus account, in the Aviary
section, according to which knowing is not “the having (hexis) of knowledge,” as some people say, but
“the possession (ktēsis) of knowledge” (197b). It could be argued that one failure of the Aviary model is
exactly that it identifies know-ing with the mere “possession” of knowledge.
198   Luca Castagnoli

Although having lost/forgotten knowledge of something certainly means not knowing

it, it need not imply possessing no knowledge at all of it. Socrates claims that when we
recollect/learn we “recover” (analambanein)50 our own (oikeian) knowledge (75e5–6)
or “rediscover” (aneuriskein) “what was there before” (huparchousan proteron) and “is
ours” (hemeteran ousan) (76e1–2); these claims can point to the possession of latent
knowledge that is temporarily unavailable to us but still “ours,” or “in us,”51 while exclud-
ing that we “hold” (echein) the knowledge (i.e., have it “at hand” to use) at birth, or at any
time thereafter until we recollect.
The role sense-perception plays in recollection is probably alluded to at the beginning
of RA, when Cebes mentions diagrammata,52 and definitely emphasized throughout
(cf. (2), (7), (11), (16)). But how can this emphasis, absent from the Meno, be reconciled
with Socrates’ exhortation to separate our reasoning from the bodily senses?53 A dis-
tinction should be drawn between the ideas (1) that the senses confer knowledge of the
Forms, (2) that we acquire knowledge of the Forms from the senses (say, by a process of
abstraction or induction in which reason operates exclusively upon sensory data), and
(3) that the senses trigger the soul’s recollection of the Forms, that is, recovery of knowl-
edge from within. Clearly there is no reason to think that the role of the senses in recol-
lection is (1); nor should we assume that (2) they provide all that is needed for our reason
to attain knowledge, as empiricist epistemologies would allow. After all, we do not
acquire knowledge of Simmias by looking at his portrait, or even by reasoning about it;
seeing the portrait triggers recollection of our existing knowledge of Simmias. If recollec-
tion starts very early on in human life, perhaps after their early triggering the senses no
­longer need to, or should, be used as we proceed to higher stages of recollection and turn
inward. Those who identify recollection only with the highest levels of philosophical
understanding, on the other hand, will need to offer a different explanation for the role
of sense perception in RA. Since philosophical recollection can presumably start from
empirically formed concepts that are not themselves sense perceptions, it must be only
indirectly that we “recollect from” the senses or are “reminded by” them.
Either way, the role played by the senses differs from the one envisaged later on in
Socrates’ autobiography. One question that had captured the young Socrates’ attention
was what the physical “organ” of thought and perception is, and whether memory and
opinion come from the senses, and knowledge from them (96b). This approach points
to a theory of cognition, and of the role that perception and memory (mnēmē) play in it,
different from the one underlying the theory of anamnēsis, and anticipating the empiri-
cist epistemology sketched by Aristotle in Metaphysics I 1 and Posterior Analytics II 19.
That it is mentioned early on in Socrates’ autobiography suggests that the theory was
later rejected by Socrates.

50 For a different reading of the verb analambanein in the discussion of recollection in the Meno
(“taking up” or “working out”) cf. Fine 2014: 150.
51 Cf. Meno 85d6–7: to recollect is to recover (analambanein) knowledge oneself in oneself (auton en
hautēi).
52 Contra Gulley 1954: 197.
53 Cf. Gulley 1954: 198–200. On perception in the Phaedo see Fine 2017.
 The Phaedo on Philosophy and the Soul   199

Although Cebes introduced RA as an argument for the soul’s immortality, what RA

purportedly proves in (17) is its preexistence.54 When Simmias objects on these grounds
(77a–b), Socrates replies that the soul’s post-death survival is also established, if we com-
bine RA with OA: if the soul preexists, and at birth comes from death to life, then it must
also exist after it has died, given that it has to be born again (77a–d). Whatever we make
of this puzzling argument,55 it confirms what we saw at the beginning of this section: OA
and RA are not merely juxtaposed, let alone competing, arguments, but are supposed to
be consistent. Whereas RA and its conclusion are referred to again several times with
approval (87a, 91e–92a, 92d), OA is not mentioned again as a convincing argument in
the rest of the dialogue. It is to further persuade Simmias and Cebes that the soul is not
the kind of thing that could be destroyed, and therefore is immortal, that Socrates intro-
duces AA and FA.56

4. Philosophy

The Phaedo is not only a piece of philosophical prose, but also a sustained and sophisti-
cated reflection upon philosophy itself. This reflection is undertaken explicitly, through
the interlocutors’ frequent remarks about the effects of philosophy in human life, and
about the methods of philosophical inquiry; but it is also implicitly suggested, through
the dialogue’s illustration of philosophy “in action,” and especially of Socrates’ use of
argument in his last conversation.
The term philosophia first occurs early in the dialogue, when Phaedo comments that
although Socrates and his friends engaged in philosophy, and philosophical logoi, the
experience was not as purely pleasurable as usual (59a). This encourages us to think of
what follows in Phaedo’s report as a philosophical conversation, and a collection of
philosophical logoi. As we have seen, Socrates first describes philosophy as a form
of mousikē, performed as an act of devotion to the gods (60e–61a).57 The idea that
philosophy—or, more precisely, philosophy performed “in the right way” (60d2, 64a4,
67e4, 69d2, 80e6, 82c3) or “worthily” (61c8), by “true” philosophers (64b9, 64e2, 66b2,

54 Recollection had been used differently in the Meno to conclude the immortality of the soul
(85d–86c).
55 For some misgivings see Gallop 1975: 135–37. See also my notes at the end of Section 2 on the
incoherence of the idea that the soul itself dies.
56 AA relies on the existence of the Forms, but in a way different from RA. FA is introduced as an
attempt to address the fact that AA leaves open the possibility that the soul, although extremely
durable, will be at some point destroyed. FA, like OA, relies on the assumption that life and death are
“opposites” (105d), and stresses its compatibility with OA against a possible objection (103b–c). FA,
unlike RA and AA, relies on the nature of the Forms in us, rather than the Forms themselves
(102b–103a).
57 Cf. Socrates’ reaction to the Delphic oracle in the Apology. For philosophy as the highest form of
mousikē see Resp. 548b and Phdr. 259b–d.
200   Luca Castagnoli

83b6)58—allows us to get as close as we can to the divine occurs time and again in the
dialogue. Not only will the philosophers enjoy the gods’ company after death; they will
also be admitted “into their race” (82b–d), because philosophy has purified their soul,
the part of them most resembling the divine (cf. AA, 80a), separating it from the body
and its concerns, and turning it toward “what is true, divine and not an object of opin-
ion” (84a–b). It is in this sense that philosophy is a “practice” or “exercise” of death (mēlēte
thanatou, 80e). This practice is not only a way of thinking, but also a way of being and
living, the only one that allows for the acquisition of true virtue (68c–69b; 82b–d).
But how does philosophy achieve the soul’s release or purification? Although argu-
ments are central to philosophy, the picture is more nuanced. The passage 82e–83d is
especially interesting in this respect: there is a cognitive element to what philosophy
does, showing that the senses cannot be trusted in the inquiry after what is real and intel-
ligible; but philosophy also “gently reassures,” “persuades,” and “encourages.” Philosophy
is almost anthropomorphized, as a friend or counsellor whispering to our soul; what
philosophy is said to do by Socrates here, and elsewhere in the Phaedo, is what Socrates
himself is portrayed by Phaedo as doing throughout the dialogue. Socrates takes care of
the souls of his friends not only by offering a set of “proofs,” giving them reason after rea-
son not to fear death, but also by encouraging, reassuring, and rallying them; trying to
“cure” their fears; pointing out their mistakes; and exhorting them to keep inquiring and
taking care of their own souls (cf. Phaedo’s praise of Socrates’ response to Simmias and
Cebes at 89a). The idea that logoi can have upon us, or at least a part of us (“the child in
us”), the same effect as “chanting spells,” just as muthoi, also brings to the fore the prag-
matic dimension of philosophical activity. Socrates’ exhortation to Cebes and Simmias
to look for another “enchanter” when he is gone, but also to “work together as you
search, because you may not easily find others more able to do this than you” (78a), also
indicates that philosophical arguments enjoy lasting persuasive force only if they are
properly internalized (cf. also Echecrates’ comment about his temporary forgetfulness
of the harmonia theory at 88d).
One way in which Socrates takes care of the souls of his friends in the Phaedo, and
Plato of the souls of his audience through the Phaedo, is by prompting them (us) to scru-
tinize the status and quality of the arguments presented and the conclusions reached,
and more generally by reminding them (us) of the limitations of human reason. In the
Synopsis I have highlighted, selectively, the striking amount of references to what is
“reasonable,” “likely,” or (to some degree) “persuasive” in Socrates’ ever-cautious intro-
duction of his views and arguments, and in his interlocutors’ reactions to them. These
Platonic pointers should be considered carefully when we examine the question of
whether Plato himself was “fully committed” to the soundness of any of the Phaedo
arguments, or to their conclusions (and whether he represents Socrates as being so). The
impression that Plato dogmatically presents a battery of “demonstrations” that are
supposed to firmly establish the immortality of the soul will not survive this analysis.

58 As opposed, for example, to the philosophy of the Pythagoreans, or of the natural philosophers
mentioned in Socrates’ intellectual autobiography? For the suggestion that Evenus incorrectly
considered himself a philosopher see Ebert 2001.
 The Phaedo on Philosophy and the Soul   201

Although philosophy is presented, and unreservedly praised, as a positive force in our

lives, Socrates frequently acknowledges its limitations. Early in his defense he stresses
several times that philosophy can free our souls only “as much as (humanly) possible”;
he also suggests that “pure” or “full” knowledge of the Forms is beyond our reach during
life, and can only be approximated even by those who use reason correctly; if full knowl-
edge is attainable, it will be attained only after death.59 Whether these cautionary notes
should be attributed only to Socrates the character, or perhaps the historical Socrates,
rather than to Plato himself is a question I cannot pursue here.
We could still think, of course, that philosophical argument can establish some truths,
including the truth of the immortality of the soul. To assuage Simmias’ and Cebes’ lin-
gering fears, despite their avowed endorsement of OA and RA, Socrates introduces AA;
but he acknowledges that the argument “of course still contains many grounds for suspi-
cion and counter-attack, at least if one is to go right through it properly” (84c). Simmias
for one thinks that it “has not seemed entirely sufficient,” and before explaining why he
says: “I think . . . that knowing the clear truth about things like this in our present life is
either impossible or something extremely difficult, but that at the same time not testing
from every angle what is said about them, refusing to give up until one is exhausted from
considering it in every way, is the mark of an extremely feeble sort of man. Because
concerning them one ought surely to achieve one of the following: either to learn or dis-
cover how things are, or, if it is impossible to do that, at least to take the best human
argument (logos)—the hardest one to disprove (dusexelenktotaton)—and to ride on that
as if one were taking one’s chances on a raft, and to sail through life in that way, unless
one could get through the journey with more safety and less precariousness on a more
solid vehicle, some divine argument (logou theiou tinos)” (85c–d). Socrates does not
comment upon Simmias’ suggestion that “perhaps” he shares the same view, but
Simmias’ words are reminiscent of Socrates’ early claims about human epistemic limi-
tations. Certainly the necessity to keep examining views in every way is one that Socrates
himself endorses not only in the Phaedo (e.g. 84c, 91c, 107b), but ubiquitously in Plato’s
corpus.60 The prospect of a safer “divine” logos and the language of sailing might allude
forward to Socrates’ “second voyage” and the hypothesis of the Forms used in it.61
However, if reference to the Forms were sufficient (and not just necessary) to make an
argument divine, this should make RA and AA “safer vehicles” than Simmias allows.
A crucial example of Socrates’ care of the soul is his warning against “misology.”
When, upon hearing Simmias’ and Cebes’ objections, all those present are sent “plum-
meting into doubt (apistian), not just about the arguments given before, but also about
what would be said later” (88c), Socrates warns them against the danger of becoming
“haters of arguments” (misologoi). Just as people become misanthropists when they
put all their trust in others without the necessary expertise in human affairs, and are

59 For discussion of these ideas cf. Fine 2016.

60 Cf. e.g. Apol. 38a; Crit. 46d; Charm. 166c–d; Grg. 457a–458a; Phlb. 41b.
61 Gallop 1975: 146–47, for example, refers to 99c–100a. Other interpreters think of some (Orphic?)
form of divine revelation (see Sedley 1995 for the view that Simmias’ raft analogy is therefore
incompatible with Platonic method).
202   Luca Castagnoli

disappointed time and again, finding them untrustworthy (89d–e), so “when someone
without expertise in arguments trusts an argument to be true, and then a little later
thinks that it is false, sometimes when it is, sometimes when it is not, and when he does
the same again with one argument after another,” he will end up blaming not himself but
the logoi, and will “spend the rest of his life hating and belittling arguments” (90b–d).
This occurs especially to those who deal “with the arguments used in disputation (anti-
logikous logous),” who think “that they alone have understood that there is nothing sound
or firm in any thing or in any argument” (90c). Misology is very different from the cau-
tious examining attitude previously suggested. Misology is the prejudicial loss of trust in
any argument, accompanied in some cases by the second-order belief that, as a matter of
fact, there are no sound arguments. To become a misologist is the greatest evil (89d): one
is “deprived of both truth and knowledge about things” (90d)62—for the necessity of
logoi in philosophy cf. 99d–100a.63
Socrates’ reaction to the dangers of misology is surprising. Rather than affirming that
of course there are sound arguments, he makes the conditional claim that “it would be a
lamentable fate if there really were some true and firm argument” and yet one were not
trusting it (90d). To avoid this, he suggests that “it will be much better to assume that we
are not sound yet,” rather than believe that “there’s probably nothing sound in arguments”
(90e), with a “pragmatic wager” that resembles the one concerning the destiny of the
soul at 114d (cf. Section 1.5 of this chapter). The inexperience of the would-be misanthro-
pists prevents them from realizing that “both the very good and the very wicked are
few in number, and that those in between are the most numerous” (90a). Socrates goes
out of his way to clarify that this failure is not relevant to misology: “arguments do not
resemble people in this way” (90b). If we take the remark at face value, Socrates’ prag-
matic wager is overcautious; but there might be some irony in the remark, and Plato
might be prompting us to reflect upon the fact that most arguments (including some of
the arguments used by Socrates in the Phaedo?) are “in between” good and bad.64
Socrates also warns that there is the danger that he tries to persuade himself of the
soul’s immortality irrespective of the truth, like those “agonistic,” “antilogic” arguers who
only aim to prevail in arguments.65 Surprisingly, Socrates even claims that this would be
beneficial, since “if I what I’m saying is actually true, then it’s quite right to be convinced;
if, on the other hand, there is nothing in store for one who has died, at least in this period
before I die I will be less of a mournful burden to those who are with me” (90b). But
this does not mean he endorses a merely therapeutic use of arguments, in which all that

62 For analysis of the misology passage, and discussion of the question of why (and how) truth
should be valued, see Woolf 2007. For other Platonic discussions of the risks of incorrect approaches to
arguments cf. Resp. 538c–539d; Tht. 168a–b.
63 Socrates’ earlier claim that the “greatest evil” is to consider the sensible world “most manifest and
true” (83c) is compatible, since not relying on logoi will lead one to rely on perception (Gallop 1975:
153–54).
64 For the suggestion that this intermediacy concerns the power of logoi to produce persuasion
cf. Miller 2015: 156.
65 For the distinction between philosophical dialectic and “eristic,” “antilogic,” or “agonistic” uses of
arguments cf. Men. 75c–d; Euthd. 271e–272b; Resp. 454a–c; 499a; Tht. 167e–168c.
 The Phaedo on Philosophy and the Soul   203

matters is to persuade others (or oneself) of something beneficial. Socrates exhorts his
friends to assess his arguments dispassionately: “you’ll give little thought to Socrates and
much more to the truth: if you think I say something true, agree with me, and if not, use
every argument to resist me, making sure that my eagerness doesn’t make me deceive
myself and you simultaneously” (91c).66
Throughout the Phaedo Socrates’ interlocutors ride an epistemic and emotional roller
coaster. Several times they declare that they are fully convinced of the strength of Socrates’
arguments, but they appear not to be really persuaded of their conclusions. They come up
with objections, and plunge into despair because of their implications, but then they are
(too?) quick to concede to Socrates’ counters, and to rally behind him. Ultimately their
behavior (Crito’s final request, everyone’s tears) could be taken to reveal that, till the very
end, Socrates’ arguments have not really taken roots in their souls.67 If this is a sober
reflection upon the limits of philosophy and philosophical argument, even in the midst
of a celebration of philosophy, and Socrates the philosopher, it is important to under-
stand what limits are being dramatized. The question of the effectiveness of Socrates’
arguments should not be conflated with the question of their soundness and apparent
shortcomings; nor should these shortcomings encourage the view that no better argu-
ments, or a deeper understanding of Socrates’ arguments, are possible.
That the release and purification of the soul from the body are an arduous and lengthy
task should come as no surprise. That the persuasion induced by philosophical argu-
ment will remain weak or fleeting until the argument, and its premises and implicit
presuppositions, have been repeatedly scrutinized and fully internalized is also only to
be expected. This is not an indictment against the power of philosophy to lead us to truth
and knowledge, or as close to them as humanly possible, as promised by Socrates. The
struggle of logos in the Phaedo is, after all, the best protreptic to keep doing philosophy. If
we still struggle to reach the dialogue’s end without misty eyes, we should not blame the
arguments, but engage with them all over again.68

Bibliography
Ackrill, J. L. 1973. “Anamnesis in the Phaedo: Remarks on 73c–75c’. ” In E. N. Lee,
A. P. D. Mourelatos, and R. Rorty (eds.) Exegesis and Argument: Essays in Greek Philosophy
Presented to Gregory Vlastos. Amsterdam: Van Gorcum, 177–95.
Annas, J. 1982. “Plato’s Myths of Judgement.” Phronesis 27, 119–43.
Apolloni, D. 1996. “Plato’s Affinity Argument for the Immortality of the Soul.” Journal of the
History of Philosophy 34, 5–32.
Bailey, D. T. J. 2005. “Logic and Music in Plato’s Phaedo.” Phronesis 50, 95–115.

66 The suggestion that Socrates risks deceiving himself (and his friends) to assuage his (their) fear
clashes with the description of Socrates’ attitude throughout the dialogue, and with his comparison of
his argument to swans’ songs at 84e–85b: these songs do not express the swans’ distress before dying,
but their delight to meet the god whose servants they are, Apollo.
67 Cf., however, Phaedo’s explanation at 117c–d: “I covered my head and wept for myself—not for
him, you understand, but for my own fortune that I’d lost such a friend.”
68 I am very grateful to the editor, Gail Fine, for her generous feedback on several drafts of this essay.
204   Luca Castagnoli

Barnes, J. 1978. “Critical Notice of D. Gallop (transl.) Plato: Phaedo.” Canadian Journal of
Philosophy 8, 397–419.
Bedu-Addo, J. T. 1991. “Sense-Experience and the Argument for Recollection in Plato’s Phaedo.”
Phronesis 36, 27–60.
Benson, H. H. 2015. Clitophon’s Challenge: Dialectic in Plato’s Meno, Phaedo, and Republic.
Oxford: Oxford University Press.
Betegh, G. 2004. The Derveni Papyrus. Cosmology, Theology and Interpretation. Cambridge:
Cambridge University Press.
Betegh, G. 2006. “Eschatology and Cosmology: Models and Problems.” In M. M. Sassi (ed.) La
costruzione del discorso filosofico nell’età dei presocratici/The construction of philosophical
discourse in the age of the Presocratics. Pisa: Edizioni della Normale, 27–50.
Betegh, G. 2014. “Pythagoreans, Orphism, and Greek Religion.” In C. A. Huffman (ed.) A
History of Pythagoreanism. Cambridge: Cambridge University Press, 149–66.
Bostock, D. 1986. Plato’s Phaedo. Oxford: Oxford University Press.
Crooks, J. 1998. “Socrates’ Last Words: Another Look at an Ancient Riddle.” Classical Quarterly
48, 117–25.
Dancy, R. 2004. Plato’s Introduction of Forms. Cambridge: Cambridge University Press.
Denyer, N. 2007. “The Phaedo’s Final Argument.” In D. Scott (ed.) Maieusis: Essays in Ancient
Philosophy in Honour of Myles Burnyeat. Oxford: Oxford University Press, 87–96.
Dimas, P. 2003. “Recollecting Forms in the Phaedo.” Phronesis 48, 175–214.
Dorter, K. 1982. Plato’s Phaedo: An Interpretation. Toronto: Toronto University Press.
Ebert, T. 2001. “Why Is Evenus Called a Philosopher at Phaedo 61c?” Classical Quarterly 51:
423–34.
Edmonds, R. G. III (ed.) 2011. The “Orphic” Gold Tablets and Greek Religion. Further along the
Path. Cambridge: Cambridge University Press.
Fine, G. 1983. On Ideas: Aristotle’s Criticism of Plato’s Theory of Forms. Oxford and New York:
Oxford University Press.
Fine, G. 2014. The Possibility of Inquiry: Meno’s Paradox from Socrates to Sextus. Oxford:
Oxford University Press.
Fine, G. 2016. “The ‘Two Worlds’ Theory in the Phaedo,” British Journal of the History of
Philosophy 24, 557–72.
Fine, G. 2017. “Plato on the Grades of Perception: Theaetetus 184–186 and the Phaedo.” Oxford
Studies in Ancient Philosophy 53, 65–109.
Fine, G. forthcoming. “Recollection and Innatism in the Phaedo.”
Franklin, L. 2005. “Recollection and Philosophical Reflection in Plato’s Phaedo.” Phronesis 50,
289–314.
Frede, D. 1978. “The Final Proof of the Immortality of the Soul in Plato’s Phaedo 102a–107a.”
Phronesis 23, 27–41.
Gallop, D. 1975. Plato. Phaedo. Oxford: Clarendon Press.
Gallop, D. 1982. “Plato’s ‘Cyclical Argument’ Recycled.” Phronesis 27: 207–22.
Gosling, J. C. B. 1965. “Similarity in Phaedo 73B seq.” Phronesis 10: 151–61.
Gulley, N. 1954. “Plato’s Theory of Recollection.” Classical Quarterly n.s. 4, 194–213.
Gerson, L. 1999. “The Recollection Argument Revisited.” Apeiron 32, 1–15.
Hackforth, R. 1955. Plato’s Phaedo. Cambridge: Cambridge University Press.
Huffmann, C. A. 1993. Philolaus of Croton. Cambridge: Cambridge University Press.
Irwin, T. 1974. “Recollection and Plato’s Moral Theory.” Review of Metaphysics 27, 752–72.
Irwin, T. 1999. “The Theory of Forms.” In G. Fine (ed.) Plato 1: Metaphysics and Epistemology.
Oxford: Oxford University Press, 143–70.
 The Phaedo on Philosophy and the Soul   205

Kahn, C. 2006. “Plato on Recollection.” In H. H. Benson (ed.) A Companion to Plato. Oxford:

Blackwell, 119–32.
Kamen, D. 2013. “The Manumission of Socrates: A Rereading of Plato’s Phaedo.” Classical
Antiquity 32: 78–100.
Kamtekar, R. 2016. “The Soul’s (After-)Life.” Ancient Philosophy 36: 115–32.
Kanayama, Y. 2000. “The Methodology of the Second Voyage and the Proof of the Soul’s
Indestructibility in Plato’s Phaedo.” Oxford Studies in Ancient Philosophy 18, 41–100.
Kelsey, S. 2000. “Recollection in the Phaedo.” Proceedings of the Boston Area Colloquium in
Ancient Philosophy 16, 91–120.
Kingsley, P. 1995. Ancient Philosophy Mystery, and Magic: Empedocles and the Pythagorean
Tradition. Oxford: Oxford University Press.
Madison, L. A. 2002. “Have We Been Careless with Socrates’ Last Words? A Rereading of the
Phaedo.” Journal of the History of Philosophy 40, 421–36.
McCoy, J. 2011. “Re-examining Recollection: The Platonic Account of Learning.” International
Philosophical Quarterly 51, 452–66.
Miller, T. 2015. “Socrates’ Warning against Misology (Plato, Phaedo 88c–91c).” Phronesis 6,
145–79.
Most, G. W. 1993. “A Cock for Asclepius.” Classical Quarterly 43, 96–111.
Nails, D. 2002. The People of Plato: A Prosopography of Plato and Other Socratics. Indianapolis,
IN: Hackett.
Johansen, T. 2017. “The Separation of Soul from Body in Plato’s Phaedo.” Philosophical Inquiry
41, 17–28.
Pakaluk, M. 2003. “Degrees of Separation in the Phaedo.” Phronesis 48, 89–115.
Robinson, R. 1953. Plato’s Earlier Dialectic. 2nd ed. Oxford: Clarendon Press, 123–45.
Rowe, C. J. 1993. Plato. Phaedo. Cambridge: Cambridge University Press.
Scott, D. 1987. “Platonic Anamnesis Revisited.” Classical Quarterly n.s. 37, 346–66.
Scott, D. 1995. Recollection and Experience: Plato’s Theory of Learning and Its Successors.
Cambridge: Cambridge University Press.
Sedley, D. 1989. “Teleology and Myth in the Phaedo.” Proceedings of the Boston Area Colloquium
in Ancient Philosophy 5: 359–83.
Sedley, D. 1995. “The Dramatis Personae of Plato’s Phaedo.” Proceedings of the British Academy
85, 3–26.
Sedley, D. 1998. “Platonic Causes.” Phronesis 43, 114–32.
Sedley, D. 2006. “Form-Particular Resemblance in Plato’s Phaedo.” Proceedings of the
Aristotelian Society 106, 311–27.
Sedley, D. 2007. “Equal Sticks and Stones.” In D. Scott (ed.) Maieusis: Essays in Ancient
Philosophy in Honour of Myles Burnyeat. Oxford: Oxford University Press, 68–86.
Sedley, D. 2009. “Three Kinds of Platonic Immortality.” In D. Frede and B. Reis (eds.) Body
and Soul in Ancient Philosophy. Berlin: De Gruyter, 145–61.
Sedley, D. and Long, A. 2010. Plato: Meno and Phaedo. Cambridge: Cambridge University Press.
Vlastos, G. 1969. “Reasons and Causes in the Phaedo.” Philosophical Review 78, 291–325.
Warren, J. 2001. “Socratic Suicide.” Journal of Hellenic Studies 121, 91–106.
West, M. L. 1983. The Orphic Poems. Oxford: Clarendon Press.
Williams, T. 2002. “Two Aspects of Platonic Recollection.” Apeiron 35, 131–52.
Woolf, R. 2004. “The Practice of a Philosopher.” Oxford Studies in Ancient Philosophy 26,
97–129.
Woolf, R. 2007. “Misology and Truth.” Proceedings of the Boston Area Colloquium in Ancient
Philosophy 23, 1–16.
 chapter 9

The R epu blic

Dominic Scott

1. A Bird’s-Eye View

Arguably the greatest of Plato’s works, and nowadays the best known, the Republic is
certainly among his most complex. From its title, the first-time reader will expect a
­dialogue about political theory, yet the work starts from the perspective of the individ-
ual, coming to focus on the question of how, if at all, justice contributes to an agent’s
­happiness. Only after this question has been fully set out does the work evolve into an
investigation of politics—of the ideal state and of the institutions that sustain it, ­especially
those having to do with education. But the interest in individual justice and happiness is
never left behind. Rather, the work weaves in and out of the two perspectives, individual
and political, right through to its conclusion. All this may leave one wondering about
the unity of the work. My purpose in this introduction is to show that, despite the enor-
mous range of topics discussed, the Republic fits together as a coherent whole.1
The need to defend justice arises in book I, when Thrasymachus claims that justice is
merely a matter of giving benefit to another at one’s own expense (343c). Deep down,
what we all really admire is not justice but the very opposite: the ideal of the tyrant,
someone who holds everyone else in his power so as to satisfy any desire he pleases and
with no fear for the consequences (344a). Thrasymachus is as forceful and memorable as
any of Plato’s characters, but Socrates eventually wears him down, and it is left to the
brothers Glaucon and Adeimantus at the beginning of book II to renew the challenge.
Even though Thrasymachus has been subdued, they still feel his underlying position
remains intact. With less heat and more light than before, they ask Socrates to show that
justice is better for the agent than injustice, even if one has the power to commit injustice
with impunity.

1 There is no shortage of books devoted to the Republic, many of which serve both as introductions
and as scholarly monographs. For example, R. Nettleship, Lectures on the Republic of Plato, 2nd ed.
(London, 1901); R. Cross and A. Woozley, Plato’s Republic [Republic] (London, 1964); N. White, A
Companion to Plato’s Republic (Oxford, 1979); J. Annas, An Introduction to Plato’s Republic [Introduction]
(Oxford, 1981). Also, chs. 11–18 of T. Irwin, Plato’s Ethics (Oxford, 1995) focus on the Republic.
208   Dominic Scott

Socrates accepts the challenge but wants to proceed more methodically than before.
First, they must define the nature of justice and begin by looking at the state before turning
to the soul. Justice, he claims, will be easier to see in the larger entity (368d–369a). This
creates the first great shift of focus in the work, from ethics to politics, and sets off an
inquiry that lasts right into the middle of book IV. Imagining themselves as the founders
of a perfectly just state, they ask what such a state would be like.
The answer turns out to be a city composed of three classes—producers, auxiliaries,
and guardians: the first to provide for the material needs of the state, the second for its
defense, and the third to rule. Each has a specific function of its own, and none is to
interfere with the others. Above all, the just city will be unified, ordered, and harmonious.
The rulers and auxiliaries, the two classes Socrates discusses at most length, will be dedi-
cated to protecting the good of the state as a whole, and every aspect of their education,
as well as the conditions under which they live, will be minutely engineered to ensure
they fulfill their roles as best they can. In a particularly famous passage, Socrates devotes
considerable attention to the arts, proposing radical censorship of the kinds of poetry
and music to which the would-be guardians and auxiliaries may be exposed (376c–398b).
The ideal state now established, Socrates returns to the individual (IV 434d). Since
justice in the city is a matter of the harmonious arrangement of its three parts, the same
ought to apply to the individual—if, that is, it also has three parts. This he shows to be the
case with an argument that divides the soul into reason, spirit, and appetite, each of
which is an analogue of one of the three classes in the state. Reason is able to direct the
soul and determine its good overall.2 Spirit is the source of such feelings as anger, pride,
and shame. Appetites are typically desires for physical pleasures, and they need to be
kept carefully in check by an alliance of reason and spirit. Justice in the soul obtains
when each part performs its own function and does not interfere with that of the others—
in particular, when neither spirit nor appetite usurps the ruling function of reason.
As such, justice is a form of psychic harmony or health. By contrast, injustice is a
condition of inner turmoil, a civil war between the parts, where one vies against another
for overall control. If this is correct, Socrates concludes, we can answer Glaucon’s
challenge: justice, as psychic health, is inherently good for the soul (445a–b); injustice,
as the polar opposite, is quickly rejected as any sort of good, losing all the appeal it had
earlier on in the work.
This is meant to provide the defense of justice demanded at the beginning of book II.
But neither Socrates nor his interlocutors want to end the conversation there: Socrates,
because he wants to establish his conclusion more firmly and clarify the nature (and
hence unattractiveness) of injustice; his interlocutors, because they have a question they
want him to answer about the ideal state (V 449c). Back at book IV 423e–424a, Socrates
had casually stated that the rulers of his state would have “wives and children in

2 “Reason” is perhaps a somewhat misleading translation of to logistikon, since it may suggest pure
reason, with no desires, which is clearly not Plato’s view; his view is that each part of the soul has its
distinctive kind of desire (580d–583b). Others may prefer “the rational part” as a translation.
 The Republic   209

common.”3 With this brief remark, he had at a stroke abolished the traditional family
unit. Now in book V, when challenged to explain what he meant, he admits that this is
only one of three proposals that will evoke consternation. Not only does he propose to
do away with the family, he would also sweep aside all barriers to the participation of
women in ruling and defending the state. Men and women differ only in respect of their
role in reproduction; there are no differences between them that are relevant to ruling
and hence no reason a woman, just by virtue of being a woman, should not be a fully
fledged member of the guardian or auxiliary classes (451c–457c).
But he has a third, even more controversial proposal up his sleeve—philosopher-rulers:

Until philosophers rule as kings or those who are now called kings and leading men
genuinely and adequately philosophise, that is, until political power and philosophy
entirely coincide . . . cities will have no rest from evils . . . nor will the human race.4

Socrates anticipates such hostility to this proposal that he spends the rest of book V and
much of book VI defending it. He then turns to consider what sort of education would
be required to produce these philosopher-rulers, which takes him all the way to the end
of book VII.
The first step in Socrates’ argument for philosopher-rulers sounds innocuous
enough: the rulers of the ideally just state will need knowledge—of justice, for a start.
But it turns out that, on its own, no amount of practical experience will yield such
knowledge. Rather, one needs to grasp the Form of Justice in itself, an entity distinct
from particular cases of justice, the essence of justice, apprehended only by intellect.
Without knowledge of the Form of Justice, one will never be able to understand what
makes any particular action or state of affairs just. In fact, the philosopher-rulers will
need knowledge of all the Forms—not only Justice but also many others, including
Beauty, Temperance, and, above all, the Good. Much of book VII focuses on the extraor-
dinary difficulty of apprehending the Forms and, in what is one of the most memorable
passages in Plato, Socrates compares the process of learning to the slow, painful ascent
from the darkness of a cave into the sunlight (VII 514a–517b).
The feeling of bewilderment described in the allegory might well be shared by some
of Plato’s readers as the book goes on. It turns out that the education of the trainee rulers
will last 30 years (and this is on top of the education in the arts that was described in
books II–III).5 Starting at the age of 20, they will have to spend an astonishing 10 years
devoted to mathematical studies before they can even begin “dialectic,” the process of
inquiring directly into the Forms. And even after 5 years of dialectic, it is not for another
15 years, involving military and administrative tasks, that they finally apprehend the
Form of the Good and assume control of the state.

3 He had already attacked the institution of private property for the guardians at 416d–417b.
4 At 473c–d. All translations are from G. Grube, Plato: Republic, rev. C. Reeve (Indianapolis,
Ind., 1992).
5 For further discussion of the arts in books II–III, see Kamtekar, chapter 25 in this volume.
210   Dominic Scott

By the beginning of book VIII, Socrates seems to have satisfied his interlocutors’
c­ uriosity about the rulers of the ideal state and returns to the project he was about to
take up at the end of book IV, which was to enumerate and describe the main forms of
vice: timocracy, oligarchy, democracy, and tyranny, each of which can appear in either
the state or the soul. In either case, timocracy aims at honor, oligarchy at wealth, democ-
racy at freedom, and tyranny at the unbridled pursuit of the worst form of appetite.
Throughout, Socrates presents this taxonomy in a narrative of decline. The political ver-
sion shows an ideal state degenerating into a timocracy, where the military hold power
(545c–548d); in time, this becomes unstable and lapses into an oligarchy, a state sharply
divided into rich and poor (550d–553a). Civil war eventually breaks out, resulting in the
victory of the poor, who impose a radical democracy where freedom and toleration rule
supreme (555b–558c). But, in time, a demagogue emerges who persuades the citizens to
let him take power; this he subsequently abuses, reducing them to the status of slaves.
The result is tyranny, the most extreme form of injustice (562a–569c).
To distinguish the four kinds of vice in the individual, Socrates recalls the theory of
the divided soul and develops it further. A timocratic soul is one in which spirit domi-
nates over reason and appetite (548d–550c). The other three types of vice are explicated
in terms of the domination of appetite: the oligarch gives priority only to those appetites
whose fulfilment is necessary for bodily health (553a–555a); the democratic individual
indulges all sorts of desires, be they for necessities or for luxuries, and he insists that all
pleasures and desires are equal in worth (558c–562a). Finally, the tyrant enslaves himself
to the worst kind of desires, “lawless” appetites (571a–580c).
When Socrates invites his interlocutors to rank the five characters in terms of happi-
ness (580a), they place the just person at the top and the other four in the order in which
they appeared. The tyrant, despite Thrasymachus’s eulogy at the beginning of the work,
turns out to be the most miserable. After giving two more arguments to justify the same
ranking, based on the pleasure that each life brings (580d–588a), Socrates ends book IX
with a graphic image that warns against allowing either the appetitive or spirited part to
take control of the soul (588b–592b).
This seems to conclude the overarching project of the Republic, the defense of justice.
But in book X, Socrates returns to some unfinished business: the status of poetry in the
ideal state. Using the psychology from book IV, he imposes further restrictions on the
kind of poetry he is prepared to allow into the ideal state. He bans all imitative poetry,
including Homer, as well as tragedy and comedy, arguing that such poets have no exper-
tise in political affairs and, further, that they corrupt the souls of their audiences, helping
the non-rational parts to grow and assume control over the rational. Finally at 608c, he
returns to the subject of justice and discusses some of its further benefits—in particular,
the rewards it brings us in the afterlife.
As I said previously, the Republic starts as a work of ethics, investigating the link
between justice and happiness in the individual. Initially, politics only appears on the
stage as a heuristic device to help define justice in the soul. But as the dialogue goes on, it
acquires an interest in its own right and, once center stage, helps broaden the scope of
the work still further, to embrace such topics as the arts in books II–III and X and
 The Republic   211

epistemology, metaphysics, and mathematics in books V–VII. But although political

theory goes on to have a higher profile than we might initially have expected, it in no
way diverts the work from the ethical course on which it started. Quite the reverse: Plato
uses his politics to deepen his ethics. In books II–V, the ideal state provided merely an
analogue for understanding justice and happiness in the individual; by the end of book
VII, it also provides the context in which the good of the individual is fully attained, for
it is only in the ideal state, with its ability to provide exactly the right kind of philosophical
education, that human nature can be perfected.
Although this chapter has to be highly selective, I try to focus on a cluster of issues
that are genuinely salient to an understanding of the work. The first section concerns the
central question of the work, about the value of justice, and examines a fundamental
objection to Plato’s argument. The second section turns from ethics to politics and
focuses on what for modern readers is the least palatable aspect of the work: its authori-
tarianism. In the third section, I focus on the nature of philosophy as discussed in the
central books.

2. The Defense of Justice

At the beginning of book II, Glaucon challenges Socrates to show that in and of itself
justice benefits the agent and that the life of justice is superior to that of injustice. Most
people, he claims, see it as a good of some sort, but only as a burden; ideally, we would
commit injustice whenever it furthered our own interests. But since, as individuals, we
usually lack the power to stop others from committing injustice against us, we accept the
restrictions of justice as an agreement to secure mutual protection. Glaucon develops
the point with the story of Gyges, who discovered a ring that would make him invisible
and used it to commit injustice with impunity and to acquire the greatest goods.
Glaucon’s point is that, if any of us had a similar opportunity, we would cast all thoughts
of justice aside. To put the challenge in its starkest terms, he imagines two people: one is
actually just but has a reputation for injustice; the other is unjust but through cunning
has acquired the greatest reputation for justice. The first is imprisoned and tortured for
his apparent injustice, while the second enjoys all the external benefits of justice without
actually having it. Glaucon ends by challenging Socrates to explain why anyone would
prefer to be the first rather than the second. His brother Adeimantus then reinforces the
challenge with a long speech of his own.
What Socrates is being asked to provide here is a definition of justice and injustice, as
well as an account that explains “what power each itself has when it’s in the soul” (358b)
or, as Adeimantus puts the point several times, “what effect each has because of itself on
the person who has it—the one for good and the other for bad—whether it remains
­hidden from gods and humans beings or not” (367e). As this last clause suggests, what
they do not want is an assessment of justice in terms of the benefits that most people
associate with it. The point of the Gyges story and the choice of lives is that such benefits
212   Dominic Scott

are “detachable” from justice per se. The good repute that usually follows from being just
depends on contingent circumstances that may sometimes not obtain. The challenge is
to argue for the inherent superiority of justice over injustice.
Notice how strong this challenge becomes by the time Glaucon sets out the choice of
lives: the just person whom he envisages may have no other good apart from justice—in
fact, not only does he lack any physical goods, he is beset by the worst kinds of physical
evil; and yet his life is still meant to be preferable to that of the unjust person who enjoys
the maximum of other goods. This is tantamount to claiming that, once you lack justice
and possess injustice instead, no amount of other goods can compensate.6
Socrates’ response is to define justice as psychic harmony and injustice as the oppo-
site: internal disorder and conflict. Once he has redescribed the two states in this way, he
expects us to reverse the preference between the just and unjust lives that seemed so
tempting in Glaucon’s original challenge. In book IX he reinforces this strategy by focus-
ing on the very extreme of injustice, the tyrant. At the end of book I, Thrasymachus had
held him up as the ideal because of his unlimited freedom to satisfy any desires he
wanted, but by the end of book IX Socrates has probed deep into his soul to reveal him
for what he is: frustrated, tormented, isolated, and enslaved. Such is the power that
injustice has “in and of itself ” on the soul (cf. 358b).
This at least is the strategy, but it faces a well-known objection.7 What Socrates was
meant to defend has been referred to as “conventional justice”: that is, a disposition to
behave in certain intuitively specified ways, such as to keep one’s compacts and prom-
ises, to honor one’s parents, and not to steal. Also, the actions or abstentions in terms of
which justice and injustice are conventionally understood seem for the most part to
concern dealings with other people. But “platonic justice,” as defined in book IV, is pri-
marily focused neither on actions nor on our relations with other people; it is essentially
an internal state, defined in terms of the relations among the parts of the soul. Just
actions are those that develop and preserve internal harmony (443e). One might imme-
diately object that Socrates has replaced the type of justice that Glaucon wanted him to
defend with a conception of his own. Therefore his argument is invalid. But Socrates is
well aware that the two conceptions differ and is not attempting to defend conventional
justice as it stands. Nevertheless, there remains a subtler objection: he has not shown
that there will be significant overlap between the kinds of action someone might per-
form according to the two conceptions. In other words, why couldn’t someone with a
harmonious soul still commit acts of conventional injustice—for instance, by breaking

6 Glaucon’s challenge has provoked considerable scholarly controversy. A notorious problem is that
he appears to contradict himself, by starting with the need to praise justice in itself and not for its
consequences, and then asking Socrates to explain what “power it has in the soul” (358b; cf. 366e),
which sounds very much like a reference to its psychological consequences. For two opposing solutions
to the problem, see N. White, “The Classification of Goods in Plato’s Republic,” Journal of the History of
Philosophy 22 (1984), 393–421, and T. Irwin, “Republic 2: Questions about Justice,” in G. Fine, ed., Plato 2:
Ethics, Politics, Religion and the Soul (Oxford, 1999), 164–85.
7 For a classic statement, see D. Sachs, “A Fallacy in Plato’s Republic,” Philosophical Review 72
(1963),141–58; reprinted in G. Vlastos, ed., Plato, vol. 2: Ethics, Politics d Philosophy of Art and Religion
(New York, 1973), 35–51.
 The Republic   213

their promises or failing to honor their parents? What is it about inner harmony that
prevents such actions?8

2.1 The Psychological Defense

For the most part, commentators have found two types of strategy to defend Plato’s
argument.9 The first is based on the psychological theory of the Republic and is at its
most explicit in book IX.10 One of the premises for this defense is the assumption that
the motivation for conventionally unjust behaviour lies in the desires associated with
appetite and spirit. Ever-increasing appetites demand more and more material resources
for their fulfillment; sooner or later, the appetitive individual will exhaust his own
­supplies and encroach on others’ property. Mutatis mutandis, a similar point can be made
about spirit, once let out of control: power and honor are limited goods; to maximize
one’s share of them, one will have reason to ill-treat one’s competitors. The second premise
in the argument is that increasing the strength of non-rational desires causes at least as
much grief to the agent of injustice as to the victim. This grief comes in various forms. For
instance, appetites are insatiable: the stronger they grow, the more urgent their demands.
This leaves the appetitive individual ever-increasingly in need—in other words, impov-
erished (579e). Furthermore, anyone in the grip of appetite will have to treat others as
mere means to his satisfaction, which makes him (at the extreme) untrustworthy and
friendless (580a), and no human being would wish to be in such a state.
All of this helps to show how the class of actions associated with each type of justice,
conventional and platonic, might coincide. Although often called psychic “harmony,”
platonic justice involves more than an inner feeling of calm; it is best thought of as psy-
chic hierarchy, where reason regulates the desires of the other two parts in accordance
with its understanding of the soul’s overall good. If the second premise of the argument
is correct, it is in the overall interest of the soul to keep non-rational desires within strict
limits and to prevent them starting on the cycle of ever-accelerating growth. So in regu-
lating desire for the overall good of the soul, reason will limit precisely those desires that
would motivate us to act out of line with conventional justice.
There is another strand in the psychological defense of justice, which relates to a spe-
cific form of non-rational desire: unnecessary appetites. These are introduced in book
VIII (558d–559c) as appetites that are not needed for physical health (indeed, they tend
to harm it) and are also such that they can be eradicated by the right training. But
Socrates also thinks that they are “harmful to reason” (559b), by which I take him to
mean that they compel it to form beliefs that conform to appetitive goals. This is illustrated

8 There is also the problem that someone might commit acts of conventional justice but suffer from
inner conflict. The oligarchic man seems a case in point (554d–e).
9 For an excellent overview of this topic, see R. Kraut, “The Defense of Justice in Plato’s Republic,” in
Kraut, ed., The Cambridge Companion to Plato (Cambridge, 1992), 311–37.
10 G. Vlastos, “Justice and Happiness in the Republic,” Platonic Studies, 2nd ed. (Princeton, N.J.,
1981), 133–34.
214   Dominic Scott

in the subsequent portrait of the democratic man, whose appetites grow so strong that
he abandons the ethical beliefs of his upbringing, an ethics of thrift, and replaces it with
one of indulgence (560e– 561b). Now, one reason the platonically just person values the
strength and autonomy of the rational part is that it helps keep the soul on course and
ensure its overall good. Since certain types of desire threaten this autonomy, they must
be restrained, if not entirely eliminated. If, at the same time, these desires also tend to
motivate conventionally unjust behavior, we have found another way to bring about a
convergence between platonic and conventional justice.

2.2 The Metaphysical Defense

The psychological defense of justice makes no use of the metaphysics of the central
books. It requires no assumptions at all about the existence of the Forms but, instead,
relies on empirical assumptions about state and soul.11 But there is a second way of sup-
porting Plato’s defense of justice that does appeal to the metaphysics and epistemology
of the central books. We know from book IV that the platonically just person must have
an understanding of the good. In the central books, this turns out to involve under-
standing the Forms, especially the Good itself, the first principle of all things. This can
be used to add a whole new dimension to the theory of book IV, where we had no inkling
that the wisdom of the just person would involve ascending such metaphysical heights.
There are two distinct ways in which the added dimension of the central books helps
support the defense of justice. First, at book VI 485d, Socrates likens the soul of the phi-
losopher to a stream whose water surges in one direction, thereby drying up other chan-
nels. Philosophers are (quite literally) enamored of the Forms, and their energy will be
so focused on understanding that their desires for physical pleasures or for honor will be
greatly weakened, if not eliminated entirely. The lines on which this argument can now
be developed are parallel to the psychological argument. As before, the sources of con-
ventionally unjust behavior lie in non-rational desire; the platonically just person,
whom we now know to be gripped by philosophical eros, will be too transfixed on the
Forms to be interested in the objects of non-rational desire and hence to commit acts of
conventional injustice. For these reasons, such a character will be a “friend and relative
of justice” (487a).
The second version of the metaphysical defense itself has two aspects. At book VI
500c, the philosopher looks to the order that exists among the Forms and uses it as a
model to create order first within his own soul and then, if circumstances require, within
the souls of his fellow citizens. One might use this text to support the defense of justice
and claim that the order thus imposed on the philosopher’s own soul would not allow

11 There is one argument at the end of book IX that does seem to allude to the metaphysics of the
central books, when it claims that the pleasures of philosophy are “more real” than those of appetite or
spirit (583b–588a). But otherwise, the Forms are completely absent from all but V 476a to VII 541b of
the work.
 The Republic   215

for the existence of rampant appetitive and spirited desires that would lead to
­conventionally unjust acts.12 The other aspect of this version of the metaphysical defense
takes its cue from the same text but focuses on a philosopher’s desire to imitate the order
and balance among the forms directly in his relationships with other people. Failures to
fulfill one’s compacts and promises or to honor one’s parents are all failures to recipro-
cate, and these disturb the balance in human relations. A philosopher who looked
toward the order of the Forms would be repelled by such disorder.13
The problem with this version of the metaphysical defense (in either of its aspects) is
that the content of the Good is very underdetermined in the Republic. (Socrates himself
appears to disavow knowledge of it at 506b–e.) So we just do not know enough about
what kind of order the philosopher would “read off ” the Forms in general, or the Good
in particular, to impose on his own desires or on his relations with others. There is no
guarantee that the order in question will typically prescribe acts of conventional justice.
This is not to say this version of the argument is doomed to fail, just that it is highly pro-
visional and, until we have a more determinate specification of the Good, we might,
instead, prefer to rely on the first version of the metaphysical defense: philosophers
will be rendered indifferent to the goals of non-rational desire by the intensity of their
intellectual eros.
But here, too, we should exercise caution. There is a well-known issue in the scholarly
literature about why the philosophers, once they have attained knowledge of the Forms,
should interrupt their contemplation to rule in the ideal state. Socrates says it is just
that they do so, while admitting they will have to be “compelled” to rule “out of necessity”
(cf. 501d, 520e, 540b): for them, ruling is something they would rather not be doing.
Some commentators have leaped on this passage as a possible counterexample to the
basic thesis of the Republic that justice benefits the agent.14 Even without delving into
this debate, we can see that the issue raises complications for the first version of the
metaphysical defense: the more one leans on the philosophers’ indifference to worldly
aspirations, including the pursuit of honor and power, the more one emphasizes their
reluctance to rule, thus seeming to exacerbate the tension between their self-interest
and the demands of justice.
There is one important point about the metaphysical defense that needs to be empha-
sized. In contrast to the psychological defense, which is the strategy that Socrates actu-
ally follows in the dialogue to meet Glaucon’s challenge, the metaphysical defense is not
one that he proffers directly; it is an attempt at “rationasl reconstruction” on the part of
recent commentators. In the books from which it has been constructed, V–VII, Socrates
is not explicitly attempting to support his reply to Glaucon as he does in VIII–IX. Rather,

12 The order the philosopher imposes on his own soul involves “thinking rational thoughts and
satisfying rationally controlled desires of his own” (J. Cooper, “The Psychology of Justice in Plato,”
American Philosophical Quarterly 14 [1977], 156).
13 This is the approach favored by R. Kraut, “Return to the Cave: Republic 519–521” [“Cave”], in
G. Fine, ed., Plato 2: Ethics, Politics, Religion and the Soul (Oxford, 1999), 247.
14 There is an extensive literature on this subject, but see, especially, R. Kraut, “Cave,” and E. Brown,
“Minding the Gap in Plato’s Republic,” Philosophical Studies 117 (2004), 275–302.
216   Dominic Scott

he has been interrupted and induced to switch his focus to a series of questions about
the rulers in the ideal state, such as what sort of expertise they need and how they are to
acquire it. What commentators have done is to take claims made within this section
and redeploy them toward a different aim: strengthening the defense of justice given in
book IV, something that Socrates himself only explicitly does in books VIII–IX.
This is not to deny that Socrates recommends pursuing such a line of argument.
At 504b (cf. 435d), he implies that the defense of justice that they have mounted so far
(i.e., the psychological defense) is a shortcut. The “longer route” would require investi-
gating the matter by reference to the Form of the Good—something that will take the
trainee guardians in the ideal state many years to achieve. In effect, this would be a much
more thorough enterprise than the very provisional line of argument I have called the
metaphysical defense.15 I return to this point at the end of this chapter.

3. Politics

Whether or not Plato can successfully bridge the gap between justice as psychic har-
mony and conventional justice, the very conception of individual justice that he created
in the Republic is one of his most enduring legacies. At the heart of this conception is the
proposal that reason should rule over the non-rational desires—regulating, limiting,
and sometimes eliminating them altogether. But when one turns to the political ana-
logue, one runs straight into one of the less attractive aspects of the work: its authoritari-
anism, the idea that it is equally appropriate for the guardian class to regulate, restrain,
and “remove” awkward citizens where necessary.
Thanks especially to Karl Popper,16 the authoritarian nature of the work—in particu-
lar, its critique of democracy—has been notorious for decades. But amid all the contro-
versy, the complexity of Plato’s critique has sometimes been missed. And complex it
certainly is, featuring at several places throughout the work, sometimes explicitly, but
often implicitly. Also, the target shifts. Sometimes the argument is against democracy in
principle, at others actual democracy, as manifested especially at Athens. The purpose of
this section is to chart a route through the Republic to separate out the different strands
of Plato’s argument.

15 On this, see D. Scott, “Metaphysics and the Defence of Justice in the Republic,” Proceedings of the
Boston Area Colloquium in Ancient Philosophy 16 (2000), 1–20.
16 K. Popper, The Open Society and Its Enemies, vol. 1: The Spell of Plato, 5th ed. [Open Society]
(London, 1966). For an excellent recent overview of this topic, see C. Taylor, “Plato’s Totalitarianism,”
Polis 5 (1986), 4–29; reprinted in G. Fine, ed., Plato 2: Ethics, Politics, Religion and the Soul (Oxford,
1999), 280–96. Another useful article is by J. Ackrill, “What Is Wrong with Plato’s Republic?” Essays on
Plato and Aristotle, 2nd ed. (Oxford, 2001), 230–52.
 The Republic   217

3.1 The Organic Theory

Before examining democracy directly, we should pause to consider a distinct but related
theme. This is the accusation that the Republic assumes what Popper called the “organic
theory.” One component of this theory is metaphysical: that the state is an entity in its
own right, a whole distinct from the sum of its parts. Popper claimed to find this implicit
in the state-soul analogy:17 if the state is sufficiently like the soul such that it can bear the
same properties of, say, courage, justice, temperance, and wisdom, it must surely have
the same metaphysical status as an individual soul and be an entity in its own right. The
other component of the organic theory is political: the interests of the state have priority
over those of its citizens, which means that individuals’ interests may be sacrificed for
the good of the higher entity, as the occasion demands.
In fact, there is no explicit or conclusive evidence for the metaphysical claim in the
Republic. The state-soul analogy cannot be used to make the case on its own, because it is
not clear how far one should press the similarities between the two sides of the analogy.
Instead of assuming that state and soul have the same metaphysical status, Socrates may
just have wanted to point to structural similarities between them.18
If the metaphysical thesis fails, one might think the political component of the organic
theory collapses with it. Commentators who reject the organic theory as an interpreta-
tion of the Republic admit there are some passages that talk of individual happiness
being sacrificed for the greater good, but they claim these need not be taken in the way
Popper supposed. Take, for example, the beginning of book IV. Adeimantus has just
complained that the rulers of the ideal state will not have much happiness, given the
­austere conditions in which they live. Socrates replies that his concern is not to focus on
the happiness of some parts of the city at the expense of others but to look to the happi-
ness of the whole. He then compares the city to a statue (420b–421c), saying it would be
ridiculous to paint the eyes with the finest color at the expense of the rest of it. Rather
than using this as evidence for the organic theory, some commentators have interpreted
this as the proposal to maximize the aggregate happiness of the citizens, which is quite a
different matter. Like utilitarianism, and unlike the organic theory, this aggregative
approach does not see the mass of citizens as forming a distinct entity. The good to be
achieved is no more than the sum of the individual happiness of each citizen.19

17 Popper, Open Society, 79; see also Cross and Woozley, Republic, 76–77 and 132.
18 For the point, see Annas, Introduction, 180.
19 J. Neu, “Plato’s Analogy of State and Individual: The Republic and the Organic Theory of the
State,” Philosophy 46 (1971), 246, compares Plato to Bentham; see also G. Vlastos, “The Theory of Social
Justice in the Polis in Plato’s Republic,” in D. Graham, ed., Studies in Greek Philosophy, vol. 2 (Princeton,
N.J., 1995), 69–103, esp. 80–84, and Annas, Introduction, 179. Ironically, utilitarianism itself might be
accused of having to assume the organic theory in order to justify maximizing aggregate happiness
without regard to boundaries between persons. See D. Gauthier, Practical Reasoning (Oxford, 1962),
126, discussed in D. Parfit, Reasons and Persons (Oxford, 1986), 331–32. If so, utilitarianism would not
be the safe refuge that Popper’s critics have assumed.
218   Dominic Scott

However, in between the full-blown organic theory and aggregative approach, there
is a third option, which I think is the correct interpretation of the Republic. Whether or
not Plato conceived of the state as an entity in its own right, he does seem to subordinate
the interests of individual citizens to a good that is distinct from the aggregate happi-
ness. There can be no doubt that the founders of the state must aim at unity in the state.
Although the point is made repeatedly throughout the work (e.g., 422e–423d), it is par-
ticularly clear at 462a:

Then isn’t the first step towards agreement to ask ourselves what we say is the great-
est good in designing the city—the good at which the legislator aims in making the
laws—and what is the greatest evil? . . . Is there any greater evil we can mention for a
city than that which tears it apart and makes it many instead of one? Or any greater
good than that which binds it together and makes it one? (emphasis added)

He does not say here that unity is a good as a means to increase aggregate happiness; it is
the state’s greatest good—period.
Another passage that makes the same point comes in book VII, just after Glaucon has
complained that compelling the philosophers to rule constitutes an injustice against
them. In his reply, Socrates says:

s. You are forgetting again that it isn’t the law’s concern to make any one class in the
city outstandingly happy but to contrive to spread happiness throughout the city
by bringing the citizens into harmony with each other through persuasion or
compulsion and by making them share with each other the benefits that each class
can confer on the community. The law produces such people in a city, not in order
to allow them to turn in whatever direction they want, but to make use of them to
bind the city together.
g. That’s true, I had forgotten. (519e–520a; emphasis added)

The references to “forgetting” here almost certainly refer back to 462a and to the statue
analogy at 420b, specifically the sentence, “We aren’t aiming to make any one group out-
standingly happy, but to make the whole city so, as far as possible.” This means that we
should take 519e–520a as an authoritative interpretation of 420b: that is, that the earlier
passage was advocating something stronger than the aggregative approach all along.
Rather, what Socrates values most of all is a feature of the ideal state—its unity—that is
independent of the aggregate well-being of its members, and, in principle, this value
could come into competition with the happiness of individual citizens. In other words,
structural properties of the state have priority over not only any one individual’s happi-
ness but also the aggregate of all individuals’ happiness.20 Though distinct from the
organic theory as Popper conceived it, this is still a deeply controversial stance to take.21

20 On this point, see Kraut, “Cave,” 244–45.

21 My interpretation is close to that of L. Brown, “How Totalitarian Is Plato’s Republic?,” in
E. Ostenfeld, ed., Essays on Plato’s Republic (Aarhus, 1998), 13–27.
 The Republic   219

3.2 The Critique of Democracy

3.2.1 The Basic Argument
In setting up the ideal state in book II, Socrates lays down the principle of specialization
(PoS): each citizen must stick to one type of work alone. Initially, this is just used to insist
that different craftsmen spend all their time and energy on a single function. But as the
discussion proceeds, the principle is used to generate the three different classes in the
state. This happens with the auxiliary class at 374d–376c: while it was standard practice
for Greek states to conscript ordinary citizens to fight in times of war, PoS requires the
ideal state to have a permanent, professional military class, distinct from the producers.
Similarly, when it comes to deciding who should manage (as opposed to defend) the
state, PoS requires us to hive off a class of professional rulers (412b–414b).
The use of PoS to create a separate guardian class implicitly rules out democracy.
Since the principle lies at the heart of the antidemocratic argument of the Republic, it
merits a closer look. Socrates adduces two reasons in its support. The first (370a) has to
do with the distribution of natural talents: “We aren’t all born alike, but each of us differs
somewhat in nature from the others, one being suited to one task, another to another.”
The second (370b) concerns efficiency: “Does one person do a better job if he practises
many crafts or—since he’s one person—if he practises one?” (Interestingly, that argu-
ment only raises a doubt about direct democracy and seems to leave representative
democracy intact.) In what follows, I shall not discuss the second of these arguments.
The first argument could only be used to mount an attack on democracy if there really
are deep natural differences relevant to political decision-making. But as Plato well
knew, there were some who denied this straight out. Well before writing the Republic, he
himself had articulated the democratic view through the voice of Protagoras. In the dia-
logue named after him, he claims that political virtue is shared by almost all humans
alike; it is the basic skill that enables us to live together in society (322d–323c). Protagoras
agrees that other functions in society require specialization because of an uneven distri-
bution of talents. But political decision-making is a special case: in terms of the Republic,
it involves a competence to which the first argument for PoS does not apply.
Although Plato had articulated the democratic position on natural talents so clearly
in the Protagoras, Socrates in Republic II–IV seems merely to assume that it is wrong and
that the qualities needed for a political leader are to be found only in a very few. In III
414b–415d, he divides the citizens into gold (guardians), silver (auxiliaries), and bronze
(producers) and claims these natures are fixed at birth. But this is more announcement
than argument, and a democrat might counter by attributing apparent differences in
natural aptitude to nurture rather than nature. If everyone was given the right kind of
education, all might turn out to have enough “gold” in their souls to participate in politi-
cal decision-making.22

22 The earlier dialogue, the Meno, does seem to suggest that knowledge (and hence virtue) is equally
accessible to all, however difficult it may be to acquire (85c–d).
220   Dominic Scott

But Socrates does lend more support to his argument in the central books, which have
a great deal to say about the moral and intellectual qualities required of the ideal rulers.
Book V (475e–480a) argues that the knowledge needed for ruling is the knowledge of
Forms, which is well out of the reach of most people (cf. VI 494a: “The majority cannot
be philosophic”). As these books continue, Socrates emphasizes again and again the
enormous difficulty involved in apprehending the Forms: on top of the poetic or musi-
cal education proposed in books II–III, the trainee guardians need 30 years of intellec-
tual and practical training before they are ready to rule. And the natural talents required
are not just intellectual: Socrates also argues that one needs specific natural tendencies
toward such qualities as temperance, justice, magnanimity, and courage (VI 487a).
Again, the likes of Protagoras will object that everyone has sufficient cognitive and
moral resources to participate in the political process and that Socrates vastly overesti-
mates and misrepresents the qualities required for such participation. But to do this,
they now have to grapple with the arguments of the central books. Whatever the out-
come of that debate, my point here is just to chart the course of Plato’s main argument
against democracy and to show that, although it begins as early as the second book, it is
not until we encounter the metaphysics and epistemology of the central books that we
really find its basis.23

3.2.2 A Critique of Democratic Athens

In the course of defending the proposal for philosopher rulers, Socrates raises objec-
tions against contemporary democracy (VI 488c–495c). His critique is at its most
explicit in the ship of state analogy (488c) and its aftermath. We are to imagine a ship-
owner who is “bigger and stronger than everyone else on board” but hard of hearing, a
bit shortsighted, and with a defective knowledge of seafaring (488a–b). The sailors all vie
with one another in persuading him to let them captain the ship, each with the intention
of plundering the goods on board for himself. For them, what makes a good navigator is
his capacity to persuade the shipowner to put him in control. In fact, the true captain
would be someone who understood sufficient astronomy and meteorology to steer the
ship on its course. But such a figure would be dismissed as a useless stargazer by the rest
of the crew.
In this analogy, the shipowner is the demos, the sailors the democratic politicians,
and the stargazer the philosopher. Socrates’ main purpose is to explain why philoso-
phers were held in such low esteem and why the proposal for philosopher-rulers would
meet with such derision. But this passage also functions as a critique of democracy as
practiced at Athens. This becomes explicit once Socrates unpacks the analogy, focusing
on the relation between the sailors and the shipowner. Such is the desire of democratic
politicians to curry favor with the demos that they do not even exert the kind of control

23 On my view, this makes for a clear contrast between the political and ethical arguments of the
Republic. In the latter case, the actual defense of justice conducted in the work makes hardly any use of
the metaphysics of the central books, whereas those books are designed precisely to develop the
political theory of the work, including the critique of democracy.
 The Republic   221

of which they would like to boast. Instead, the power relation is very much the other way
around: they spend all their time and energy trying to discover the values of the demos
and, through a process of assimilation, gradually adopt those values for themselves
(490a–495c; cf. esp. 492c). There is a double insult against Athenian democracy here.
The demos itself is ill equipped to run its own affairs, and the politicians merely ingrati-
ate themselves to it; if they ever did have any finer qualities, they lost them in the process
of winning power.
This critique differs from the main argument by focusing not on the problems that
democracy has in principle with leadership but on an actual democracy and its “lead-
ers.” Lest there be any doubt that Athenian democracy is in Plato’s sights throughout
this discussion (488c–495c), we need do no more than mention the obvious allusions
to such figures as Alcibiades (494c–495b), the sophists (493a), and, not least, Socrates
himself (496c).

3.2.3 Democracy in Book VIII

Finally, we come to the most explicit critique of democracy in the Republic: book VIII
(557a–558c). The democracy portrayed here is a very extreme form, but this should not
cause us any surprise, because Socrates’ procedure in analyzing all the degenerate con-
stitutions in books VIII–IX is to isolate the main forms of political vice. In different
combinations with each other, they could form an infinite number of constitutions, but
Socrates wishes to focus on the primary colors of political theory, as it were.
Pure democracy is based on the ideal of freedom, just as timocracy was based on
honor and oligarchy on wealth. The freedom in question is the freedom to do as one
chooses: even convicted criminals are allowed to roam the streets unchecked (558a).
There is no condemnation of any way of life or value; it is a society of complete tolera-
tion. In this spirit, the democratic state elects its leaders by lot, refusing to admit merit or
character as requirements for office. As a result, the state changes complexion according
to who holds power on any particular day: it is a political chameleon, a supermarket of
constitutions, barely a distinctive constitution in its own right.
For Socrates, the nub of the problem with democracy lies in its instability—at least to
the extent that it seeks to realize its core values: freedom and toleration. For as well as
being basic to democracy, these are also the source of its undoing. Extreme toleration
allows criminal elements to grow strong, create their own bodyguards, and then use
them to replace democracy with tyranny. All constitutions are unstable to some degree:
even the ideal state would not last forever but would ultimately degenerate into a timoc-
racy. But democracy is especially unstable and, to make matters worse, lies only one step
away from tyranny, the very worst kind of constitution.
This is not primarily a critique of any existing democracy, certainly not Athens, which
was no model of toleration (witness Socrates’ fate at its hands). Nor should we have
been expecting him to have any actual target in mind. His methodology in books
VIII–IX commits him to finding a constitution that is the political equivalent of a chem-
ical element—perhaps a highly volatile substance that might rarely be found in practice,
if at all. This critique also differs from what I have called the “main” argument against
222   Dominic Scott

democracy. Both can be called arguments against democracy in principle, but one
focuses on the democrat’s attitude to leadership and its qualifications, the other on the
democrat’s most basic values and of the dangers of following them through in practice.
In this way, the two arguments against democracy in principle are independent of
one another.

4. Philosophy in the Central Books

We have just seen how the basic argument against democracy finds support in a number
of claims developed in the central books: that the knowledge required for political
­decision-making is philosophical, that such knowledge is extraordinarily difficult to
acquire, and hence that those actually qualified to rule in the ideal state would be very
few and far between. In this section, I explore these claims further, by looking more
closely at how books V–VII distinguish the philosopher from the non-philosopher, and
how they characterize the nature and difficulty of acquiring philosophical knowledge.
The discussion of philosopher-rulers in these books can be broadly divided into two
sections. The first defends the claim that only philosophers should rule (V 475e–VI 502a);
the second describes their education (VI 504a–VII 540c). Socrates starts the first section
by arguing that only philosophers have the requisite knowledge for ruling, and then
turns to their moral qualities, claiming that these coincide with the qualities required
for political office. At 487b, Adeimantus objects to Socrates’ proposal on the grounds
that most people would think philosophers the last people suited to running the state.
Surprisingly, Socrates feigns agreement, but his real point is that this only applies to
those popularly conceived of as philosophers. Genuine philosophers would be utterly
different from the useless cranks who have usurped the title, and, if people understood
a philosopher’s true nature, they would not be so outraged by Socrates’ proposal.
The second main section of books V–VII, on the guardians’ education, begins with
three famous images: the sun, the line, and the cave. The first is an attempt to sketch the
nature and power of the supreme object of study, the Form of the Good; the second pres-
ents a classification of the four cognitive stages through which their education should
apparently pass; and the cave allegory illustrates their ascent from the sensible world to
the intelligible and their subsequent return to the world of practical affairs. The rest of
book VII goes through the educational curriculum in order, including 10 years of math-
ematics and 5 of dialectic.
Because the cave allegory (514a–517a) is central to our interests, I need to rehearse
some of the essentials. Socrates asks his interlocutors to imagine a group of people sit-
ting at the bottom of a cave, chained to their seats (a fact of which they are quite
unaware), with their backs to the entrance. Behind them is a fire, and in front of the fire
people are moving artifacts and puppets above a wall. This creates shadows in front of
the prisoners, which they think constitute reality. Next, we have to imagine what would
 The Republic   223

happen if one of them were released from his chains and turned around. He would be
initially dazzled by the fire, finding it difficult to make out the puppets, and would much
rather return to look at the shadows. The same pain would occur if he was dragged out of
the cave into the sunlight, and he would only be able to look at the shadows or reflections
of the objects. Eventually, however, his eyes would become accustomed, and he would
be able to see everything clearly. In the allegory, the outside of the cave represents the
intelligible world and the inside the perceptible world, the fire within it being equivalent
to the sun. The prisoners bound to their seats are “like us” (515a): they stand for the
­normal cognitive condition of most humans.

4.1 The Definition of Philosophy

At V 474b–c, immediately after announcing the third wave of paradox, Socrates prom-
ises to define what a philosopher is, a project that appears to have been completed by the
beginning of VI (484a). The definition depends on the distinction between Forms, uni-
tary in nature and apprehended only by thought, and their perceptible manifestations
in, for example, colors, shapes, and sounds. Using the Form of Beauty as his example,
Socrates contrasts the philosopher, who apprehends the Form itself, with the “lovers of
sights and sounds,” a group of people who pride themselves on having acquired as much
experience of beauty as possible (e.g., by frequenting cultural and dramatic festivals).
Their problem, according to Socrates, is that they do not apprehend the Form of Beauty
itself. What they call “beauty” they conceive of as a range of different sensible properties,
differing from one context to another: they would say, for example, that what makes an
object beautiful in some cases is bright color and, in others, dark color.24
But they deny that there is a single, distinct entity that underlies and explains these
perceptible phenomena, an entity apprehended only by thought. Because they confuse
Beauty itself with its instances or likenesses, Socrates compares them to people in a
dream: they “think that a likeness is not a likeness but rather the thing itself that it is like”
(476c). Accordingly, he says that such people do not have knowledge of beauty, merely
opinion about it. It is only the philosophers who have knowledge, because they grasp the
nature of the unitary Form in itself and distinguish it from its instances. The cave alle-
gory confirms this distinction, insofar as the prisoners think that the shadows, which
are ultimately mere likenesses of the objects outside the cave, constitute reality. As the
objects outside the cave are the equivalent of the Forms, we find the same point that the
prisoners, who are “like us,” are in a dreamlike state, thinking that a likeness is the very
thing of which it is the likeness.25

24 For this reading of the passage, see G. Fine, On Ideas (Oxford, 1993), 58–59.
25 The allegory also complicates the distinction set out in book V, because there are different levels
of similarity in the cave: the shadows are images of the puppets, which are themselves images of the
objects outside. Unfortunately, there is not space in this essay to unpack the complex imagery of the
cave in any detail.
224   Dominic Scott

What this shows is that in books V–VII Socrates defines philosophy by reference to its
objects. Philosophy is not simply a matter of abstracting away from concrete particulars
to apprehend general patterns and structures26 but involves a commitment to a peculiar
metaphysical stance—that there exist unchanging, nonsensible objects of which ­sensible
properties are mere likenesses.
In articulating the distinction between philosopher and non-philosopher, Socrates
leaves no room for doubt on one point: the non-philosopher has no awareness of a non-
sensible reality. This much is beyond dispute. Nevertheless, it leaves room for the follow-
ing question: Might the non-philosophers have a latent grasp of the Forms (cf. 518b–c),
which they use unconsciously to structure their thought about the sensible particulars
around them? Although they do not possess explicit knowledge, they still manage to
classify just things as just and beautiful things as beautiful, and they have a vocabulary
that they use in making such classifications. (Recall that their classification of just and
unjust actions is not so far off the mark: in fact, the Platonic conception of justice
­converges to a striking degree with the ordinary one.) The question arises of how they
manage to do this, and the suggestion just made would provide an answer.27
This interpretation, however, is controversial. Nowhere in the Republic does Socrates
even allude to non-philosophers using their latent resources to form opinions or classify
objects of perception, which we would expect given the importance of the idea. So an
alternative view is that non-philosophers may indeed have a latent grasp of the Forms,
but this remains entirely inactive. They acquire their concepts and opinions from expe-
rience—even about beauty, goodness, and the virtues. If one is skeptical that Plato would
ever have subscribed to such a view, one should turn to Phaedo 68d–69b, where Socrates
criticizes non-philosophers for having (what is in effect) a naturalistic conception of the
virtues. Most people, he complains, conceive of temperance as abstaining from one
pleasure for the sake of a greater one in the future, or of courage as facing one danger to
avoid facing a greater one. In each case, the virtue is “cashed out” by reference to experi-
ence—that is, in terms of the feelings of pleasure and pain. He states that a parallel
account of justice could be given, without saying what it is (69b), though Glaucon’s
social contract theory could provide such an account, insofar as it analyzes justice as a
balance between expected benefits and burdens. All this suggests that for Plato it is quite
possible to form evaluative concepts without any recourse to the Forms, latent or
explicit; and as far as non-evaluative properties are concerned, we should note that

26 Contrast the view of E. Havelock, Preface to Plato (Oxford, 1963), who thinks that for Plato a
philosopher “is at bottom a man with the capacity for the abstract” (282). For a wide-ranging and
innovative treatment of Plato’s attempt to define philosophy, see A. W. Nightingale, Genres in Dialogue
(Cambridge, 1995).
27 For this view, see M. Ferejohn, “Knowledge, Recollection, and the Forms in Republic VII,” in
G. Santas, ed., The Blackwell Companion to Plato’s Republic (Oxford, 2006), 214–33; see also V. Harte,
“Language in the Cave,” in D. Scott, ed., Maieusis: Studies in Honour of M. F. Burnyeat (Oxford, 2007),
195–215.
 The Republic   225

Republic VII 523a–524c clearly states that sight sees such properties as largeness and
smallness; what it cannot do is understand the nature of each.28
Although there is not space to pursue this controversy any further here, it is worth
mentioning because it invites us to press the question of exactly how sharply Plato draws
the distinction between philosopher and non-philosopher: What degree of continuity
exists between the two perspectives? How the controversy is resolved might also
rebound on our assessment of the argument against democracy: If non-philosophers
have a latent grasp of Forms that they actually use in their everyday thinking, shouldn’t
Socrates be prepared to take their contributions to political debate more seriously
than he does?

4.2 The Difficulty of the Transition

Even on the more optimistic of these interpretations, there remains a substantive gap
between philosophical and non-philosophical perspectives. We now turn to the diffi-
culty of bridging this gap, a point reinforced throughout the central books. In discussing
the lovers of sights and sounds, Socrates claims not only that they fail to apprehend the
Form of Beauty but also that they could not be led to do so even if someone actually tried
to help them (476b–c, 479e). The point is immediately generalized: “There are very few
people who would be able to reach the Beautiful itself and see it by itself ” (476b). So it is
no surprise that later on, at VI 493e, he claims that “the many” cannot accept the reality
of the beautiful itself, and cannot become philosophers.
The cave illustrates the difficulty of making the transition by its use of the light meta-
phor. Both when the prisoner first turns around to the fire and when he steps outside the
cave he is dazzled. Initially, he just wants to return to his seat and stay there. The allegory
shows how apprehending the different levels of reality is not just a matter of looking but
of accustoming oneself, a process that is both arduous and time-consuming. Specifically,
the allegory brings out the problem of moving from the sensible world to the intelligible,
the movement from inside to outside the cave. Socrates talks of the prisoners being
chained to their seats. At first, one might be tempted to view this in a political light, per-
haps seeing them as the prisoners of the puppeteers. Thoughts of 1984-style scenarios
readily come to mind. But the kind of imprisonment in question is epistemological, not
political. At Phaedo 81b– 83e, Socrates also talks of the soul as a prisoner, explicitly of the
bodily senses and feelings. What he means is that when a person perceives something or
feels physical pleasure, he also thinks that what he is experiencing is “most real” (cf. 83c).
The vividness of the perception or the intensity of the pleasure imposes a perspective
that makes it almost impossible to understand how there could exist a nonsensible real-
ity, of which the sensible is merely a copy. The prison of the cave allegory is Plato’s way of
representing a similar point in the Republic.

28 On this, see D. Scott, Recollection and Experience (Cambridge, 1995), 83.

226   Dominic Scott

Another passage that emphasizes the difficulty of making the transition to the Forms is
the section discussing the role of mathematics in the education of the would-be guardians.
They must spend 2 years each in the study of arithmetic, plane geometry, solid geometry,
astronomy, and harmonics: 10 years in all. Why? At several points Socrates states that
arithmetic and geometry are useful in warfare (521d, 522e, 525c, 526d, 527c) and also
remarks that people who study arithmetic tend to be mentally sharper as a result (526b).
But neither of these reasons can begin to explain the length and intensity of the rulers’
mathematical studies. To find a better explanation, we might point to a brief but significant
remark made at 531c: that harmonics is useful for studying the Beautiful and the Good. The
point may be that harmonics seeks to understand the principles underlying the order and
beauty of audible phenomena. The same applies to astronomy: Plato bids us to leave
behind the visible beauty of the heavens and move back to its mathematically explicable
order (530b). The more general point is that the perceptible world exhibits goodness and
beauty, which we can begin to understand if we investigate the underlying mathematics.
But, ultimately, mathematics itself requires explanation, and this will only be provided by
knowledge of the Good, the Beautiful, and the other Forms.29 Nevertheless, mathematics
is worth studying because it cultivates an intellectual appreciation of beauty and order,
thus preparing us for understanding the nature of the Forms.
Still, one might ask why we should have to go through such long mathematical
­training. Once we have become aware of the mathematically intelligible order underlying
sensible phenomena, why not advance as quickly as possible to dialectic and investigate
the Form of the Good without further delay? Again, the answer lies (partly) in the pecu-
liar difficulty of moving from the sensible world to the intelligible, the movement from
inside to outside the cave. Although, as we have just seen, this point had already been
made in the Phaedo, what is new in the Republic is the role of mathematics as the bridge
to help us make the extraordinarily difficult transition from perceptible to intelligible.
When we start mathematical study, we start to think about the perceptible world in more
abstract terms. But this really is just a start and is quite different from believing that
there exists a distinct intelligible realm to which the sensible is merely an approximation.
It is only after a long period of mathematical study that such a mindset starts to form.

4.3 Pretenders to Philosophy

In the Republic, therefore, philosophy involves a complete reorientation toward a
­distinctive subject matter, a process that goes directly against the cognitive grain. This,
in turn, helps explain a theme that runs throughout the central books: that true philoso-
phers are so few and far between. There are, of course, many who style themselves as

29 On the role of mathematics in the philosophers’ education, see M. Burnyeat, “Platonism and
Mathematics: A Prelude to Discussion,” in A. Graeser, ed., Mathematics and Metaphysics in Aristotle
(Bern, 1987), 213–40, and Burnyeat, “Plato on Why Mathematics Is Good for the Soul,” in T. Smiley, ed.,
Mathematics and Necessity: Essays in the History of Philosophy, Proceedings of the British Academy 103
(2000), 1–81.
 The Republic   227

­ hilosophers, but for Socrates they are in one way or another only pretenders to the title.
p
The lovers of sights and sounds claim expertise in beauty and are said to be “like philoso-
phers” (475e). Yet, for all their experience, they fail to grasp Beauty in itself. In book VI
(495a–496a), Socrates complains that those who genuinely have a talent for philosophy
tend to be deflected from it, and those who attempt to take it up instead are for the most
part sophists. But philosophy, being distinguished by the nature of its objects, requires
much more than logical agility and an appetite for abstraction. Socrates makes a similar
point at VII 537d–539d, complaining that, as things are, people engage in (what they take
to be) dialectic too early in life. He is referring to young men who debate about the just
and the beautiful, asking and answering definitional questions. The problem is that they
enjoy the battle of argument more than the pursuit of truth. Again, despite their ability
to operate at a relatively abstract level, such people are not true philosophers, and their
debates have only the appearance of true dialectic.
But if most of what passes for philosophy is just a poor imitation, and true philoso-
phers are so thin on the ground, what are we to say about Socrates’ interlocutors—or,
indeed, ourselves as readers of the Republic? As we follow the arguments of the work,
what sort of activity are we engaged in? Is it, in fact, philosophy in the true sense of
the word?
I argue above that the defense of justice actually mounted in books II–IV and VIII–IX
makes almost no reference to the Forms. For the most part, it appeals to empirical
assumptions directly about human psychology and politics. In the central books, there
is talk about Forms, about the knowledge based on them, and about the way in which
such knowledge can be attained. But this is not dialectic in the true sense—that is, direct
inquiry into the nature of the Forms. We are just peeping outside the cave but then find-
ing the light too strong; or perhaps we have not even been that far, but someone else,
who has stepped outside, is telling us what it is like. The central books are best described
as meta-philosophy and the remaining books as preparation for philosophy. By its own
lights, the Republic is not really a work of philosophy.

Bibliography
Ackrill, J. “What Is Wrong with Plato’s Republic?” Essays on Plato and Aristotle, 2nd ed.
(Oxford, 2001), 230–52.
Annas, J. An Introduction to Plato’s Republic (Oxford, 1981).
Brown, E. “Minding the Gap in Plato’s Republic,” Philosophical Studies 117 (2004), 275–302.
Brown, L. “How Totalitarian Is Plato’s Republic?” in E. Ostenfeld, ed., Essays on Plato’s Republic
(Aarhus, 1998), 13–27.
Burnyeat, M. “Platonism and Mathematics: A Prelude to Discussion,” in A. Graeser, ed.,
Mathematics and Metaphysics in Aristotle (Bern, 1987), 213–40.
Burnyeat, M. “Plato on Why Mathematics Is Good for the Soul,” in T. Smiley, ed., Mathematics
and Necessity: Essays in the History of Philosophy, Proceedings of the British Academy 103
(2000), 1–81.
Cooper, J. “The Psychology of Justice in Plato,” American Philosophical Quarterly 14 (1977),
151–57.
228   Dominic Scott

Cross, R., and Woozley, A. Plato’s Republic (London, 1964).

Ferejohn, M. “Knowledge, Recollection, and the Forms in Republic VII,” in G. Santas, ed., The
Blackwell Companion to Plato’s Republic (Oxford, 2006), 214–33.
Fine, G. On Ideas (Oxford, 1993).
Gauthier, D. Practical Reasoning (Oxford, 1962).
Grube, G., trans. Plato: Republic, rev. C. Reeve (Indianapolis, Ind., 1992).
Harte, V. “Language in the Cave,” in D. Scott, ed., Maieusis: Studies in Honour of M. F. Burnyeat
(Oxford, 2007), 195–215.
Havelock, E. Preface to Plato (Oxford, 1963).
Irwin, T. Plato’s Ethics (Oxford, 1995).
Irwin, T. “Republic 2: Questions about Justice,” in G. Fine, ed., Plato 2: Ethics, Politics, Religion
and the Soul (Oxford, 1999), 164–85.
Kraut, R. “The Defense of Justice in Plato’s Republic,” in R. Kraut, ed., The Cambridge
Companion to Plato (Cambridge, 1992), 311–37.
Kraut, R. “Return to the Cave: Republic 519–521,” in G. Fine, ed., Plato 2: Ethics, Politics, Religion
and the Soul (Oxford, 1999), 235–54.
Nettleship, R. Lectures on the Republic of Plato, 2nd ed. (London, 1901).
Neu, J. “Plato’s Analogy of State and Individual: The Republic and the Organic Theory of the
State,” Philosophy 46 (1971), 238–54.
Nightingale, A. W. Genres in Dialogue (Cambridge, 1995).
Parfit, D. Reasons and Persons (Oxford, 1986).
Popper, K. The Open Society and Its Enemies, vol. 1: The Spell of Plato, 5th ed. (London, 1966).
Sachs, D. “A Fallacy in Plato’s Republic,” Philosophical Review 72 (1963), 141–58; reprinted in
G. Vlastos, ed., Plato, vol. 2: Ethics, Politics and Philosophy of Art and Religion (New York,
1973), 35–51.
Scott, D. “Metaphysics and the Defence of Justice in the Republic,” Proceedings of the Boston
Area Colloquium in Ancient Philosophy 16 (2000), 1–20.
Scott, D. Recollection and Experience (Cambridge, 1995).
Taylor, C. “Plato’s Totalitarianism,” Polis 5 (1986), 4–29; reprinted in G. Fine, ed., Plato 2:
Ethics, Politics, Religion and the Soul (Oxford, 1999), 280–96.
Vlastos, G. “Justice and Happiness in the Republic,” Platonic Studies, 2nd ed. (Princeton, N.J.,
1981), 111–39.
Vlastos, G. “The Theory of Social Justice in the Polis in Plato’s Republic,” in D. Graham, ed.,
Studies in Greek Philosophy, vol. 2 (Princeton, N.J., 1995), 69–103.
White, N. “The Classification of Goods in Plato’s Republic,” Journal of the History of Philosophy
22 (1984), 393–421.
White, N. A Companion to Plato’s Republic (Oxford, 1979).

Afterword:
In addition to the items listed in the bibliography above, there are also some i­mportant
­collections of articles on the Republic, which came out around the time I wrote this c­ hapter, or
afterward:
Ferrari, G. ed. The Cambridge Companion to Plato’s Republic (Cambridge, 2007).
McPherran, M. ed. Plato’s Republic: A Critical Guide (Cambridge, 2010).
 The Republic   229

Notomi, N. and Brisson, L. eds. Dialogues on Plato’s Politeia (Republic): Selected Papers from
the Ninth Symposium Platonicum. International Plato Studies 31 (Sankt Augustin, 2013).
Santas, G. ed. The Blackwell Guide to Plato’s Republic (Oxford, 2006).
Vegetti, M., Ferrari, F., and Lynch, T. eds. The Painter of Constitutions: Selected Essays on Plato’s
Republic. (Sankt Augustin, 2013).
I have also written at some length about the methodology of the Republic in D. Scott, Levels of
Argument: A Comparative Study of Plato’s Republic and Aristotle’s Nicomachean Ethics (Oxford,
2015), chs 1–5.
 chapter 10

Pl ato’s Pa r m en ides
A Reconsideration of Forms

Sandra Peterson

The Parmenides’s reputation as Plato’s most challenging,1 indeed enigmatic,2 work

continues.

1. Audience, Narrators, Dates

The dialogue is a narration by a Clazomenean Cephalus, unknown to readers today, to

unspecified listeners at an unspecified date. Cephalus relates that he and unnamed
­companions visited Athens, seeking Glaucon and Adeimantus, to ask their half-brother
Antiphon to recite arguments (logous, 126b8) Antiphon heard from Pythodorus and
memorized (126c). Pythodorus witnessed the arguments when Socrates, Zeno, and
Parmenides engaged in dialegesthai (dielechthêsan, 126c1). Dialegesthai is conversation,
especially conversational contest in which a questioner gets answers from a respondent
and then makes inferences upon which both parties agree. Expertise at dialegesthai is
dialektikê, dialectic.3 (Contestants are sometimes inexpert.) Plato depicts youths as
­typically interested in engaging in or observing such contests.4

1 S. Scolnicov, Plato’s “Parmenides” [Parmenides] (Berkeley, 2003), 1; G. Priest, “The Parmenides.

A Dialethic Interpretation” [Parmenides], Plato Journal 12 (2012), 1–63, 1; C. Kahn, Plato and the
­Post-Socratic Dialogue [Plato] (Cambridge, 2013), 1.
2 M. L. Gill, “Introduction” to M. L. Gill and P. Ryan, Plato. “Parmenides” [Parmenides]
(Indianapolis, Ind., 1996), 1. Kahn Plato, 2; S. Rickless, “Plato’s Parmenides,” [Parmenides] Stanford
Encyclopedia of Philosophy, revision of July 30, 2015), 1.
3 A dialektikos, the person that has dialektikê, has the skills of questioning and answering (Cratylus
390c). M. Frede, “Plato’s Arguments and the Dialogue Form,” Oxford Studies in Ancient Philosophy 9
Suppl. (1992), 201–19 describes the role and commitments of questioner and answerer. See Aristotle,
Topics 159a15–35 on the job of the questioner and the answerer in competitive dialegesthai.
4 Theaetetus 146b; Gorgias 485a; 486c; Lysis 204a; 211b; Charmides 154e–155a; Euthydemus 275a–c;
Alcibiades 106b; Protagoras 336c; 348b; Sophist 232c–233b. Xenophon Memorabilia 1.2. 40–46 mentions
Pericles’s youthful interest.
232   Sandra Peterson

Cephalus says his unnamed companions are “very much philosophoi.”5 We discern
that philosophoi are eager to hear (126e5) memorized dialegesthai. But because Plato’s
other dialogues give multiple and quite different accounts of philosophia,6 while the
Parmenides gives none, we are uncertain what Cephalus’s word philosophoi means.
Pythodorus witnessed the dialegesthai in 450 or 454 b.c., at one of the quadrennial
Panathenaic Festivals. Antiphon recited his second-hand report to Cephalus long
afterward, at perhaps age 40, so in about 382 b.c.7
The Parmenides uniquely depicts Socrates as perhaps 15 or 19 years old. Plato’s other
dialogues show an older Socrates that prefers dialegesthai.8 He uses it to engage youths
and to test professionals and pretenders.9

2. Overview

Cephalus’s recitation of Antiphon’s recitation of Pythodorus’s recitation begins with

Socrates’ flurry of objections to a lecture of Zeno’s. Zeno had concluded that there
cannot be many things, arguing that they would “undergo impossibilities” (127e8).
Although Socrates calls Zeno’s conclusion paradoxical in the extreme (“against every-
thing ordinarily said,” 127e9–10), Socrates declares himself unimpressed (129b). Socrates
would, however, be surprised if someone showed (130a) that kinds (genê) or forms (eidê)
themselves (129c1), items such as likeness and unlikeness, have opposite attributes
(129e), “in all kinds of ways” (130a).
Parmenides then questions Socrates about forms. Socrates falls into absurdities and
admits confusion. Parmenides diagnoses that Socrates tries to mark off forms too soon,
before having gotten exercised (135c). Parmenides recommends exercise (135d4) of
which Socrates is to do many repetitions (135e–136a; 136b-c) if Socrates is to discern the
truth (136c).
Because Socrates does not understand (136c) Parmenides’s detailed directions
(135e–136c) for such exercise, Parmenides labors through (137b) a sample game (137b1,
paidian; 135d5, adoleschias) of 195 arguments questioning a youngster, Aristoteles.
Aristoteles forcefully agrees to the totality of conclusions (166c, “Most true”). One
conclusion—that the one, the form, both is and is not “all things in all ways” (166c1–4)—
achieves exactly what Socrates said would surprise him.

5 Translations from the Parmenides are mine, from M. Migliori and C. Moreschini, Platone.
Parmenide (Milan, 1994).
6 On Plato’s multiple explanations of philosophia see S. Peterson, “Plato’s Reception of Socrates: One
Aspect,” in C. Moore (ed.), The Brill Companion to the Reception of Socrates (Leiden, 2019).
7 D. Nails, The People of Plato (Indianapolis, Ind., 2002) s.v. “Antiphon.”
8 Symposium 194d. Laches 186e–187c.
9 Laches 187e–188a. See J. Brunschwig, “Aristotle on Arguments without Winners and Losers,”
Wissenschaften Jahrbuch 1984/85: 31–40 on contests aiming not to gratify a winner but to clarify an issue.
 Plato’s Parmenides   233

3. Socrates’ Ill-Assorted Barrage

At 127e Socrates reviews one of Zeno’s arguments:

Suppose there are many.

Then they must be both like and unlike.
It is impossible for unlike things to be like or like things unlike.
So there are not many.

Socrates’ ensuing objections range from personal insult to subtle witticism to common
sense to explicit logical devastation.
The insult is that Zeno tries to deceive (128a7) his audience into thinking that he says
something different from his intimate Parmenides. The witticism is that Zeno has given
many arguments against the thesis that there are many things (127e11–12).
Socrates’ common-sense objection is that an explicit premise of Zeno’s is false: it is
false that what is both like and unlike “undergoes impossibilities” (127e8): it is, rather, an
unsurprising fact that visible things are both like “in this way and to such an extent” and
also unlike (129b–e). (More later on what “in this way” means.)
Socrates gives his logically devastating objection in a single question:

Don’t you acknowledge (nomizeis) that there is itself by itself (auto kath’ hauto) a
form (eidos ti) of likeness, and again something else opposite (allo ti enantion) to
that sort (tô(i) toioutô(i)), . . . what unlike is: and that in these two beings you and I
and the other things we call many share? (128e6–129a3).

The young Socrates (127c5) displays brash adolescent combativeness by initiating a

question-answer contest with the veteran professional Zeno. But Socrates has a decisive
objection, namely that Zeno’s argument relies on an unstated premise that entails the
very proposition that Zeno purports to reduce to absurdity. In saying that to be both like
and unlike is to “undergo impossibilities” (127e8), Zeno presupposes that there are the
two opposites likeness and unlikeness.
Commentators appear entirely to overlook that decisive objection.10 It deserves
reflection.
“Form” is the standard and unavoidable rendering of Socrates’ word eidos (129a).
Eidos, literally “that which is seen,” or “aspect,” indicates a feature or attribute via which
something may be viewed. I may view or consider my teacup in various aspects: as
vessel, pottery, blue. . . . The English “form,” from the Latin translation forma for eidos,
is unhelpful for those not thinking in Latin. “Form” most usually means either

10 See Meinwald, Parmenides, 5; Gill Philosophos (Oxford, 2012), 19; Rickless, Plato’s Forms in
Transition [Forms] (Cambridge, 2007), 99; Priest, “Dialethic,” 1.
234   Sandra Peterson

“shape-outline” (as in “form-fitting”) or “variety” (as in “What form of welcome did you
expect?”). Instead Socrates here means a noticeable feature.
Forms such as likeness and unlikeness, as Socrates later says, are “grasped in
­reasoning” (130a2). That is, they are intelligible content. Without the presupposition
that there are the two opposites likeness and unlikeness Zeno’s statement that to be both
like and unlike is to suffer impossibilities would have no content. Zeno’s attempted
reduction to absurdity incoherently relies on a hidden premise entailing exactly what
he targets for reductio.
Many interpreters hold a different view of Socrates’ objections to Zeno. They hold
that Socrates’ word “form,” and his phrase, “itself by itself,” are here technical vocabulary
from a complex theory of Plato’s11 of which Plato sketches versions in Phaedo (75d,
76d7–9, 96a101e, 100b) and Republic (475b–484a, 523a–525b, 596a–597e).12 Here are
some reasons to reject that view of Socrates’ objections.
First, with his word nomizeis, meaning primarily “recognize a customary usage,”
Socrates treats Zeno’s implicit premise as common belief, not as technical innovation.
Second, the word “itself” is primarily a topic-focusing device. It directs the hearer to
focus on the aspect, likeness, by itself, that is, considered without any further description
or qualification such as “likeness in age” or “likeness in hometown.” (See a topic-focusing
“thirst itself ” at Republic 437e4).13 The augmented phrase “itself by itself ” can simply be
a more emphatic version of “itself.”14 By itself the word itself is no evidence that Socrates
uses it as a technical phrase trailing theory behind.
Third, the view that Socrates imports a technical theory includes the claim that
Socrates’ phrase “itself by itself ” indicates that a form such as unlikeness is separate
from the many unlike items, the participants in the form.15 Although very soon (130b2–5)
Parmenides’s questions extract that claim from Socrates, to suppose that Socrates’
phrase “itself by itself ” at 128e6 already conveys that forms are separate from partici-
pants unjustifiably charges Socrates with inept reasoning. I will explain.
Socrates introduces the word separately into the conversation a bit later:

if someone first distinguishes separately (chôris) the forms themselves by themselves

of the things I was talking about just now, such as likeness and unlikeness . . .and
all such (129d6–e1).

11 K. Sayre, Parmenides’ Lesson [Parmenides] (Notre Dame, 1996), 65. Scolnicov, Parmenides, 48–48.
See note 10 of this chapter.
12 See T. Irwin, “The Theory of Forms,” in G. Fine (ed.), Plato I: Metaphysics and Epistemology
(Oxford, 1999), 143–70 and D. Sedley, “An Introduction to Plato’s Theory of Forms” [“Introduction”]
Royal Institute of Philosophy Supplement 78 (2016), 3–22.
13 See T. Penner, The Ascent from Nominalism [Ascent], (Dordrecht, 1987) 91–95. See 32–37 of
S. Peterson, “The Language Game in Plato’s Parmenides” [“Game”], Ancient Philosophy 20 (2000), 19–51,
and F.-G. Hermann, Words and Ideas: The Roots of Plato’s Philosophy [Roots], (Swansea, 2007), 8–9.
14 G. Fine, “Separation,” Oxford Studies in Philosophy 2 (1984), 31–87, 61; Scolnicov, Parmenides, 48;
Hermann, Roots, 14–17.
15 M. L. Gill, “Introduction” to Parmenides, 16. Rickless, Forms, 18–20, cites G. Vlastos, “Separation
in Plato” [“Separation”], Oxford Studies in Ancient Philosophy 5 (1987) 187–96, 162. See also Fine
“Separation,” 58–59 and R. Allen, Plato’s “Parmenides” (New Haven, Conn. 1997) 98–99.
 Plato’s Parmenides   235

“Just now” refers back to the distinction at 128e–129a2 between the two forms likeness
and what unlike is (“a form itself by itself, of likeness, and another, opposite to this”).
“Separately” recalls that earlier distinction, the reason for which was clearly that
opposition entails distinctness.16 For Socrates to imply at 129d with “itself by itself ” that
likeness and unlikeness are separate in whatever way from extra items a and b respec-
tively would give no reason to distinguish likeness from unlikeness.

4. Socrates’ Decisive Objection

Parmenides’s later rephrasing of Socrates’ objection provides further indication that the
objection does not rely on technical assumptions. Having questioned Socrates,
Parmenides says:

If someone, Socrates, looking at all these things just now <shown to be difficulties>
and others such, will in return not allow there to be forms among beings, and will
not mark off (horieitai) a form of each <item that is> one, wherever he will turn he
will not get hold of <a> thought (dianoian) . . . and thus he will destroy altogether the
possibility of dialegesthai. But then you seem to me to have perceived this sort of
thing even more (kai mallon). (135b5–c1)

Parmenides here recalls and affirms Socrates’ observation: thought and dialegesthai
require distinct forms.
I emphasize this local setting that illuminates Socrates’ question. All interpreters of
text consider setting. If the words of Socrates’ question float by in our bowl of alphabet
soup, none of us looks for supporting theory. Here the setting including Parmenides’s
respect for the force of Socrates’ simple question about forms sheds more light than
Plato’s other dialogues.17
Socrates’ clumsy long-windedness, to which Plato calls attention (130a7, “when he
paused”) is partial cause of commentators’ overlooking the crucial objection. Were
Socrates expert at dialegesthai, he would have stopped talking and awaited an answer
from Zeno immediately after the one-sentence question that struck the lethal blow.
(Compare the older Socrates (Protagoras 329d–333e), reputed for skill at dialegesthai
(336c), who questions briefly.) In violating convention and strategy for dialegesthai,
Socrates gives Parmenides material for counter-questioning.

16 Similarly Parmenides (149e8–11) extracts Aristoteles’s agreement that if largeness and smallness
are opposites, there are the two forms largeness and smallness.
17 C. Meinwald, “What Do We Think We’re Doing?,” [“What”], Plato Journal 17 (2017), 9–20,
discusses how much we should assume from Plato’s other writings as we study a particular dialogue.
236   Sandra Peterson

5. Socrates’ Challenge and the

Separateness of Forms

With some self-centered effusions that amuse Parmenides and Zeno (130a) Socrates
challenges Zeno to surprise him. He says:

If he should demonstrate this thing itself, what one is, to be many, or again, the
many to be one, at this I’ll be in wonder. (129b6–c1)
If someone first distinguishes separately (chôris) the forms themselves by them-
selves . . . and then shows that in themselves (en hautois) they can mix together and
be divided apart (diakrinesthai)18 . . . I would indeed admire it wonderfully.”
(129d6–e4)

See also 129b1–3, 129c2–3, and 129e5–a2.

Socrates has these expressions, among others, to indicate forms.

(i) “a form itself by itself of likeness” (homoiotêtos) (128e6–129a1)

(ii) “what unlike is” (ho estin anomoion) (129a2)
(iii) the likes themselves (auta ta homoia); the unlikes (129b1–2)
(iv) the one (129d5)

In (i) the “itself ” is the topic-narrowing device noted previously plus the intensifying,
“by itself.” Here the noun “likeness” specifies the form under consideration. Parmenides
later uses an adjective to specify the form: “a form of just itself by itself ” (130b7–8).
With (ii) Socrates uses the phrase “what unlike is” in parallel to “likeness itself by
itself.” The English translation “what unlike is” is perfectly ordinary speech. We may say,
for example, “If that greyhound and that Pekinese aren’t unlike each other, I don’t know
what unlike is.” The Greek is apparently also ordinary. (Compare at Theaetetus 147a the
question about mud “what it is” (hoti pot’ estin).) The ordinary intelligibility of locutions
of the pattern “what F is” helps us understand the phrases Socrates treats as its equiva-
lents, namely “the F itself ” and “the F,” which sometimes lack clear sense in English.
(I will often remind the reader of that equivalence.) It would be unreasonable to attempt
discussing what Socrates says if we cannot give him phrasing we understand.
The phrase “what F is” reminds us of the question “What is <the> F?” that the Socrates
of other dialogues asks (as at Euthyphro 5d7, “What . . . is the pious?”). The question asks
for an explanatory analysis, a definition. The answer to “What is mud?” is (Theaetetus
147c) “earth mixed with water.”

18 Kahn’s translation (Plato, 3). Gill/Ryan’s Parmenides translates “separate.” Since the Greek verb has
no part resembling the important adverb “separately” (chôris), I prefer translation that does not suggest
connection with chôris.
 Plato’s Parmenides   237

As to (iii), Socrates uses the plural, “the likes themselves” (129b1), as equivalent to the
previous “likeness itself by itself.” We may use a plural similarly by saying: “If those two
aren’t like each other, I don’t know what likes are.”
Socrates’ phrase “the one” at (iv), illustrates a grammatical point: the phrase consist-
ing of neuter article to and neuter adjective is equivalent to an abstract substantive
noun. To homoion, for example, literally “the like,” is equivalent to “likeness” (129a1,
homoiotêtos).
Parmenides takes up Socrates’ word “separately” (chôris at 129d7–8), asking:

Have you distinguished (diê(i)rêsai) . . . forms such as likeness itself separately

(chôris) . . . from the likeness we have? (130b)

Socrates assents, despite not having made such a distinction previously in this con-
versation. Recall that when Socrates introduced the word separately (129d7), he omitted
to say in what way or from what forms are separate.19 For Socrates’ devastating objection
to Zeno the relevant completion to “separate” was clearly “separate from other forms”:
unlikeness, in being other (allo, 129a1) than likeness, is in that way distinguished sepa-
rately from it. In Socrates’ nonstop monologue, moreover, he not only distinguished
opposites such as likeness and unlikeness from each other, he also deemed as separate
opposites from various pairs: “separate (chôris, 129d7) . . . plurality and oneness, rest and
motion and all such” (129e1). When Socrates challenges Zeno to show how all of these
“in themselves” can “mix together” and “be divided apart” (diakrinesthai, 129e3), he
clearly has in mind that forms are separate from one another. The simplest reason to
suppose for Socrates’ distinguishing likeness from rest is that we grasp likeness and rest
as different intelligible contents (“grasped in reasoning” (130a2)).
Socrates’ assent at 130b gives him new commitments that had no role in his objec-
tions to Zeno:

Likeness itself is also separate from things that partake of it.

Likeness itself is separate from the likeness we have.20

Parmenides’s next questions provoke Socrates to doubt that there is a form of human
being “separate from us” (130c1). Socrates resists a form of mud “separate” and “other
than anything we touch with our hands” or a form for anything “totally undignified and
worthless” (130d). But Parmenides immediately discourages such doubts. Presumably
Parmenides notices that Socrates’ implied commitment to a distinct (allo) form for each
separate item grasped in reasoning (129d6–e1; 130a2) commits him to forms of mud and
of human being.

19 See Kahn Plato, 4.

20 Scholars dispute what such separation is. Is it that (1) Forms are merely distinct from
participants?; (2) Forms can exist with no participants ? (Fine, “Separation”; Vlastos, “Separation.”) (3)
Forms exist only when there are participants but not in participants? (Rickless, Forms, 20 note 12)
238   Sandra Peterson

6. Socrates and Parmenides

on Sharing

Parmenides now takes up Socrates’ word share. Socrates had first used the word at 127d,
questioning Zeno:

<Do you not acknowledge> . . . also that you and I and the other things that we call
many share (metalambanein: also “participate”) in these two items that there are
<likeness and unlikeness>? And that things sharing in likeness become like in this
way and to such an extent as they share, and the <things sharing in> unlikeness
unlike, and in both, both?” (128e6–129a6)

There is reason both for and against supposing that the word share here is a technical
apparatus from a theory. On the one hand, “share” has ordinary nontechnical use. For
example, Gorgias 468a says: “sitting, making sea voyages, and sticks sometimes share
what is good and sometimes what is bad.” That means that these items are sometimes
good and sometimes bad. Moreover, Socrates’ first question’s preface, “Do you not
acknowledge?,” indicating humdrum ordinary belief, possibly also introduces Socrates’
next question. If Socrates again invokes ordinary belief, his qualifications “in this way
and to such an extent” merely indicate awareness that the adjectives like and unlike
require a completion such as “in respect of age” or “in respect of name” to be true or false
of anything. On the other hand, Socrates’ qualification “in this way” possibly refers to
something in the immediate vicinity, in which case it could amount to “by sharing.”
Then possibly Socrates implies some explanatory role for sharing, with “share” as
technical vocabulary from an unordinary theory.21
Whether “share” is ordinary or technical when first introduced, Socrates soon
commits himself to unordinary additions when he accepts from Parmenides as sole
options that what shares a form (1) shares the whole of it or (2) shares a part of it. Both
options prove unacceptable (130e–131e): sharing the whole makes the form “separate
from itself ” (131b); sharing a part implies that the form is no longer one (131c).
Plato has previously informed us—though not using Socrates’ verb share to convey
it—that Socrates, Zeno, and Parmenides share in a conversation, that Antiphon and
Glaucon share a mother, and that Antiphon and his grandfather share a name and an
interest in horses. If we accept the question “sharing whole or part?” for these different
cases, we would answer it variously. Conversationalists share the whole by having a part;
people share a mother or a name by each having the whole; one might share either all or
part of an interest. Socrates might have considered that for sharing a form or aspect such
as squareness (=what square is), a natural answer to “whole or part?” is that squares
share what a square is by each satisfying the whole account of what a square is.22

21 Gill Philosophos, 20.

22 C. Meinwald, “How Does Plato’s Exercise Work?” [Exercise], Dialogue 53 (2014),
477 suggests understanding participation as “closely related to what we might call satisfying an
account”; see Meinwald, Plato 266–68.
 Plato’s Parmenides   239

7. New Concession: One over Many

Socrates has made clear that he believes each form is one (131c). He confirms that when
he concedes to Parmenides a premise of an argument called “the Third Man” after the
terminology of Aristotle (e.g., Metaphysics 990b17; 1079a13).23 Scholars call the initial
premise that Parmenides proposes “the one over many” principle:

<O>n this sort of <ground> you think that each form is one: whenever some many
things seem large to you, there perhaps (isôs) seems to be some one form (idea: liter-
ally, “aspect”), the same over <them> all as you see (idonti) them, whence you
believe the large (to mega) to be one (132a2–4).

Parmenides envisages several large visible items (“see”). For definiteness consider a large
greyhound, a large Pekinese, and Mount Everest. Large is what they all are.
Parmenides’s phrase, “the same over all,” conveys that the viewer expects each form,
the F (=what F is), to have one account that provides the same explanation of how each F
item is F. For example, every square is square for the same reason. (Compare Meno 72c8:
the various virtues “all have some one form, the same, by which they are virtues.” See
also Euthyphro 6d: “the form itself by which all the pious things are pious.”)
Socrates agrees here to:

Each form is one. (132a1).

But what he agrees to is unclear because the role of Parmenides’s word, whenever, is
uncertain. Alternatives are:

(i) Whenever many things seem F . . ., there is one form of the F over them.
(ii) There is one form of the F such that whenever many things seem F, . . . it is
over them.

(i) says that a form of F-ness over several F items is countable as one item, as each
greyhound is one greyhound.24 (ii) means that there is exactly one form of the F over any

23 G. Vlastos, “The Third Man Argument in the Parmenides” [“Third Man”], Philosophical Review 63
(1954), 319–49; reprinted in R. E. Allen (ed.), Studies in Plato’s Metaphysics [Studies] (London, 1965),
231–63 inspired continuing response, for example: P. T. Geach, “The Third Man Again” [“Third Man”]
in Allen, Studies, 265–77; C. Strang, “Plato and the Third Man,” Proceedings of the Aristotelian Society 37
(suppl. 1963), 147–64; S. M. Cohen, “The Logic of the Third Man” [“Logic”] Philosophical Review 80
(1971), 448–75; S. Peterson, “A Reasonable Self-Predication Premise for the Third Man Argument”
[“Third Man”], Philosophical Review 82 (1973), 451–70; Penner, Ascent, 251–99 and passim. See also
G. Vlastos, “Plato’s ‘Third Man’ Argument (Parm. 132a1–b2: Text and Logic,” Philosophical Quarterly
(1969), 298–311, reprinted in G. Vlastos, Platonic Studies, 2nd ed. (Princeton, N.J., 1981), 342–60.
24 Gill, “Introduction,” 29–32. Rickless, Forms, 38; D. Bailey, 666–71 of “The Third Man Argument”
[“Third Man”], Philosophy Compass 4 (2009), 666–81.
240   Sandra Peterson

and all F items, for example, a unique form of the large.25 There has been controversy
how to formulate Parmenides’s proposed one-over-many premise.26 But we should per-
haps not expect Parmenides’s questions to Socrates to convey either alternative clearly.
Parmenides is exploring Socrates’ own clarity.
With the phrasing, “on this sort of ground you think,” Parmenides suggests that the
one-over-many gives Socrates’ (main?) reason for believing there is one form over sev-
eral large items. But that suggestion is unexpected: Socrates has not previously cited
observing several F items as a reason for his belief in the form, the F (=what F is). Recall
his resistance to forms of human being and mud. Rather, Socrates’ reason was that talk
of opposites requires forms (128e6-129a2) and that separate topics of reasoning are
forms (129d7–e1; 130a2). Perhaps Parmenides overheard a one-over-many premise in
Socrates’ earlier conversation about forms with young Aristoteles (135c9–d2)?27
Whether “one” here conveys “one form of the F over each of some several F’s” or
“exactly one form of the F over all F’s,” the reader now wonders why Socrates forecasted
his surprise should anyone show that the many (=what many is) is one (129b): he already
believes that each form is one somehow.

8. Self-Predication

Parmenides asks Socrates:

What about the large itself and the other large things? If you look over them all in
your mind in the same way will not some one large <item> appear by which all these
necessarily seem large? (132a6–8)

“In your mind” signals the sort of seeing appropriate for forms. “By which” means
“because of which” or “explaining why.” It clarifies “over.” (Note that our initial several
large items all seem F for many features F. For example, they are all sublunary, separate,
and small in comparison with the universe. Many forms over them variously explain
their many shared features. Parmenides mentions only a form that explains their
largeness.)
“All. . . appear large” and “the large itself and the other large things” (132a) commit
Socrates to this:

The large itself is large.

25 Rickless Forms 38, 65–66 thinks 132a1 does not claim uniqueness. See later on in this chapter.
26 See for example, Cohen, “Third Man,” and Fine, Ideas, 204–11.
27 Forms are a popular topic still at the time of Aristotle’s Topics, which gives advice for questioning
those who undertake as answerers to defend a statement about forms (hoi tithemenoi ideai: 113a24–32;
137b3–13; 143b25–32; 147a5–11; 148a14–22).
 Plato’s Parmenides   241

Scholars label such statements “self-predications.”28 I use the label to indicate a pattern
of sentence: its predicate-expression derives from its subject-expression. So understood,
the label leaves truth conditions undetermined.29 However, some scholars assume truth
conditions with the label.30
Claims of this syntax that have a form-expression as subject and a predicate-expression
applicable to ordinary things seem strange if one thinks of forms as unfamiliar technical
inventions belonging to a special theory of Plato’s found, for example, in the Phaedrus
(247c–e). Such forms are “outside the heavens” and accessible only to a special few.31 For
such items, it seems strange if the impious itself is impious, the mortal itself is mortal,
and mud itself is mud.
A first step toward recognizing that sentences whose mundane predicates are attached
to form-expressions as subjects are not strange is to reflect that many familiar and
acceptable claims share that syntax, for example, “Charity suffereth long,” from the
Bible, or “The tiger is a carnivore” from our encyclopedias.32
A second step toward seeing the plausibility of self-predications is to remember that
the different expressions Socrates uses to speak of forms allow several equivalent formu-
lations for “The F itself is F.” The equivalent formulation, “What F is is F,” sounds not only
reasonable, but inevitable. Consider the form, the square itself (= what square is, or what
a square is). What <a> square is, as specified in the analysis or definition of the square, is:
<an> equal-sided rectangle. Since what <a> square is = the square itself, then the square
itself is an equal-sided rectangle. So of course the square is square. A similar pattern of
argument works for any form the F (=the F itself, =what F is).33
Parmenides proposes:

Therefore (ara) another (allo) form of largeness (eidos megethous) will appear,
having come to be alongside the large itself and the things sharing it; and another
(heteron) again over these, by which all of these will be large; and no longer will each
one of the forms be one for you, but unlimited (apeira) in multitude. (132a10–b2)

“Another . . .and another” clearly begins an infinite regress. Most interpreters take the
infinite regress, or its beginning, the second form, as the sole consequence Socrates
shuns.34 That second form conflicts with his conviction that each form is one

28 See Vlastos, “Third Man,” 236; J. Malcolm, Plato on the Self-Predication of Forms (Oxford, 1991);
D. Apolloni, The Self-Predication Assumption in Plato (Lanham, Md., 2011).
29 See Peterson, “Third Man”; Penner, Ascent, 291–99.
30 For example, D. Bailey, “Third Man” advocates (668) the new label “self-instantiation,” explaining:
“any Form itself has the very property it causes in its sensible particulars.”
31 Though the Phaedrus passage lacks the words eidos or idea, scholars identify the items of 247c2 as
technical Forms. (C. J. Rowe, Plato. Phaedrus (Oxford, 1986) 179 ad loc.)
32 See Geach, “Third Man,” 267–71; Peterson, “Third Man,” 455–59.
33 Meinwald, “Third Man,” 374–75 argues differently for the reasonableness of the pattern “The F
itself is F”; Rickless, Forms, 32–36 deduces it from axioms of the theory he thinks Socrates holds; Fine,
Ideas, 52–53 and 62–64 broadens the applicability of the predicate. See also Sedley, “Introduction,” 13.
34 For different views see Sayre, Parmenides, 81–82; Rickless, Forms, 68–75; Bailey, “Third Man,” 676–77.
242   Sandra Peterson

­(131b5–6; 131e11).35 Socrates immediately makes amendments in order to retain that

conviction (132b3–4).

9. A Missing Premise. Review of the

Third Man Argument

Socrates does not notice that despite Parmenides’s “therefore,” the objectionable second
form of largeness does not follow from the one-over-many principle and self-predica-
tion. Readers that think Plato’s Socrates would not assent to an unwarranted inference
nor would Parmenides as depicted have encouraged an invalid inference, supply an
implicit premise to warrant the inference. A possible supplement that would enable the
inference is:

Nothing is large because of itself.

I label this or its generalization, “Nothing is F because of itself,” the “non-self-explanation”

claim.

I represent the third man argument, in brief and general form, thus:36
Several (visible) items seem F to you. (fact)
Whenever several items seem F to you, you think there is one form,
the F itself, because of which they are F. (one over many)
The F itself is F. (self-predication)
Several F items plus the F itself over them seem F to you. (fact)
Nothing is F because of itself. (supplied non-self-explanation)
<You think that> there is a second form of the F over the several F items
plus the F itself over them. (by one over many)

Our supplied premise is, however, faulty, as the author Plato presumably recognizes.37
To see its fault, consider the square itself (=what a square is). The square itself is square
precisely because of itself, that is, because of what a square is—namely an equal-sided
rectangle. Socrates’ youthful confusion by this point in the conversation is a reason to
supply the premise, despite its flaw, to explain his accepting the inference.

35 Cohen, “Logic,” 472–73, says Socrates should protest: “Another the Large Itself? . . .That doesn’t
make any sense! . . . ? How could it be different from the first?” Cohen does not here explain his protest,
but suggests that uniqueness is “built into” Plato’s way of referring to Forms.
36 Scholars represent it variously. Vlastos, “Third Man,” supplying what he considered the simplest
premise, gave the argument contradictory premises. See for repairs and some other differences Geach
(“Third Man”), Cohen (“Third Man”), Peterson (“Third Man”), Fine (Ideas 205–11).
37 Kahn Plato, 11–12 deems non-self-explanation “an unPlatonic premise.” Gill, Philosophos, 37–38
explains why young Socrates might accept it.
 Plato’s Parmenides   243

10. Possible Responses to

the Third Man Argument

The Third Man Argument has captured commentators’ attention. Given one or another
valid and interesting formulation of the argument, scholars have then asked whether
Plato’s own views of forms as presented in other dialogues can avoid the Third Man
objection.38
The logically available responses to the argument are to reject one or more of its prem-
ises or to accept its conclusion. One might reject non-self-explanation.39 One might
reject self-predication.40 S. Marc Cohen urges rethinking the one-over-many premise:

What <the Third Man> shows is that to keep uniqueness, the one-over-many
principle will have to be abandoned or modified, for it is an application of that
principle to the set consisting of large particulars and the Large itself that generates
a second form.41

Finally, one might respond to the argument by accepting the conclusion that the form of
largeness (or any form of F-ness) is not unique.42
I revisit the Third Man later in this chapter.

11. Parmenides’s Diagnosis

of Socrates’ Mistake.
The “Greatest” Difficulty

I skip over two arguments, known in the literature as the “forms as thoughts” argument
and the “likeness regress,” with which Parmenides confronts Socrates.43 However, in
carrying out my next project I discuss the final difficulty that Parmenides presents,
which he calls “the greatest” (133b4).

38 Fine, Ideas, ch. 16: “Is Plato Vulnerable to the Third Man Argument?”
39 See earlier in this chapter; Fine, Ideas, 231–38. 40 For references see Meinwald, Plato, 273.
41 Cohen, “Third Man,” 473. Fine, Ideas, 204–11.
42 Rickless, Forms, 240. See the Appendix of this chapter.
43 On forms as thoughts see Gill, “Introduction,” 38–42; Rickless Forms 75–80; and 20–23 of
M. Burnyeat, “What Descartes Saw and Berkeley Missed,” Philosophical Review 91 (1982), 3–40. On the
likeness regress see M. Schofield, “Likeness and Likenesses in the Parmenides,” in C. Gill and
M. McCabe (eds.), Form and Argument in Late Plato (Oxford, 1996), 49–77. See also H. Peacock, “The
Third Man and the Coherence of the Parmenides” [“Third Man”], Oxford Studies in Ancient Philosophy
52 (2017), 113–76.
244   Sandra Peterson

My next project is twofold: first, to identify Parmenides’s diagnosis of what produces

Socrates’ difficulties and second, to consider the benefits Parmenides claims the recom-
mended exercise will have for Socrates.
In diagnosis Parmenides speaks of Socrates’ “marking off ” forms. Having presented
several difficulties, Parmenides asks:

“Do you see, Socrates, how great is the impasse (aporia: also “lack of resources”;
less literally, “difficulty”) if someone marks off thoroughly (diorizêtai) beings as
forms themselves by themselves?” (133a8-9)

“Mark off ” (horizô, in “mark off thoroughly”) means literally, “put a boundary around.”44
Parmenides predicts worse from marking off, using another derivative of horizô:

<Y>ou almost do not yet touch on how great the aporia is if, sharply marking off
something (aphorizomenos), you will always posit one form for each of the things
that are. (133a11–b2)45

Socrates has not previously used a relative of the verb “mark off ” (horizein) for marking
off forms. (His sole previous use, (diorisasthai, 131e7) meant “figure out.”) Parmenides’s
varied references to marking off apparently allude to Socrates’ previous talk of forms
being distinguished ((diairêtai) separately (chôris) themselves by themselves (auta
kath’hauta) (129d7–8).46
Parmenides then begins his “greatest” (133b4) difficulty by getting Socrates’ agree-
ment to these two proposals:

As many of the forms as are what they are (133c10, eisin hai eisin) in relation to one
another, these have being (133c10, ousian) in relation to themselves (pros hautas)
(133c10), but not in relation to the things alongside us (pros ta par’hêmin: 133d1).
Those things alongside us in turn . . .are <what they are> in relation to themselves
but not in relation to the forms. (133d2–4)

Parmenides’s examples are: “Mastery itself is what it is of slavery itself ” (133e3) and
“The particular master is master of the particular slave” (133d–e3).
Socrates does not notice that Parmenides’s two proposals and his examples at 133d
are ambiguous. The source of the ambiguity is the phrase “what they are.” The phrase
can be a term of art, such as the phrase “what it is” that indicates the analyzing answer to
the question “What is it?” But the phrases “what they are” and “what it is” can also make
ordinary attributions, as in “Large is what they all are.” Thus, “Mastery itself is what it is
in relation to slavery itself ” has these two meanings:

44 “Thoroughly” captures the intensifying prefix dia-.

45 “Sharply” captures the ap- that prefixes the recurrent root horizô.
46 The verb “mark off ” (horizô) enough resembles an unrelated verb chôrizô (“separate”), which is
genuinely related to chôris (“separately”) to make us suspect Parmenides of wordplay.
 Plato’s Parmenides   245

(i) The master itself is to be defined by mention of the slave itself

(ii) The master itself is master of the slave itself.

The former claim is true and the latter, as an ordinary attribution, is false.
Similarly, “The particular master is what he is in relation to the particular slave,” has
these two meanings:

(i) The particular master is what he is (i.e., is to be defined) in relation to the

­particular slave.
(ii) The particular master is what he is, that is, a master, of the particular slave.

The latter is true. The former is false because there is no definition or explanatory
analysis of what this particular master is. A definition completely marks off, or distin-
guishes from all else, the narrowly focused-on item the F itself (=what F is). The latter’s
account explains why each of several F things is F. The particular master is not such as to
have an account that completely defines him, and that explains why each of many
masters is a master.47
Missing the ambiguity, Socrates mistakenly grants that Parmenides’s premises imply
that we do not have knowledge itself, and hence that none of the forms is known by us
(134b9–12). Socrates then grants that “Knowledge—a kind itself—is much more precise
than the knowledge that belongs to us” (134c), and that the gods have the most precise
knowledge and most precise mastery (134c11). The gods’ knowledge therefore cannot be
of (frequently messy) human affairs nor can their mastery be over us (134d9–e7), results
“yet more dreadful” (134c3).48 Again Socrates is mistaken: it is not the case that knowledge
itself—what knowledge is, unqualified by any further description—is precise, any more
than mud itself is much stickier than any mud we encounter.
Parmenides reiterates that marking off creates such difficulties:

“These things, Socrates, . . .and yet altogether many others in addition to these it is
necessary that the forms have if there are forms themselves (hautai hai ideai) and one
will mark off (horieitai, from horizô) each form as something itself.” (134e10–135a3)

Finally Parmenides explicitly diagnoses Socrates’ marking off “too soon” as a fault:

“Socrates, you are trying to mark off for yourself (horizesthai—from horizô) something
fine and just and good and each of the forms too soon, before getting exercised
(gumnasthênai).” (135c9–d1)

47 Meinwald, “Exercise,” 487: sensible items do not answer the “What is it?” question.
48 See more later on in the chapter on the “greatest” difficulty. See also S. Peterson, “The Greatest
Difficulty for Plato’s Theory of Forms,” Archiv für Geschichte der Philosophie 63 (1981), 3–16; B.-U. Yi and
E. Bae, “The Problem of Knowing the Forms in Plato’s Parmenides,” History of Philosophy Quarterly 15
(1998), 271–83; M. Duncombe, “The Greatest Difficulty at Parmenides 133c–134e and Plato’s Relative
Terms” [“Greatest”], Oxford Studies in Ancient Philosophy 45 (2013), 43–61.
246   Sandra Peterson

“Too soon” and “before getting exercised” strongly suggest that exercise will help
Socrates to mark off forms. Parmenides stresses that marking off is crucial:

If someone will not allow forms to be among beings, and will not mark off (­ horieitai)
a form of each <item that is> one, he will not, wherever he will turn, get hold of <a>
thought, not allowing that there is a form among beings of each <item marked off>,
always the same, and thus he will altogether destroy the possibility for dialegesthai.
(135b5–c3)

12. Possible Benefits from

the Exercise

Urging Socrates to exercise (135d), Parmenides suggests that without exercise, certain
good results will not ensue. He does not guarantee them.

(i) The exerciser will be able to mark off forms not too soon (135c–d1).
(ii) He will discern (diopsesthai: thoroughly see) the true (to alêthes) strictly (kuriôs)
(136c5–6).
(iii) It will be possible for him, when he hits upon the truth, to have understanding
(noun schein) (136e1–3).

“Strictly” (kuriôs, 136c6) in item (ii) deserves comment. Other possible translations
are “in the chief way” or “authoritatively.” The adverb might modify either the verb see or
the adjective true. I understand “to discern the strictly true.” I take it that Parmenides
implies that exercise enables strictly accurate speech.

13. Pre-exercise Review

In discussing Parmenides’s exercise, I will keep in mind some earlier points.

1. Socrates did not consider that several F items might share the whole of the F
when each satisfies the whole account of what F is.
2. The one-over-many premise of the Third Man was unclear and unexpected.
3. Socrates’ apparent implicit acceptance of non-self-explanation was mistaken.
4. Socrates did not see an ambiguity in the claim that a form is what it is only in
­relation to another form.
5. Parmenides affirmed Socrates’ insight that dialegesthai requires marking off forms.
 Plato’s Parmenides   247

14. Parmenides’s Directions

for Exercise

Parmenides advocates an exercise that resembles Zeno’s arguments in considering

“wandering” (planên, 135e2–4). “Wandering” indicates that a result such as “the subject
is F” will be matched by a result such as “the subject is the opposite of F” or “The subject
is not F.” But Parmenides requires that in Socrates’ exercise the subject that wanders will
be “among those things that one might most grasp in speech and might consider to be
forms” (135e2–4).
Parmenides makes a two-part stipulation. Socrates must “not only consider the con-
sequences from the hypothesis that each <item> hypothesized is”; he must also “hypoth-
esize if that same thing is not” (135e8–136a2). Socrates is to do multiple rounds (“each
<item> hypothesized”) of exercise. Each round considers consequences of a hypothesis
concerning a form or forms (135e1–3) and then also considers consequences of its
negation.
Parmenides makes a second two-part stipulation: the exercise must consider “what
on each hypothesis will follow for the very things hypothesized and for the others”
(135b1–4). For example, an exercise on likeness will consider not only likeness but also
forms other than likeness.
Parmenides makes a third two-part stipulation. I italicize some recurrent phrases:

<You must consider them> in relation to themselves and in relation to each other.
(136b3–4)

He gives several examples, each of which includes slight variants of these phrases “in
relation to (pros) themselves” and “in relation to (pros) each other.” Parmenides clearly
draws attention to these phrases.49 Parmenides repeats the new phrases, with slight
­variations, in his summary:

And in a word, concerning whatever you might hypothesize as being or as not being
or as suffering any other affection, you must examine the consequences in relation
to itself and in relation to each one of the others, . . . and in relation to several and in
relation to all in the same way, and then you must examine concerning the others
in relation to themselves and in relation to an other—whichever you select; and <you
must do all this> whether you hypothesize <it> as being or whether as not being.
(136b6–c4)

49 Meinwald, Parmenides, 177 note 1 commends K. Sayre, “Plato’s Parmenides: Why the Eight
Hypotheses Are Not Contradictory,” Phronesis 23 (1978), 133–50 for early observation that these phrases
generate sections of argument. See also K. Sayre, “The Method Revisited” [“Revisited”] in A. Havlicek
and F. Karfik, Plato’s “Parmenides.” Proceedings of the Fourth Symposium Platonicum Pragense.
(Prague, 2005), 125–40, section 3.
248   Sandra Peterson

Though Parmenides makes these pros phrases conspicuous by repetition in his

directions for exercise, he does not use them systematically within the exercise as sign-
posts nor does he ever explain them. He does, however, repeat a version of them in his
midway summary (160b2–3) and his concluding summary of results (166e5). The reader
must decide what guidance the phrases give.

15. The Hypothesized Item of

Parmenides’s Sample

Parmenides reluctantly (136e8–137a8) enacts a sample exercise. He offers to begin from

“my own hypothesis” (137b3), “hypothesizing about the one itself ” (137b3–4). One might
object that Parmenides reports his “own hypothesis” inaccurately. Earlier Socrates (with
Zeno’s approval at128e3–4) summarized Parmenides’s claim as, “All is one” (128b1–2).
Socrates also deemed Parmenides’s view pretty much the same as Zeno’s view, “there are
not many <things>”(128b4–6).
But Parmenides is accurate. He has agreed (135b5–c1) to Socrates’ observation that to
speak of items being like and unlike presupposes the forms likeness and unlikeness
(128e6–129a3). So Parmenides realizes that to say that all is one—as opposed to many—
similarly presupposes that there are the forms what one is and what many is. His own
hypothesis was indeed about the one itself.

16. The Look of the Exercise

In Parmenides’s sample exercise four clearly announced sections (hereafter, as in Gill

and in Rickless, “Deductions”)50 first derive results from the hypothesis (hereafter H)
that the form, the one (=what one is) is.51 Four more clearly announced Deductions get
results from the negation (hereafter not-H) of the hypothesis. Under the hypothesis
H Deduction 1 produces negative consequences of the pattern “The one is not F.”

50 I do not count separately the end of Deduction 2. It has some appearance of being a new (hence
ninth) unit. See Meinwald, Parmenides, ch. 6; Peterson, “Game,” 47–48; Gill, Philosophos, 48–49; 64–65.
51 Parmenides’s phrasing of the hypothesis varies. Deduction 1 begins from H formulated as hen
estin (137c4). A possible translation would be “it is one,” which takes the lone adjective hen as predicate.
A lone adjective may, however, serve as a subject-expression rather than a predicate. In Plato see
Republic 476a1 (“beautiful is opposite to ugly”). See Phaedo 74a9–12: “we say equal (ison) is
something . . . the equal itself (auto to ison).” In English we have, “Well-begun is half done” and Ogden
Nash’s “Too clever is dumb.” For translation and textual issues see Meinwald, Parmenides, 39–45 and
Gill, Parmenides, 65–66 and “Design,” 509. Gill argues for the translation, “it is one.”
 Plato’s Parmenides   249

Deduction 2 produces positive consequences (e.g., “The one is G”) from H. (By
­“consequence,” I mean the consequent of each conditional “if H, then C” or “if not-H,
then R.”) Deductions 3 and 4 from H likewise segregate positive and negative conse-
quences for the others. Deductions 5, 6, 7, and 8 under not-H likewise segregate positive
and negative results. The Deductions from H follow this plan:

Deduction 1 (from 137c): If the one is, then the one is not many, not a whole, . . .
Deduction 2 (from 142b): if the one is, then the one is unlimited, like, unlike, . . .
Deduction 3 (from 157b): If the one is, then the others are . . .
Deduction 4 (from 159b): If the one is, then the others are not . . .

The Deductions from not-H look like this:

Deduction 5 (from 160b): If the one is not, then the one is . . .

Deduction 6 (from 163b): If the one is not, then the one is not . . .
Deduction 7 (from 164b): If the one is not, then the others are . . .
Deduction 8 (from 165e): If the one is not, then the others are not . . .

Results between Deductions appear to contradict one another. For example, Deduction
1 yields that the one is not many (137c). Deduction 2 yields that the one is unlimited in
multitude (145a). Deduction 5 yields that there is knowledge of the one (160d3–5).
Deduction 6 yields that there is not knowledge of the one (164b1–3). Results within
Deductions appear at least to oppose one another. For example, Deduction 2 proves that
the one is in motion and at rest (146a), and that the one is the same as itself and different
from itself (146a–b).
The exercise appears to get conflicting results both from H and from not-H, thus
reducing both to absurdity and leaving us with the contradiction that both not-H and
not-not-H (i.e., H).
Scholars that I lack space to discuss have various reactions to that appearance:
(1) both hypotheses of the exercise are ill-formed nonsense;52 (2) newly introduced
conflicting premises within the various Deductions yield the contradictions;53
(3) the exercise is a parody;54 (4) the exercise shows problems for Eleatic monism, the
metaphysical view that Plato views as his main rival.55 Differently, it has been proposed
that (e) Parmenides wants us to embrace a contradiction.56

52 G. Ryle, 132 of “Plato’s Parmenides,” reprinted in Allen, Studies, 97–147.

53 G.E.L. Owen, 98 of “Notes on Ryle’s Plato,” reprinted in G. E. L. Owen, Logic, Science, and
Dialectic (Ithaca, N.Y., 1986), 85–103.
54 A. E. Taylor, ch. 14 of Plato (New York, 1952). See Gill, Philosophos 50, n.11.
55 Peacock, “Third Man.”
56 Priest, “Dialethic,” 2: “Plato is indeed suggesting that the one has contradictory properties.” For
yet other reactions see the survey in Sayre, Parmenides, ix–xx.
250   Sandra Peterson

17. Meinwald’s Account

I take as a basis for my discussion Meinwald’s account of the exercise. It seems to me, for
reasons that will emerge, the most successful of the several worthwhile accounts that
scholars offer.
Parmenides’s three twofold stipulations led us to expect eight (2 × 2 × 2) sections of
exercise, but they did not obviously predict segregation of positive from negative conse-
quences. Meinwald proposes that the emphasized phrase “with reference to itself ” (pros
heauto) governs the sections with negative consequences in that we should understand
that each item in an assemblage of negative consequences bears an implicit pros heauto;
the emphasized phrase “with reference to the others” (pros ta alla) attaches implicitly to
each result of a positive section.57 Meinwald thus explains how segregation of positive
from negative results ensues from Parmenides’s directions.58
On Meinwald’s interpretation, when “The F is G” is a statement pros heauto, it
articulates what it is to be F: the statement contributes to the analysis that completely
and permanently defines what F is.59 For example, the pros heauto truth, “The square is
an equal-sided rectangle,” completely defines the form, distinguishing it from all others.
Pros heauto statements “articulate the structure of fundamental reality” (“Exercise,” 467).
(Meinwald speaks of “tree predication” (“Exercise,” 467). She represents relations
between a form and the constituents of its definitions in a tree diagram leading from a
genus to its species.)60 Deduction 1’s result, “The one is not many” (137d), meaning on
Meinwald’s account, “It is not the case that pros heauto the one is many,” tells us that the
analysis of what one is does not involve what many is. “The one is not the same as itself ”
(139d1), to be understood as “It is not the case that the one is pros heauto the same as
itself,” says that oneness is not to be analyzed via sameness with itself. 61
The phrase pros ta alla, “with reference to the others,” according to Meinwald implicitly
attached to each result in Deductions with positive consequences, indicates a truth that

57 Meinwald, Parmenides, chs. 3–8; Meinwald, “Exercise,” 465–81. Scolnicov, Parmenides, 27 takes
the phrases to govern the positive and negative deductions, but understands the phrases differently
from Meinwald. In contrast, Rickless e.g. Parmenides, 15 and M. L Gill, “Design of the Exercise in
Plato’s Parmenides,” [“Design”], Dialogue 53 (2014), 495–520, 499, and 506, do not believe that each
Deduction is governed by one of the pros phrases. They take Parmenides’s final principle of division
that gives us the eight main Deductions to be to segregate positive consequences from negative.
58 Meinwald (“Exercise,” 474) observes that Rickless’s and Gill’s view that one or another pros phrase
attaches, in its ordinary sense, to each result in each Deduction implies that “Parmenides’ description
of the exercise represents it as a study only of relational (two-place) predicates.” That implication conflicts
with the exercise’s consideration of one-place predicates also (e.g. 145b, “has shape”; 159b, “all things”).
59 Meinwald, “Exercise,” 478.
60 Scolnicov, Parmenides 22 n.86, objects to the word “predication” as “too Aristotelian.” But we can
understand it mildly as “asserted connection.”
61 Gill, Philosophos, 63–63 implying that being is “inside the nature” of the one, perhaps
distinguishes nature from what appears in a definition. On our ordinary notion of definition being
would not be mentioned in the definition of the one any more than in the definition of mud or of what
a square is.
 Plato’s Parmenides   251

does not articulate analysis or definition.62 Pros ta alla statements concern “displays of
features” and “The subjects are instantiating or exemplifying something expressed by
the predicate term” (Meinwald “Exercise,” 467). The fact that the one satisfies the defini-
tion of sameness is the ground for the truth of the pros ta alla statement about display or
instantiation, “The one is the same as itself ” (146c). The statement is a pros ta alla (“with
reference to others”) claim in that another (allo) form, sameness, explains how the one is
the same.
Meinwald’s distinction between pros heauto and pros ta alla sections at one stroke
removes the appearance that results of positive sections contradict results of negative
sections.63
Meinwald dispels the appearance of intra-Deduction conflicts such as within the pros
ta alla Deduction 2 by supplying appropriate qualifications.64 For example, the one is
both like in one way and unlike in another way (148a–d). The qualifications needed are
sometimes perfectly obvious and sometimes not.65
Meinwald’s account thus absolves Parmenides of the unlikely intent to argue for
contradictions.

18. The Third Man Argument Again

Meinwald argues as follows that her disambiguation of self-predications via her

­distinction of types of predication disables the Third Man Argument that Parmenides
and Socrates envisage in Part I of the Parmenides. The first several large things that
Parmenides asks Socrates to consider are several visible large things (132a3, “sees”). All
these items would be large pros ta alla, since pros ta alla predications are the only
predications true of visible items.66 The large itself is large in a way (pros heauto) that is
different from the (pros ta alla) way in which the many large visible items are large. So
according to Meinwald the several large visible items plus the large itself (Parmenides,
155–156) require no form to explain how they are all large.67

62 See Meinwald, “Exercise,” 482 on possible misunderstandings of “definitional.”

63 An important but I believe incorrect objection to Meinwald’s account is that the phrases pros
heauto and pros ta alla, which already have ordinary uses to indicate any relation to self or any relation
to others, sometimes occur in the exercise clearly with their ordinary meaning. See M. Miller, Plato’s
“Parmenides” (Princeton, N.J., 1986), 226–27; Sayre, Lesson, xviii with its note 48, and chapter 1; Rickless
Forms, 102–06; Gill “Design” pp. 504–05 with note 16 p. 504 with references to others who make the
same objection. Not granting the continuing availability of the ordinary uses, Rickless Parmenides, 16,
objects that Parmenides should not have consequences about the one in relation to itself in Deduction 2,
a pros ta alla section. Against such objection, see Peterson, “Game,” 44–47.
64 See Meinwald, “Exercise,” 485–88 responding to Gill (references in Meinwald’s footnote 47 p. 485).
65 Peterson, “Game,” 48–50.
66 Meinwald, “Exercise,” 487: displaying is all sensibles can do.
67 Meinwald, Parmenides, 156 (“not . . . a new group of large things”); Plato 269–70 (“no new group”).
252   Sandra Peterson

It is important to distinguish (1) Meinwald’s explanation why the one-over-many

requires no new form to explain why the several visible large things plus the large itself
are all large from (2) the reason the one over many requires no form of the seal over
them all to explain why the seals in our local zoo and the Great Seal of the United States
are all seals. The one-over-many clearly does not aim to generate a single form for a
genuinely ambiguous predicate such as “seal.” Since the word seal is ambiguous, there is
no one thing that is what a seal is. In contrast, the predicate “large” in the second step of
its Third Man Argument is not ambiguous. There is just one explanation of what large is.
Rather, according to Meinwald, it is the whole sentence, “The large is large” that has two
readings. It can convey two types of predications that have two different types of truth
conditions (Meinwald, Plato, 265; Meinwald, “Exercise,” 467).68
Our formulation of the Third Man above does in fact, differently from the version
Meinwald assumes, require a form over a collection of items one of which is F pros heauto
while the others are F pros ta alla. However, Plato can and presumably would block the
inference to multiple forms of the F by rejecting the faulty non-self-explanation premise.
Even though Meinwald’s distinction is not needed to block the inference for our
formulation of the Third Man, her distinction has considerable relevance for our formu-
lation. Her distinction illuminates self-explanation. It draws to our attention that there
are two ways in which the F itself may be F because of itself. That is, there are two ways in
which the account of what F is may explain how something is F. First, the explanation of
any pros heauto self-predication, “The F itself is F” is that the account of what F is analyzes,
that is, completely defines, the F itself. And second, for a case such as sameness itself, for
which additionally and untypically the pros ta alla self-predication “Sameness is the
same” is true, the form satisfies its account as other items do. (Sameness is of course not
actually other (allo) than itself. But the phrase pros ta alla is a tag for attributions of
instantiation or display because such attributions (“Socrates is pale,” “Temperance is
difficult”) typically involve a relation to others (alla) than the displayer.)
Moreover, we can now see that rejecting non-self-explanation disables two varieties
of the Third Man Argument that Parmenides and Socrates do not consider.69 One
variety involves an initial several items that are pros ta alla F and a covering form that is
pros ta alla F (as well as, of course, pros heauto F): for example, sameness, as well as many
visible items, is the same as itself. A second variety of regress arises for pluralities all of
whose members are F pros heauto. For example, the several virtues, justice, piety, and
courage require the form over them, virtue itself.70 The three virtues plus virtue itself all

68 Meinwald’s view is also distinct from the proposal that Plato holds that the predicate, for example,
“just”, though not ambiguous, has a much wider range of application than we might expect, and can be
true of the form. See, e.g., Fine, Ideas, 62–63 for the latter proposal and Rickless, Forms, 34–35.
69 B. Frances, “Plato’s Response to the Third Man Argument in the Paradoxical Exercise of the
Parmenides,” Ancient Philosophy 16 (1996), 47–64 studies types of third man arguments Meinwald
Parmenides omits to discuss. Rickless Parmenides, 25 views the omission as a serious difficulty for
Meinwald; Frances views the gap as minor and reparable. J. Pelletier and E. Zalta, 169–70 of “How to
Say Goodbye to the Third Man” [“Third Man”], Nous 34 (2000). 115–202 make repairs.
70 Pelletier and Zalta, “Third Man,” distinguish a second version of the One-over-Many premise for
groups all of whose members are F pros heauto (172–73).
 Plato’s Parmenides   253

are virtues pros heauto. But once we give up the non-self-explanation premise we require
no new form to explain either why virtue itself is virtuous (pros heauto) or why sameness
is the same (pros ta alla). Each form the F is F because of itself, whether (as always) it is F
pros heauto or whether (as untypically) it is also F pros ta alla.71
Because Meinwald attributes to Plato a theory of forms “still alive” beyond the
Parmenides (Plato, 274), it is useful to see how these versions of the Third Man that do
not come up in the Parmenides fare given her distinction.

19. Some Merits of

Meinwald’s Account

So far, Meinwald’s account verifies self-predication. It clarifies self-explanation and the

varieties of the Third Man Argument. It takes account of the conspicuous recurrence of
the pros phrases in Parmenides’s directions. It accounts for the segregation of positive
and negative results in Parmenides’s sample exercise. It removes the appearance that
Parmenides’s exercise argues for a contradiction.
It also deals with the “greatest” difficulty. We noticed already that Socrates and
Parmenides discuss it without distinguishing the two meanings of “X is what it is in
relation to Y.” “The master X (form or particular among us) is master in relation to Y”
(suggested by 133d7–134a2) can mean: (i) X (as master) is analyzed as master in relation
to Y, or (ii) X (as master) instantiates or displays mastery in relation to Y. (i) corresponds
to Meinwald’s pros heauto predications. (ii) corresponds to her pros ta alla predications.
For forms, only statements with the meaning (i) are true (strictly true); for sensible
objects, only statements with the meaning (ii) are (strictly) true.72
A further merit of Meinwald’s account is that it explains how repeated exercise on any
hypothesized form as subject yields the benefits Parmenides predicts. A negative pros
heauto Deduction 1, for example, correctly marks off (135c9–d1) the form hypothesized
from other forms that do not figure in its definition. Moreover, the distinction between
pros heauto claims and pros ta alla claims is a step toward discerning the strictly or pre-
cisely true (136c5–6). It is strictly true, for example, that the one is marked off pros heauto
from the same, but also strictly true that it can be combined with sameness pros ta alla.
Further, the Deductions enhance our understanding: the huge network of demarcations
(pros heauto) and connections (pros ta alla) among forms73 shows the “wandering”
without which it is impossible to understand a truth that one encounters (136e1–3).

71 Meinwald, Plato, 270–71. Pelletier and Zalta, “Third Man” 174 and 181 distinguish two versions of
non-self-explanation and find both false.
72 Meinwald “Exercise,” 467: a pros heauto statement that A is B is not true unless A is a form.
73 See Peterson, “Game,” 28–31 and Peterson, “New Rounds,” 271–72.
254   Sandra Peterson

It might seem a defect of Meinwald’s account that Parmenides’s sample exercise yields
no positive pros heauto statement, no analysis of its subject, the one.74 Meinwald’s expla-
nation is that Parmenides’s hypothesized subject, the one, is so foundational that other
forms cannot analyze it.75 The entirely negative results of Deduction 1 are an effect of the
chosen form hypothesized (Meinwald, “Exercise,” 485). In contrast, Deduction 1 of an
exercise on the hypothesis that the square itself is might arrive at the pros heauto positive
truth that the square is an equal-sided rectangle. But it might also have many of the same
negative results as Parmenides’s exercise on the one: it is not the case that pros heauto the
square is many, or a whole, or the same, or different, . . .
On Meinwald’s account the exercise does in fact give us some extremely important
information about the one itself (=what one is). For example, Deduction 3 asks us
in thought to take away from each of the others their oneness (158b1–c6). Doing so,
“we find them totally undone and no longer capable of being identified as individuals
at all.”76 This result shows the great importance of the one in comparison to most other
forms. Meinwald, “Exercise,” 485: “Consider how little it affects the others than the cat if
we consider them on their own.” We lose catnaps and tomcats and the like, but there is
no cosmic obliteration.
An important dimension I lack space here to consider is Meinwald’s claim that the
arguments Parmenides gives in the exercise are all valid.77

20. The Parmenides and

a Theory of Forms

Scholars question whether the difficulties that Parmenides poses for the adolescent
Socrates’ admissions about forms seriously damage what Plato’s other dialogues say
about forms.78 Scholars give various answers, for example: Plato thinks none of the
objections were worthwhile as objections to his views about forms;79 the objections rec-
ord Plato’s “honest perplexity” about forms;80 Plato is admirably self-critical in revealing
difficulties for his earlier views.81

74 Gill, “Design,” 507 finds that the entirely negative results of Meinwald’s pros heauto deductions are
“scarcely productive.”
75 Meinwald, “Exercise,” 483–85.
76 Meinwald, Parmenides, 133, giving a condensed summary of pages 133–39.
77 Meinwald Parmenides, 4, 26–27. Meinwald, Parmenides analyzes a few arguments convincingly.
Sayre, Parmenides and Rickless, Forms, treat all, in distinctive ways often congenial to Meinwald’s
overall account. Peterson, “Seven Arguments,” analyzes seven.
78 For example, Fine, Ideas, ch. 16 asks, “Is Plato vulnerable to the Third Man Argument?”
79 P. Shorey, The Unity of Plato’s Thought (New York, 1980) especially 86.
80 Vlastos, “Third Man,” 254–55. 81 Sedley, “Introduction,” 12.
 Plato’s Parmenides   255

Some scholars propose that the exercise defends a theory of forms from the difficulties.
(Understand “theory” as authors that cite a Theory of Forms mean it.) Rickless finds that
the exercise recommends rejecting some axioms of Plato’s earlier theory of forms.82
Meinwald holds that the exercise gives us a way to develop a theory underspecified in
earlier work.83 Peacock holds that the defense consists in showing difficulties at least
equally great for the main rival to Plato’s theory.84
The assumption that Plato’s other dialogues intend a theory with which the
Parmenides must be reconciled is natural.85 The assumption’s implications are worth
pursuing. It is also worthwhile to defer that assumption, say as an experiment.
I find additional reasons to defer it. Some bits of the purported theory of forms in
earlier dialogues seem to me highly unlikely to have passed the scrutiny of a careful
thinker such as Plato, for example, the instantiation interpretation of self-predication.86
Further, the proposals about forms that Socrates speaks in dialogues such as Republic
and Phaedo that scholars take as sources for Plato’s views of forms are not obviously
convictions of the Socrates that speaks them—for example, the forms’ accessibility
only to a select few. Rather, in Republic and Phaedo, Socrates speaks in order to meet
the demands of interlocutors that gave him speaking assignments.87 If the character
Socrates does not straightforwardly state his convictions, there is little reason to think
he asserts Plato’s. Nothing compels us to deem proposals about forms from other
­dialogues somehow present in the Parmenides.
Deferring the assumption of a multi-dialogue collection of assertions about forms,
I will accordingly view the Parmenides by itself and ask whether it makes reasonable
proposals about forms. It is striking that Parmenides admires the spare insight that
Socrates’ first question to Zeno contains: we acknowledge forms if we make distinctions
in dialegesthai. I share Parmenides’s admiration. I am inclined to take the insight as a
message from the Parmenides.
Readers that believe Plato intends a theory have compared Plato’s putative theory to
later theories.88 For example, the one-over-many principle has fascinated scholars by its
similarity to a too generous principle, unrestricted comprehension, which leads to the
famous Russell’s paradox for sets. If Parmenides’s one-over-many premise is similarly
generous, it requires a form over all items that are not self-exemplifiers. But if that form
exemplifies itself, it does not; and if it does not, it does. If so, the one-over-many would
generate an impossible object.

82 Rickless, Forms, 240. 83 Meinwald, Parmenides, 170–72; “Good-bye”, 372, 389–91; Plato 274.
84 Peacock, “Third Man.”
85 G. Vlastos, Socrates: Ironist and Moral Philosopher [Socrates], (Ithaca, N.Y., 1991), 117, n.50 calls
attention to a similar assumption, his own “grand methodological hypothesis.”
86 Self-predication understood as self-instantiation is widely ridiculed, but nevertheless attributed to
Plato. See Meinwald, “Goodbye,” n. 3 for some citations; Bailey, “Third Man,” 666–68; Gill Philosophos, 24.
87 See S. Peterson, Socrates and Philosophy in the Dialogues of Plato [Socrates] (Cambridge, 2011) on
Republic and Phaedo. Socrates’ speeches in Symposium and Phaedrus are persuasive display, not credos.
88 Penner, Ascent, x and 93–94: “<T>he Theory of forms is the first systematic theory of abstract
objects in the history of Western thought.”
256   Sandra Peterson

Pelletier and Zalta find the Russell paradox so potentially troublesome to Plato’s
theory of forms that they say (“Third Man,” 182),

<T>he first and foremost worry for a theory of Forms is to avoid the Russell
paradox.89

I propose that Plato’s depicted Parmenides who urges exercise need not share that worry,
because he has no unrestricted form-generating principle. He apparently does not trust
the one over many. (Nor will the adolescent Socrates, if he heeds Parmenides’s advice.)
For Parmenides urges Socrates to treat every assertion about a form as a hypothesis
(135e9; 136b7–8): for each case Socrates is to consider (skopein) the consequences if the
form is. Presumably if Socrates reaches a contradiction as he considers the hypothesis
that there is a form of the non-self-exemplifying, he will reject that hypothesis. He would
discover that there is no such form as quickly as we can discover that there is no barber
that shaves all and only those people who do not shave themselves.
The one itself, the form that Parmenides hypothesizes for his sample exercise, ­survives
unscathed his Deductions. Indeed, it surpasses them. For Parmenides also advises that
Socrates consider in each case the hypothesis that the candidate form the F is not. When
Deduction 8 considers the hypothesis that the one is not, it establishes that if the one
is not, not one thing is (166c1).90 That decisively reduces to absurdity the hypothesis
that there is no such item as what one is. It proves that there is the form the one itself
(=what one is). That form is no longer merely hypothetical.
In the light of their concern to avoid the Russell paradox, Pelletier and Zalta say (182):
“Theorists . . . owe us a consistency proof of their respective Theories of Forms.” Plato’s
Parmenides seems to me not a theorist with that obligation. His advice to hypothesize
each form and consider consequences shows that he takes instead a tentative and piece-
meal approach to discover what forms there are. He resembles a cook who doubles her
pudding recipe and observes the results, without thinking that she first owes a proof of
the consistency of arithmetic.
The creation of a system that generates forms and avoids set-theoretic style paradoxes
seems an intriguing spark struck from Plato rather than life-saving intervention for an
enterprise he cherished.

21. Appendix: Uniqueness

Rickless (Forms, 240) holds that “the fundamental lesson of the Parmenides” is to reject,
among other axioms of Plato’s previous theory of forms, the principle that each form is
unique. Rejecting Uniqueness eliminates what conflicted with the regress of the Third

89 See also Penner, Ascent, 44–48 and 282–91, and Cohen, “Third Man,” 469, n. 33.
90 Meinwald, “Exercise,” 485. Gill (Philosophos, 46, 70; “Design,” 496) stresses this result.
 Plato’s Parmenides   257

Man (Forms 73–75, 187). I would say in contrast to Rickless that Plato has the strong
r­ eason not to discard the Uniqueness principle that its denial is unintelligible.91 The F
itself (=what F is) is identified by its unique analysis. One account picks out a single form
or aspect, for example, what a square is. A different form would have a different account
and could not be what a square is, which explains why any square is square.
Further supporting his view that Plato now discards Uniqueness, Rickless argues that
Deduction 2 establishes directly and soundly (Parmenides, 17, 19) that Uniqueness is
false. According to Rickless (Forms 73–75 and 140–142) an argument at 142d9–143a3
(D2A3) elegantly shows that if the one is, the one is infinitely many.
I agree that the argument is elegant and sound. I do not agree that its conclusion
says that there are infinitely many forms of the one in such a way as to conflict with
Uniqueness.
142d9–143a3 generates the results that if the one is, there is also the oneness of the
being of one, the oneness of the being of being, the oneness of the oneness of being, the
oneness of the being of the oneness of being, and so on. There is in that way an infinity of
forms of oneness. However, contrary to Rickless’s view, none of these are duplicate forms
of oneness itself.
Oneness itself (=what one is) is what one, with no other qualifications, is by itself.
The oneness-of-being, having a description qualified by “of being” is not then what one
(itself by itself) is. What one is is that by which each single canine, each single raindrop,
each single integer, and every single thing, is one. These are not one by the oneness-
of-being (=what it is for being to be one). If the oneness of being is a form, it is over
exactly one item. That item is being-which-is-one. Possibly the oneness of being and the
oneness of the oneness of being and its ilk have the status of subtypes of oneness, as
the armadillo and the barracuda are subtypes of the animal.

Bibliography
Allen, R. E. Plato’s “Parmenides” (New Haven, Ct., 1997).
Allen, R. E. (ed.), Studies in Plato’s Metaphysics [Studies] (London, 1965), 231–63.
Apolloni, D. The Self-Predication Assumption in Plato (Lanham, Md., 2011).
Bailey, D. “The Third Man Argument” [“Third Man”], Philosophy Compass 4 (2009), 666–81.
Brunschwig, J. “Aristotle on Arguments without Winners and Losers,” Wissenschaften Jahrbuch
1984/1985: 31–40.
Burnet, J. Platonis Opera vol. 2 (Oxford, 1901).
Burnyeat, M. “What Descartes Saw and Berkeley Missed,” Philosophical Review 91 (1982), 3–40.
Cohen, S. M. “The Logic of the Third Man,” Philosophical Review 80 (1971), 448–75.
Duncombe, M. “The Greatest Difficulty at Parmenides 133c–134e and Plato’s Relative Terms,”
Oxford Studies in Ancient Philosophy 45 (2013), 43–61.
Fine, G. On Ideas: Aristotle’s Criticism of Plato’s Theory of Forms (Oxford, 1993).
Fine, G. “Separation,” Oxford Studies in Ancient Philosophy 2 (1984), 31–87.

91 Recall Cohen’s question, note 35 of this chapter. Worth noting: the formal development of Pelletier
and Zalta, “Third Man,” 176, yields uniqueness.
258   Sandra Peterson

Frances, B. “Plato’s Response to the Third Man Argument in the Paradoxical Exercise of the
Parmenides,” Ancient Philosophy 16 (1996), 47–64.
Frede, M. “Plato’s Arguments and the Dialogue Form,” Oxford Studies in Ancient Philosophy 9
(suppl.) (1992), 201–19.
Geach, P. T. “The Third Man Again,” in Allen, Studies, 265–77.
Gill, M. L. ed. and transl. and P. Ryan, transl. Plato. “Parmenides” (Indianapolis, Ind., 1996).
Gill, M. L. “Design of the Exercise in Plato’s Parmenides,” Dialogue 53 (2014), 495–520.
Gill, M. L. “Philosophos”: Plato’s Missing Dialogue (Oxford, 2016).
Havlicek, A. and F. Karfik, Plato’s “Parmenides.” Proceedings of the Fourth Symposium Platonicum
Pragense. (Prague, 2005).
Hermann, F.-G. Words and Ideas: the Roots of Plato’s Philosophy (Swansea, 2007).
Irwin, T. H. “The Theory of Forms” in G. Fine, (ed.), Plato I: Metaphysics and Epistemology
(Oxford, 1999), 143–70.
Kahn, C. H. Plato and the Post-Socratic Dialogue (Cambridge 2013).
Malcolm, J. Plato on the Self-Predication of Forms (Oxford, 1991).
Meinwald, C. “Good-bye to the Third Man,” in R. Kraut (ed.), Cambridge Companion to Plato
(Cambridge, 1992), 365–96.
Meinwald, C. “How Does Plato’s Exercise Work?,” Dialogue 53 (2014), 465–94.
Meinwald, C. Plato (New York, 2016).
Meinwald, C. Plato’s “Parmenides” (Oxford, 1991).
Meinwald, C. “What Do We Think We’re Doing?,” Plato Journal 17 (2017).
Miller, M. Plato’s “Parmenides.” (Princeton, N.J., 1986).
Migliori, M. and C. Moreschini. Platone. Parmenide (Milan, 1994).
Nails. D. The People of Plato (Indianapolis, Ind., 2002).
Owen, G. E. L. “Notes on Ryle’s Plato,” reprinted in G. E. L. Owen, Logic, Science, and Dialectic
(Ithaca, N.Y., 1986), 85–103.
Peacock, H., “The Third Man and the Coherence of the Parmenides,” Oxford Studies in Ancient
Philosophy 52 (2017), 113–76.
Pelletier, J. and E. Zalta, “How to Say Goodbye to the Third Man,” Nous 34 (2000), 165–202.
Penner, T. The Ascent from Nominalism (Dordrecht, 1987).
Peterson, S. “A Reasonable Self-Predication Premise for the Third Man Argument,”
Philosophical Review 82 (1973), 451–70.
Peterson, S. “New Rounds of the Exercise of Plato’s Parmenides,” Modern Schoolman 80 (2003,
245–78.
Peterson, S. “Plato’s Parmenides: A Principle of Interpretation and Seven Arguments,” Journal
of the History of Philosophy 34 (1996), 167–92.
Peterson, S. “Plato’s Reception of Socrates: One Aspect,” in C. Moore, ed., The Brill Companion
to the Reception of Socrates (Leiden, 2019).
Peterson, S. “The Greatest Difficulty for Plato’s Theory of Forms,” Archiv für Geschichte der
Philosophie 63 (1981).
Peterson, S. “The Language Game in Plato’s Parmenides, Ancient Philosophy 20 (2000), 19–51.
Peterson, S. Socrates and Philosophy in the Dialogues of Plato (Cambridge, 2011).
Priest, G. “The Parmenides: A Dialethic Intepretation,” Plato Journal 12 (2012) 1–63.
Rickless, S. “Parmenides,” Stanford Encyclopedia of Philosophy, revision of July 30, 2015, at
https://plato.stanford.edu/entries/plato-parmenides/.
Rickless, S. Plato’s Forms in Transition (Cambridge, 2006).
Rowe, C. Plato. “Phaedrus” (Oxford, 1986).
 Plato’s Parmenides   259

Ryle, G. “Plato’s Parmenides,” reprinted in Allen, Studies, 97–147.

Sayre, K. Parmenides’ Lesson: Translation and Explication of Plato’s “Parmenides” (Notre Dame,
Ind., 1996).
Sayre, K. “Plato’s Parmenides: Why the Eight Hypotheses Are Not Contradictory,” Phronesis 23
(1978), 133–50.
Sayre, K. “The Method Revisited: Parmenides, 135e9–136c6,” in A. Havlicek and F. Karfik (eds.),
125–40.
Schofield, M. “Likeness and Likenesses in the Parmenides,” in C. Gill and M. McCabe (eds.),
Form and Argument in Late Plato (Oxford, 1996), 49–77.
Scolnicov, S. Plato’s “Parmenides” (Berkeley, Calif., 2003).
Shorey, P. The Unity of Plato’s Thought (New York, 1980 reprint of Chicago, 1903).
Strang, C. “Plato and the Third Man,” Proceedings of the Aristotelian Society 37 (suppl. 1963),
147–64.
Taylor, A. E. Plato (New York, 1952).
Vlastos, G. “Plato’s ‘Third Man’ Argument (Parm. 132a1–b2: Text and Logic,” Philosophical
Quarterly (1969), 298–311, reprinted in G. Vlastos, Platonic Studies, 2nd ed. (Princeton, N.J.,
1981), 342–60.
Vlastos, G. “Separation,” Oxford Studies in Ancient Philosophy 5 (1987), 187–96.
Vlastos, G. Socrates: Ironist and Moral Philosopher (Ithaca, N.Y. 1991).
Vlastos, G. “The Third Man Argument in the Parmenides,” Philosophical Review 63 (1954),
319–49.
Yi, B.-U. and E. Bae. “The Problem of Knowing the Forms in Plato’s Parmenides,” History of
Philosophy Quarterly 15 (1998), 271–83.
 chapter 11

The Thea etet us

Mi-Kyoung Lee

Is it possible to explore and settle questions about the nature and possibility of
­knowledge without also considering what the possible objects of knowledge are? That is,
can epistemology be done independently of metaphysics? Or must epistemology always
go hand in hand with consideration of what kinds of things there are, and of what can be
said about them and how?
This question is raised most vividly for readers of Plato when assessing the central
epistemological claim of the Republic: that knowledge is impossible unless one grasps
the Forms, and that those who do not recognize the existence of the Forms can at
best achieve “opinion” (Rep. V. 475e–480a). It may then come as a surprise, when one
turns to Plato’s late dialogue the Theaetetus, which is devoted to the question “What is
knowledge?,” that Plato nowhere explicitly makes or even considers this claim, that
knowledge is not possible without the Forms. For one thing, the dialogue is filled with
examples of knowledge where the objects known include ordinary, mundane, individual
objects such as Theaetetus, Theodorus, oxen, wagons, and stones, as well as colors, smells,
and sounds. Socrates and his interlocutor discuss examples of knowledge, including
knowing that a stone is white, knowing that so-and-so is guilty of such-and-such a
crime, knowing that this person standing here is Theaetetus.
Of course, this by itself does not indicate a change in view—if Plato thought that
Forms are required for knowledge, but are not the only possible objects of knowledge,
then it would be possible to have knowledge of things other than Forms.1 However, it is
striking that the Forms are nowhere explicitly mentioned in the dialogue. One might
even suppose that the Theaetetus offers evidence that Plato eventually gave up the theory

1 Cf. Gail Fine, “Knowledge and Belief in Republic V,” Archiv für Geschichte der Philosophie 60
(1978), 121–39, and Gail Fine, “Knowledge and Belief in Republic 5–7,” in Stephen Everson (ed.),
Epistemology (Cambridge: Cambridge University Press 1990), 85–115. For a recent defense of the “Two
Worlds” interpretation of Plato, according to which Plato argues in Republic V that the only possible
objects of knowledge are the Forms and the only possible objects of opinion sensible particulars, see
Francisco J. Gonzalez, “Propositions or Objects? A Critique of Gail Fine on Knowledge and Belief in
Republic V,” Phronesis 41 (1996), 245–75.
262   Mi-Kyoung Lee

altogether;2 or, at any rate, one might arrive at the impression that Plato has decided, in
the Theaetetus, to make a fresh start by considering what knowledge is, while remaining
agnostic on the question of what the possible objects of knowledge might be.
This view is, in a way, both right and wrong. In the Theaetetus Socrates does not
assume the existence of the Forms. Indeed, he tends to use premises about the nature of
reality that are incompatible with the Forms. In that sense, the Theaetetus is free of the
Forms. But this does not mean the Forms have been abandoned. Plato adopts a complex
strategy for examining the nature of knowledge in the Theaetetus: he sometimes has
Socrates examine a conception of knowledge purely on its own terms. In other parts of
the discussion, he has Socrates examine a proposed conception of knowledge with the
help of substantial metaphysical claims about the nature and kind of objects that are
known—claims that conflict with what Plato argues for elsewhere. Plato’s strategy
appears to be to allow a fairly generous set of assumptions about the nature of the objects
of knowledge, assumptions that are introduced for dialectical reasons, and not because
he endorses them himself. He explores various conceptions of knowledge without
assuming that the only things that can be known are Forms or that knowledge is not
possible unless there are Forms. Nevertheless, although Plato does not prove the impos-
sibility of knowledge for one who does not acknowledge the existence of Forms,3 some of
the problems do appear to come from premises belonging to a Forms-free metaphysics.
But before we examine these issues in detail, let me briefly consider the main contours
of the dialogue and of Plato’s method of argumentation in it. After the introductory
section (Tht. 145c–151d), in which a preliminary attempt at defining knowledge is
rejected,4 three further definitions of knowledge (K1–K3) are proposed, examined in
detail, and then rejected. At the end of the dialogue, Socrates and his interlocutors
express puzzlement about what knowledge is, but Socrates declares Theaetetus better
prepared now to take up these questions again on a future occasion (Tht. 210bd). The
Theaetetus is thus an aporetic dialogue, sharing that form with early “Socratic” dialogues
such as the Euthyphro.
According to the first definition (K1), knowledge is the same as perception (Tht. 151e).
This leads to a long and extended attempt to spell out what exactly this amounts to and

2 McDowell tends to favor this as a working hypothesis; see the passages cited at John McDowell,
Plato: Theaetetus [Theaetetus] (Oxford: Clarendon Press, 1973), 159.
3 Contrast Cornford, who holds that the implicit moral of the Theaetetus is that “True knowledge
has for its object things of a different order—not sensible things, but intelligible Forms and truths about
them” (Francis Macdonald Cornford, Plato’s Theory of Knowledge: The Theaetetus and the Sophist of
Plato Translated with a Commentary [Theaetetus] (London: Routledge and Kegan Paul, 1935), 162). For
a discussion of some shortcomings of Cornford’s thesis and how it might be improved, see Gökhan
Adalıer, “The Case of ‘Theaetetus’ ” [“Case”], Phronesis 46/1 (2001), 2–3; Timothy Chappell, Reading
Plato’s Theaetetus [Theaetetus] (Indianapolis, Ind.: Hackett, 2005), 20–21.
4 This section includes Socrates’ famous comparison of himself to a midwife (Tht. 148e–151d); for
discussion, see Myles F. Burnyeat, “Socratic Midwifery, Platonic Inspiration,” Bulletin of the Institute of
Classical Studies of the University of London 24 (1977), 7–16; Myles F. Burnyeat, The Theaetetus of Plato
[Theaetetus], trans. M. J. Levett (Indianapolis, Ind.: Hackett, 1990); David Sedley, The Midwife of
Platonism: Text and Subtext in Plato’s Theaetetus [Midwife] (Oxford: Clarendon Press 2004).
 The Theaetetus   263

how one might support it. Plato has Socrates introduce a number of metaphysical t­ heses,
belonging to the so-called Secret Doctrine,5 in support of the definition. In the end, the
supporting metaphysical theory is rejected as incompatible with the proposed defini-
tion (Tht. 181c–183c); furthermore, the definition is rejected on its own for ­independent
reasons (Tht. 184b–186e). According to the second definition (K2), knowledge is true
opinion or judgment (doxa) (Tht. 187b).6 Plato has Socrates explore this definition by
seeing whether it is possible to explain how false belief is possible. Their repeated failure
to be able to explain how it is possible to think about something, and at the same time
make a mistake about it, strongly implicates the definition of knowledge itself (though,
admittedly, Socrates does not make the connection clear). The definition is also rejected
in a more straightforward fashion by pointing to contexts in which we would clearly
want to say that true judgment is not sufficient for knowledge (Tht. 200d–201c).
Finally, the third proposed definition of knowledge (K3) states that knowledge is true
judgment with an account (logos) (Tht. 201cd). Socrates and his interlocutors explore
this definition in two stages. First, they explore it in terms of a “Dream theory,” which
includes various metaphysical assumptions about the kinds of objects that can and
cannot be known, along with various reasons some things can be known and others can-
not (Tht. 201d–203c). The thesis of asymmetry in the knowability of objects is first
rejected by making explicit use of those metaphysical assumptions of the Dream theory
(Tht. 203e–205e), and then rejected in a more straightforward fashion (Tht. 206ac). Second,
they examine what an “account” (logos) is (Tht. 206c–210a). Each notion of account
they examine encounters problems, and in the end it is not clear whether and what kind
of account is necessary for knowledge—but their failure is partly due to the kinds of
assumptions retained from the Dream theory, and indeed from earlier parts of the
dialogue, about what kind of objects can be known and what can be said about them.
In what follows, I pursue two themes concerning the relation between epistemology
and metaphysics in the Theaetetus. The first theme concerns Plato’s methodology: some-
times Socrates examines a thesis on its own, and sometimes he examines it by assuming
premises on behalf of that thesis. These ancillary theses introduce metaphysical ideas
and commitments that are meant to describe sufficient conditions under which the
proposed definitions would be true. But they turn out to create problems for the very
definitions they were meant to support. In the case of K1 and the Secret Doctrine,
Socrates argues that K1 is not true if those metaphysical ideas are true. And in the case of
K3 and the Dream theory, Socrates argues that if one makes certain apparently reasonable
assumptions about the nature of things, then it is not possible to maintain, as K3 does,
that some things can be known and others cannot.
The second theme of the discussion concerns the kinds of objects of knowledge under
consideration. For the reader of the Theaetetus, the theory of Forms is the elephant in
the room—Socrates never mentions it, but that does not mean it has gone away. Socrates
nowhere argues that knowledge requires a grasp of the Forms—much less that knowledge

5 Socrates introduces this as a doctrine Protagoras taught his students “in secret” (Tht. 152c).
6 For a discussion of these and other possible translations of doxa, see Burnyeat, Theaetetus, 69–70.
264   Mi-Kyoung Lee

can only be had of the Forms. But the repeated failure to arrive at a definition of
­knowledge when assuming a metaphysics incompatible with the Forms suggests
(though it does not require) that progress could be made if we admitted certain assump-
tions characteristic of Plato’s metaphysics of Forms.

1. Knowledge Is Perception (K 1)

The first definition of knowledge Plato examines in the dialogue is the thesis that

(K1) Knowledge is perception (aisthêsis). (Tht. 151e)

Now this definition is of great interest because it articulates two ideas about knowledge
and perception that get short shrift in other dialogues such as the Phaedo and the
Republic: the idea that our senses are accurate and informative about their proper
objects—that is, colors, sounds, smells—and the idea that knowledge depends on and is
built up from perception. It becomes clear that Plato is particularly interested in explor-
ing the first idea, for he quickly connects the definition of knowledge as perception with
another thesis, Protagoras’s measure doctrine, according to which “man is the measure
of all things, of what is that it is, of what is not that it is not” (Tht. 152a). This is construed
as the claim that

(P) Whatever appears to be the case to one is the case for one.

Claim (P) appears to have been introduced for the following reason: Plato understands
Protagoras as focusing on what is the “measure” or criterion of truth. Protagoras’s
­thesis—that each of us is the measure of truth—derives its plausibility from the fact that
the senses are the criterion of what is and what is not, at least in the case of things such as
hot, cold, sweet, bitter, and so on.7 Thus, the senses are a criterion of truth, and what
they perceive in the case of sensible qualities is true. But if Protagoras’s measure claim is
true in the case of sensible qualities, such as hot, cold, and the like, then it follows
that perception “is always of what is, and free from falsehood” (Tht. 152c), and from this
we are to infer that it must be the same as knowledge. This, then, represents one version
of the empiricist claim: since perception is infallible with respect to the sensible quali-
ties, it should be regarded as a kind of knowledge. What the senses tell us is always true,
and hence their claims to knowledge should not be dismissed. Furthermore, K1 implies
that every instance of knowledge is a case of perception and that the senses do tell us
about everything; thus, it implies that nothing exists that is not perceived. Thus, (P),

7 For further discussion, see Mi-Kyoung Lee, Epistemology after Protagoras: Responses to Relativism
in Plato, Aristotle, and Democritus [Epistemology] (Oxford: Clarendon Press, 2005), 8–29.
 The Theaetetus   265

when restricted to sensible qualities, implies and is implied by (K1), at least on a certain
interpretation.8
Plato proceeds to explore the Theaetetan-Protagorean proposal (K1 and P) by working
out in detail the kind of metaphysical assumptions that would make the thesis true. He
introduces a set of Heraclitean theses, which include the idea that everything is in
motion and changing, and the thesis that, if something is F, then it also is or will be its
opposite, not-F (Tht. 152de). Drawing on this set of ideas, he shows that you can describe
a world in which perception is always true and that whatever appears to be the case in
perception is the case for one (Tht. 153d–160e). In this world, the object of perception
and the perceiving organ together generate perceptible properties and perceivings that
are unique to each encounter and that are each “of ” the other (Tht. 156a–157c). That is,
when I perceive a stone, I perceive the whiteness that was generated together with my
perception. One obvious question is why we should suppose that these two “offspring”
should always be generated together and why it’s not possible to have one without the
other. The answer is that these are assumptions that are simply brought in—ad hoc or
otherwise—under the rubric of the Heraclitean doctrine in order to make good on
Theaetetus’ and Protagoras’s claims.
The exact nature of the connection between (K1) and (P), on the one hand, and the
Secret Doctrine, on the other, is controversial. There are, in fact, two related issues: first,
how exactly to interpret (P); second, how (K1) and (P) are related to the metaphysical
theses in the Secret Doctrine. There are a number of possible answers to the first question:
Protagoras’s measure doctrine can be variously interpreted as (1) the thesis of relativism
about truth, according to which truth is relative, and nothing is true absolutely; (2) the
thesis of infallibilism, according to which all beliefs and appearances are true simpliciter;
and (3) relativism of fact, according to which whatever appears to be the case to one is
the case for one—a position that resembles (1) in emphasizing the importance of the
relativizing move but is more like (2) insofar as it is noncommittal on the question of
whether truth itself is to be relativized.9
As for the second issue, two general lines of interpretation are possible:

1. (K1), (P), and the Secret Doctrine are connected by relations of mutual entailment,
so that each one requires and is required by each of the others. On this view, there
is no way to maintain Protagoras’s measure doctrine without also being committed
to a radical doctrine of flux.

8 On some interpretations of (K1), this entailment does not hold true; see R. M. Dancy, “Theaetetus’
First Baby: Tht. 151e–160e,” Philosophical Topics 15/2 (1987), 61–108.
9 Point (1) can be found in Myles F. Burnyeat, “Idealism and Greek Philosophy: What Descartes Saw
and Berkeley Missed,” Philosophical Review 91/1 (1982), 3–40. Point (2) can be found in Gail Fine,
“Protagorean Relativisms,” in J. Cleary and W. Wians (eds.), Proceedings of the Boston Area Colloquium
in Ancient Philosophy, vol. 19 (Lanham, Md.: University Press of America, 1996), 211–43, and in Fine,
“Conflicting Appearances: Theaetetus 153d–154b,” in C. Gill and M. M. McCabe (eds.), Form and
Argument in Late Plato (Oxford: Oxford University Press, 1996), 105–33. I have given arguments for
point (3) in Lee, Epistemology, 30–45; see also Sarah Waterlow, “Protagoras and Inconsistency,” Archiv
für Geschichte der Philosophie 59 (1977), 29–33.
266   Mi-Kyoung Lee

2. (K1), (P), and the Secret Doctrine are not connected by relations of mutual
­entailment but by a narrower set of relations: the Secret Doctrine is sufficient for
the truth of (P), which, in turn, is sufficient for the truth of (K1). On this view,
Plato does not think, or argue, that the relations among the three positions is one
of entailment; rather, Plato is trying to characterize, on the basis of the Secret
Doctrine, a world of which (P) and thus (K1) hold true.10 All this requires him to
do is to find metaphysical assumptions on which the truth of (K1) and (P) would
follow—and these are found in the Secret Doctrine.

The two issues—concerning the exact interpretation of (P), and its connections with
(K1) and the Secret Doctrine—are not unrelated. If one adopts the view that interpretation
1 describes Plato’s strategy, then we have a problem since relativism about truth does not
appear to commit one to the metaphysical doctrine of flux, and, indeed, seems to be
incompatible with it, since the doctrine of flux would appear to be presented in the
Theaetetus as being true simpliciter, whereas relativism about truth denies that there are
any such truths. But perhaps (P) is not (1) relativism about truth but, rather, infallibilism,
(2) above. On this view, all appearances and beliefs are true simpliciter. But if contradictory
beliefs are true together without relativization, then doesn’t this imply that contradic-
tory states of affairs are simultaneously true? It is in order to save (P) from this problem
that Plato introduces the flux doctrine: contradictory beliefs can simultaneously be true
without contradiction because they turn out to be true of different things. On the other
hand, if one adopts the view that interpretation 2 describes Plato’s strategy, then Plato is
not arguing that (K1) or (P) commits one to any metaphysical thesis at all, but only that,
on certain metaphysical assumptions, (K1) and (P) turn out to be true—or so it seems.
On interpretation 2, the interpretation I favor, Plato argues that if one accepts the
“Heraclitean” metaphysical doctrine, then Theaetetus’ and Protagoras’s claims follow. So
do the dual theses in fact find support in the Heraclitean doctrine that Socrates intro-
duces? His own answer is yes and no. At first glance, the theory of perception Socrates
works out on the basis of the Secret Doctrine seems effective in showing how Theaetetus’
and Protagoras’s claims could be true. It can even handle problem cases such as sickness
and dreaming; as Socrates notes, if perceptions are always generated together with
perceptible properties, and these “offspring” are generated in different ways over time,
then there is no reason to suppose that someone who is supposedly awake is more
authoritative about her perceptions than one who is asleep (Tht. 157e–160e).
But in the end, they have to conclude that the dual theses are insupportable on the
Heraclitean hypothesis, for the reason that the thesis that “everything is changing”
implies not just the truth of Protagoras’s and Theaetetus’ claims but the opposite as well
(Tht. 181c–183b). For if everything is changing, then it will certainly be the case that for
every perception, there is a perceptible property matching it that will come into being
with it. But ex hypothesi the perception and perceptible property themselves will be

10 Sufficiency here is a one-way entailment, which is weaker than mutual entailment: if the Secret
Doctrine is true, then (P) and (K1) are true, but (P) and (K1) do not imply that the Secret Doctrine is true.
 The Theaetetus   267

undergoing change (Tht. 182de). For example, whiteness and the perceiving of white—that
is, the “twin births” in the Heraclitean story of perception that Socrates tells on
Protagoras’s behalf—themselves are constantly changing, so that whiteness becomes
not-white and perceiving becomes not-perceiving. What exactly this means is not clear;
at the very least, it suggests that a person’s perceiving that the stone is white is no more
true than it is false.
The Secret Doctrine is a metaphysical doctrine that leaves no room for the existence
of anything such as Forms. It says that everything is both F and not-F, that everything is
always changing, that everything is what it is relative to something else (Tht. 152de). The
Forms, by contrast, are never both F and not-F (e.g., Phaedo 74bc; Republic 478e–479e).
The Forms do not undergo change. And the Forms are not what they are relative to
something else—for example, relative to a perceiver—but are whatever they are in
themselves. Its total lack of stability—such as would be provided by the theory of Forms,
if one accepted it—is part of the reason that the Secret Doctrine comes to grief. For it
says that things such as perceptible properties and the perceptions of those properties
themselves do not remain stable. Thus, on this view, nothing can be said to be white
(as opposed to not-white) or to be perceiving, since whiteness itself and perceiving are
always changing.11
When Socrates says that the Secret Doctrine tells us that “whiteness” itself is
becoming not-white (Tht. 182d), it is unclear whether he has in mind the universal
color white or a particular instantiation of white. Either way, the lesson remains the
same: one cannot make perceptions true in a world that lacks the kind of stability that
Forms would provide. For in a world without that stability, things such as whiteness and
perceiving themselves undergo constant change, such that someone who is perceiving
something as white cannot be said to have a true perception—since even if they are
right, because there is something white out there, they are at the same time wrong,
because what is out there is at the same time not white. This argument cannot stand
alone as a proof for the existence of Forms, for nothing has yet been said about why
there must be entities that never change at all—it only shows that there must be some
necessary truths or that the nature of what it is to be white cannot change. (For example,
whiteness is necessarily white, and it is impossible for perceiving to become something
other than perceiving.) So (K1) and (P) can only be true in a world in which some such
limits have been placed on the extreme thesis of flux and opposites found in the Secret
Doctrine—limits such as can be found in Plato’s own view about the place of Forms in
a world of flux.
Theaetetus’ definition (K1), then, cannot be given any support by the Secret Doctrine.
Socrates proceeds to examine Theaetetus’ definition on its own, independently of the

11 On interpretation 1, according to which Protagoras is committed to the theory of flux, it is

difficult to explain why Protagoras is committed to such an extreme thesis of flux. Why couldn’t he
simply say that things are sometimes changing in some respects—not everything in every respect? On
interpretation 2, however, flux is introduced as a part of a metaphysical doctrine that is meant to
provide sufficient conditions for (P). On this view, Plato is not arguing that Protagoras is committed to
the radical thesis of flux.
268   Mi-Kyoung Lee

doctrine of opposition and flux, at Theaetetus 184–86.12 This argument is of interest

because it comes the closest of Plato’s arguments in the Theaetetus to ground familiar
from Republic book V: as in the Republic, Plato has Socrates argue here that perception is
not sufficient for knowledge because it cannot “get at being.” But whereas in the Republic
the reason for this has to do with the nature of the objects of perception—namely, the
fact that they are changeable and variable—the reason given here has to do with the
nature of perceptual states and activity.
At 184–86 Socrates argues that perception is infallible about the proper objects of
sense: sight is authoritative about colors, hearing about sounds, and so on. But unlike
Epicurus, Plato does not make much of this fact, if indeed he thinks it’s true. Though he
thinks that the senses are authoritative about their own objects, he is interested rather in
their limitations: they cannot tell us about anything beyond their own special objects.
Thus, for example, the sense of sight is capable of determining about colors but not about
sound, much less about anything such as “being,” which is needed to determine truth.
And if the senses are not capable of getting at “being” or any of the other common
objects of thought, and “being” is necessary for truth, which, in turn, is necessary for
knowledge, then it follows that perception is not sufficient for knowledge.
What does it mean to say that perception cannot get at “being,” which is necessary for
truth and for knowledge? Perception consists of a bare sensory awareness and lacks even
the basic ability to apply concepts and form judgments. But grasping being is, in the first
instance, saying what things are. What we do with the senses does not even rise to the
level of making a judgment of the form “x is F,” such as “the stone is white”; the senses do
not “say” anything at all. Perception by itself has no propositional content; all that we do
with the senses is to apprehend some color, experience some texture, and the like.
Knowledge, however, requires at the very minimum the propositional complexity
involved in making judgments.13

12 Socrates also examines (P) independently of the doctrine of opposition and flux in the celebrated
refutation of Protagoras at Theaetetus 169e–171d. The classic treatment of this argument is
Myles F. Burnyeat, “Protagoras and Self-Refutation in Plato’s Theaetetus,” Philosophical Review 85
(1976), 172–95.
13 This accords with one of the two interpretations offered in John M. Cooper, “Plato on Sense-
Perception and Knowledge (Theaetetus 184–186),” Phronesis 15 (1970), 123–46, and is endorsed by
Myles F. Burnyeat, “Plato on the Grammar of Perceiving,” Classical Quarterly 26 (1976), 44–45.
However, Plato does not consistently adhere in this passage to a neat distinction between perception as
sensory awareness and the mind’s conceptualizing activity; for one thing, it seems that perception
includes not just sensory awareness but some use of concepts in order to label things as “sweet” “hot,”
and so on, which, in turn, suggests that use of concepts does not necessarily require the use of einai.
For this reason, Cooper prefers the second interpretation, on which perception is taken to be the
activity of the mind in apprehending things by means of the senses. Perception then does attain the use
of sensory concepts and can determine by itself whether something is white or red or sweet or hot
(Tht. 184d–e). But it fails to be knowledge because it does not attain objective validity. For whereas it is
possible to “read off ” from sensory data whether something is hot, cold, wet, or dry, it is not possible to
determine in the same way whether something is (really) beneficial or valuable, same, different, and so
on. On this interpretation, what the senses do falls short of what is needed to determine what is really
the case. For problems with this interpretation, see Sedley, Midwife, 106–07 n.29.
 The Theaetetus   269

Socrates then rejects the claims of perception to be knowledge because, on his newer
and narrower understanding of perception,14 perception constitutes bare sensory
awareness and never attains the level of making judgments about how things are. This
argument does not imply that knowledge is not possible without the Forms. But at the
same time, it doesn’t tell us that knowledge is possible without the Forms—nothing here
suggests that Plato has renounced the view that knowledge requires grasping the Forms.
What the argument here tells us is that knowledge is not possible without grasping
“being,” where that includes not merely the ability to make judgments and claims about
how things are (which is why perception falls short) but also the ability to make expert
judgments about what is true, an ability that requires the grasp of objective standards for
each subject matter (Tht. 186bc).15 And this point is consistent with the claim in the
Republic (though, again, it does not imply) that only the person who grasps the Forms is
in a position to know whereas the person who is ignorant of the Forms is not.
What the argument here at Theaetetus 184–86 does leave open is that it might be
possible to know things about the objects of perception. For the argument here focuses
not on the unsuitability of the objects of perception for being objects of knowledge but,
rather, on how perception relates to its objects. Since perception fails to count as knowledge
because it doesn’t even rise to the level of making statements about them, this leaves
open the possibility that sensible objects could be objects of knowledge. This could
explain why, in the next section where Socrates examines a new definition of knowledge
as true judgment, he consistently uses examples of knowledge about sensible objects to
illustrate his points.

2. Knowledge as True Judgment (K 2)

The second proposed definition—according to which knowledge is true belief or

judgment—flows from the rejection of the first definition of knowledge as perception.
Perception is rejected as insufficient for knowledge since it is limited to the apprehension
of proper sensibles (such as hot, cold, red, sweet, etc.) and lacks propositional content
and so cannot even get at “being.” That is, perception cannot by itself deliver judgments
about what is the case. But that then raises the question whether true judgments about
what is the case might be sufficient for knowledge (187bc).
The proposal is not innocuous. It says that getting something right—making a correct
judgment about something—is enough to count as having knowledge about it. But Plato
already pointed out in the Meno that there is a significant difference between true
belief and knowledge: true belief is presumably just as good as knowledge as long as it

14 For the argument that Theaetetus 184–86 signals a change in Plato’s conception of perception, see
Michael Frede, “Observations on Perception in Plato’s Later Dialogues,” Essays in Ancient Philosophy
(Minneapolis: University of Minnesota Press, 1987), 3–8.
15 This is why perception falls short, according to the second interpretation argued for by Cooper.
270   Mi-Kyoung Lee

“stays put.” But true belief is easily dislodged; someone who only has true belief and not
knowledge will easily be persuaded of the falsity of her belief and will quickly change her
mind. What is needed to make one’s belief stay put is an “account of the reason why”
(aitias logismos) that will make the belief stable (Meno 98a).
Another way of putting this is to say that true belief is not a capacity or ability. You can
make a true judgment once, by accident or randomly, for the wrong reasons or perhaps
because you made a good guess. Judging correctly about something on one occasion is
quite compatible with making mistakes about it on other occasions. Saying that true
belief is sufficient for knowledge thus violates the idea of knowledge as expertise, the
idea that knowledge is a capacity that makes one the source of authoritative and infalli-
ble judgments about a thing. Surely someone who knows something can’t also make
mistakes about it.
To say, then, that knowledge is the same as true judgment is to maintain that true
judgment itself is the (sole) source of authority and infallibility with respect to knowledge.
Getting things right—no matter how one manages to do so and how reliably one is able
to do so—is enough in itself to count as having knowledge. Note that this was implied by
the earlier proposal that perception is knowledge; perception was deemed to be knowledge
because perception is infallible, and truth is sufficient for knowledge (152c).16 This
definition of knowledge as true judgment gives Plato the opportunity to examine the
assumption contained in the definition of knowledge as perception—an idea otherwise
taken for granted up to now—that what makes anything a suitable candidate for knowledge
is its getting things right.
Like the definition of knowledge as perception, the definition of knowledge as true
judgment is examined in two phases. First, Plato has Socrates examine it indirectly, by
seeing whether it is possible to explain how false judgment is possible, if we suppose
that true judgment is sufficient for knowledge. Socrates’ repeated failure to explain
the possibility of false judgment—five attempts in all at (1) Tht. 188ac, (2) 188c–189b,
(3) 189b–191a, (4) 191a–196c, and (5) 197a–200d—is an indirect indictment of the defini-
tion of knowledge on which the discussion depends.17 The fundamental problem is that
it doesn’t seem possible to explain how one can be thinking of something (as opposed to
something else) and make a mistake about it (say, by misidentifying it as something else
entirely, Y). The source of the problem is the very definition of knowledge as true judg-
ment itself, for on that view, as we noted earlier, knowledge is not an enduring capacity

16 The connection of this section with Protagoras is reinforced by the fact that much of this section
is devoted to the apparent impossibility of false belief; the Euthydemus (285e9–286c9) attributes the
denial of falsity to Protagoras. In antiquity, Proclus also thought that this section was a continuation of
the discussion of Protagoras in the first part of the dialogue (In Plat. Prm. 657.5–10; cf. David Sedley,
“Three Platonist Interpretations of the Theaetetus,” in C. Gill and M. M. McCabe (eds.), Form and
Argument in Late Plato [Oxford: Oxford University Press, 1996], 82 n.3; Sedley, Midwife, 119).
17 See also Gökhan Adalıer, “Materialism in Plato’s Theaetetus” [“Materialism”] (PhD dissertation,
Duke University, 1999), and Adalıer, “Case.” Gail Fine, “False Belief in the Theaetetus” [“False Belief ”],
Phronesis 24 (1979), 70–80, also thinks the definition of knowledge as true judgment is implicated in
the failure to explain how false judgment is possible.
 The Theaetetus   271

that can be exercised or not on various occasions but, rather, consists simply of true or
correct judgment, whenever it happens to occur.
Thus, if knowledge is the same as true judgment, then whenever you make a correct
judgment, you have knowledge. But to think about something, you must be thinking
about it and not something else, and so you must be judging it truly (about what it is)—
but then by definition you know it. And if you know it, it seems impossible to make a
mistake about it, since it is not possible to know and not know the same thing. In other
words, if even thinking of something requires that one know what one is thinking of,
then it seems to follow—at least according to the line of thought Socrates and Theaetetus
pursue18—that it is not possible to think of something and make a mistake about it at
the same time. This is the fundamental obstacle that Socrates and Theaetetus keep
confronting and trying to find a way around.
For example, consider the fourth attempt to explain false judgment, the “Wax Block”
model of thinking (Tht. 191a–196c), according to which there is a wax block or tablet in
our souls onto which we imprint our perceptions, thereby gaining the ability to call up
those thoughts long after the sensory affection has passed. The wax block itself seems to
represent the faculty of memory and of thought. Socrates introduces it to solve the prob-
lem that it seems impossible to think of something as X and at the same time to think of
it as something else, Y. He solves it by finding a way to have something in mind without
thinking of it as X: by perceiving it. Socrates says that we perceive things and then
imprint the images of those things into the “wax block” in our minds—that is, in our
memory. Thus, having a wax imprint of something in our minds represents the capacity
to call up an image of, and think of, that thing. But when one perceives an object, the
object is presented to one without one’s being aware of what it is.19 Thus, it is possible to
have something in mind and to misidentify it without knowing it. One can have an
object presented to one in perception (e.g., Theaetetus), and when one matches this to
the wrong imprint in one’s wax block (e.g., that of Theodorus), one is effectively making
a mismatch without being guilty of knowing Theaetetus and making a mistake about
him. For the reason is that the wax block allows one to perceive Theaetetus, but in per-
ceiving Theaetetus, one is not perceiving Theaetetus as Theaetetus; in other words, one
can perceive an object without having any thought whatsoever about what it is.
So according to the proposal, one has a wax block in one’s mind, which allows one to
have an imprint of X, which represents the thought or memory of X, and one is also
capable of having perceptions of X (though one does not perceive X as X), and in making
judgments about things, one either successfully or unsuccessfully matches perception
with imprint. A mismatch then represents the thought that “this is Y,” where “this” in
fact refers to some X that is not the same as Y. This appears to be a successful explanation

18 This follows only if we assume that all judgments are identity statements (see note 20 of this
chapter), or if we assume that knowing what one is thinking of is knowing everything about what one
is thinking of (cf. the “all-or-nothing” view of knowledge as acquaintance discussed in Fine, “False Belief ”).
19 This passage continues to assume, in line with Tht. 184–86, that perception has no propositional
content, though perhaps it differs from it in allowing that what we see are objects like Theaetetus, not
just the special sensibles.
272   Mi-Kyoung Lee

of false judgment. But it is rejected because it is unable to explain how false judgment
can occur in cases where perception is not involved (Tht. 195b–196c). And it is clear that
false judgment occurs even about things that we grasp by means of the mind; for example,
one can make a mistake about 12, thinking that it is the same as 5 + 6.
One noteworthy feature of the entire section on false judgment is that Socrates
focuses almost exclusively on judgments of identity about particular things, such as
judging Theaetetus (i.e., judging who he is) or judging Socrates (i.e., judging who he is).20
In my view, the focus on identity statements does not vitiate the argument; even if it does
not cover all judgments, such as judgments like “Socrates is snub-nosed,” Socrates is still
quite right to think that it is a problem, given their initial assumptions, to explain how
one can know something and make a mistake about it. More significant, in my view, is
the fact that they tend to focus on judgments about particular things. Here, as elsewhere in
the dialogue, we find an ecumenical tendency toward the question of what kind of
objects can be known.
Though the problems of explaining false judgment cannot be laid at the doorstep of
this focus on judgments about unique particulars,21 we will see that it will later give rise
to a problem for the final definition of knowledge as true judgment with an account. The
problem raised there is that anyone with true judgment already appears to be in possession
of an account, if having an account is having the distinguishing mark that sets off what
one knows from everything else; hence, adding “with an account” adds nothing to true
judgment that it didn’t already have (Tht. 208d–210a). As we’ve already seen, if having a
true judgment about X consists of having X in mind, and no one or nothing else, then it
does seem that true judgment already carries with it the ability to distinguish X from
everything else. But this problem disappears, as I argue later on in the chapter, if one
takes kinds, rather than unique individual objects, as the objects of knowledge.
Besides the indirect examination of the definition of knowledge as true judgment,
Plato also has Socrates examine it directly: Socrates dispatches it fairly quickly, in an
argument that takes barely two paragraphs (Tht. 201ac). As Socrates notes, juries can be

20 At Tht 188c5–7, Socrates draws the conclusion that it is impossible to judge that one thing is
another—that is, that false misidentifications are impossible; he then infers that one cannot make any
false judgments (188c7–8). Since there seem to be more forms of judgments than identifications, such
as misdescriptions, it would appear that the inference does not follow (cf. Burnyeat, Theaetetus, 70–123,
esp. 70–73, for a statement of the problem). Either the argument is meant to be limited to identity
judgments (Cornford, Theaetetus, 113; Frank Lewis, “Two Paradoxes in the Theaetetus,” in J. M. E.
Moravcsik (ed.), Patterns in Plato’s Thought (Dordrecht: D. Reidel, 1973), 123–24; Nicholas White,
Plato on Knowledge and Reality (Indianapolis, Ind.: Hackett, 1976), 164; McDowell, Theaetetus, 195),
or more than identity judgments are considered (C. F. J. Williams, “Referential Opacity and False
Belief in the Theaetetus,” Philosophical Quarterly 22 (1972), 298–99; Fine, “False Belief,” 74; and David
Bostock, Plato’s Theaetetus [Theaetetus] [Oxford: Clarendon Press, 1988], 172–73). A Cornford-style
approach has recently been argued for by Adalıer, “Materialism,” who thinks these assumptions are
characteristic of a position Plato is arguing against, one which assumes that all judgments are
judgments of identity, precisely because it does not admit Forms and therefore the possibility of
predication.
21 Unless one supposes that this entire section implicitly presupposes an ontology of particulars; cf.
Adalıer, “Materialism,” and Adalıer, “Case.”
 The Theaetetus   273

correctly convinced that certain events occurred—for example, that the defendant
­committed a murder at a particular time—even though they did not themselves witness
the event. But only someone who has actually seen the event could be said to know that
it occurred. That is, in order to have knowledge, one needs proper evidence or justification
for one’s belief. The members of the jury could be said to have, at best, correct judgment,
not knowledge.22 The striking thing about this argument, for our purposes, is that it
clearly implies that knowledge is possible for things such as particular facts and events
and, furthermore, that perception may have a role to play in acquiring the proper
­evidence or justification required for knowledge. Here again is evidence that Plato is
prepared to entertain a wide range of possible objects of knowledge, though the argu-
ment still leaves open the possibility that Plato thinks that knowledge of the Forms is
necessary even to know, for example, that some event occurred.

3. Knowledge as True Judgment

with an Account (K 3)

The refutation of the definition of knowledge as true judgment shows that true ­judgment
by itself is not sufficient for knowledge: one needs something additional, playing the role
that firsthand witnessing of an event plays in the case of knowing what happened on a
particular occasion. This point leads Socrates and Theaetetus to their final proposal
­concerning knowledge, (K3), that knowledge is true judgment with an account (logos)
(Tht. 201cd). This definition is the most likely to be endorsed by Plato himself, since
there are many passages in other dialogues where something like K3 is endorsed—most
famously, the statement in the Meno that “true beliefs are not worth much until one ties
them down by reasoning about the cause” (aitias logismoi; Meno 98a).23
The central idea in Theaetetus’ definition of knowledge is that “things of which there’s
no account are not knowable . . . whereas things which have an account are knowable”
(Tht. 201d). This introduces an asymmetry between things that do and do not have an
account (call this “asymmetry of logos,” or “AL”), which together with the requirement
that everything known must have a logos (call this “knowledge requires a logos,” or “KL”),
gives rise to an asymmetry between things that can be known and things that cannot be
known (call this “asymmetry of knowledge,” or “AK”).24 K3 explicitly says that some

22 For further discussion, see M. F. Burnyeat, part 1, and Jonathan Barnes, part 2, of “Socrates and
the Jury: Paradoxes in Plato’s Distinction between Knowledge and True Belief,” Proceedings of the
Aristotelian Society 54(suppl.) (1980), 173–91 and 193–206.
23 See also Phaedo 76b5–6, 97d–99d2; Symposium 202a5–9; Republic 534b3–7; Timaeus 51e5. For
further discussion, see Taylor, chapter 18 in this volume.
24 Cf. M. F. Burnyeat, “The Material and Sources of Plato’s Dream” [“Dream”], Phronesis 15 (1970),
101–22; Gail Fine, “Knowledge and Logos in the Theaetetus,” Plato on Knowledge and Forms: Selected
Essays [“Knowledge”] (Oxford: Clarendon Press, 2003), 225–51, originally published in Philosophical
Review 88 (1979), 366–97.
274   Mi-Kyoung Lee

things can be objects of knowledge (namely, those things that have an account) and that
other things cannot be objects of knowledge (namely, those things of which there is no
account). But as it stands, K3 is extremely abstract; it is unclear what exactly a logos or
account is, and why certain things admit of an account, whereas others do not.
Socrates begins his examination of this definition with a move familiar from the
­perception section of the Theaetetus: he introduces another thesis—or, rather, a set of
theses that are meant to illustrate and support the proposed definition. That is, he examines
the definition by offering a set of ideas that is sufficient for the truth of the definition; he
introduces what he refers to as a “dream” to show how Theaetetus’ definition could be
true (Tht. 201d–203d).
In particular, the “dream” is meant to answer the second question posed previously,
why some things can’t be given a logos and hence can’t be known, whereas others can.
According to the Dream theory, the asymmetry exists because things fall into two dif-
ferent kinds: “primary elements (stoicheia), as it were, of which we and everything else
are composed” (Tht. 201e–202a) and those things that are composed out of them. It is
unclear what these primary elements are and how they figure as constituents in every-
thing else. We are simply told that (1) elements can only be named; (2) that one cannot
say anything else of an element—such as “is,” “is not,” “itself,” “that,” or “each”—since
that would be to add something to it which does not belong to it alone; and (3) that an
element can be perceived, not known. By contrast, things composed out of elements
(a) can be given an account, (b) which consists of names woven together, and (c) can be
known (201e–202b).
Like K3, the Dream theory is abstract and open to multiple interpretations. Are the
primary elements material stuffs, or are they parts out of which other material objects
are constituted? Such an interpretation is encouraged by the fact that Socrates talks of
primary elements “out of which we and everything else are composed” (201e), as well as
by his later remark that elements “have no account and are unknowable, but they’re
perceivable” (202b). Or is the Dream theory a theory about meaning, where the primary
elements and things that are composed out of them are the meanings of our words and
meanings of sentences or propositions constructed out of them? One could cite in sup-
port of this Socrates’ speaking of the elements being “woven together” into a complex,
just as the names are woven together into a logos or account (202b). So understood, there
is a resemblance between the Dream theory and Wittgenstein and Russell’s Logical
Atomism, a resemblance noted by Wittgenstein himself.25 There are no doubt other
possible spheres of application—and perhaps this is a sign of how potentially powerful
the theory is. But, in my view, none of these do justice to Plato’s intentions. Plato delib-
erately leaves it open what the “primary elements” are,26 for he only wants to focus on
certain features of ontology and language and not others. Plato leaves many features of

25 Cf. Burnyeat, Theaetetus, 149–64; Chappell, Theaetetus, 208–11, offers a reply to Burnyeat’s
objections to the “Logical Atomist” interpretation of the Dream theory.
26 This is a theme of Burnyeat’s discussion of the third definition of knowledge (Burnyeat,
Theaetetus, 129, 131–32, 164).
 The Theaetetus   275

the Dream theory vague in order to make it sufficiently general and hence widely
applicable to a variety of possible objects of knowledge.
The Dream theory focuses on the following three features of ontology and language:

1. It tells us that things fall into two kinds: elements and those things that are com-
posed or “woven together” out of them. That is, the distinction between things that
can and can’t be known appears to correlate with a distinction between things
that are ontologically basic and others that are made up out of them.
2. Elements can only be named (not given an account), and they can only be
­perceived (not known). That is, their ontologically basic status gives rise to the
fact that they cannot be given a logos (AL) and thus the fact that they cannot be
known (AK); they are spoken of and grasped through other means.
3. A logos says of a thing what is proper to it. This effectively restricts all logoi to
identity statements—that is, statements or definitions of what a thing is.27 The
Dream theory continues to assume, as in the false judgment section, that all judg-
ments are judgments of identity about particular objects.

Next, Socrates considers what the Dream theory would say about the following case:
take letters to be primary elements and syllables to be complexes made up out of them
(Tht. 202e). Thus, the first syllable of Socrates’ name, “SO,” is a complex, and the letters
“S” and “O” are the elements out of which the syllable is composed. The account of the
complex “SO”—given in answer to the question “What is ‘SO’?”—would be that it is “S”
and “O.” However, “S” and “O” cannot themselves be given an account; as Theaetetus
says, “How could one express in an account the elements of an element? In fact, Socrates,
‘S’ is one of the unvoiced consonants, only a noise, which occurs when the tongue hisses,
as it were” (Tht. 203b). One will wonder, of course, why what Theaetetus has just said
about the letter “S” could not count as an account of an element. But the reason is evi-
dently that it does not refer to the parts of a letter, because a letter has no parts ex hypothesi.
We can thus infer that in the Dream theory we are to assume that the account of a thing
is simply an enumeration of its elements (EE) and, furthermore, that the elements are a
thing’s parts. The Dream theory is thus reductionist because it takes a thing to be no
more than its parts and therefore to be wholly analyzable into its parts.28 But why should
we assume that?

27 The oddity of this stricture—that one should not, in general, say anything of a thing that doesn’t
belong to it, and that one should only say of a thing what belongs to it alone—has historically put
people in mind of Antisthenes, partly because Aristotle seems to suggest that Antisthenes had the
strange view that the only way you can talk about a thing is to name it; hence, both subject-terms and
predicate-terms in sentences serve the same function—that is, to name—and a sentence itself is
nothing other than an extended name. Whether or not Plato has him in mind is not clear, partly
because we know so little about Antisthenes. For an even-handed judgment on this matter, see
Burnyeat, “Dream,” and Burnyeat, Theaetetus, 164–73.
28 Cf. A. E. Taylor, Plato, the Man and His Work (London: Methuen, 1926), 344–46; K. Sayre, Plato’s
Analytic Method (Chicago: University of Chicago Press, 1969), 120–30; Adalıer, “Materialism,” 207–48,
esp. 234–41; Sedley, Midwife, 158.
276   Mi-Kyoung Lee

To answer this, let’s look at Socrates’ two refutations of the Dream theory—or more
precisely, his refutations of the thesis of asymmetry in knowledge between elements and
complexes. One of these assumes this controversial point, that the relation between an
element and a complex is that of part to whole and that the whole is the same as the sum
of its parts (call this “WP”); the other does not. In the first argument, WP is assumed in
order to argue that the thesis of asymmetry in knowledge between element and complex
is untenable: either they are equally knowable, or they are equally unknowable (not-AK)
(Tht. 203d–205e). The second argument does not assume WP; it simply points out that
our experience in coming to know things is the opposite of AK: far from it being the case
that the elements are unknowable, we find that in our own experience of learning, the
elements are better known than those things that we know by means of their elements
(Tht. 205e–206b). For example, when we learn to read, we concentrate on learning the
letters first and only later on recognizing the syllables constructed out of them. As
Socrates says, “the class of elements admits of knowledge that is far clearer, and more
important for the perfect grasp of every branch of learning, than the complex” (Tht. 206b).
Plato’s strategy in offering two different arguments against AK seems to be to start
both from premises (such as WP) that he would probably not accept and from premises
that he might accept. The advantage of this strategy is that it covers his bases; insofar as
WP is widely accepted, even if not by Plato himself,29 an argument showing that WP is
incompatible with AK would strongly suggest that AK should be rejected. The second
version clinches the argument, showing that we have good reason to reject AK even if we
do not accept WP.
What follows if we reject AK, the thesis of asymmetry of knowledge between
­elements and complexes? Either elements and complexes are likewise knowable or like-
wise unknowable. That we are meant to conclude the former is suggested by the second
argument Socrates gives against AK, in which he says that our experience of learning
our letters suggests that, far from it being the case that we have no knowledge of the ele-
ments, knowledge of the elements of a subject matter is fundamental in coming to learn
it. Supposing that elements and complexes are both knowable (not-AK), what should we
say about KL, AL, and EE, since KL, AL, EE, and not-AK are inconsistent?30

A. One option is to reject KL: not everything requires a logos to be known.31 For
example, one might suppose that certain Forms—in particular, the Form of the
Good—will figure in Plato’s answer to the question of what the elements of

29 Cf. Burnyeat, Theaetetus, 191–209, esp. 199–201; Verity Harte, “Plato’s Problem of Composition,” in
John J. Cleary and Gary M. Gurtler (eds.), Proceedings of the Boston Area Colloquium in Ancient
Philosophy 2001, vol. 17 (Leiden: Brill, 2002), 1–26.
30 Cf. Fine, “Knowledge,” 236.
31 Some argue that knowledge for Plato requires some kind of nondiscursive, intuitive grasp of its
objects (e.g., I. M. Crombie, An Examination of Plato’s Doctrines, 2 vols. (London: Routledge and Kegan
Paul, 1962, 1963), 2:1131–34; Richard Robinson, “Forms and Error in Plato’s Theaetetus” in his Essays in
Greek Philosophy (Oxford: Clarendon Press, 1969), 52–55). Alternatively, one might think that the point
of giving up KL is to acknowledge that not everything can be defined without circularity, and hence at
least the most basic Forms must be known by some other way (Stephen Menn, “Collecting the Letters”
[“Collecting”], Phronesis 43/4 [1998], 201).
 The Theaetetus   277

everything are but that the Form of the Good itself cannot be given an account,
since it is the most fundamental of all. This option would be particularly com-
pelling if one thought that Plato was committed to EE as a model of what an
account is—that is, if one thought that accounts can only be one-directional,
from the more complex to the simpler, from explanandum to explanans. (Note that
EE does not commit one to WP, since the elements in terms of which one gives
an account of a thing need not be parts of that thing.) Arguably, Aristotle took
this option, since he distinguishes first principles or elements that are known by
other means than demonstrative knowledge32—namely, by nous.
B. Another option is to retain KL and reject AK and AL: both elements and com-
plexes can be known, and since knowledge requires an account, both elements
and complexes can be given accounts—albeit accounts of different sorts. The key
here is to reject EE, according to which an account is an enumeration of a thing’s
elements.33 The reason the Dream theory gave for denying accounts to simples
was that simples don’t have parts. But Theaetetus’ own reply when explaining
that the letter “S” does not have an account—that it has no account because there
are no letters in a letter, that it is simply an unvoiced consonant—shows that one
could give a different kind of account of “S,” one that did not analyze a thing in
terms of its parts, but that placed it in a classification scheme relative to other
letters and sounds: vowels versus consonants, voiced versus unvoiced, and so on.
So elements could receive accounts not in terms of their parts—since they don’t
have any—but rather in relation to other elements and ultimately in relation to
the whole field to which they belong.

Deciding which option Plato intends us to go for would be too large and complex an
undertaking for this chapter.34 For our purposes, it is sufficient to note that whether one
thinks we are supposed to carve out a different conception of knowledge besides the
kind of knowledge that is true judgment with an account, as in option A, or to defend
the viability of K3 as a definition of knowledge by jettisoning some of the problematic
features of the Dream theory, as in option B, there is no reason to suppose that Plato is
committed to the assumption that a whole is identical to the sum of its parts (WP).
According to option A, Plato retains the idea that a logos is, fundamentally, an enumera-
tion of a thing’s elements—though the elements of what is known are not necessarily a
thing’s parts (i.e., not WP). Instead of expanding the conception of logos, we’re supposed
to realize that knowledge does not require a logos. Whatever the most basic items of

32 Aristotle says that knowledge is always “with an account” (Posterior Analytics II 19. 100b10;
Nicomachean Ethics VI 6. 1140b33) but distinguishes demonstrative knowledge with the self-explanatory,
undemonstrated knowledge of first principles (Posterior Analytics I 3. 72b19–24, II 19. 99b20).
33 “The endorsement of KL, coupled with the rejection of AK, suggests that AL and,
correspondingly, EE are also to be rejected: since elements are as knowable as compounds, and since all
knowledge requires accounts, there must be accounts of elements” (Fine, “Knowledge,” 237).
34 Recommended readings include Fine, “Knowledge”; Bostock, Theaetetus; Burnyeat, Theaetetus;
Adalıer, “Materialism”; Sedley, Midwife; Chappell, Theaetetus; Thaler, “Taking the Syllable Apart”; and
Broadie “The Knowledge Unacknowledged.”
278   Mi-Kyoung Lee

ontology are, they are not going to be known by means of a logos but, instead, will be
­perceived or known in some other way. According to option B, Plato rejects WP, and
also EE, as imposing an unnecessarily restrictive conception of logos on the definition of
knowledge as true judgment with an account (which then points forward to Theaetetus
206c–208b, especially 208ab, where he makes this point explicitly). If one expands
one’s conception of logos, then one might think that both elements and complexes can
have logoi, and, correspondingly, both elements and complexes can be known, albeit in
different ways.
What we ultimately think of the definition depends on getting clear about what a
logos is, as is made clear in the final section of Socrates’ examination of K3, where he
considers three different conceptions of logos and raises problems for each one
(206c–210a). The first proposed interpretation of logos—according to which it is simply
“speech”—is quickly dismissed, since presumably adding speech to true judgment
doesn’t get one anything more than true judgment (Tht. 206ce). The second account
of logos holds that one gives a logos of a thing when one goes through its elements
(Tht. 206e). For example, to give a logos of a wagon is to name the parts it has—for example,
“wheels, axle, body, rails, yoke” (Tht. 207a). Though the Dream theory did not explicitly
state what it is to give a logos, this account articulates what was assumed in the Dream
theory: that the logos of a thing is an enumeration of a thing’s elements. For this reason,
Socrates suggests that “our man”—the one who came up with the Dream theory—would
scoff at one if one gave as the account of a wagon “wheels, axle, body, rails, yoke” (207ab).
He would say that these were no more the elements of a wagon than syllables are of
a name; rather, a proper account of a wagon would refer to the wagon’s “hundred
­timbers”—that is, the many individual elements out of which it is made.
This way of conceiving of a logos is clearly inadequate. For example, no mention is
made of the importance of structure or arrangement or the relationship between
the parts of a thing for understanding what it is. Presumably any account of the syllable
“SO” should mention not only the letters that make it up but also their order and
arrangement—after all, “SO” is different from “OS.” An account of a wagon should name
not only its parts but also their order and arrangement; a wheelbarrow and a wagon
could conceivably be made up out of the same materials and parts but be distinguished
by different arrangements of those materials.
Plato does not say this; his point is deeper. The objection he has Socrates raise to this
way of thinking of logos shows that a logos cannot consist simply in being able to name
the elements or parts of a thing (207d–208b). Socrates objects that someone might be
able to go through Theaetetus’ name, spelling it correctly and giving a correct account of
all the letters making it up. But that person might at the same time make a mistake about
the first syllable of Theodorus’ name, spelling it “TE” instead of “THE.” And similarly, he
might make a mistake about the syllable “AI” when he finds it in another word, though
he spelled it correctly in the name “THEAITETOS.” Such a person does not know how
to spell Theaetetus’ name.
The point, then, is that knowledge of something does not simply consist of enumerating
the elements of a thing; one must also be able to recognize those elements as such when
 The Theaetetus   279

they occur elsewhere. The mistake the novice speller makes when he spells Theaetetus’
name correctly, but misspells Theodorus’, shows that the speller doesn’t have a system-
atic grasp of the rules of spelling and of phonetics. Indeed, Plato uses the same example
of spelling and letters in the Philebus to illustrate the methods of collection and division
(Phlb. 18bd). The novice speller is unable to “collect” the letters correctly: he does not
recognize letters and syllables as of the same kind when they are found in different
words, as when letters and syllables have been combined in different ways.35 This objec-
tion then points toward a different conception of what it is to have a logos: it is to have the
capacity to recognize the parts of an individual thing (e.g., letters in a particular word) as
its elements, where those elements can only be identified in terms of a larger, interre-
lated system characterizing an entire field or genus, one which can explain, for example,
how all the words of a language should be spelled. Such a grasp of an entire field or genus
is the province of the expert.36
For option A—according to which we are supposed to give up the requirement of
logos for knowledge because some things are known by means of a logos of their ele-
ments, whereas the elements themselves are known, but not by means of a logos—this
comes as a welcome amendment to the conception of logos at work in the definition of
knowledge as true judgment with a logos. That is, we are to understand that giving the
logos of a thing in terms of a thing’s elements does not simply consist of naming its parts.
And it will insist that the elements themselves do not have logoi and are known in a dif-
ferent way. Option B—according to which we should retain K3 as the definition of
knowledge, and reject AK and AL—can also admit this amendment to the conception of
logos at work in the definition. For the objection helps to make the point that having a
logos of a thing should not be thought of as simply being in possession of a list of the
parts of a thing but, rather, as having the ability to locate and recognize the relevant ele-
ments for a thing, an ability that would require one to relate that thing to other things of
the same kind.
Finally, Socrates considers a third conception of logos according to which having a
logos consists of “being able to state some mark by which the thing one is asked for dif-
fers from everything else” (208c). For example, the sun is “the brightest of the heavenly
bodies that go round the earth” (208d). An account must, then, “get hold of the differen-
tiation of anything, by which it differs from everything else, whereas as long as you grasp
something common, your account will be about those things to which the common
quality belongs” (208d).
Socrates raises the following difficulty for the definition of knowledge that results
with this meaning of “logos” (208e–210a): it would seem that even in order to judge

35 Cf. Burnyeat, “Dream”; Burnyeat, Theaetetus; Fine, “Knowledge”; Menn, “Collecting.”

36 The importance of this point in reading this part of the Theaetetus is emphasized in Burnyeat,
“Dream”; Julia Annas, “Knowledge and Language: The Theaetetus and the Cratylus,” in Malcolm
Schofield and Martha Craven Nussbaum (eds.), Language and Logos: Studies in Ancient Greek
Philosophy Presented to G. E. L. Owen (Cambridge: Cambridge University Press, 1982), 95–114; and
Alexander Nehamas, “Episteme and Logos in Plato’s Later Thought,” Archiv für Geschichte der
Philosophie 66 (1984), 11–36.
280   Mi-Kyoung Lee

correctly about Theaetetus, one has to have in mind and grasp Theaetetus as he is different
from everyone else. After all, if one has in mind those features that Theaetetus shares
with anyone else, then one will no more be thinking of Theaetetus than anyone else.
Hence, even in correct judgment about a thing, one must already grasp the features that
distinguish it from everything else. But then this seems to render “with an account”
empty; one will not have added anything to true judgment when one adds an account
to it. Thus, adding an account of how something differs from something else to a
­correct judgment will not add anything informative to what was already contained in
the judgment itself.
One might insist that adding an account consists of getting to know rather than
­judging the differentness. But this will not help, because adding “knowledge of the dif-
ferentness” to a true judgment would simply make the definition of knowledge circular:
knowledge is true judgment about a thing plus knowledge of how it differs from every-
thing else (209e–210a).
What are we supposed to make of this conception of logos, as well as of Socrates’
reasons for rejecting the resulting definition of knowledge? On the one hand, one might
think that there are reasons for regarding it with some suspicion. Socrates’ marking
something as “what most people would say” (208c) is never a recommendation in favor
of the proposal.37 Furthermore, this “popular” conception of knowledge seems to
assume that knowledge is always of unique individual objects, like the sun, Theaetetus,
and so on, and that knowing a thing means being able to say how it differs from other
unique individuals. For this reason, Socrates slips comfortably back into the language of
the wax block when he discusses this proposal. He says that one won’t have Theaetetus in
one’s judgment “until precisely that snubness [of Theaetetus’] has imprinted and depos-
ited in me a memory trace different from those of the other snubnesses I’ve seen, and
similarly with the other things you’re composed of. Then if I meet you tomorrow, that
snubness will remind me and make me judge correctly about you” (209c). What allows
one to judge that this is Theaetetus and not someone else is the fact that Theaetetus’
unique individual qualities—for example, the particular snubness of his nose—have
been imprinted in Socrates’ memory, his wax block, so that on future occasions, he is
able to make a correct judgment by matching the person he perceives with an imprint
possessing exactly those features that he happens to have.
On the other hand, one might think that the problem lies not in the idea itself but in
its application. One might argue, for example, that the idea that to give an account of a
thing is to give the sêmeion or distinguishing mark of it is very close to what Plato says in
other dialogues38 but that the way this interpretation is applied in the Theaetetus is prob-
lematic and the cause of the difficulty. For it is assumed that the things for which we are
to give sêmeia are particular individual objects; thus to know Theaetetus is to be able to
recognize him and distinguish him from everyone else. And here, giving the sêmeion
does seem to be something one who has a true judgment about Theaetetus should
already be capable of doing, if they are judging correctly about him as opposed to

37 Burnyeat, Theaetetus, 191, draws attention to this remark.

38 Fine, Knowledge, 228; Sedley, Midwife.
 The Theaetetus   281

someone else. That is, if one is thinking about Theaetetus, as opposed to someone else,
one must be picking him out of the crowd not by means of general features that
Theaetetus shares with others but by means of particular features that are unique to him.
However, if one gives up the assumption that the objects of knowledge are unique
individual objects, there is no reason to think that one would be capable of giving
the sêmeion if one had a true judgment about a kind of thing. For example, one might be
able to recognize and make true judgments about zebras without being able to give the
sêmeion of zebras—to say how zebras are different from other species. After all, that is
the special province of the expert in biology. In other words, the sêmeion of a particular
individual object serves to distinguish that object from other objects—in particular,
objects of the same species. Hence, what one looks for to distinguish Socrates from
Theaetetus is, for example, the particular bend of his nose, or the color of his skin, or the
height and weight of the individual—or some combination of these features. But the
sêmeion of a kind of thing—of human beings, of justice, or of beauty—serves to distin-
guish it from other kinds of things. (Think, for example, of Aristotle’s conception of defi-
nition: to give a definition of a thing, you have to give the genus plus its differentiae.39
The differentiae are obviously not the sêmeion or the distinguishing mark of an individ-
ual particular concrete object but, rather, what distinguishes a species from other things
belonging to the same genus.) Thus, the third conception of logos is only vulnerable to
Socrates’ objection if one assumes that what we have knowledge of are particular things
such as Socrates or Theaetetus. If one focuses on knowledge of kinds, there is no reason
to think that being able to judge truly about a kind carries with it the ability to give any
kind of account of what distinguishes that kind from others; for example, even if I cor-
rectly judge my neighbor’s tree to be an oak, there is no reason to think I can also give an
account of what distinguishes oak trees from all the other kinds of trees that there are.
Hence, the definition of knowledge as true judgment with an account of the distinguish-
ing mark remains a promising contender as an account of knowledge.
Although Plato gives no indication here in the Theaetetus that this is what he has
in mind, it is consistent with his interest in genus-species hierarchies in other late
­dialogues.40 And if we apply the third conception of logos in this way to kinds, rather than
particulars, then it is also consistent with Plato’s claim in the Republic that all knowledge
requires a grasp of the Forms; here the point of the definition would be that understand-
ing something requires that one be able to give an account of it in terms of what it is to be
that kind of thing, which, in turn, requires one to relate it to and distinguish it from
other kinds that belong to the same genus. Again, nothing here requires a commitment
to Plato’s theory of Forms specifically. But it does suggest—along with other passages in
the dialogue—that a metaphysics consisting entirely of particulars, with no room for
kinds of things to which these particulars belong, would have less chance of success in
sustaining what would otherwise seem to be a promising definition of knowledge,
namely (K3).

39 Topics VI 4. 141b26: “A correct definition must be given through the genus and the differentiae,
and these are better known without qualification and prior to the species.”
40 Taylor, chapter 18 in this volume.
282   Mi-Kyoung Lee

4. Conclusion

Plato examines three definitions of knowledge in the Theaetetus using a variety of methods.
One feature of his strategy is to examine a particular definition of knowledge using
two different methods—one of which explores the definition by offering metaphysical
premises in support of that definition, the other of which explores and refutes the defini-
tion on its own terms, without any such metaphysical commitments. Thus, for example,
Plato tests and examines the thesis that knowledge is perception both on its own
(at 184–86) and in conjunction with a number of metaphysical theses, including the thesis
that everything is always changing and the thesis that everything is always characterized
by opposites (Tht. 152c–183c). Similarly, Plato tests and examines one aspect of the defi-
nition of knowledge as true judgment with an account both on its own (206ac) and in
conjunction with a number of metaphysical theses, contained in the Dream theory—in
particular, the thesis that a thing is nothing more than the sum of its parts (203c–205e).
It is beyond the scope of this chapter to try to determine what Plato intended us to
conclude (Taylor discusses some options in Chapter 18). Instead, I conclude with a more
limited point: Plato introduces various metaphysical theses in order to provide support
for a definition of knowledge. But they are not necessarily ones that he himself would
endorse. Plato’s purpose in assuming metaphysical premises that are incompatible with
the Forms is analogous to the role of the hedonist hypothesis in the Protagoras
(353c–354e). There, Socrates assumes the truth of hedonism—that the good is the same
as pleasure—in order to show that there is no such thing as being overcome by pleasure.
He puts this forward as a working assumption, on behalf of the ordinary folks whom he
and Protagoras are addressing (353a, 354b), which will help Socrates to show that no one
is ever overcome by pleasure: in particular, that reason cannot be outweighed by
­pleasure. In my view, neither the character Socrates nor Plato endorses the hedonist
hypothesis.41 Rather, Socrates’ purpose in introducing the hedonist hypothesis seems to
be to convince those people who are already committed to hedonism—which would
presumably include most readers and, perhaps, most people in Socrates’ audience—
that it’s never the case that anyone is overcome by pleasure. But even if one is not com-
mitted to the hedonist hypothesis, one can see that Socrates could in principle offer
another argument along the same lines that doesn’t depend on that assumption—as
indeed the Stoics would much later.
Similarly, in the Theaetetus, Plato repeatedly introduces metaphysical doctrines
he does not himself endorse on behalf of epistemological theses he wishes to explore.
Thus, he introduces the flux doctrine in order to flesh out a picture of a world in which
knowledge is the same as perception and in which all appearances are true. As it turns out,

41 This is not uncontroversial; some think that Plato accepted hedonism at the time of the Protagoras
(see, e.g., T. Irwin, Plato’s Ethics (Oxford: Oxford University Press, 1995), 85–92).
 The Theaetetus   283

the flux doctrine implies both that these are true and that they are not true. Hence, the
conclusion is, minimally, that the flux doctrine cannot provide support for Theaetetus’
definition of knowledge as perception after all. More robustly, Socrates and his inter-
locutors conclude that the flux doctrine is incompatible with any kind of knowledge.
This hardly constitutes a proof for the existence of the Forms. But it does suggest that
there need to be limits on the extent of flux—such as are provided by Plato’s theory
about Forms.
In the case of the Dream theory, Plato assumes the thesis that a thing is the same as the
sum of its parts (WP) on behalf of the definition of knowledge as true judgment with an
account. Plato goes on to show that the thesis of asymmetry in knowledge is untenable.
But he also argues that the thesis of asymmetry is untenable even if one does not assume
WP. And he then goes on to show what is wrong with conceiving of a logos as a list of the
parts of a thing, which, in turn, suggests that we shouldn’t conceive of things as being
nothing more than the sum of their parts. Again, this hardly constitutes a proof for the
existence of the Forms. But it does suggest the shortcomings in theories that locate
knowledge in one’s grasp of the parts of a particular thing—rather than in the under-
standing of what it is to be that kind of thing, an understanding involving a grasp of the
systematic relations it has with other kinds of things, such as can be found in the concep-
tion of knowledge advocated in other Platonic dialogues.
As we have seen, Plato does not offer anything quite as straightforward as an a­ rgument
that knowledge is impossible without the Forms. But three aspects of the metaphysical
theories he introduces on behalf of the various definitions of knowledge he considers in the
Theaetetus prove to be problematic. First, the theses of flux and opposition—according
to which everything is changing, and everything is F and not-F—are ultimately deemed to
be incompatible with the first definition of knowledge and indeed with any conception
of knowledge. In particular, these problems result if whiteness is always “flowing” or
coming to be not white, and if perceiving is always coming to be not perceiving—that is,
if the nature of things is subject to change. Acknowledging that some things, such as
the nature of things, cannot change may not yet commit one to the theory of Forms, but
it certainly resembles the claim Plato often makes elsewhere that the Forms do not
admit of their opposites. Second, Socrates tends in the dialogue to suppose that what is
known are sensible particulars rather than kinds to which sensible particulars belong.
Expanding the range of possible objects of knowledge to include kinds does not, of
course, commit one to the theory of Forms. But it takes one in a direction that is more
congenial to the theory of Forms than to an ontology exclusively composed of material
particulars. Finally, Socrates adopts the viewpoint in the final section of the dialogue
that a thing is nothing other than the sum of its parts and, therefore, that to say what a
thing is is to say what it is made out of. His rejection of this kind of reductionism again
does not commit him to the theory of Forms. But it does suggest that understanding
what a thing is depends not on finding out what a thing is made out of but on finding out
how it relates, in a system, to other kinds of things—an idea that Plato goes on to explore
in other late dialogues.
284   Mi-Kyoung Lee

Bibliography
Adalıer, Gökhan. “The Case of ‘Theaetetus,’ ” Phronesis 46/1 (2001), 1–37.
Adalıer, Gökhan. “Materialism in Plato’s Theaetetus” (PhD dissertation, Duke University,
1999).
Annas, Julia. “Knowledge and Language: The Theaetetus and the Cratylus,” in Malcolm
Schofield and Martha Craven Nussbaum (eds.), Language and Logos: Studies in Ancient
Greek Philosophy Presented to G. E. L. Owen (Cambridge: Cambridge University Press,
1982), 95–114.
Barnes, Jonathan. “Socrates and the Jury: Paradoxes in Plato’s Distinction between Knowledge
and True Belief (Part 2),” Proceedings of the Aristotelian Society 54(suppl.) (1980), 193–206.
Bostock, David. Plato’s Theaetetus (Oxford: Clarendon Press, 1988).
Broadie, Sarah. 2016. “The Knowledge Unacknowledged in the Theaetetus,” Oxford Studies in
Ancient Philosophy 51: 87–117.
Burnyeat, Myles F. “Idealism and Greek Philosophy: What Descartes Saw and Berkeley
Missed,” Philosophical Review 91/1 (1982), 3–40.
Burnyeat, Myles F. “The Material and Sources of Plato’s Dream,” Phronesis 15 (1970), 101–22.
Burnyeat, Myles F. “Plato on the Grammar of Perceiving,” Classical Quarterly 26 (1976), 29–51.
Burnyeat, Myles F. “Protagoras and Self-Refutation in Plato’s Theaetetus,” Philosophical Review
85 (1976), 172–95; reprinted in Stephen Everson (ed.), Epistemology (Cambridge: Cambridge
University Press, 1990), 39–59.
Burnyeat, Myles F. “Socrates and the Jury: Paradoxes in Plato’s Distinction between Knowledge
and True Belief (Part 1),” Proceedings of the Aristotelian Society 54(suppl.) (1980), 173–91.
Burnyeat, Myles F. “Socratic Midwifery, Platonic Inspiration,” Bulletin of the Institute of
Classical Studies of the University of London 24 (1977), 7–16.
Burnyeat, Myles F. The Theaetetus of Plato, trans. M. J. Levett (Indianapolis, Ind.: Hackett, 1990).
Chappell, Timothy. Reading Plato’s Theaetetus (Indianapolis, Ind.: Hackett, 2005); originally
published, Sankt Augustin, Germany: Academia Verlag 2004.
Cooper, John M. “Plato on Sense-Perception and Knowledge (Theaetetus 184–186),” Phronesis
15 (1970), 123–46.
Cornford, Francis Macdonald. Plato’s Theory of Knowledge: The Theaetetus and the Sophist of
Plato Translated with a Commentary (London: Routledge and Kegan Paul, 1935).
Crombie, I. M. An Examination of Plato’s Doctrines, 2 vols. (London: Routledge and Kegan
Paul, 1962, 1963).
Dancy, R. M. “Theaetetus’ First Baby: Tht. 151e–160e,” Philosophical Topics 15/2 (1987), 61–108.
Fine, Gail. “Conflicting Appearances: Theaetetus 153d–154b,” in C. Gill and M. M. McCabe
(eds.), Form and Argument in Late Plato (Oxford: Oxford University Press, 1996), 105–33.
Fine, Gail. “False Belief in the Theaetetus,” Phronesis 24 (1979), 70–80.
Fine, Gail. “Knowledge and Belief in Republic V,” Archiv für Geschichte der Philosophie 60
(1978), 121–39.
Fine, Gail. “Knowledge and Belief in Republic 5–7,” in Stephen Everson (ed.), Epistemology
(Cambridge: Cambridge University Press, 1990), 85–115.
Fine, Gail. “Knowledge and Logos in the Theaetetus,” Plato on Knowledge and Forms: Selected
Essays (Oxford: Clarendon Press, 2003); originally published in Philosophical Review
88 (1979), 366–97.
Fine, Gail. “Protagorean Relativisms,” in J. Cleary and W. Wians (eds.), Proceedings of the
Boston Area Colloquium in Ancient Philosophy, vol. 19 (Lanham, Md.: University Press of
America, 1996), 211–43.
 The Theaetetus   285

Frede, Michael. “Observations on Perception in Plato’s Later Dialogues,” Essays in Ancient

Philosophy (Minneapolis: University of Minnesota Press, 1987), 3–8.
Gonzalez, Francisco J. “Propositions or Objects? A Critique of Gail Fine on Knowledge and
Belief in Republic V,” Phronesis 41 (1996), 245–75.
Harte, Verity. “Plato’s Problem of Composition,” in John J. Cleary and Gary M. Gurtler (eds.),
Proceedings of the Boston Area Colloquium in Ancient Philosophy 2001 vol. 17 (Leiden: Brill,
2002), 1–26.
Irwin, Terence. Plato’s Ethics (Oxford: Oxford University Press, 1995).
Lee, Mi-Kyoung. Epistemology after Protagoras: Responses to Relativism in Plato, Aristotle, and
Democritus (Oxford: Clarendon Press, 2005).
Lewis, Frank. “Two Paradoxes in the Theaetetus,” in J. M. E. Moravcsik (ed.), Patterns in Plato’s
Thought (Dordrecht: D. Reidel, 1973), 123–49.
McDowell, John. Plato: Theaetetus (Oxford: Clarendon Press, 1973).
Menn, Stephen. “Collecting the Letters,” Phronesis 43/4 (1998), 291–305.
Nehamas, Alexander. “Episteme and Logos in Plato’s Later Thought,” Archiv für Geschichte der
Philosophie 66 (1984), 11–36.
Robinson, Richard. “Forms and Error in Plato’s Theaetetus,” Essays in Greek Philosophy
(Oxford: Clarendon Press, 1969), 39–73; originally published in Philosophical Review 59
(1950), 3–30.
Sayre, K. Plato’s Analytic Method (Chicago: University of Chicago Press, 1969).
Sedley, David. The Midwife of Platonism: Text and Subtext in Plato’s Theaetetus (Oxford:
Clarendon Press, 2004).
Sedley, David. “Three Platonist Interpretations of the Theaetetus,” in C. Gill and M. M. McCabe
(eds.), Form and Argument in Late Plato (Oxford: Oxford University Press, 1996), 79–103.
Taylor, A. E. Plato, the Man and His Work (London: Methuen, 1926).
Thaler, Naly. “Taking the Syllable Apart: The Theaetetus on Elements and Knowledge,” Oxford
Studies in Ancient Philosophy 41 (2012): 201–28.
Waterlow, Sarah. “Protagoras and Inconsistency,” Archiv für Geschichte der Philosophie 59
(1977), 29–33.
White, Nicholas. Plato on Knowledge and Reality (Indianapolis, Ind.: Hackett, 1976).
Williams, C. F. J. “Referential Opacity and False Belief in the Theaetetus,” Philosophical
Quarterly 22 (1972), 289–302.
 chapter 12

The Ti m a eus on th e
Pr i ncipl es of
Cosmol ogy

Thomas Kjeller Johansen

1. Principled Knowledge

It is a common view in Greek philosophy that in order to master a body of knowledge

one needs to understand its principles (arkhai). The principles are the basic propositions
that explain the other propositions in the body of knowledge while they themselves
are not similarly explained by any of the other propositions. Or, if one thinks of the
body of knowledge in terms of the entities that are known, then the principles are
those basic entities that are responsible for the being or coming into being of the other
entities, while their own being or becoming is not similarly dependent on any of the
other entities.1
If one also holds that there are different bodies of knowledge, it is natural to think
that those bodies of knowledge also have different principles. Indeed one may think
that part of what makes those bodies of knowledge different is that they rest on differ-
ent principles. Aristotle is an exponent of this view: “different sciences have different
principles, for example arithmetic and plane geometry” (De Anima 402a21–22). This is
not to say that different sciences may not also share some principles. As Aristotle makes
clear in the Posterior Analytics, there are some principles that related sciences such as
arithmetic and geometry share, but they are not the ones that define arithmetic as such
or geometry as such.2

1 Cf. Aristotle, Metaphysics V.1 “It is common, then, to all [arkhai] to be the first point from which a
thing either is or comes to be or is known.”
2 Cf. Posterior Analytics I.9–10.
288   Thomas Kjeller Johansen

Plato too seems to hold that grasping a body of knowledge requires a grasp of its
­ rinciples. One example is Republic VI (510b2–9), where Socrates explains the image of
p
the line. He has divided the line into two sections, the intelligible and the perceptible,
each of which he then subdivides. The intelligible section is now divided into the Forms
and the objects of hypotheses such as those made by mathematicians.3 For Socrates
knowledge of what is hypothesized in mathematics requires a grasp of the principle
(arkhê) on which those hypotheses depend. However, mathematicians commonly make
a mistake, not in using hypotheses—Socrates accepts that there are hypotheses appropriate
to each discipline—but in behaving as if they were known as first principles that do not
themselves require further explanation.4 The real first principle, Socrates says, is pro-
vided by the form of the good. He refers to this principle as the principle of everything
(hê tou pantos arkhê, 511b7). The form of the good is the principle of everything in the
first instance because it is the principle of all intelligible beings. But it is also the principle
of everything else insofar as everything else is an image, at one or more removes, of the
intelligible beings. At the highest level, the form of the good is studied through dialectic,
which allows us to ascend systematically to the most basic principle of all.
Later in the Republic (521c–534e) Socrates contrasts those disciplines that pull one
toward the study of being, and ultimately the form of the good, with those that remain
focused on coming-into-being. He presents a new view of how the mathematical disci-
plines are to be practiced. They should be used to lead the mind away from the concern
with perceptible matters to the study of being itself. The mathematical sciences, includ-
ing astronomy, should be studied in preparation for the study of being in dialectic.
Perceptible objects should be considered only to the extent that they provide the intel-
lect with the opportunity, through problems or examples, to account for the forms. The
mathematical sciences provide the first step in the intellect’s upward journey from the
world of becoming to that of being, an ascent that is completed in the grasp of the form
of the good, the last verse, as it were, in “the song of dialectic” (532a1–2).5
The Timaeus, like the Republic, emphasizes the need for us to grasp the proper princi-
ple of our disciplines of study. As Timaeus says in his opening speech, “Now in every
subject it is of utmost importance to begin (arxasthai) according to the natural principle
(arkhê)” (29b2–3).6 But what is the natural principle of cosmology? Timaeus’s cosmol-
ogy concerns, we are told (27a5–6), the coming-into-being of the cosmos, down to and
including the nature of man. So our question becomes: What is the natural principle of

3 It is a matter of dispute whether such hypotheses are restricted to special mathematical objects.
For the debate and arguments against this restriction, see G. Fine, “Knowledge and Belief in Republic
V–VII” [“Knowledge”], in S. Everson (ed.), Epistemology (Cambridge 1990), 85–115, reprinted in Fine,
Plato on Knowledge and Forms. Selected Essays [Plato] (Oxford 2003), 85–116.
4 Cf. M. F. Burnyeat, “Platonism and Mathematics: a prelude to discussion”, in A. Graeser (ed.),
Mathematics and Metaphysics in Aristotle, Bern and Stuttgart 1987, 213–40; reprinted in M. F. Burnyeat,
Explorations in Ancient and Modern Philosophy, vol.2 (Cambridge 2012), 145–72 at 150–52.
5 Cf. M. F. Burnyeat, “Plato on Why Mathematics Is Good for the Soul,” in Proceedings of the British
Academy 103 (2000), 1–81 at 22.
6 A proverb quoted also at Republic 377a12.
 The Timaeus on the Principles of Cosmology   289

the study of the coming-into-being of the cosmos? As we shall see, the principle is a
principle of coming-into-being, not of being.7 Timaeus accepts that there may be more
fundamental principles of everything, but cosmology, as he understands it, does not
provide the appropriate method for approaching such ultimate principles.8 The subject
of cosmology is the world as it has come into being and its method one that is appropri-
ate to this subject matter. Unlike dialectic in the Republic, cosmology, as Timaeus under-
stands it, is not concerned with the principle of absolutely everything.9 Cosmology is
not concerned with being as such, its ultimate principle is not the ultimate principle of
being, and its method is not that of dialectic, in the Republic’s sense.10
It may still turn out that what is ultimately responsible for the coming-into-being of
the cosmos is the same as what is ultimately responsible for all being according to the
Republic, that is, the form of the good. In that case there would be a sense in which
the principle of becoming and the principle of being are the same. However, since in the
Timaeus, as in the Republic, being is categorically different from becoming, what it is for
something to be a principle of becoming is different from what it is for something to be a
principle of being. Being responsible in the manner of a principle for the coming into
being of something serves a different function from being responsible for the being of
something. So even if the same thing answers to both job descriptions, we would still
have to say that this thing qua principle of coming-into-being differs from it qua princi-
ple of being.
The fact that the explanatory subject of the Timaeus is the coming-into-being of the
cosmos is reflected in a different conception of the mathematical sciences from that of
the Republic. Timaeus does not justify the study of astronomy and mathematical cos-
mology by their contribution to the study of being. This is not because he does not
value the study of being, for it is clear that he, like Socrates, places it above the study of
coming into being (29b–c). However, he sees the study of being as distinct from that of
the cosmos. As he says, the study of coming-into-being is meant to provide a break
from the study of being (59c–d). For cosmology, as we shall see, does not aim to raise
questions about being as such; rather it aims to understand coming-into-being. We
are facing a different explanatory task in the Timaeus from that of the astronomers in
the Republic.

7 Cf. 28b6–7, 29e4. 8 Cf. 48c2–d4.

9 Other than in so far as to pan means “the universe,” but that is a different sense of pan from that
of the Republic.
10 I thus agree with S. Broadie’s criticism of the view of cosmology as a “gateway to metaphysics,”
Nature and Divinity in Plato’s Timaeus [Nature] (Cambridge 2012), 61. Contrast C. J. Rowe, “The Status
of the ‘Myth,’ ” in C. Natali and S. Maso (eds.), Plato Physicus [Physicus] (Amsterdam 2003), 21–31 at 30,
who presents cosmology, “in the spirit of the Republic,” as an indirect account “of the truth itself, how
things really are,” where by “how things really are” we should think of the being of the forms. In my
opinion, the aim of cosmology is not directly or indirectly to illuminate the forms but to make coming
into being intelligible to the extent this is possible.
290   Thomas Kjeller Johansen

2. Principles in the Timaeus

My aim in this chapter is to articulate the distinctive principles of cosmology, as Plato

conceives of them in the Timaeus, and so provide the starting point for an understand-
ing of this work. The principles of cosmology emerge in particular from two passages of
the Timaeus, 27d5–30c1 and 47e3–53c3, which set out the assumptions underlying the
study of the cosmos. We therefore need to look at these passages in some detail.
In the first passage Timaeus begins (27d5–28b2) by articulating three premises, as
follows:

1) Being is that which is graspable by intelligence with an account (logos); becoming

that which is graspable by opinion (doxa) with unreasoning (alogos) perception.
2) Everything that comes into being has a cause (aition).
3) When a craftsman uses an eternal model his product is necessarily fine (kalon); if
he uses a generated one, the product is not fine.

Timaeus then (28b2–29b2) applies these premises, serially, to the cosmos.11 First, he
argues that the world belongs to what comes into being rather than what is because it is
perceptible:

T1 One should consider first about the entire heaven (or cosmos, or let’s call it by
whatever other name one might prefer to call it by) a question one should consider
about everything in the beginning (en arkhêi), namely, whether it always was, hav-
ing no beginning (arkhê) of coming-into-being (genesis) or whether it has come into
being, having started from some principle (arkhê). It has come into being: for it is
tangible and has a body, and all such things are perceptible, and perceptible things,
being graspable by opinion with perception, appeared to be coming into being and
generated. (28b2–c1)12

Applying Premise 1, T1 argues that since the world is perceptible it must have come
into being. As part of the argument, Timaeus links having come into being with having
a beginning or principle of coming-into-being. In contrast, what always was has no such
principle. So, if the cosmos has coming-into-being (genesis), it must have a p
­ rinciple. But
what does Timaeus mean by “principle of coming-into-being”? As for coming-into-being,
I take it that the dominant sense in T1 is that of coming-into-existence.13 For Timaeus

11 For a fuller analysis of the argument see T. K. Johansen, “Why the Cosmos Needs a Craftsman:
Plato's Timaeus 27d5–29b”, Phronesis 59 (2014) 297–320.
12 Translations my own unless otherwise noted.
13 The coupling in Premise 1 (28a3) of “coming-into-being” with “being destroyed” (apollumenon)
suggests that Timaeus there also, if not exclusively, is thinking of coming-into-being in the sense of
coming-into-existence. For the interpretation of “genesis” and “gignomenon” in Premise 1, see M. Frede,
“Being and Becoming in Plato,” with A. Code, “Reply to Michael Frede’s ‘Being and Becoming in
Plato,’ ” in Oxford Studies in Ancient Philosophy, Supplementary Volume 1988, 37–52 and 53–60.
 The Timaeus on the Principles of Cosmology   291

assimilates what has come into being with what has been “generated” or “born” (gegonos
with gennêtos) at 28b1 and 28c1–2. We are here primarily talking about genesis in the
sense of generation or coming-into-existence.
The term arkhê, meanwhile, occurs three times in T1. In its first use, the arkhê is
clearly the beginning or starting point of inquiry. Such a starting point can be under-
stood temporally, or simply as the first in an order of items that need to be addressed,
like the first item on a shopping list. In its second use, however, arkhê is linked to
­coming-into-being. Insofar as we think of coming-into-being as a process in time, it is
natural to think that the arkhê also represents a temporal beginning.
However, this temporal reading of the coming-into-being of the world has been dis-
puted since antiquity. Some have taken the temporal description of the cosmogenesis as
a mere didactic device, which serves to set out eternal causal relations as if they were a
series of events in time.14 On this view, the claim that the world “has come into being”
(gegonen) means no more than that it is now and always in a state of having come into
being. The principle of its coming into being is the standing cause of its continuous or
continual coming into being.
For several reasons, I do not think this is a viable reading. First, the temporal under-
standing of the principle seems supported by Timaeus’s use of tenses. What has come
into being (perfect tense), contrasts with what always was, in having started from some
principle.15 Second, Timaeus associates the principle of the world’s coming into being
with its cause (aition) and describes the cause in ways that strongly suggest that it is the
cause of a temporal process:16

T2 Again, we say that it is necessary for what came into being to have come into
being by some cause (aition). It is a big job to find the maker (poiêtês) and father
(patêr) of this universe, and having found it, it is impossible to state it to everyone.
(28c2–5)

14 A reading referred to, though not espoused, by Aristotle, De Caelo 279b32–280a2. See also
Plutarch, On the Generation of the Soul in the Timaeus, 4. For a recent defense of this non-temporal
reading, see M. Baltes, “Γέγονεν (Platon Tim. 28B6). Ist die Welt real entstanden oder nicht?”
[“Γέγονεν”] in K. A. Algra, P.W. van der Horst, and D. T. Runia, (eds.), Polyhistor. Studies in the History
and Historiography of Ancient Philosophy (Leiden 1996), 76–98.
15 The supporters of the non-temporal reading point out that the perfect tense in Greek is used
aspectually to say something about the present rather than temporally to refer to a past action, cf.
F. M. Cornford, Plato’s Cosmology [Cosmology] (London 1937), 24–25 and Baltes, “Γέγονεν”, 91–4. Baltes
writes as if the temporal reading could only be firmly established if Timaeus used a finite form of the
aorist rather than the perfect. However, Laws 781e–782a shows that Baltes is wrong to suggest that we
would expect the aorist rather than the perfect if Timaeus had meant to say that the world came into
being in the past. Note also (with D. Sedley, Creationism and Its Critics in Antiquity [Creationism]
(Berkeley 2007) Critias 106a4, where Timaeus expressly says that the world has came into being a long
time ago, using again the perfect tense (γέγονεν). However, even if Baltes were right about the perfect,
his argument would fail since the imperfect tenses that Timaeus also uses are sufficient to locate the
creation in the past. Sensing this, Baltes dismisses ἦν at 28b6 as “conversational laxness” for what
strictly speaking, he says, is meant as ἐστί. But if ἦν is laxness, the repeated imperfects απηργάζετο and
ἔβλεπεν (29a1,3) would be outright carelessness.
16 Timaeus’s use of arkhê thus illustrates Aristotle’s observation that a standard sense of arkhê is
“cause” (aition), cf. Metaphysics Δ.1, 1013a16–17.
292   Thomas Kjeller Johansen

Premise 2, we are told, applies to the cosmos, as we would expect given that it has come
into being. So the coming-into-being of the cosmos has a cause. T2 leaves us in ­suspense
as to what exactly the cause is. However, Timaeus does tell us something important
about the cause: it is the maker and the father of the universe. Now poiêtês, like the
English “poet,” may have connotations with the making of verse.17 However, as the
Symposium reminds us,18 we should think more generally of anybody who is responsi-
ble for bringing what was not into being as a “poet.” The temporal aspect of poiêsis is
made quite explicit by the Sophist: “we say . . . that every power is productive (poiêtikê)
which becomes responsible for those things that were not earlier (proteron) coming into
being later (husteron).”19 The Visitor’s examples, an animal growing from a seed or a
plant from a root, show that “earlier” and “later” are meant temporally. Meanwhile, the
notion of the cause as “the father of the cosmos” is one that Timaeus uses throughout his
cosmology and reinforces through the language of generation.20 On Timaeus’s own
account of fatherhood,21 the process of fathering is clearly a temporal one, involving
stages of maturation before the coming-into-being is complete.22 The process of sowing
seeds for growth mirrors Timaeus’s account of the actions of the demiurge in creating
human souls: the demiurge creates the immortal souls and sows them in the “instruments
of time,” before handing them over to the created gods to nurture them (41c–d, 42d).23
Procreation, human as divine, is clearly a temporal process. Both as “the father” and
as “the maker,” it seems, then, that the cause initiates the coming-into-being of the
­cosmos in time.
So far, then, I have suggested that the principle of coming-into-being should be
understood both as a causal and temporal principle.24 It is that which initiates the com-
ing-into-being of the cosmos, where the coming-into-being is to be understood as
occurring in time.
The designation of the cause as “maker” is developed next:

T3 But again we need to consider the following point about it [sc. the cosmos]: in
relation to which of the models did the builder complete it, in relation to what is the
same or in relation to what has come into being. If this world is beautiful and its

17 Cf. Aristotle, Poetics 47b13–16, 51b27–29, with S. Halliwell, Aristotle’s Poetics (London 1998), 56–59.
18 Symposium 205b. 19 Sophist 265b9–10, referring back to Sophist 219b4–6.
20 Cf. gennaô (“to generate”) and cognates, e.g., at 28c2, 32c1, 34a7, b9, 37a2, 38b6, c4, e6, 41a5,
55b4–5, 68e4, 69c4.
21 Cf. 91d1–5. 22 Cf. 91d5: ἀποτελέσωσιν γένεσιν. Fatherhood does not end with impregnation.
23 Sowing the seeds is here, as at 91d, both an agricultural and procreative notion.
24 The strongest objection leveled at the temporal reading is that it presupposes that there was time
before and during the creation, but for Timaeus time only arises with the creation of the planets (cf.
37c–38c). The best replies to the objection are those inspired by G. Vlastos, “Creation in the Timaeus: Is
It a Fiction?” in id., Socrates, Plato, and Their Tradition, edited by D. W. Graham (Princeton 1995),
265–79 at 271–75. We should distinguish between time as an ordered and measurable succession of
past, present, and future, and time as a succession of before and after. The creation of the planet
introduces the former notion of time, not the latter, and the latter is sufficient to make sense of the
temporal succession of chaotically moving appearances in the precosmos, cf. Sedley, Creationism,
104–05.
 The Timaeus on the Principles of Cosmology   293

craftsman good, it is clear that he was looking toward the eternal model. But if that
which it is sacrilegious even to say holds, then he was looking toward a model that
has come into being. Surely, it is clear to all that he was looking toward the eternal
one, for the world is the best of the things that have come into being, and the crafts-
man is the best of causes. Having come into being in this way it has been crafted in
relation to what is graspable by reason and wisdom and is the same. (28c5–29b1)

T3 applies Premise 3 to the cosmos. The question is now, given that the world has come
into being (T1) and given that it has a cause, namely, its maker (T2), which of the two
possible models the maker employed for the coming-into-being of the world. The
answer, that the craftsman of the world looked to an eternal paradigm, confirms our
expectations of craftsmanship from other Platonic works. So in Republic X a genuine
craftsman was contrasted with a pseudo-craftsman in looking toward an eternal rather
than a generated paradigm (597d–598b). We would also expect the choice of an eternal
model given that this is the way to make the model fine (Premise 3). For in other dia-
logues Socrates emphasizes that craftsmanship seeks to make its product as good as pos-
sible. So in the Gorgias it is characteristic of a craft such as medicine, as opposed to mere
knacks, that it considers what is better or worse for those it affects (501c). In Republic I,
the craft considers the benefit of that of which it is the craft (342c). We expect, then, that
a craftsman will do what is required to craft the best possible product, including choos-
ing the right model.
So the world is modeled on an eternal paradigm because the craftsman chose this
model, and he chose this model because he is the best craftsman, and as such sought to
make the product as good as possible. Now in terms of identifying the principle or start-
ing point of the creation of the cosmos, this makes the eternal paradigm secondary to
the craftsman. Strictly speaking, the forms as a paradigm are not a principle of the
coming-into-being of the cosmos, for they are not the first factor in the order of explaining
or causing the coming-into-being of the cosmos; rather, the maker is. Alternatively, if we
allow for more or less fundamental principles,25 the formal paradigm is a principle but
only a secondary one. It may after all seem proper to count the forms as a principle of the
cosmos insofar as the cosmos is created by being modeled on them. However, since
there is a further cause of why the cosmos is modeled on the forms (i.e., the maker), the
forms appear only to be a secondary principle.
Note in this context that the paradigm has been selected by the craftsman within the
larger kind of eternal beings. So in Premise 3 Timaeus said that “when the craftsman
looks towards what is always the same and uses some such thing as a paradigm, etc.,” that
is to say, the craftsman looks to the eternal and selects something from that class or kind
as his model.26 When we are told more precisely what the paradigm is we are told that it

25 A parlance licensed, perhaps, by 53d4–7.

26 D. Zeyl, Plato Timaeus (Indianapolis/Cambridge 2000), gets it right: “So whenever the craftsman
looks at what is always changeless, and using a thing of that kind as his model, . . .,”; D. Lee (old Penguin
translation) doesn’t: “Whenever, therefore, the maker of anything keeps his eye on the eternally
unchanging and uses it as his pattern . . ..” (emphases added).
294   Thomas Kjeller Johansen

is a living being or animal (zôion), and more specifically a living being containing all the
other living beings within it (30c–31b). Given the range of forms on offer in other
Platonic dialogues, it seems unlikely that the paradigm living being includes all the
forms there are. The formal paradigm is then a specific instance of the wider kind of
eternal beings, chosen by the maker among the various eternal beings for its suitability
as a paradigm for making the best possible world.
So far I have argued that Timaeus’s opening speech is concerned with the identifica-
tion of the first principle (arkhê) of cosmology. Since the cosmos has come into being,
the principle is a principle of coming-into-being, rather than a principle of being. The
principle of coming-into-being is implicitly identified with the cause (aition) of coming-
into being. This cause is variously designated as “the maker” and “the father.” Both terms
seem to confirm that the principle of coming-into-being should be understood not just
as a causal one, but also as a temporal one, as a cause that initiates a process of coming-­
into-being in time. As a genuine craftsman, moreover, the principle is an intelligent
one seeking to make the best work possible, and therefore working according to an
­eternal model.

3. The Standards of Cosmological

Argument

In the next passage, however, Timaeus speaks of another arkhê:

T4 These things being so, again there is every necessity for this cosmos here to be a
likeness of something.27 Now in every subject it is of utmost importance to begin
according to the natural principle (arkhê), and so, on the subject of a likeness and its
model, we need to make the following distinction. The accounts (logoi) are of the
same kind as the very things of which they are interpreters. So the accounts of that
which is stable and certain and transparent to rationality (nous) are stable and
unchanging—insofar as it belongs to accounts to be irrefutable and invincible, they
should not fall short of this— while the accounts of that which is made as a likeness
by reference to that thing [sc. what is stable, etc.], [these accounts] being of what is
a likeness are themselves likely and stand in an analogy to those accounts [sc. the
accounts of what is stable], namely, as being stands to coming-into-being, so truth
stands to conviction (pistis).28 (29b1–c3)

27 “These things being so, etc.” introduces a new paragraph, pace the punctuation of the Oxford
Classical Text; cf. Cornford Cosmology, 23, n.1.
28 Pistis refers here not to the subject state of being convinced but the object state of the account’s
being convincing or probative, cf. the similar use of the term in Parmenides B l, line 30 with the
interpretation of J. Bryan, Likeness and Likelihood in the Presocratics and Plato (Cambridge 2012), 92.
 The Timaeus on the Principles of Cosmology   295

Now the second sentence of T4 may seem to suggest that Timaeus will only now tell us
the first principle of the cosmos.29 However, Timaeus is not saying that he is now going
to introduce the principle of the coming-into-being of the cosmos. Rather he is making
a general claim, that on any subject matter we need to start with the natural principle, a
claim that he then illustrates by the particular subject matter of a likeness and its model.
While confirming, then, Timaeus’s general concern with identifying the proper princi-
ple of any discipline, T4 does not offer us a new principle of coming-into-being, but
rather the principle for the particular subject matter of a likeness and its paradigm.
The principle is that “the accounts (logoi) are of the same kind as the very things of
which they are interpreters.” From this principle, it follows that if an account is of being,
and being is “stable and certain and transparent to rationality,” the account should have
the same or similar attributes; whereas if the account is of a likeness then the account
should be likely. We should note the normativity of the principle, made explicit at
29b9–c1. For accounts of being, “insofar as it belongs to accounts to be irrefutable and
invincible, should not fall short of this.”30 Similarly, likelihood for accounts of a likeness
is both the maximum and the minimum to which they should aspire.
T4 is emphatic that cosmology is not to be held to the standards of the study of being,
as exemplified by dialectic in the Republic. The lower argumentative standards are a
direct consequence of the subject matter of cosmology, a world that has come into
being, albeit as a likeness of being.31 In a later passage, Timaeus makes reference to this
­standard, the likely account, to explain why it is inappropriate through cosmological
argument to seek the most basic principles of everything:

T5 We tend to posit them [sc. earth, water, fire and air] as the elemental “letters” of
the universe and tell people that they are its “principles” on the assumption that they
know what fire and the other three are. In fact, however, they shouldn’t even be
compared to syllables. A person with even a modicum of wisdom would not make
such a comparison. So let me now proceed with my treatment in the following way:
I cannot state “the principle” or “principles” of all things, or however else I think
about them, for the simple reason that it is difficult to show clearly what my view is
if I follow my present manner of exposition. (Tim. 48b7–c6, transl. after Zeyl)

Timaeus is clear that the “present manner of exposition” means arguments based on
likelihood (48d2–3). Such arguments are unsuited to demonstrating the principle of all
things. I take the contrast to the Republic’s notion of dialectic as concerned with ascend-
ing to the principle of everything to be deliberate. The alternative offered by dialectic, to
ascend to the principle of everything, whereby our grasp of the four bodies might be
imagined to turn into the highest form of knowledge, is not an option in the Timaeus,

29 So, Proclus tells us (in Tim. 337.20–23), the sentence was read by some in antiquity as referring
forward to the “most proper principle” at 29e4.
30 Cf. M. F. Burnyeat, “ΕΙΚΩΣ ΜΥΘΟΣ”[“ΕΙΚΩΣ”], Rhizai 2, no.2 (2005), 143–66 at 150–52.
31 For a detailed reading of the passage see Burnyeat, “ΕΙΚΩΣ” My own interpretation is set out in
Johansen (Natural), 48–64.
296   Thomas Kjeller Johansen

since we are methodologically bound to the likely account by our subject matter, the
coming into being of the cosmos. Our assumptions about the coming into being of
­bodies therefore have a lower cognitive status than if they were explained in relation to
the principle of everything.
As an illustration, consider Timaeus’s description of the geometrical principle of the
construction of the simple bodies:

We posit (hupotithemetha) this principle (arkhê) of fire and the other bodies advanc-
ing according to the likely account with necessity.32 But god and of men he who is a
friend to god know the principles still higher than these. (53d4–7)

Since Timaeus is concerned with the principles lying behind the geometrical construc-
tion of the simple bodies, it is likely that the higher principles would be, at least in the
first instance, mathematical. These principles may in turn, if the image of the line still
applies, rely on further principles, including ultimately the form of the good. However,
Timaeus thinks it is justified to leave these higher principles aside in the current context
because they do not enter directly into the account of the construction of the simple
bodies. From the point of view of cosmology, this is as it should be: our interest in prin-
ciples only takes us as far as the principles of what comes into being in this world.33 That
there are higher principles than that, principles such as the Republic’s principle of every-
thing, is assumed, but these do not relate directly to the generation of the world as a like-
ness and are, therefore, appropriately not dealt with by the method of the likely account.

4. The Most Proper Principle

of Coming-into-Being

T4 concludes what Socrates refers to as the “prelude” (prooimion, 29d5). The prelude set
out certain basic distinctions that have allowed us to outline the nature of the subject
matter and the sorts of accounts we should expect to be given of it. Timaeus now takes
up the account of the creation of the cosmos proper. But first he offers a more precise
description of the principle of coming-into-being and the cosmos:

T6 Now let us say for what reason (di’ hêntina aitian) the constructer constructed
coming-into-being and this universe. He was good, and nobody good ever has any

32 Cf. also hupotithêmi at 48e6, 55e4, 61d3.

33 It is a matter of debate whether, according to the image of the line, the cognitive states in principle
are confined to their corresponding objects or whether, as Gail Fine “Knowledge” argues, it is possible,
for example, to have understanding (noêsis) of the objects of mathematical hypotheses or thought
(dianoia) of the forms. Timaeus 53d, indirectly at least, appears to lend support to Fine’s position. For
Timaeus implies that if the mathematical principles used in the construction of the simple bodies were
related to the higher principles they would have a stronger cognitive status than the likelihood they
have on our current account.
 The Timaeus on the Principles of Cosmology   297

envy (phthonos) about anything. Being without envy, he wanted all things to become
as similar to himself as possible. This indeed is in the most proper sense the princi-
ple of coming-into-being and the cosmos (geneseôs kai kosmou . . .arkhên kuriôtatên)
which someone would be most correct in accepting from wise men. (29d7–30a2)

T6 reads at first as a corrective. In T1-2 Timaeus gave the impression that the principle of
coming-into-being was the cause (aition) of the cosmos. But now we are told that the
most proper (kuriôtaten) principle of coming-into-being answers not to the cause
(aition) of coming-into-being, the maker, but rather to something referred to as the aitia
of coming-into-being. However, since Timaeus raises the question of the aitia as a new
and distinct question, it is clear that he does not think that the question of the aitia has
been answered by the identification of the aition in T2. The aition of the coming-into-
being of the cosmos was the maker and father, but the aitia, we are now told, is his desire
to make the world as similar to him as possible.34
“Most properly” (kuriôtatên) here suggests that we were not wrong to think that the
aition was the principle, but that it is not the most proper or precise way of under-
standing the principle. We can see why that would be so. Being told that somebody
made the universe still leaves a question as to why or in what capacity he made the
universe. The aitia states what exactly it was about the aition that made it produce the
universe. It is the aitia, then, which most properly is the principle of the coming-into-
being, not the aition.
One way of thinking about the relationship between the aition and the aitia is to see it
as involving the difference between what after Aristotle we call the “efficient” and the
“final cause.”35 The efficient cause initiates the change,36 while the final cause is the goal
toward which the change is aimed. Aristotle holds that, where both are at work, efficient
causes are generally, that is, both in art and nature, secondary to final causes, since by
identifying the goal we can understand why the efficient cause is at work.37 With the end
or purpose of a building in mind we can understand why the builder initiates the series
of changes leading to a building. Similarly, we may think that when Timaeus in T2 refers
to the father and maker of the cosmos he has in mind that which initiates the
­coming-into-being of the world—the efficient cause, in Aristotle’s money—but when he

34 Timaeus seems to be observing a distinction between aition and aitia, which can be traced in
other Platonic dialogues, cf. M. Frede, “The Original Notion of Cause,”, in id. Essays in Ancient
Philosophy (Oxford 1987), 129. For a clear statement of the distinction in the Phaedo, see J. Lennox,
“Plato’s Unnatural Teleology” in id., Aristotle’s Philosophy of Biology. Studies in the Origins of Life
Sciences (Cambridge 2001), 280–302 at 283; first published in D. O’Meara (ed.), Platonic Investigations
(Washington D.C. 1985), 195–218. I. Mueller, “Platonism and the Study of Nature”, in J. Gentzler (ed.),
Method in Ancient Philosophy (Oxford 1998), 67–90 at 85–86, rightly finds the distinction also in the
Timaeus. The uses of aitios at 61b6, 63e8, 76c6, 80a1, 87e5 (cf. sunaitios 46d1, 76d6, and summetaitios,
46e6) can all reasonably be taken to refer to the thing responsible for the causation.
35 Cf. Proclus, in Tim. 335.28–357.23, suggests that this was a standard way of reading the text in
antiquity.
36 Cf. e.g. Physics 194b29-32, Posterior Analytics 94b7, Generation of Animals 778b1, De Anima
417b20.
37 Cf. Parts of Animals I.1 639b-40a.
298   Thomas Kjeller Johansen

refers to the aitia he has in mind that for the sake of which the world was made, the “final
cause.”38 Moreover, when Timaeus ascribes priority to the aitia over the aition, calling
the aitia the “most proper” principle, he would, again like Aristotle, be ascribing priority
to the final cause over the efficient. The reason he ascribes priority to the aitia would be
just that it explains why the aition is at work: his desire to make the world as good as
­possible (the aitia) tells us why the god (the aition) is moving the world in the first place.
So we need to understand the aitia before we can understand why the maker is at work.
Yet this account of the relationship between aitia and aition needs to be modified. We
cannot take the aitia simply to be the goal of the maker. For the aitia is not simply the
good that the god aimed at but his desire to bring about this good, and on Aristotle’s
plausible analysis of the causation of action, the desire counts as the efficient cause. So it
might be said that the aitia, while it mentions the goal, is in effect a restatement of the
efficient cause. So, the aitia states the respect in which the aition works as an efficient
cause, by desiring a certain outcome for the world. David Sedley, denying that Plato
in the Timaeus has the concept of a final cause, thus speaks of the god’s intelligence as
“a goal-directed efficient cause.”39
It is clearly right that in stating the aitia, the goal cannot be detached from the god’s
desire. However, that does not mean that we should reduce the goal to an aspect of the
efficient cause. This would be to ignore the explanatory priority that the goal has in
explaining the way the efficient cause works. It is after all because it appears good to the
maker that the world should be thus and so that he desires it to be in this way. There is a
sense, then, in which it is the good as it appears to the god that is the cause of his desire
for the world to be so. To be sure, the apparent good works here not as another efficient
cause of the god’s desire, but rather as a final cause. That is to say, the goal works as a
cause by informing and directing the god’s desire. But if the goal, in this way, has causal
priority over the god’s desire, it cannot be right simply to see it as an aspect of this desire.
Rather than reducing the god’s goal to an aspect of the efficient cause, it seems better to
speak, as does Gail Fine, of the god’s desire to make the best possible world as an efficient
cause with a final cause constituent.40
We should note Timaeus’s exact formulation of this final cause constituent. He says
not simply that the maker wanted to make the universe as good as possible, but that he

38 The father is of course Aristotle’s standard example of the efficient cause, cf. Phys. 194b30.
39 Sedley, Creationism, 114, n.47.
40 G. Fine, “Forms as Causes: Plato and Aristotle”, in A. Graeser (ed.), Mathematics and Metaphysics
in Aristotle (Bern 1987), 69–112, reprinted in Fine, Plato, 350–96 at 375. On this interpretation of the
Timaeus, Aristotle’s refusal in Metaphysics Α.7 (988b6-8) to recognize final causes in the Timaeus seems
understandable if uncharitable: understandable, because Aristotle would want to keep efficient and
final causes distinct in natural causation (cf. On Generation and Corruption I.7 324b13–18);
uncharitable, because Aristotle presumably himself accepts that in cases of conscious agency final
causes are embedded in efficient causes; cf. De Anima III.10, in particular 433a15–20. See further
R. Bolton, “The Origins of Aristotle’s Natural Teleology in Physics II” in M. Leunissen (ed.), Aristotle’s
Physics: A Critical Guide, Cambridge 2015, 107–20 and T. K. Johansen, “Should Aristotle Have
Recognised Final Causation in Plato’s Timaeus?”, in R. Mohr and B. Sattler (eds.), Plato’s Timaeus Today
(Parmenides 2010), 179–200.
 The Timaeus on the Principles of Cosmology   299

wanted to make the world as like himself as possible. Of course, Timaeus has just
emphasized that the maker is good, so saying that he wants to make the world like him-
self implies that he would also thereby make it as good as possible. However, it might
reasonably be thought that goodness can come in different varieties, so it is not obvious
that by wanting to make the world as good as possible the demiurge should also want to
do so in the way that makes the universe as like himself as possible. In fact, it does seem
that the goodness that the god as an eternal, intelligible being (37a1) enjoys is not quite
the same as the goodness enjoyed by the cosmos as a perceptible, created being. That the
world is made good by instantiating somewhat different properties from eternal being is
clear, for example, from the creation of time (37c–38b). The god created time as a mov-
ing likeness of eternity, where eternity is a simple unity, but time has parts, that is, past,
present, future. The goodness of the world does not lie in its being a mere carbon copy of
its eternal paradigm; rather it lies in its having properties analogous to those of its eter-
nal paradigm, but also appropriate to coming-into-being.
Consider now the god’s first actions:

T10 ‘For god wanted for everything to be good, and nothing, if possible, to be bad,
and so when he took over everything that was visible in a state of unrest, moving
discordantly and without order, he brought it into order from disorder, believing
that order was in every way better than disorder. Now it wasn’t permitted (nor is it
now) that one who is supremely good should do anything but what is best.
Accordingly, the god reasoned and concluded that in the realm of things naturally
visible no unintelligent thing could as a whole be better than anything which does
possess intelligence as a whole, and he further concluded that it is impossible for
anything to come to possess intelligence apart from soul. Guided by this reasoning,
he put intelligence in soul, and soul in body, and so he constructed the universe. He
wanted to produce a piece of work that would be as excellent and supreme as its
nature would allow. This, then in keeping with our likely account, is how we must
say divine providence brought our world into being as a truly living being, endowed
with soul and intelligence.’ (30a6–c1, Zeyl transl.)

It was already implied in T3 that the maker was characterized by intelligence (noêsis).
For he made the world by looking to an eternal paradigm that, as Premise 1 told us, is
graspable by intelligence. Elsewhere Timaeus makes it clear that the maker is or has
intelligence (nous).41 By investing the world with nous, as his first task, the god has
therefore not just made the world as good as possible; he has also made it as good as
­possible in the way that maximizes its likeness to himself.42

41 Cf. 39e7–9 and 48a2. If god is nous is that compatible with god’s having desires, as claimed in T8?
Yes, for as the argument for the tripartition of the soul in Republic IV shows, nous is capable of
generating its own desires; cf. e.g. T. Irwin, Plato’s Ethics (Oxford 1995), 215; H. Lorenz, “The Analysis of
the Soul in Plato’s Republic” in G. Santas (ed.), The Blackwell Guide to Plato’s Republic (Oxford 2006),
146–65 at 154–57.
42 I tend to agree with the view that god is nous without soul, where soul is required only for those
things that have or come to have nous, cf. R. Hackforth, “Plato’s Theism” in R. E. Allen (ed.), Studies in
300   Thomas Kjeller Johansen

The principle of coming-into-being explains the degree of similarity that obtains

between the god and the paradigm: the paradigm has been chosen in order to allow the
world as far as possible to be like the god. Scholars have often noted these similarities,
sometimes with a view to assimilating or identifying the demiurge with the paradigm.43
However, this assimilation gets Timaeus’s point exactly the wrong way around. It is not
that the demiurge is similar to the formal paradigm because he just is the same as the
paradigm, but rather that the formal paradigm has been chosen by the demiurge
because it is in the relevant respects similar to him. The principle of the creation is the
god’s desire to make the world as similar to him as possible and the choice of the eternal
paradigm is instrumental to that end.
That the god should make another thing like himself makes good sense also in view of
that fact that he is not only the craftsman but also the father of the universe (T2).
Normally, a craftsman does not make another thing such as himself. A carpenter does
not make another carpenter but, say, a chair. However, in generation the cause is like the
offspring. As the Symposium says, generation works by producing another thing such as
oneself.44 Or as Aristotle puts it, “man generates man.” As the father of the cosmos, the
god makes another thing like himself: a unitary, immortal, intelligent god.45
If, finally, we ask why god, the father, should want to make a world as like himself as
possible, Timaeus’s answer in T6 is that it is because he lacks envy (phthonos). But what
does lacking envy have to do with wanting something else to be like oneself? Aristotle’s
definition of “envy” (phthonos) in Rhetoric II.10 suggests an answer. Envy, for Aristotle,
is “a kind of pain in respect to one’s equals for their apparent success in things called
good, not so as to have the thing oneself, but [solely] on their account” (1387b23–25).46
Envy is thus a vice specifically concerned with the goods that others who might be con-
sidered one’s equals would enjoy. In his goodness, the demiurge, however, is so far
removed from envy that he wants to create something that as far as possible is his equal
such that it can enjoy, as far as possible, the same goods as he.47 Given his complete lack

Plato’s Metaphysics [Studies] (London 1965), 444–45; S. Menn, Plato on God as Nous (Carbondale and
Edwardsville 1995), 19–24; F. Karfik, Die Beseelung des Kosmos (Leipzig 2004), 199–200. If so, the
demiurge’s first creative act also confirms my point that the maker seeks to maximize the likeness
between himself and the creation in a way that it is appropriate to something in the category of coming
into being: while he himself is nous, his creation has nous in the manner of a rationally moving soul.
43 Cf. Baltes, “Γέγονεν”, 88; E. D. Perl, “The Demiurge and the Forms: A Return to the Ancient
Interpretation of Plato’s Timaeus,” Ancient Philosophy 18 (1998), 81–92; F. Ferrari, “Causa paradigmatica
e cause efficiente: il ruolo delle Idee nel Timeo,” in Natali and Maso, Physicus, 83–96 at 88–91.
44 Symp. 207d3–4, 208a6–b2. For the idea of similarity between parent and offspring, cf. also
Republic 506e3.
45 Cf. 38b6–9, 41a7–b6, 92c8–9.
46 Transl. from D. Konstan, The Emotions of the Ancient Greeks: Studies in Aristotle and Greek
Literature, Toronto 2006, 112–13 (emphasis added).
47 Cf. also Broadie, Nature, 13–14. For an alternative view of envy in T6, see F. G. Hermann, “Φθόνος
in the World of the Timaeus” in D. Konstan and N. K. Rutter (eds.), Envy, Spite and Jealousy. The
Rivalrous Emotions in Ancient Greece (Edinburgh 2003) 53–83.
 The Timaeus on the Principles of Cosmology   301

of envy,48 it makes sense that the god’s creative desire should take the form of wanting to
create another being like himself to enjoy the good that he enjoys.
So far I have argued that Timaeus’s introduction is concerned with identifying the
proper principle of cosmology. This principle is a principle of coming-into-being for
what is the principle of the cosmos as something that has come into being. The principle
is associated with the maker of the creation, but in its most proper sense Timaeus identi-
fies it with the maker’s desire for the world to be as similar to himself—and hence as
good—as possible. It is this principle that prescribes the choice of an eternal model, and
specifically, the choice of a living being, for the creation of the cosmos.

5. Necessity: Another Principle

Takes the Stage

In 47e2–48e1, a passage brimming with references to arkhai, Timaeus advises us that we

need another principle or starting point (arkhê) for our cosmology.49 What is the second
principle and why is it called for? At 47e2–48a5 Timaeus tells us that intelligence for the
most part guided necessity toward the best in the creation of the universe and that we
have so far considered the universe, with a few exceptions,50 insofar as intelligence so
guided necessity. That intelligence had to persuade necessity to create the world but only
managed to do so for the most part shows how necessity is a causal principle that is inde-
pendent from intelligence.51 But what does Timaeus mean by “necessity”?
Timaeus refers to “necessity” also as “the wandering cause” (48a6–7). The two desig-
nations may appear incompatible: necessity suggests causal determinacy and regularity,
“wandering” indeterminacy, and irregularity. However, “wandering” is to be under-
stood in contrast to the goal-directedness of the intelligent cause, and not to determi-
nacy or regularity: the intelligent cause does what it does for the sake of the good,
whereas the wandering cause brings about its outcome without regard to its goodness.
Calling necessity “the wandering cause” does not mean that it is indeterminate or irreg-
ular, only that its outcome is not determined by considerations of what is good.
“Wandering” underlines that necessity is not goal-directed.52
Already 46d–e suggested that there was a notion of necessity that belonged to bodies.
Now we are seeking to understand necessity as a principle independent of the intelligent
cause. We are therefore invited to consider what the bodies and their affections were

48 Note the strongly emphatic negation, oudeis peri oudenos oudepote . . . phthonos, 29e1–2.
49 See, in particular, 48a7–b3.
50 I take one of the exceptions to be the account of mirror images at 46a2–c6, cf. the reference to
“necessity” at 46b1.
51 On the sense of persuasion, see G. Morrow, “Necessity and Persuasion in Plato’s Timaeus” in
Allen, Studies, 421–37, and Johansen, Natural, 99–103.
52 See further Johansen, Natural 93–95, and A. S. Mason, “Plato on Necessity and Chaos,”
Philosophical Studies 127 (2006) 283–98.
302   Thomas Kjeller Johansen

before the intelligent cause ordered them in the creation. That is to say, we are going to
look at the “nature and affections” of earth, water, fire, and air (48b3–5) before the intel-
ligent cause acted on them in order to see what properties bodies necessarily give rise to
in and of themselves. It is to meet this request that Timaeus introduces a third kind of
entity, alongside the two we know from Premise 1. The third kind is of a mysterious
nature, having no inherent properties, yet serving an indispensable role as the “receptacle”
of the coming-into-being of bodies.53
Remember that we are interested in the nature of fire, earth, air, and water before the
coming-into-being of the cosmos. The receptacle is introduced first (49b–50b) because of
a certain difficulty in talking about fire, earth, air, and water as being anything. The problem
is that they always seem to be changing into each other and never appear to be the same.
So it is difficult to say which of them is which. Later Timaeus refers to the precosmic
bodies as “traces” (ikhnê, 53b2) of earth, fire, water, and air.54 While the term traces sug-
gests at least a superficial similarity with the cosmic version of the four bodies, it also
suggests that they are not the real thing. We can see why: not yet having been invested by
god with forms and numerical order, the precosmic bodies did not have the proper
nature of fire, earth, air, and water. In these circumstances, Timaeus says (50b1–2), it
seems safer when asked what each of them is to refer to the receptacle, just as if when
asked what some gold was that is constantly changing shape it would be safest to say that
the thing was gold.
It may seem natural to take the comparison with the shapes in gold to imply that the
receptacle is the matter out of which the precosmic bodies are composed. However, later
Timaeus refers to the receptacle as the “space” (chôra) that provides a place for all the
things that come to be (52a9–b1). The issue whether the receptacle should be under-
stood as the matter out of which bodies are constructed or the space in which they are
located has been intensely discussed. My own view, which I cannot argue for here, is that
the two views are not incompatible: we can understand the receptacle as space,55 and at
the same time allow that there is a sense in which space is that out of which bodies,
understood as geometrical figures, are constructed.56 It is worth noting, however, that as
far as the god’s creation of the four basic bodies is concerned, Timaeus presents, not the
receptacle itself, but the traces as that out of which the bodies are created (ex hautôn,
53a7, cf. 53b1–6). One might, of course, say that since the traces are in some sense just
qualifications of the receptacle, the god also orders the four bodies out of the receptacle,
but if so, that is only indirectly. Moreover, saying that the receptacle itself is ordered is
problematic since it is supposed to remain without any inherent properties if it is to
serve properly as a receptacle (50d–51b).

53 For detailed interpretation, see D. Miller, The Third Kind in Plato’s Timaeus (Göttingen 2003).
54 I take no position here in the debate whether the appearances at 49a6–50a4 should be understood
as phenomenal particulars, or recurrent types. For a summary of the debate with further bibliography,
see Zeyl, Plato Timaeus, lvi–lxiv.
55 For reasons for prioritizing the spatial description of the receptacle, see Johansen, Natural, 118–36.
56 See V. Harte, Plato on Parts and Wholes (Oxford 2002), 247–64, in particular 250–51; cf. also Zeyl,
Plato Timaeus, lxi–lxiv.
 The Timaeus on the Principles of Cosmology   303

The description of the four bodies as ever-changing appearances in the receptacle

suggests a condition that is specific to the precosmos. Yet it is clear that Timaeus is
seeking to set up a description of the precosmos that allows us to draw certain parallels
with the behavior of the bodies in the cosmos. So, in describing the transformations
of bodies, he refers in the present tense to what we see (49d4) or think that we see
(49b8–c1), and at 52e4 he refers to the three kinds as having been also before the world
came into being. Later (54b–c) he says that the appearance that all the four bodies
change into each other was wrong, since earth, given its different geometrical composi-
tion, stands apart from the transformations. If this claim is to count as a correction,
then it seems that Timaeus cannot have meant earlier to refer merely to the precosmic
state of the four bodies.
Timaeus seems then to have adopted a view point on the precosmos that to some
extent carries over to the cosmos. The account of the motions of the bodies in the recep-
tacle is a good example. Timaeus tells us (52e–53a) that in the precosmos the four bodies,
that is, their traces, shook and were in turn shaken by the receptacle, somewhat like a
winnowing basket separating grains. The result was that the bodies joined with their
kindred bodies in different parts of the receptacle. It is clear that this process continues
in the cosmos (57c). The difference is that the distribution according to likeness and
unlikeness now reflects the geometrical natures of the four bodies. Similarly, while the
bodies changed into each other also in the precosmos, they continue to do so in the cos-
mos, albeit now according to rules of geometrical composition.
There is then a viewpoint from which one can describe the precosmic and cosmic
bodies in the same way, namely the viewpoint that abstracts from the mathematical
order whereby the god shaped the bodies (53b4–5). This is how we tend to see the bodies
now, absent such mathematical knowledge. It is instructive for Timaeus to adopt this
viewpoint because he can thereby show the properties that the materials of the creation
contribute to the bodies in the cosmos: already in the precosmos, bodies come into
being as likenesses of forms, possessed with tendencies to move toward kindred bodies
and to change into each other. The materials continue to contribute such properties to
the bodies in the cosmos.
Timaeus understands the way the precosmic bodies contribute properties to the cos-
mic bodies as a matter of necessity. His concern in 47e–68d as a whole—a section some-
times referred to as the “Works of Necessity”—is to explain the properties of bodies
from the bottom-up, starting with the most basic constituents of the creation. As I
understand it, necessity here attaches to the consequences of bodies’ having a certain
material composition. Properties that bodies have by necessity are those properties that
the bodies take over because of their material constituents.
Three examples illustrate the way Timaeus uses necessity to explain the properties of
bodies from the bottom-up, from the point of view of their material constituents. First,
if a body has depth, he says, it is necessary that it is comprised by surfaces (53c–d).
Second, if the faces of earth are made of isosceles triangles, it is necessary that earth be
stable and immobile (55e). Third, if the head is fleshy, then it is necessary, given the char-
acter of flesh, that it lacks sensitivity (75a–b). Necessity here attaches to the consequence
304   Thomas Kjeller Johansen

of having certain properties given the properties of the materials at a lower level of
­composition. On this understanding, the so-called works of necessity are staggered in
such a way that a body at each level n+1 has certain properties necessarily because of the
properties of its constituents at level n.
Understanding necessity in this way, as relative to a certain level of composition, helps
us with two problems. One is that the passage at 53c–68d cannot be meant to account for
the “works of necessity” to the exclusion of intelligence. The geometrical composition of
the simple bodies is explicitly assigned to god at 53b4–5 and again at 56c3–7. It cannot
then be Timaeus’s intention to account for the composition of bodies and their com-
pounds by reference to necessity to the exclusion of intelligence. On my suggestion, in
contrast, we may allow for intelligence to operate alongside necessity at each level of
composition. Necessity is a separate explanatory principle at each level because it refers
back to those properties the materials have in virtue of the lower level of composition.
But this is compatible with the materials having been composed at the lower level by the
gods: recall here the case of flesh bringing certain necessary properties to the head
because of the way the gods had composed flesh.
Another problem, rarely addressed by commentators, is that if the receptacle on its
own is meant to account for necessity it is curious that there is not a single mention of
necessity during the entire passage discussing the receptacle. Rather, after 48a the first
mention comes again at 53c5–6, with, as we saw, the discussion of the geometrical com-
position of the simple bodies. The omission is to be expected, however, if necessity
attaches to the consequences of the materials’ having certain properties at level n for the
composition of a body at level n+1. Since the properties and movements of the traces in
the receptacle represent level 0 from the point of view of composition of bodies, we
should not expect necessity to enter our account at this level but only at the next level,
that of the composition of the simple bodies, as in fact we found at 53c. Put differently,
necessity as the modality of bottom-up causation does not apply to items at the bottom
level itself but only to the relationship between items at the bottom level and those at the
next level up and subsequent levels.

6. Conclusion: Bringing the

Two Principles Together

At the end of the account of “the Works of Necessity” (68e–69b), Timaeus considers the
relationship between the necessary and the intelligent or “divine” causes. It is clear that
explanatory priority rests with the divine cause: the necessary cause was introduced so
that we could better grasp the divine. For the necessary cause was used by the divine
cause, as an auxiliary cause, in bringing about his fair design. Timaeus is clearly thinking
of the necessary cause in extension of the craftsmanship model: the necessary cause
 The Timaeus on the Principles of Cosmology   305

­ rovides the means by which the divine craftsman realizes his design. Necessity may
p
present an obstacle to design, as we saw in the case of the human head, which could by
necessity not both be covered in flesh and be sensitive, even if that would have been
preferable. However, we may also, Timaeus is suggesting, think of necessity not as an
obstacle but as an aid to the intelligent cause. For if there were not materials that neces-
sitated certain outcomes, then the materials could not reliably help the craftsman realise
his design. If heating up metal did not necessarily make it pliable, then the furnace
would be of less use to the blacksmith. We need to understand necessity, then, both as a
constraining and as an enabling cause in relation to the divine cause. Plato’s cosmology
is based on these two causal principles, which from opposite ends of the creation
together allow us to see why this, rather than any other, world came into being.
In the end, then, we are left in cosmology, not with one highest principle of being as in
the Republic, but with two principles both of which are specifically geared to explaining
the world as something that has come into being.57 The specificity of the principles to
coming into being is obvious in the case of necessity, since this principle attaches to bod-
ies as such. However, the highest principle, the god’s benevolence, is also a principle of
coming-into-being, not a principle of being. As we have seen, the highest principle is, in
the manner of the craftsmanship and fatherhood, essentially a cause of coming-into-being
of what was not before. To understand the Timaeus properly, we need to read the princi-
ples of cosmology as fundamental and specific to the study of the visible world. In the
Timaeus, the notion of cosmology as a mere means to the study of being has been left
well behind.58

Bibliography
Baltes, M., “Γέγονεν (Platon Tim. 28B6). Ist die Welt real entstanden oder nicht?” [“Γέγονεν”]
in Algra, K. A., van der Horst, P. W., and Runia, D. T. (eds.), Polyhistor. Studies in the History
and Historiography of Ancient Philosophy (Leiden 1996), 76–98.
Bolton, R., “The Origins of Aristotle’s Natural Teleology in Physics II” in M. Leunissen (ed.),
Aristotle’s Physics: A Critical Guide (Cambridge 2015) 107–20.
Broadie, S., Nature and Divinity in Plato’s Timaeus (Cambridge 2012).
Bryan, J., Likeness and Likelihood in the Presocratics and Plato (Cambridge 2012).
Burnyeat, M. F., “Platonism and Mathematics: A Prelude to Discussion,” in A. Graeser (ed.),
Mathematics and Metaphysics in Aristotle (Bern and Stuttgart 1987) 213–40; reprinted in
M. F. Burnyeat, Explorations in Ancient and Modern Philosophy, vol.2, (Cambridge 2012)
145–72.

57 The argument of this chapter lends indirect support to D. Sedley’s thesis in “The Origins of Stoic
God” (in D. Frede and A. Laks (eds.), Traditions of Theology (Leiden 2002), 41–83) that the Stoics could
have derived their two-principle physics from a reading of the Timaeus. For the later ancient answers to
the question, how many principles of cosmology in the Timaeus?, see D. Runia, “Plato’s Timaeus, First
Principle(s) in Philo and Early Christian Thought,” in G. Reydams-Schils (ed.), Plato’s Timaeus as
Cultural Icon (Notre Dame 2003), 133–51.
58 I am grateful to Gail Fine for many helpful comments on this chapter.
306   Thomas Kjeller Johansen

Burnyeat, M. F., “Plato on Why Mathematics Is Good for the Soul,” in Proceedings of the British
Academy 103 (2000), 1–81.
Burnyeat, M. F., “ΕΙΚΩΣ ΜΥΘΟΣ”[“ΕΙΚΩΣ”], Rhizai vol. 2, no.2 (2005), 143–66.
Code, A., “Reply to Michael Frede’s ‘Being and Becoming in Plato’, ” Oxford Studies in Ancient
Philosophy, Supplementary Volume 1988, 53–60.
Cornford, F. M., Plato’s Cosmology (London 1937).
Ferrari, F., “Causa paradigmatica e cause efficiente: il ruolo delle Idee nel Timeo,” in Natali, C.
and Maso, S. (eds.), Plato Physicus [Physicus] (Amsterdam 2003), 83–96.
Fine, G., “Forms as Causes: Plato and Aristotle,” in Graeser, A. (ed.), Mathematics and
Metaphysics in Aristotle (Bern 1987), 69–112; reprinted in Fine, Plato on Knowledge and
Forms. Selected Essays [Plato] (Oxford 2003), 350–96.
Fine, G., “Knowledge and Belief in Republic V–VII” [“Knowledge”], in Everson, S. (ed.),
Epistemology (Cambridge 1990), 85–115; reprinted in Fine, Plato, 85–116.
Frede, M., “The Original Notion of Cause,” in Frede, M., Essays in Ancient Philosophy (Oxford
1987), 125–50.
Frede, M., “Being and Becoming in Plato,” Oxford Studies in Ancient Philosophy, Supplementary
Volume 1988, 37–52.
Hackforth, R., “Plato’s Theism” in Allen, R. E. (ed.), Studies in Plato’s Metaphysics [Studies]
(London 1965), 439–47.
Halliwell, S., Aristotle’s Poetics (London 1998).
Harte, V., Plato on Parts and Wholes (Oxford 2002).
Hermann, F. G., “Φθόνος in the World of the Timaeus” in Konstan, D. and Rutter, N. K.
(eds.), Envy, Spite and Jealousy. The Rivalrous Emotions in Ancient Greece (Edinburgh
2003), 53–83.
Irwin, T., Plato’s Ethics (Oxford 1995).
Johansen, T. K., Plato’s Natural Philosophy [Natural] (Cambridge 2004).
Johansen, T. K., “Should Aristotle Have Recognised Final Causation in Plato’s Timaeus?,” in
R. Mohr and B. Sattler (eds.), Plato’s Timaeus Today (Parmenides 2010) 179–200.
Johansen, T. K., “Why the Cosmos Needs a Craftsman: Plato’s Timaeus 27d5–29b,” Phronesis
59 (2014) 297–320.
Konstan, D., The Emotions of the Ancient Greeks: Studies in Aristotle and Greek Literature
(Toronto 2006).
Karfik, F., Die Beseelung des Kosmos (Leipzig 2004).
Lennox, J., “Plato’s Unnatural Teleology,” in Lennox, J., Aristotle’s Philosophy of Biology. Studies
in the Origins of Life Sciences (Cambridge 2001), 280–302; first published in D. O’Meara
(ed.), Platonic Investigations (Washington D.C. 1985), 195–218.
Lorenz, H., “The Analysis of the Soul in Plato’s Republic” in Santas, G. (ed.), The Blackwell
Guide to Plato’s Republic (Oxford 2006), 146–65.
Mason, A. S. “Plato on Necessity and Chaos,” Philosophical Studies 127 (2006) 283–98.
Menn, S., Plato on God as Nous (Carbondale and Edwardsville 1995).
Miller, D., The Third Kind in Plato’s Timaeus (Göttingen 2003).
Morrow, G., “Necessity and Persuasion in Plato’s Timaeus” in Allen, Studies, 421–37.
Mueller, I., “Platonism and the Study of Nature,” in Gentzler, J. (ed.), Method in Ancient
Philosophy (Oxford 1998), 67–90.
Rowe, C. J., “The Status of the ‘Myth’,” in Natali and Maso, Physicus, 21–31.
 The Timaeus on the Principles of Cosmology   307

Runia, D. “Plato’s Timaeus, First Principle(s), and Creation in Philo and Early Christian
Thought,” in Reydams-Schils, G. (ed.), Plato’s Timaeus as Cultural Icon (Notre Dame, Ind.
2003), 133–51.
Sedley, D., “The Origins of Stoic God” in Frede, D. and Laks, A. (eds.), Traditions of Theology
(Leiden 2002), 41–83.
Sedley, D., Creationism and its Critics in Antiquity, (Berkeley, Calif. 2008).
Vlastos, G., “Creation in the Timaeus: Is It a Fiction?” in Vlastos, G., Graham, D. W. (ed.),
Socrates, Plato, and Their Tradition (Princeton N.J., 1995), 265–79.
Zeyl, D., Plato Timaeus, (Indianapolis/Cambridge 2000).
 chapter 13

The Sophist on
Statem en ts,
Pr edication, a n d
Fa l sehood

Lesley Brown

Among several striking features of Plato’s late dialogue, the Sophist, two stand out. First,
it divides clearly into two very different parts. In the Outer Parts (216–236d; 264b9–end),
the main speaker, a nameless visitor from Elea in Italy (hereafter ES, for Eleatic Stranger)
embarks on a discourse ostensibly designed to say what a sophist is. Using the so-called
Method of Division, the ES offers no fewer than seven accounts of what the sophist is.
Interrupting the seventh attempt, the Middle Part (236d9–264b8) provides a striking
contrast. There the ES undertakes a lengthy discussion—sparked by problems arising
from defining a sophist as a maker of images and purveyor of false beliefs—which, for
most readers, is of far greater philosophical interest and value.1 Though such an ostensi-
ble “digression” is not unprecedented in Plato—one may think of the central books of
the Republic—the disparity between the two parts is arresting.2
A second striking feature is the markedly didactic approach. At the start, Socrates
asks the ES (217a) to tell the inquirers what the people of Elea think about the issue in
hand—namely, the relation between sophist, statesman, and philosopher: Are they
three different kinds, or two, or just one? This approach is not the more usual “Let’s dis-
cuss this matter together.” The ES opts to present his material via question and answer

1 N. Notomi, The Unity of Plato’s Sophist [Unity] (Cambridge, 1999), from whom I take the labels
Outer Part and Middle Part, ch. 1 usefully compares other Platonic “digressions” with that of the
“Middle Part.”
2 It is especially hard to envisage how the work was received by anyone who was introduced to it at a
reading, unaware of the surprise in store halfway through the work and of the different degree of
difficulty and abstractness of the Middle Part.
310   Lesley Brown

with the intelligent Theaetetus but makes it clear that this is just a presentational device,
not a true open-ended investigation.3 Plato has something he wants to convey.
Both features highlight some of the key enigmas of the dialogue: What is the relation
between the Outer and Middle Parts? How seriously are we to take the Outer Parts, and
is there a genuine, and successful, attempt to say what the sophist is? The fact that the ES
offers seven alternative definitions, each purporting to be of the sophist (and not, as we
might expect, of different types of sophist) gives us pause, as does the quirkiness of the
“definitions,” not least the final one.4 On my unorthodox reading, we are not intended to
regard any of the definitions as correct, especially since the search has assumed some-
thing that Plato cannot have accepted: that sophistry is an expertise, a technē (denied at
Gorgias 464d).5 Nonetheless, Plato ensures that we learn plenty from the dialogue about
the many differences between sophistry and philosophy, but also that we note their
common ground, especially their shared interest in puzzles, aporiai.6 This will be a
theme of the subsequent discussion.
This essay focuses on two key problems discussed and solved in the Middle Part: the
Late-learners’ problem (the denial of predication), and the problem of false statement.
I look at how each is, in a way, a problem about correct speaking; how each gave rise to
serious philosophical difficulty, as well as being a source of eristic troublemaking; and
how the ES offers a definitive solution to both. As I said above, the Sophist displays an
unusually didactic approach: Plato makes it clear that he has important matter to impart,
and he does so with a firm hand, especially on the two issues I’ve selected.

1. Lead-in to the Middle

Part and Synopsis

Defining the sophist as a maker of images and falsehoods leads us—so the ES
­proclaims—into matters full of long-standing problems: “How one should express
­oneself in saying or judging that there really are falsehoods, without getting caught up in

3 From 217c–e. At d8, ES regrets he will not have a genuine exchange with Socrates. Cf. M. Frede,
“The Literary Form of the Sophist” [“Literary Form”], in Form and Argument in Late Plato, ed. C. Gill
and M. M. McCabe (Oxford, 1996), 138–39.
4 Resumé of first six at 231c–e; cf. 265a. Seventh “definition” at 268c ff: ES: “An imitator, of the
contradiction-making sort of the dissembling part of conceit-imitation, of the semblance-making
kind of image-making, who’s marked off in the human (not the divine) portion of production a
magic-trickery with arguments—if someone says such is the lineage and blood of the one who really is
a sophist, then I think they’ll be speaking the very truth.”
5 L. Brown, “Definition and Division in the Sophist,” in Ancient Theories of Definition, ed. D. Charles
(Oxford, 2010).
6 For different views, see C. C. W. Taylor, “Socrates the Sophist,” in Remembering Socrates, ed.
L. Judson and V. Karasmanis (Oxford, 2004), 157–68; Notomi, Unity.
 The Sophist on Statements, Predication, and Falsehood   311

contradiction by such an utterance: that’s extremely difficult, Theaetetus.”7 The puzzle is

not (contra Notomi, Unity, 193) “Do falsehood and appearance really exist?” but “How
should we express ourselves when saying they do, since to do so involves postulating
that not being is?”8 The ES then develops an exquisite series of aporiai about the expres-
sion “what is not/not being.”9 He goes on to lard his remarks with pointers to “uttering
things correctly,” “correct speaking,” and so forth and ironically exclaims: “Don’t look to
me for correct speaking (orthologia) about what is not.”10
The Middle Part proceeds by developing a wealth of problems, then systematically
solving them.

(i) Problems about not being or what is not (237d–241c)

Resolve: to show that what is not is in some respect, and what is is not in a way
(241d–242a)
(ii) Problems about being (242b–251a)
Upshot: we’re in as much difficulty about what is as we are about what is not (250e).
(iii) A new problem: the Late-learners’ prohibition on saying that one thing is many
things (251a–c)
(iv) ‘‘Partial mixing’’ must be the correct one of three possible theories, since we can
rule out ‘‘no-mixing’’ (Late-learners) and ‘‘total mixing’’ (251d–253b)
Greatest Kinds: a four-point program laid out (254b–d2)
(v) Five ‘‘Greatest Kinds’’ selected and proofs offered that they are five (i.e., points
1 and 2 of the four-point program) (254d–255e)
(vi) Points 3 and 4: the Communion of Kinds—investigation of how change com-
bines with the other four kinds; demonstration that change is and is not being;
and that being is, in a way, not being (255e–257a)
(vii) Negation, negative expressions, not being and the parts of difference (257b–258e)
Upshot: we have shown that, and what, not being is (258e–259e)
Remaining tasks: to show what statement is and that falsity in statement, judg-
ment, and “appearing” is possible (260a–261b)

7 From 236e4. I reject the emendation ad loc in the 1995 Oxford Classical Text, Platonis Opera I, ed.
E. A. Duke, W. F. Hicken, W. S. M. Nicoll, D. B. Robinson, and J. C. G. Strachan (Oxford, 1995). Cf.
Frede, “Literary Form,” 143–44.
8 This alludes to the locution “say/judge what is not” to mean “make a false statement/judgment.”
See Section 3.5 of this chapter.
9 On these aporiai, see especially G. E. L. Owen, “Plato on Not-Being,” in Plato 1, ed. G. Fine
(Oxford, 1999), 431–38. In (i), the term mē on can’t be applied to anything without contradiction; in (ii),
nothing that is—such as number—can be applied to it, so that ascribing either the number one (by the
appellation to mē on, “what is not”) or plurality (by the label ta mē onta, “things that are not”) involves
self-contradiction; in (iii), even the charge that “not being is inexpressible, unsayable and so forth” itself
falls foul of the prohibition on treating it as something that is. See L. Brown “Aporia in Plato’s
Theaetetus and Sophist” in Aporia in Ancient Philosophy ed. G. Karamanolis and V. Politis (Cambridge,
2018). For a detailed and rigorous discussion see P. Crivelli, Plato’s Account of Falsehood [Falsehood],
(Cambridge 2012), ch 2.
10 From 239a8, 239b4, 239b9; cf. 239d1.
312   Lesley Brown

(viii) What statement (logos) is; the difference between “names” and “verbs” and
between naming and saying (261c–262e)
(ix) True and false statements (262e–263d)
(x) False judgment and false “appearing” (263d–264b)

2. The Late-Learners’ Problem and

Its Solution in the Demonstration
of Communion of Kinds

2.1 The Late-Learners’ Problem: Summary

and Rival Diagnoses
In these stretches, Plato unveils a problem at (iii), and solves it, after setting up a consid-
erable apparatus, at (vi). He does so using some complex analyses, and this is where the
issue of speaking correctly comes in—or, rather, of understanding correctly what has
been said. He will tell us that we must not be disturbed by certain ways of speaking,
when we say, of two kinds K and L, that “K is L and K is not L,” and we will accept this
once we recognize the different ways in which each conjunct is said (256a10–b4). So the
ES promises a disambiguation, but what is it? A long-standing debate concerns whether
his diagnosis of the problem and his solution turn crucially on pinpointing different
meanings or uses of “is” (or rather, Greek esti).11
There are two major schools of interpretation, those I’ll call “optimists” and
“­pessimists.” The optimists, who include Ackrill, Vlastos, and others, argue as follows.12
The puzzle that Plato attributes to certain unnamed people, who are rudely labeled
“Late-learners” (Soph 251–52), depends on the refusal by these awkward thinkers to rec-
ognize that in sentences of the form “A is B,” “is” can have two meanings or uses: that of
identity (is the same as) and that of the copula, the “is” of predication. Plato (according
to the optimists) diagnoses their difficulty as the failure to recognize the two meanings

11 The debate is often conducted in terms of different meanings of “is,” following Frege. M. Frede,
“Plato’s Sophist on False Statements” [“False”], in Cambridge Companion to Plato, ed. R. Kraut
(Cambridge, 1990), 397–424, argues for a weaker claim, that Plato distinguishes uses but not meanings
of “is,” since different meanings would correspond to different forms, while Plato recognizes only one
form of being. Frede’s position was developed first in Prädikation und Existenzaussage (Göttingen, 1967).
For the purposes of this essay, I do not distinguish between the two claims but treat them as
interchangeable. In discussing other scholars’ views I use “meaning” or “use” in accordance with
their usage.
12 J. L. Ackrill, “Plato and the Copula: Sophist 251–59” [“Copula”], in Plato 1: Metaphysics and
Epistemology, ed. G. Vlastos (Garden City, N.Y., 1971), 210–22; G. Vlastos, “An Ambiguity in the
Sophist” [“Ambiguity”], Platonic Studies (Princeton, N.J., 1981), 288 n.44; J. van Eck, “Plato’s Logical
Insights: Sophist 254d–257a” [“Insights”], Ancient Philosophy 20/1 (2000), 71–74.
 The Sophist on Statements, Predication, and Falsehood   313

or uses of “is,” and later (at vi) displays them, by the device of different paraphrases for
“is” or esti. Triumph! Plato anticipated the great Gottlob Frege. The pessimists accept
this distinction between different uses of “is” and agree that it is needed to dissolve the
difficulty of the Late-learners. But they sorrowfully declare that the passage where
Ackrill and others find Plato making this key discovery can’t be read in that way; that,
alas, Plato did not solve the problem correctly: did not discover the distinction between
the two meanings of “is.”13 The optimists and the pessimists share a common premise: if
Plato distinguished these two meanings or uses of “is,” then he made an important dis-
covery; and if he didn’t, he missed making that same discovery. But this assumption is
the one I’m going to challenge.
I accept that Plato does not distinguish these two meanings or uses of “is.” But (unlike
the pessimists), I’ll show that he solved the problem in a perfectly adequate way, by dis-
tinguishing what I’ll call “identity sentences” from predications. Indeed, following other
writers, I dissent from the tradition (deriving from Frege’s “On Concept and Object”) of
accepting a special “is” of identity.14 My reading credits Plato with a successful solution
to the “Late-learners’ problem,” one that does not appeal to the rather dubious distinc-
tion between the meanings of “is.” Our task is to examine the texts and to give as faithful
an interpretation as we can; it will be a bonus if, as a result, we can vindicate Plato’s
­so-called logical insights.
At 251a5–6, the Stranger turns to the problem of how we call the same thing by many
names (pollois onomasi tauton touto . . . prosagoreuomen) and describes the views of the
so-called opsimatheis, Late-learners.15

str. Well, when we speak of a man we name him lots of things as well, applying
colors and shapes and sizes and vices and virtues to him, and in these and thou-
sands of other ways we say that he is not only a man but also good and many other
things. And so with everything else: though we assume that each thing is one, by
the same way of speaking [logos] we speak of it as many and with many names.
tht. What you say is true.
str. This habit of ours seems to have provided a feast for the young and some old
folk who’ve taken to studying late in life. For anyone can weigh in with the quick
objection that it is impossible for what is many to be one and for what is one to be
many, and they just love not allowing you to call a man good, but only the good
good and the man a man. I dare say, Theaetetus, that you often meet people who

13 Pessimists include D. Bostock, “Plato on ‘Is-Not’ (Sophist 254–9)” [“Is-Not”], Oxford Studies in
Ancient Philosophy 2 (1984), 89–119; J. Gosling, Plato (London, 1973), ch. 13.
14 For arguments against isolating an “is” of identity, see F. Sommers, “Do We Need Identity?”
Journal of Philosophy (1969), 499–504; M. Lockwood, “On Predicating Proper Names” [“Predicating”],
Philosophical Review (1975), 471–98 (who also argues for the interpretation of Sophist 255e–256e, which
I favor); C. Kahn, The Verb “Be” in Ancient Greek (Dordrecht, 1973), e.g.at 372, 400; and B. Mates,
“Identity and Predication in Plato,” Phronesis 24/3 (1979), 211–29. Cf. the discussion in F. A. Lewis, “Did
Plato Discover the Estin of Identity?” [“Did Plato”], California Studies in Classical Antiquity 8 (1975),
113–42.
15 For discussion of who the Late-Learners represent, see F. M. Cornford, Plato’s Theory of
Knowledge [Theory] (repr. London, 1960), 254.
314   Lesley Brown

are keen on that sort of line. Some of them are getting on in years, and their
intellectual ­bankruptcy makes them marvel at that sort of thing and suppose that
in this they have made an exceptionally clever discovery.

So this is their position: (1) they object to calling one thing many and with many names
(251b3); (2) they don’t allow you to legein agathon anthrōpon (251b8–c1) (either “to call a
man good” or “to say the man is good”); and (added later) (3) they don’t allow you to call
anything something different, since they don’t accept that anything has communion
with the attribute of another thing (252b9–10, paraphrase).
Presumably they forbid both using a compound description “good man” and saying
“the man is good.” And presumably this is because they assume that the only function of
a word is to name, so they rule out both “good man” and “the man is good” as “making
one many” (by naming two things, man and good). They refuse to accept that it is harm-
less and indeed useful to speak of something “as many and with many names”: that is, to
apply a number of attributes, as in one of the above locutions.
So much for what the Late-learners don’t allow. What do they allow? Here there is a
controversy. On some interpretations Plato tells us that they don’t allow any sentences at
all, but only names or namings.16 I disagree. I think we are told that the Late-learners do
allow some sentences, provided that in whatever you utter you don’t “make one thing
many”; provided you only call a thing itself, not something else. A sentence may be
permitted in which you say that a thing is itself, if the many names it uses are for the
same thing. “You must only say a thing is itself, you mustn’t say it is something else”
(cf. 252b9–10).
“They only allow you to say ‘the man is a man’ but not ‘the man is good.’ ” Must this
be read as charging them with a failure to understand “is,” with not allowing an “is”
of predication, in a sentence such as “the man is good”? Not necessarily. It may just be
that they make a mistake about the whole locution—in particular, about the role of what
comes after the “is.” The Late-learners assume that its role is to name the very same thing
as the subject term names. On the same ground they would reject the appellation “good
man,” with the thought that, since both words are names, and are not synonymous, then
two things, not one, would be named by that expression. They do not accept predication,
because they deny any underlying methexis or sharing-in. And it is to answer them that
the following sections are written, in which the sharing or koinōnia of kinds is described.
On this point Ackrill—in my view—is quite correct; but not when he reads Plato as
identifying the mistake made by the Late-learners in terms of a mistake about “is.”
Confirmation of my diagnosis comes from a later source, the account of the views
of the Megarian Stilpo in Plutarch’s Adversus Colotem.17 Stilpo apparently, like the
Late-learners, rejected statements such as “the man is good” but also statements such as
“the horse runs.” In other words, even sentences without “is” were rejected, presumably

16 J. M. E. Moravcsik, “Being and Meaning in the Sophist,” Acta Philosophica Fennica 14 (1962),
57–59; Gosling, Plato, 219–20.
17 Quoted in N. Denyer, Language, Truth and Falsehood [Language] (Cambridge, 1991), 34–35.
 The Sophist on Statements, Predication, and Falsehood   315

because the second term did not name the same thing as the first term. Stilpo’s difficulty,
then, does not concern the role of “is.” Rather, it is a refusal to accept that parts of logoi
are used not to name but to predicate, or to attribute, something to the subject.
To sum up: the Late-learners allow only identity sentences, and their mistake is the
mistake of not understanding predication, or its metaphysical basis: “sharing in.”18 Some
earlier arguments in the dialogue had gone wrong because they treated predicates like
names and so treated predicative sentences as identity sentences.19 Plato’s task is to
explain the notion of predication in order to show that the following is possible: K is L
(because it shares in L), and K is not L (because K is different from L). A thing can be
what it also is not: this is what Section 2.2 of this chapter is designed to show, in answer to
the mistaken view of the Late-learners. As I have argued, we don’t have to construe the
problem as a problem about meanings or uses of “is” but, rather, as a problem about
types of sentence: identity sentences versus predications. And so to credit Plato with
logical insight, we don’t have to read his solution as distinguishing different meanings or
uses of esti—which is a good thing, because he doesn’t do so, as we shall see.

2.2 The “Communion of Kinds” as Offering

the Solution to the Late-Learners Problem
We fast-forward through the Middle Part, omitting sections (iv) and (v) in which—inter
alia—the ES introduces the notion that dialectic involves investigating the relations of
kinds, and draws an analogy between letters of the alphabet and kinds, such that some
kinds operate in the way vowels do, enabling the joining of letters while being them-
selves one type of letter. We omit also the first two points of the four-point program, the
introduction of the five so-called Greatest Kinds—kinēsis (change), stasis (stability),
being, same, and different—and the intriguing proofs of their distinctness from one
another.20 We resume where the ES promises to fulfill the remaining points: (3) to see
what power of combination they have with one another (4) in order to get hold of to on
(being) and to mē on (not being) and to show that it is safe to say that to mē on really is
mē on (not being really is not being).

2.2.1 The “Communion of Kinds” (255e–256e): Plato’s “Four Quartets”

It is vital to the understanding of Plato’s aims in this section to see how systematically the
passage is organized, as many earlier commentators have shown. The kinds are taken

18 I discuss in Section 2.4 of this chapter Frede’s alternative view that the key distinction is between
the uses of “is” in self-predications (which the Late-Learners allow) and in other-predications (which
they forbid). Section 2.4 also discusses Crivelli’s view, Falsehood 107–08, which is close to but different
from Frede’s.
19 Those at 243d–244b and 250a8–d3. These arguments are designed to be parallel and to be
fallacious: the second ends in a contradiction, and the reader is clearly invited to discern what has gone
wrong, then to connect it with the Late-Learners’ aporia.
20 For a full and subtle discussion see Crivelli, Falsehood, 117–49.
316   Lesley Brown

kath’hen, one by one. One is chosen, change, and its interrelations with each of the other
four kinds are examined in turn. I call these groups of sentences the “four quartets”
because in a typical group there are four distinguishable propositions linking change
with the other kind under discussion.

Group 1: Change and stability

1a Change is different from stability (255e10)

So 1b Change is not stability (e14)
But 1c Change is (256a1)
Because 1d Change shares in being (a1)
Group 2: Change and the same

2a Change is different from the same (256a3)

So 2b Change is not the same (a5)
But 2c Change is the same (a7)
Because 2d Change shares in the same (a7,b1)
Group 3: Change and different

3a Change is different from different (256c5)

So 3b Change is not different (c8)
But 3c Change is different (c8)
[because Change shares in different (not in text)]
Group 4: Change and being

4a Change is different from being (d5)

So 4b Change is not being (d8)
But 4c Change is being (d8–9)
Because 4d Change shares in being (d9)

It is clear that Groups 2, 3, and 4 have the same pattern, viz:

a K is different from L
So b K is not L (denial of identity between K and L, since it
follows from a)
But c K is L (L is predicated of K, as shown by paraphrase at d)
Because d K shares in L

Because Group 2 is the first to exemplify this pattern, Plato treats it at length, taking
pains to explain why the apparent contradiction between 2b and 2c is not a real one.
He explains that 2b asserts what 2a asserts, and thus does not contradict 2c, which is
equivalent to 2d. The same point is made more briefly for Group 3, and at greater length
in Group 4, the target of the exercise. The apparent contradiction between the b and c
sentences is made possible because the names of the three kinds concerned—same,
­different, and being—can function both as abstract nouns (as required in b) and as
adjectives (as required in c). I return to this point later.
 The Sophist on Statements, Predication, and Falsehood   317

We have noticed a pattern common to the later three groups.21 What is Plato up to in
this carefully worked passage? What are his aims and achievements?

2.2.2 Common Ground to All Interpretations

Plato aims to give a careful account of the connections between the sample kind change
and the four other kinds, in turn; and to do so by offering analyses, in terms of “sharing
in” (metechein and similar expressions), of key sentences expressing these connections,
sentences that take the form “K is L” or “K is not L” where K stands for change and L for
one of the other kinds, in order. He does this to show why conjunctions of the form “K is
not L and K is L” are not, despite appearances, contradictory, and why each conjunct can
be true, when properly construed.
In particular, he aims to show that “change is not being and change is being” is not a
contradiction, that both conjuncts are true, and thus to vindicate the status of not being.
This is Group 4, the one it was all building up to.

2.2.3 Accepted by Most but Not All (Frede and Crivelli Dissent)
Plato uses the device of analysis in terms of metechein (sharing in) to distinguish statements
of identity from predications. More precisely, he shows that “K is not L and K is L” can both
be true provided that “K is not L” denies that the kind K is the kind L—that is, denies the
identity of K and L—while “K is L” is a predication or attribution of L to K (in other words,
says that K shares in L). I call this the “minimal interpretation” of the section.
Now the important question: How does Plato hope to achieve this?

2.2.4 The Optimists’ View: Distinguishing Meanings or Uses of Esti

The crucial lines are 256a10–b4. In these lines Plato makes the ES explain why the two
previous claims, 2b and 2c (change is not the same and change is the same) must both be
admitted.

Str. Change, then, is both the same and not the same—we must agree and not
­dispute it. For when we said [it was] the same and not the same, we were not
speaking in a similar way, but when [we say it is] the same, we say that because of
its ­sharing in the same in relation to itself, but when [we say it is] not the same,
that, by contrast, is because of its communion with the different, through which it
is separated from the same and isn’t it but different, so that once again it’s rightly
said to be not the same.

21 Group 1 is different at 1c, since the ES has insisted that change cannot in any way share in stability:
252d2–11, 255ab, esp. a11–b1. The text at 256b6–8 considers the counterfactual “if change were to share
in stability in some way,” clearly implying that this is impossible, despite our expectation that change, as
a Form, must be stable. Crivelli, Falsehood, 149–66, offers a rather strained interpretation of 256b6–8
whereby Plato is allowing that there is a reading on which it is true that change shares in stability and
can therefore be called stable.
318   Lesley Brown

Note, “we were not speaking in a similar way” (ou . . . homoiōs eirēkamen). The
­ ptimists argue that this draws attention to an ambiguity, and we may agree. They argue
o
further that the ambiguity in question must be that of the verb “is,” since they hold, in
the Frege tradition, that this is the correct account. But a major problem is that in these
key lines Plato does not draw attention to the word esti; worse, he actually omits it in the
crucial sentence. We must indeed supply it, but still, if he had really been signaling an
ambiguity in esti, surely he would not have omitted it at the vital moment.22
The optimists have a reply here. Even if Plato omitted it, he must still have located the
ambiguity in the esti. They argue as follows. Consider the three pairs of contradictory
propositions (2b+c, 3b+c, and 4b+c). The esti is the only constituent common to these
pairs that could account for the ambiguity in each quartet.23 Now I agree that we should
look for an account of these lines that can also serve as an explanation of the other
groups as well, since Plato evidently constructed the passage carefully and means his
account of Group 2 to serve also for the two later groups.24 But optimists are wrong to
claim that the only element common to all three that could explain the ambiguity is the
verb esti. The three pairs share the same form, and the ambiguity may be due to that, not
to the occurrence of a given word (“is”) used in two ways.

2.3 Plato’s Solution: What Ambiguity Is He Pointing Out?

There are two alternative solutions that I prefer to the claim that Plato locates the
­ambiguity in “is.”25

2.3.1 Solution 1
Solution 1 locates the ambiguity in what follows the esti. In other words, Plato points to a
difference in “the same” between 2b “Change is not the same” and 2c “Change is the
same,” as suggested by Owen.26 And it is quite correct that the words “the same” play
these different roles in the two sentences! Even those who accept two meanings or uses
of esti must agree that there are also two meanings or uses of tauton. To say change is not

22 Defending the “optimist” line, van Eck, “Insights,” 71–74, argues that Plato does “distinguish a
non-predicative sense of ‘is’ at 256b3–4,” albeit using gegone rather than esti. We may agree that gegone
here means “is, as a result of,” and that in gegonen ouk ekeino all’ heteron (isn’t it but something
different) the “isn’t it” denies identity between change and tauton. But it doesn’t follow that Plato is
distinguishing a sense of “is.”
23 Vlastos, “Ambiguity,” 291 n.46.
24 For this reason, we may reject a different interpretation (Gosling, Plato, 218–19) by which the
solution is to add different completions to “the same” in the two conjuncts. Such a reading, though
possible for Group 2, will not allow an equivalent solution for Groups 3 and 4, which Plato clearly
intends.
25 For a number of suggestions about what Plato’s solution is, see Lewis, “Did Plato.” His preferred
solution (134–36) has Plato invoking a special sense of not found in 2b, change is not tauton, and also in
3b and 4b. Crivelli, Falsehood, 149–66, discusses different solutions and opts for a distinction between
“ordinary” and “definitional” readings of the relevant sentences. See 2.4 of this chapter for discussion.
26 Owen, “Not Being,” 258 n.63; Lockwood, “Predicating,” 479 n.12.
 The Sophist on Statements, Predication, and Falsehood   319

tauton is to say it is not the kind, sameness. And the same goes for 3b (change is not the
kind different) and 4b (change is not the kind being). It may be helpful to compare the
uses of the word blue in the following sentences:

s1 The sky is blue. (“Blue” used as an adjective, to attribute blueness to the sky)
s2 The color of the sky is blue. (“Blue” used to designate the color, blue)

The crucial item is the word or phrase that follows esti; that is, in Group 2, tauton
(the same).
Now if Plato were being accurate, he should write to tauton “the the same” in 2b, to
show that the phrase is being used as an abstract noun to refer to the kind sameness.
And he should write to heteron in 3b, and to on in 4b. It is only because he doesn’t do so
that he is able to produce apparent contradictions. If he had written, at 4b, “change is not
to on,” that evidently does not contradict Kinēsis estin on, which means “change is a being
(is a thing that is).” One reason he does not use these forms is that Greek, where possible,
avoids the definite article after the verb “to be,” so Plato felt free to leave it out—in order
to achieve his apparent contradictions.27 To repeat, the word tauton plays two different
roles, adjectival in 2c (change is tauton) but as an abstract noun in 2b (change is not
tauton). Should we not give Plato credit for pointing out this difference of role, when he
offers the elucidation in the key lines 256a10–b5? After all, he does seem to lay the
emphasis on tauton in the crucial sentence.
To support this interpretation, I make one philosophical point and one appeal to the
text. The philosophical point, hinted at above, is that we must admit that there is a dual
use of the words the same, whether or not we accept, with Frege and others, a dual use of
the word is. Usually a different form of the word will be used where the sentence is an
identity sentence; for instance, we will say “change is different” (adjective) but “it is not
difference” (abstract noun). But where there is the same form (as in the pair “the sky is
blue” and “the color of the sky is blue”), we have to assume a different function (once as
an adjective, once naming the color blue).
The textual point in support of this interpretation is drawn from some curious lines
that follow Group 2.

str. And if this very thing, change, were to participate in any way in stability, it
would not be at all odd to call it stable (stasimon).
tht. Very true, if we are to agree that some of the kinds are willing to mix with one
another and others are not. (256b6–10)

These lines have puzzled commentators. Why are they here? Is something missing
(as, e.g., Cornford believed)?28 At any rate, the ES is evidently not asserting that change
does share in a way in stability. Rather, the sentence is a counterfactual: if change were to

27 This point about Greek usage (cf. Lewis, “Did Plato”) answers Bostock’s objection (“Is-Not,” 93).
28 Cornford, Theory, proposed a lacuna after 256b7.
320   Lesley Brown

share in any way in stasis, it would not be odd to call it stasimon (or, to say it is stasimon,
stable). Now, we know that change doesn’t share in stasis, for it has been emphasized
­several times (cf. n.21). Why does the Stranger revert to it? If I am right that he has just
pointed out the different roles for tauton in 2b and 2c, then perhaps he is underlining the
adjectival role of tauton, where “is tauton” means “shares in tauton,” by displaying the
parallel with “shares in stasis,” which becomes “is stasimon.” This—drawing attention
to the adjectival form, stasimon, as parallel to the adjectival function in 2c, 3c, and
4c—would partly explain this otherwise out-of-place remark.

2.3.2 Objection to My Proposal and Reply

It has been objected that “the names tauton, heteron and on cannot vary in sense within
any of the three sentences, for . . . the meaning is fixed unambiguously by Plato’s assump-
tion that each name refers to the identical Form within both of the apparently contradic-
tory conjuncts.”29 Reply: Not so, and here is an argument to show it. Suppose Plato had
chosen to pursue the Communion of Kinds with the assertions that 5a being is different
from stability, so 5b being is not stability, but 5c being is stable. There’s no danger here of
an apparent contradiction, but still Plato could analyze 5b and 5c on the lines of 2b and
2c, analyzing 5c as “being shares in stability.” No one would claim that the sole function
of “stable” (Greek stasimon) in our imaginary 5c is to refer to the form or kind stability,
though that is part of its function. Its evidently adjectival form would make it obvious
that its role was different from that of “stability” in 5b. And presumably Plato could have
made this point about “the same” as used as an adjective in 2c, despite Vlastos’s claims.

2.3.3 Solution 2: More Modest

Perhaps we are wrong to think that Plato identifies one element as the locus of ambiguity.
Just because he offers a paraphrase does not mean we must attach each element of the
paraphrase to an element of the original sentence. Perhaps instead he simply notes, quite
correctly, and shows by means of paraphrase, that in each pair one sentence functions to
deny identity between change and the other kind (Change is not the kind being), while
the second sentence predicates that kind of change (Change is a being). This would be a
holistic solution, rather than an atomistic one. If we seek the correct account of why the
pairs of sentences are not contradictory, in spite of appearances, this may be the safest
answer. The ambiguity depends on the whole sentence forms, not on any one element. If
that is all Plato wishes to convey on the matter, then it is perfectly adequate, in my view.

2.4 Alternative Interpretations: Frede and Crivelli

Two further lines of interpretation remain to be discussed, those of Michael Frede and
Paolo Crivelli. Both reject the reading whereby Plato is indicating a difference between
identity statements and predications, and both find Plato invoking a distinction between

29 Vlastos, “Ambiguity,” 291 n.46. Bostock, a “pessimist,” uses the same argument (“Is-Not,” 97).
 The Sophist on Statements, Predication, and Falsehood   321

different kinds of predication: roughly, between essential and nonessential predication.

Each locates his favored distinction in different texts of the Sophist from the Communion
of Kinds passage we are considering, and each then tries to discern his favored distinc-
tion in the Communion of Kinds passage. This much their interpretations have in com-
mon. But in other respects they differ from one another, so I take them in turn.
Frede (but not Crivelli) assumes that Plato is pointing out a distinction in uses of “is.”
I label his line “superoptimist” since he too holds that Plato is adverting to a crucial dis-
tinction in uses of “is”/esti, but, unlike the optimists, he holds that this single distinction
is the only one needed to solve all the problems in the Sophist. (Optimists such as Ackrill,
however, hold that at other points Plato is also distinguishing the existential “is.”) The
key distinction, for Frede, is the one between the use of “is” to say what a thing is in itself
or by itself, and the use of “is” to say what a thing is by standing in the appropriate relation
to something else.30 He illustrates the distinction with reference to two uses of “is white.”
Socrates is white by standing in a relation to a color (i.e., second use of “is”). The color
white, however, is white by being this feature, not by having it (i.e., first, “in itself ” use
of “is”). Like the optimists, Frede takes his favored distinction to be the key to both the
Late-learners’ problem and the Communion of Kinds passage. The Late-learners—on
his view—allow only “in itself ” predications and disallow the second kind, the kind we
use when we say “Socrates is white.”
A full discussion of Frede’s rich position is beyond the scope of this essay.31 In brief,
I find his account of the Late-learners’ position highly plausible, equally plausible with
the account I favor, according which Late-learners allow (in effect) identity statements
but disallow predications. Each interpretation is compatible with the prohibition on
calling anything something different. Frede’s view would prefer, as the Late-learners’
slogan, “you can say what a thing is in itself, but not what it is in some other way,” while
the identity view would imagine the slogan “you can say a thing is itself, but not any-
thing else.” Each fits what we are told about the Late-learners’ theory.
However, we want to find Plato demonstrating in the Communion of Kinds section
just the distinction that the Late-learners refused to accept, and here—in this later pas-
sage—I find Frede’s interpretation less plausible. Why? Consider the opening lines of
Group 1—1a, change is different from stability; 1b, change is not stability. This pattern is
repeated, for the first two sentences of the next three groups: K is different from L, so K is
not L. We must surely expect to interpret 2b, Change is not the same, 3b, Change is not
­different, and 4b, Change is not being along the lines of 1b. But 1b is surely a denial of
identity. Frede’s interpretation wants 2b, 3b, and 4b to be read as denials of “in itself

30 Frede, “False,” 400. At 401 he explains that his discerning of this distinction “crucially rests on the
assumption that Plato in 255c12–13 distinguishes these two uses of ‘is,’ ” sc. with the labels “auta
kath’hauta” and “pros alla.”
31 In L. Brown, “Being in the Sophist: A Syntactical Enquiry,” in Plato 1, ed. G. Fine (Oxford, 1999),
474–76, I discuss Frede’s claim that this distinction features in the proof of the nonidentity of different
and being at 255c–d. At 470–71, I outline the interpretation for which Section 2.3 of this chapter gives a
fuller argument.
322   Lesley Brown

predication,” not as denials of identity.32 Thus 2b is to be understood as “Change is not,

in itself, the same,” and so on for the remainder. But the equivalent reading of 1b cannot
succeed. If Plato had wanted 1b to be a denial of “in itself predication,” then he would
have written “Change is not by its own nature stable” (cf. 250c6–7, where the point is
made that being, by its own nature, is neither changing nor stable). Since I find it impos-
sible to read the negative claims in the four quartets in any other way than as denials of
identity, I cannot accept Frede’s reading.
I now turn to the somewhat similar interpretation offered by Crivelli. I argue that it
faces the very same difficulty in its reading of the Communion of Kinds passage. Here is
his take on the distinction Plato makes the ES allude to when he says that “we were not
speaking likewise” when we called change the same and not the same. The distinction
Crivelli appeals to is between what he labels “definitional” and “ordinary” readings of
sentences.33 He previously identified this distinction as crucial in understanding cer-
tain earlier arguments (122ff, discussing the argument at 255a4–255b7). It is beyond the
scope of this chapter to evaluate the meticulous way in which Crivelli argues for its
importance at other points in the Sophist’s argumentation. Here I consider his defense of
finding his favored distinction also at the heart of the Late-learners’ problem and in
the Communion of Kinds passage. Crivelli writes that the distinction he identifies
“preserves the distinctive intuition” of Frede’s interpretation, without being identical
to it. Where Frede found Plato distinguishing two uses of “is,” viz the one used to say
what something is in itself, and the other to say that a thing stands in a relation to
something else, Crivelli finds two readings of sentences (definitional and ordinary), or,
in a formulation he treats as equivalent, two kinds of predication, essentialist and
ordinary predication.34
Here is his reading of the Late-learners’ difficulty: “to refute the late-learners Plato
should draw a distinction between predications whereby the predicate-expression is
taken to provide a full description of the nature or essence of the item signified by the
subject expression, on the one hand, and predications whereby what is said is simply
that the item signified by the subject-expression instantiates the kind signified by the
predicate-expression, on the other” (108). Because they don’t recognize the second kind

32 Frede, “False,” 422: his chief reason for denying that in this section sentences of the form “X is not Y”
are nonidentity sentences is that 263b11–12, which seems to refer back to 256e6–7, must concern
denials of predication, not denials of identity. Hence his wish to read this section as also featuring
denials of (in-itself) predication. But in my view, this solution to what is a real problem comes at too
high a price.
33 Crivelli, Falsehood, at 122 introduces and explains the distinction with reference to the argument
at 255a4–255b7. At 161 he applies it to the Communion of Kinds passage under discussion here. Unlike
Frede (see note 30 of this chapter) Crivelli does not tie his distinction to the distinction at 255c12–13
between things said to be auta kath’hauta and pros alla.
34 Crivelli writes of the distinction between essentialist and non-essentialist predication when
discussing the Late-Learners at 107–09. At 109 he seems to equate that with his favored distinction
between definitional and ordinary readings of sentences such as “Change is stable.” The label
“definitional reading” is perhaps surprising given that he exemplifies it most often in sentences such as
“Largeness is large” “Change changes” and so on (123). A true definition such as “Soul is change capable
of changing itself ” (cited at 124) is of course also to be given a “definitional reading.”
 The Sophist on Statements, Predication, and Falsehood   323

of predication, but only the first, the Late-learners’ problem (on this reading) is their
refusal to allow any predication other than essentialist predication. This explains why
they will allow you to say “the man is a man” but not “the man is good.” As I remarked
about Frede’s similar account of the Late-learners’ problem, I find this as plausible as the
more common account, according to which the Late-learners allow identity statements
but disallow predications. Both give a plausible account of the Late-learners passage,
but, as with Frede’s reading, there is a major difficulty for Crivelli’s attempt to find his
favored distinction at work in our Communion of Kinds passage (255e–256e).
How is Crivelli’s distinction between definitional and ordinary readings of sentences
supposed to play out in the Communion of Kinds passage? Take the sentences 2b and 2c,
the point at which the ES remarks that we should accept both, since “we weren’t speak-
ing in the same way” when we called them the same and not the same. Where most crit-
ics take these to be equivalent to 2b Change is not the same as the same but 2c Change is
(predicatively) the same [as itself], Crivelli takes both sentences to be predications, such
that 2b is true on its “definitional” reading, while 2c is true on its “ordinary” reading.
But the very same difficulty arises for this reading as for Frede’s. It is hard to square
with the overall construction of our passage, starting as it does with the pair of sentences
1a Change is different from stability and 1b Change is not stability. Note that there is an
abstract noun, stability, in complement position in 1b. When Plato makes the ES infer 1b
Change is not stability from 1a Change is different from stability, he surely indicates that
both sentences are to be read as denying identity between change and stability.35 And so
we are surely meant to infer that the negative sentences of the later pairs (2b, 3b, and 4b)
are also to be read as denials of identity, not (with Frede and Crivelli) as denials of some
special kind of predication, whether “in itself ” or “definitional” predication. So I find it
impossible to accept Crivelli’s understanding of 2b Change is not the same as a denial
that change is definitionally the same.
On one point, however, Crivelli’s and my interpretation are in agreement. In order to
identify what Plato means when the ES says “we were not speaking likewise” in asserting
2b and 2c, we do not need to locate the ambiguity Plato has in mind in a single word
(such as estin, i.e., “is”) or in a single element of the sentence. Crivelli’s suggestion that
Plato is indicating two readings of sentences is (in this respect) in keeping with what was
suggested at the end of Section 2.3. In the terms used there, Crivelli is proposing a holis-
tic solution whereby Plato, in indicating how each of the pair of apparently inconsistent
sentences can be paraphrased, notes the different function of each sentence but without
picking out a single word as the locus of ambiguity. I have argued, however, that the two
readings of sentences proposed by Crivelli—definitional and ordinary readings—do
not fit the text as well as the two proposed at the end of the previous section, whereby the
negative sentence of each pair (such as 4b change is not being) denies the identity of
change with another kind while the positive one (4c change is [a] being) predicates that
kind of change.

35 I cannot agree with Crivelli, Falsehood, 164 that we can “ignore the difference between ‘stability’
and ‘stable.’ ”
324   Lesley Brown

2.5 Two Remaining Questions

To conclude this discussion of Plato’s treatment of the Late-learners’ problem, I raise and
reply to two questions. First: “The so-called problem of the Late-learners is so silly that
we can’t imagine anyone being seriously bothered by it. Did Plato really need to go to
such lengths to refute so absurd a view?” In reply, I endorse Ackrill’s claim: the thesis was
put forward not only by elderly jokers but also by serious thinkers who felt themselves
obliged to maintain it for what seemed to them compelling theoretical reasons.36 We
have already seen (Section 2.1 of this chapter), that it was also maintained by the
Megarian thinker Stilpo. And, as Denyer has shown, variants on it have appealed to phi-
losophers such as Bradley, who worried about saying that a lump of sugar is sweet and
white and hard: “A thing is not any of its qualities, if you take that quality by itself; if
‘sweet’ were the same as ‘simply sweet,’ the thing would clearly not be sweet. And again,
insofar as the sugar is sweet it is not white or hard; for these properties are all distinct.”
And so on.37 One or both of the following may prompt the thesis: a metaphysical view
about what the world ultimately consists in, or a view of language that sees naming as
the only function of bits of language. Bizarre though it may seem to us, we cannot dis-
miss it as a mere sophism unworthy of serious attention from Plato.
The second question asks why, when the Late-learners’ puzzle concerns statements
about particulars such as “the man is tall and handsome,” the solution in the Communion
of Kinds section concerns statements about kinds. The answer, I think, is this. In each
case (a predication about a particular and one about a kind) we have, in effect, a claim
that one thing is many things, and that it is what it also is not. Now the claim that partic-
ulars, such as Socrates, are many things isn’t so troubling for Platonic metaphysics. But
the claim, of a Form or kind, that it must be many things, and must be what it also is not,
needed more defense. Republic V had claimed that Forms always are, and in no way are
not. But in the Communion of Kinds section, the ES shows not just the difference
between identity statements and predications in general but how even Forms or kinds
can be spoken of in both these ways. The upshot (in Group 4) is the demonstration that a
kind such as change both is a being and is not being (i.e., it shares in being even though
it is not the Form being). Thus, we have the first place in which the Resolve is fulfilled
(see synopsis in Section 2 of this chapter): showing that what is not being (i.e., is not the
kind being) nonetheless is in a way (i.e., it is a being, and thus lots of other things
besides). In other words, we have been shown not only what the Late-learners denied—
how it is legitimate and true to say that one thing is many things, and is what it also is
not—but also how a kind (other than being) can be a being and yet not be being itself.
Understanding just what is being said in these apparently contradictory locutions is the
key to resolving them.

36 Ackrill, “Copula,” 215.

37 F. H. Bradley, Appearance and Reality, 2nd ed. (Oxford, 1932), 16, quoted in Denyer, Language, 44.
 The Sophist on Statements, Predication, and Falsehood   325

3. The Account of False Statement

Once again we fast forward, omitting discussion of the most puzzling section (vii) of the
dialogue. I say a little about it later; for now we note that it concludes with the declara-
tion that the inquirers have found what the form of not being is.38 But that, we are told, is
not the end of the inquiry. By means of a carefully placed series of signposts (from 260b
onward) the ES stresses that fulfilling the Resolve is not enough for demonstrating the
possibility of false statement.39 He emphasizes that showing that kinds mix was neces-
sary but not sufficient to solve all their problems and, in particular, was insufficient to
solve the problem of falsehood. To do that, they must also investigate what statement
and judgment (logos and doxa) are, to see if they can be false (to see if “not being can mix
with them” (260b10–c4)). Theaetetus repeats the point (261ab), and it’s made a third
time by the ES (261c). Plato was evidently concerned that the reader should see that a
fresh topic has been broached and that they are moving to a new discussion.
By almost universal agreement, the section in which the ES explains what a logos is
and how there can be false ones is one of the most successful and important of the whole
dialogue. Though the account is well known, I here outline it once again, discuss how it
should be understood (3.2–3.4), then ask what is most valuable in the account (3.5).40

3.1 The Account of What a Statement Is

The key to understanding how a logos can be false lies in first understanding what a logos
is. Here Plato proceeds with the utmost care. He scripts a scene in which Theaetetus first
misunderstands (261d7) and leaps to a wrong conclusion. This allows the ES, in correct-
ing him (262b2), to emphasize the novelty of his new point, which is this. With words as
well as with kinds, “partial mixing” is the order of the day, if statements are to eventuate.
But the ES informs Theaetetus that the “partial mixing” of words he is about to expound
must not be confused with the “partial mixing” of kinds discussed in the four-point pro-
gram. Words (onomata) come in two varieties: names and verbs (onomata and rhēmata:
thus onoma has both a general and a more specific meaning). Not any concatenation of

38 “Having demonstrated what the nature of the different is, and that it’s parcelled out over all the
things that are, set against each other, we’ve dared to say that the part of it set against the being of each
thing—that very thing really is not being” (258d). For a fuller discussion of Sophist 257–59 see L. Brown
“Negation and Not-Being: Dark Matter in the Sophist” in Presocratics and Plato Festschrift at Delphi in
Honor of Charles Kahn, ed. R. Patterson, V. Karasmanis, and A. Hermann (Las Vegas, NV, 2012).
39 “Statement” is the best translation for logos in this section. It has a range of meanings that include
reason, speech, and definition.
40 My account owes much to that of Michael Frede, “False,” sec. III, though I dissent from his
understanding of one major issue: how to understand the reference to what is different in the
paraphrase the ES offers of what it is for a statement to be false. See also Crivelli, chapter 20 in this
volume.
326   Lesley Brown

words makes a logos; rather, a logos must combine a name with a verb, where “verb” is
the designation used of actions and “name” is the designation used of the doers of those
actions.41 Neither a string of verbs (such as “walks runs sleeps”) nor a string of names
(such as “lion deer horse”) makes a logos. A logos is a special kind of interweaving;
someone who interweaves a verb with a name doesn’t only name but succeeds in saying
something (262d2–6).42
Plato here makes a crucial point. Saying something—what the utterer of a statement
does—is different from merely naming. To achieve this “saying something” a logos needs
two parts with different functions: “one part whose function is to name, refer to, identify
a subject, and another part by means of which we say something, state something, pred-
icate something of or about the subject.”43 As Frede’s terminology in the sentence just
quoted shows, we may think of the distinction in a variety of ways. Perhaps the key idea
is the distinction between the part of the statement used to refer to the subject (the
onoma, name, or subject-expression) and the part used to predicate something of the
subject.44 And we can agree that if Plato intends to distinguish word classes, the claim
that each logos has a noun and a verb picks out only a subclass of statements, whereas he
seems to want to characterize simple statements more generally “and really is looking
for syntactical categories.”45 With the distinction between naming and saying, and with
the recognition that a statement is essentially structured, as a special weaving together of
parts with different functions, certain puzzles found in earlier dialogues—notably
Euthydemus—denying the possibility of false statement or judgment and of contradic-
tion are finally put to rest.46 What the puzzles had in common was that they treated a

41 “An expression we apply to actions we call a verb” (262a2). The word order, together with the use
of legein rather than kalein, indicate that this is not intended as a strict definition of rhēma. Cf.
M. Hoekstra and F. Scheppers, “Onoma, rhēma et logos dans le Cratyle et le Sophiste de Platon,”
L’Antiquité Classique (2003), 69, who insist, plausibly, that the major point of the passage is not the new
assignation of familiar words for words (onoma, rhēma) to distinct roles but the recognition that a
special kind of fitting together (harmottein) is involved in any logos.
42 Interweaving, plegma; cf. sumplekōn (weaving together) at 262d4. 43 Frede, “False,” 413–14.
44 I cannot agree with D. Sedley, Plato’s Cratylus (Cambridge, 2003), that in all this Cratylus
prefigures Sophist. Sedley claims Plato in Cratylus uses the terms onoma and rhēma to focus on “the
two linguistic acts . . . of naming and predication” and that Socrates shows “awareness that onomata and
rhēmata are functionally disparate items within the statement.” Denyer, Language, 148–50, correctly
remarks that in Cratylus (as elsewhere in Plato outside this stretch of Sophist) rhēma typically means
phrase, group of words, as opposed to onoma, a single word. Contra Sedley, Crat 399ab and 399b7, and
421d–e are best explained in this way. Cf. Sophist 257b6–c2 (before the official demarcation and
identification of onoma and rhēma): in support of Denyer, note that at 257b6 mē mega (not large) is
called a rhēma, but at c6 the ES speaks of the onomata which follow the “not” in expressions such as
“not large.”
45 Frede, “False,” 413.
46 Frede, “False,” 413–17, and M. Burnyeat, “Plato on How Not to Speak about Not-Being” [“How
Not To”] in Le Style de la pensée, ed. M. Canto and P. Pellegrin (Paris, 2002), 40–65. Burnyeat holds that
in the earlier works Plato hints at the vital distinction between the subject of a logos and what’s said
about it, but concedes that prior to Sophist there is no “hint of the grammatical or syntactic distinction
drawn there between the part of an assertoric sentence that refers to the subject and the part that
ascribes to that subject a predicate such as flying or sitting” (45).
 The Sophist on Statements, Predication, and Falsehood   327

logos as an unstructured whole; many of them portrayed saying and/or judging like
naming, using a “scandalous analogy” (Burnyeat) between judging and touching.

3.2 The Account of True and False Statements

After stressing that a logos is a special kind of structured whole, only one of whose
parts has the function of referring to the thing it is about, the ES can at once get
Theaetetus to agree that both “Theaetetus sits” and “Theaetetus flies” are, by the above
account, statements.47 Then he proceeds smartly to explain the truth of the one and the
falsity of the other. He does so twice over, in what I shall call the two “Final Formulae
for Falsehood,” the first (which has been much discussed) at 263b4–11, and the second
(relatively neglected) at 263d1–4.

First Final Formula for Falsehood

(A) The true one says of things that are about you that they are; while (B) the false
one says different things from the things that are; that is, (C) it says, of things that
are not, that they are. In a problematic sequel the ES continues, in a highly elliptical
manner: (D) (but it says) things that are, but are different (from what is) about you.48

To understand all this, we must first eliminate the plural forms, a stylistic device loved
by Plato but highly confusing to the reader. The chief warrant for doing so is this: the
sample true logos “Theaetetus sits,” which plainly says one thing about Theaetetus, is
described as saying ta onta, things that are.49 Replacing plurals with singulars, and leav-
ing to one side for now a host of problems with this stretch, I recast (A) and (C) in what
follows, but postpone discussion of the controversial (B) and (D) until later.50

47 “Plato quite pointedly lets the Eleatic Stranger settle the question of reference for the sample
statements discussed before he lets him go on to consider their truth or falsehood.” Frede, “False,” 418.
48 For a full discussion, including a proof that hōs estin at 263b4 must be translated “that they are,”
not “as they are,” see D. Keyt, “Plato on Falsity” [“Falsity”], in Exegesis and Argument, ed. E. Lee,
A. Mourelatos, and R. Rorty (Assen, 1973), 287–91. D. Sedley, The Midwife of Platonism (Oxford, 2004),
133 n.19, proposes a reassignment of speakers in the problematic lines b9–11.
49 Further support for replacing plurals with singulars comes at 263d1–4, where the statement
“Theaetetus flies” is said to be “a synthesis of verbs and names” when it is plainly a synthesis of one verb
and one name. This licenses us to rewrite the entire sentence replacing plurals with singulars, as
discussed.
50 Those who favor the so-called Oxford interpretation, discussed in Section 3.3, cannot agree that
the use of the plurals is merely a stylistic device, for they invoke the plural in (B) 263b7 to indicate that
it is correct to import a universal quantifier into the translation. Thus: “Plato could have said in
263b3–4 that the true statement says of something that is that it is. But he wants to get a reference to the
whole class of things that are, relative to a given subject, into the characterization of the true statement,
as this will be needed to get an adequate characterization of the false statement. This corresponds to the
need for a universal quantifier in a proper characterization, first of the use of ‘ . . . is not . . . ’ along Plato’s
lines, and then of falsehood, a need several commentators have rightly insisted on” (Frede, “False,”
420). I dispute this line of argument in fn 53 of this chapter.
328   Lesley Brown

The true one (A) says, of something that is about you (viz., sitting) that it is. The false
one (C) says, of something that is not (viz., flying) that it is. (Probably we must under-
stand “about you” here, too.) In other words, the false one is false because it says, about
Theaetetus, that what is not (viz., flying) is about him. Now, if we confine our attention
pro tem to (A) and (C), we recognize how elegantly they dispose of the idea that a false
statement simply “says what is not” where this is supposed to be like (the impossible)
touching what is not. Plato has distinguished “saying something about something” from
“naming.” Each statement names Theaetetus, each is about (peri) him—that is, is about
something that is, and thereby secures its reference—and each says something about him.
Plato can now allow the ES to say, without fear, that the false one says, about Theaetetus,
what is not, but says that it is about him. Even confining ourselves to (A) and (C), we find
a fully satisfactory account of true and false statements, at least if we confine ourselves to
simple assertions.51

3.3 How to Understand “Different” in the Formulae

for Falsehood? Three Readings
Now we turn to (B) and (D).

(B) The false one says different things from the things that are. (263b7)

We have seen this is used as equivalent to “saying things that are not.” And I argued
above that we may and should replace plurals by singulars, giving

(Bs) The false one says something different from what is.

How should this be understood? One difficulty is immediately evident; I label it the
“Problem.” Suppose Theaetetus is sitting, and suppose I state, “Theaetetus is talking.”
Then I have said about Theaetetus something that is different from something that is
about him—viz., sitting. But of course he may be talking as well as sitting, in which case
my statement is true. But (B) was supposed to characterize a false statement. So, on the
simplest interpretation, it is a nonstarter.
Two main readings of (B) have gained support, each of which avoids the Problem.
Following Keyt, we call them the “Oxford interpretation” and the “incompatibility

51 Contra Frede (“False,” 418), I agree with J. McDowell, “Falsehood and Not-Being in Plato’s
Sophist” [“Falsehood”], in Language and Logos, ed. M. Schofield and M. Nussbaum (Cambridge, 1982),
133 n.35, that, as they stand, the Formulae cover only true and false affirmative statements. Nonetheless,
it is clear how they can be adapted for negative truths and falsehoods.
 The Sophist on Statements, Predication, and Falsehood   329

interpretation.”52 However, I reject them both and defend a less popular, but
­increasingly supported, reading, as the correct one.

Reading 1, the Oxford interpretation. The false logos, “Theaetetus flies,” says, about
Theaetetus, something that is different from everything that is about him. On this
reading the Problem is solved. “Theaetetus is talking” will indeed be false if talking
differs from everything that is about him (i.e., which is true of him).
Reading 2, the incompatibility interpretation. The false logos, “Theaetetus flies,” says,
about Theaetetus, something that is incompatible with what is about him. This too
solves the Problem. If I ascribe an attribute incompatible with what is about
Theaetetus, I must indeed be making a false statement about him. While talking is
merely different from sitting, flying is—or was before the invention of the airplane—
incompatible with sitting.

But while both of these solve the Problem, neither of them can easily be extracted from
what Plato wrote. The Oxford interpretation faces the objection that there is no good
reason to supply, in (B) and (D), that universal quantifier—the “every”—that is so
­crucial.53 An even more serious obstacle is the wording of the Second Formula for
Falsehood, at 263d1–4, which is rarely discussed.54 There we are told that in the false
statement, “concerning you, different things are said to be the same, and not beings are
said to be beings.” Once again we may substitute singulars: “something different is said
to be the same, and something that is not is said to be something that is.” Now the Oxford
interpretation requires different supplements, as follows: in the false statement “some-
thing different [from everything that is] is said to be the same [as something that is].”55
And this is impossibly awkward. My verdict on the Oxford interpretation is that, though
it gives an adequate account of what it is for a statement to be false, it is not what Plato

52 Keyt, “Falsity,” 294–95. He discusses four alternative readings in all, but not my preferred one,
Reading 3. Crivelli, Falsehood 6.2 and Chapter 20 of this volume defends the so-called Oxford
interpretation. B. Hestir, Plato on the Metaphysical Foundation of Meaning and Truth (Cambridge,
2016), 196, proposes “an alternative to the Oxford interpretation and the ‘incompatibility range’
interpretation that preserves what I think both get right.”
53 Cf. n.50. An appeal to the plural in ta onta “things that are” at b7 is illegitimate. In 263b4–5 we
read: “The true one (‘Theaetetus sits’) says the things that are that they are about you.” Since ta onta
evidently refers to just one thing/verb, “sits,” it must there be understood as “what is.” We cannot, with
the Oxford interpretation, suddenly read it to mean “everything that is” two lines later.
54 As noted in J. Szaif, Platons Begriff der Wahrheit [Wahrheit] (Munich, 1996), 492, with whose
overall reading I am in considerable agreement.
55 The need for a different supplement is concealed in the formulation of J. van Eck: “Things that are
different from what is the case concerning him (viz. flying) are described as the same (as what is the
case about him)” (“Falsity without Negative Predication” [“Falsity”], Phronesis 40/1 (1995), 40). But, as
we saw, supporters of Reading 1, including van Eck, have to understand “what is” (or here: “what is the
case”) differently in the two supplements. In a response to the 2008 version of this chapter, J. van Eck
proposes a new interpretation, defending Plato’s use of plurals in a novel way in his “Plato’s Theory of
Negation and Falsity in Sophist 257 and 263. A New Defense of the Oxford Interpretation” Ancient
Philosophy 34 (2) 2014, at section IV.
330   Lesley Brown

intended. It is hard to find in the wording of the First Final Formula, and impossible to
read in the Second Final Formula for Falsehood at 263d1–4.
Reading 2, the incompatibility interpretation, also fails for textual reasons, though it
has the great strength that the sample statements do indeed feature incompatibles, “sits”
and “flies.” Sitting and flying, as we noted, are not merely different but incompatible;
they exclude each other. Many objections to the incompatibility interpretation have
been made on philosophical grounds, and I discuss one of these later in the chapter. The
overwhelming difficulty, however, is not a philosophical one, but that it requires that
Plato intend a change of meaning in heteron, which up to now has meant “different.” Can
it be that now, without warning, he uses it to mean “incompatible”? This must be avoided
if possible. And we can avoid it with the third interpretation, which is a variant on the
“incompatibility interpretation.”

Reading 3, the incompatibility range interpretation. Reading 3 allows us to preserve

what is good in each of the above—that is, it allows us to keep heteron to mean
“­different,” and it takes account of the fact that Plato’s sample statements feature
incompatible attributes.56

I introduce it with the help of an important but difficult text from earlier in the Sophist at
257b1–c3, where the ES explains the meaning of negative expressions. There he distin-
guishes between what is contrary (enantion) and what is “only different” (heteron
monon), and in so doing, he introduces the idea of a range of incompatible attributes
such that what is not F has one of the other attributes from the range in question (though
not necessarily the contrary of F).

(1) ‘‘Whenever we speak of not being, we don’t speak of something contrary to

being, but only different.’’ ‘‘How so?’’ (2) ‘‘For example, when we say ‘not big’ do you
think we signify small by that expression any more than equal?’’ ‘‘No.’’ (3) ‘‘So when
it is said that a negative signifies a contrary, we shan’t agree, but we’ll allow only this
much—the prefixed word ‘not’ merely indicates something other57 than the words
following the negative, or rather, other than the things which the words uttered after
the negative apply to.’’

The illustration in (2), where the ES is explaining what “not large” means, makes it clear
that while “small” is the contrary of large, “equal” is “only different” (see (1)). Plato’s point
is this: if we think of A’s size in relation to B, A may be not large, without being small
(the contrary of large), since A may be equal in size to B; so when I say that A is not large,

56 Compare M. Ferejohn, “Plato and Aristotle on Negative Predication and Semantic

Fragmentation,” Archiv für Geschichte der Philosophie 71 (1989), 262 ff. To Ferejohn’s list of adherents of
this Reading (n.9), we may now add Szaif, Wahrheit, 487–99, esp. 491, and M.-L. Gill, Philosophos,
Plato’s Missing Dialogue (Oxford 2014), at 160–61. She prefers the label “incompatibility set,” to
accommodate sets such which do not form continua. Crivelli (who opts for Reading 1) labels my
Reading 3 “quasi-incompatibility” (chapter 20 in this volume and Falsehood 239). In Section 3.4 I
discuss twentieth-century versions of the same thesis.
57 Literally: one of the others, tōn allōn ti.
 The Sophist on Statements, Predication, and Falsehood   331

I am not saying that it is small (in relation to B).58 Here we are introduced to the idea of a
range of incompatible properties or attributes F, G, and H, such that what is not F is
either G or H.59 The range may have any number of members; we may think of colors,
shapes, and so on. With this in mind, we can retain the translation “different” for heteron
but recognize that an attribute different from F taken from that range will be incompati-
ble with F. In support of this interpretation, think how laughable it would have been if in
(2) the ES had chosen a random attribute different from large, and said (for instance),
“When we say not big, do you think we signify small any more than yellow?” Being yel-
low does not rule out being large, so appealing to it in the explication of what “not large”
means would be ridiculous.
Using the help offered by 257b–c, where, as I’ve shown, the account of “not large”
invokes the idea of a range of incompatible properties when it labels “equal in size”
merely different from “large,” we can return to defend the incompatibility range inter-
pretation of the Final Formulae for Falsehood, starting with the First Formula.

(B) The false one says different things from the things that are. (263b7)
I’ve argued that this is equivalent to
(Bs) The false one says something different from what is.

We have noticed that this section also features incompatibles, sitting and flying. So we
may read (Bs) as follows: The false one says something different [taken from the relevant
range of incompatible properties] from what is about you (because it says you are flying,
which is a different one of the range of locomotive properties from the one that applies
to you—namely, sitting). And we can now read the Second Formula (263d) in a far more
natural way than the Oxford interpretation allowed.

Different things are said to be the same, and not beings are said to be beings.
Again I replace plurals by singulars, yielding
Something different [from what is about you] is said to be the same [as what is about
you].

Now we have the same supplement both times, avoiding the intolerable awkwardness
required by the Oxford interpretation. Once again, the different thing, flying, is chosen
from the range of incompatible locomotive attributes, so that if I attribute a different
thing from what is, I am bound to say something false.60

58 Precisely what his positive account here is is a controversial issue that we needn’t go into here.
59 In the next section I discuss an objection to this account of negation.
60 Van Eck, “Falsity,” 26–27, rejects the incompatibility range interpretation (which he numbers 4i)
with the protest that the supposed restriction of “different” predicates to ones from a range of
incompatible properties is “unannounced in the text” at 257b and at 263. But sentence (2), 257b6–7
comes close to announcing it, as I explained previously. Van Eck offers further opposing arguments in
sec. III of his “New Defense.” Crivelli, Falsehood, 247–48, suggests that (rather than being a
reformulation of the First Formula) what I have called the Second Formula for Falsehood is in fact a
332   Lesley Brown

In addition to these two strong indications that Plato has in mind a range of
i­ ncompatible properties, one may also cite the account of “other-judging” in Theaetetus
189b and following, where a similar idea may be at work.61

3.4 Objection to Reading 3, and Reply

It may be objected that any account of falsehood that makes an essential reference to
incompatibility (as both Readings 2 and 3 do) suffers from such a serious flaw that char-
ity requires us to avoid attributing it to Plato. The flaw is this: such an account gives at
best a sufficient, but not a necessary, condition for a false statement. (The same objection
applies to the equivalent accounts of negation.) For example, the objector points out, it
can be false that virtue is square (and true that virtue is not square) without it being the
case that virtue is some shape other than square. In reply, I concede that the flaw is
indeed serious. Must it have been so obvious to Plato that he could not have held the the-
ory? No. Indeed, such a theory of negation continued to attract leading philosophers
into the twentieth century.
It was maintained by J. Mabbott and G. Ryle, two of three contributors to an
Aristotelian Society Symposium on Negation in 1929. Ryle wrote that “when a “predi-
cate” is denied of a “subject,” that predicate must always be thought of as one member of
a disjunctive set, some other member of which set (not necessarily specified) is asserted
to be predicable of the subject.”62 H.H. Price, the third contributor, made the above
objection, insisting that statements such as “virtue is not square” and “the soul is not a
fire shovel” were both meaningful and true, in spite of resisting analysis in terms of a
range or set of incompatible properties. And this objection is correct. But though the
account that invokes a range of incompatible properties to explain negation and/or
falsehood is indeed flawed, it was an attractive candidate, and, as we have seen, Plato’s
remarks about the meaning of “not large” at 257b–c require it, while his Formulae for
Falsehood are best explained by appeal to it—in particular, the comparatively neglected
second formula at 263d.

3.5 Which Feature of the Account of Falsity

Is More Important?
We have now established (by the defense of Reading 3) how Plato means us to under-
stand the reference to the different, which glosses “not being” (ta mē onta) in the two

description of a new kind of falsehood, viz “the falsehood of affirmative sentences about kinds
according to their definitional reading.” This unusual interpretation depends on Crivelli’s accepting the
emendation peri . . . tou (about something) at 263d1 in place of the manuscripts’ peri sou (about you),
Falsehood 235 n.45.
61 Cf. P. Crivelli, “Allodoxia,” Archiv für Geschichte der Philosophie 80 (1998), 15–16; Szaif, Wahrheit,
495–96.
62 G. Ryle, “Negation,” Proceedings of the Aristotelian Society IX (suppl.) (1929), 86.
 The Sophist on Statements, Predication, and Falsehood   333

Final Formulae for Falsehood. So we can return to take stock of Plato’s achievement
in this section. We have seen that the account of false statement contains two main
­elements: (1) the insistence that a logos, true or false, is about something, and (2) a para-
phrase glossing “not being” by “different.” Discussions tend to focus on (2), partly
because it is the harder to interpret but also because a focus on the problem of not being
set the scene originally for the Middle Part. But I contend that (1) both is and is repre-
sented by Plato as the major contribution to the account of falsehood. To do so, I must
counter the objection of John McDowell.63 He argues that the Sophist’s revelation of the
subject-predicate structure of a logos is not the key to the solution of the problem of false
statement. Rather, Plato clearly indicates that the salient error lay in a mistake about not
being, and that the solution is to demolish the Eleatic mistake about negation.64
McDowell points out that in the section that develops puzzles about not being/what is
not, there are two early definitions of falsehood. I label them “Def A” and “Def B.”

Def A. False saying/judging is saying/judging what is not. (240d9)

Def B. False saying is (i) saying what is not is or (ii) saying what is is not. (240e–241a)

Def B is a double-barreled definition, of which the first part covers false positive
­statements such as “grass is red” and the second false negative statements such as “grass is
not green.”65 McDowell refers to Def B as the disjunctive definition and to Def A as the
definition that conveys “the crude position” about statements.66 Now, as we have seen, the
account of logos and of false and true logos at 260–63 is well fitted to dispose of the “crude
picture” of statements and false statements. For once we insist that a logos says something
about something, it is at once unproblematic how, having thereby got a grip on reality by
that reference, it can go on to say something false about the thing in question.
Why does McDowell reject the familiar account, which sees 260–63 as putting to rest
the crude picture by insisting on the need for a subject of a logos, as well as something
said about the subject? His answer: this does not address the much subtler Def B of false-
hood (the one he labels the disjunctive characterization). Someone who claims to find
the locution “is not” puzzling (as importing some contrary of being) will not back down
when we add “about you.”67

63 McDowell “Falsehood,” esp. sec. 6, 132–34.

64 Likewise, Sedley, Midwife, esp. 113–14, 134, holds that it is the account of not being in terms of
difference that is the key advance of Sophist, rather than the analysis of a logos into onoma and rhēma,
which (as discussed at n. 44) he thinks is not a new discovery in Sophist.
65 The double-barreled definition can cover positive and negative existential statements and
judgments, as well as predicative ones. But it should not be understood as confined to existentials.
Where I use the terms “single-barreled” and “double-barreled,” Crivelli, Falsehood, 61 calls them the
“unipolar” and “bipolar” descriptions of false beliefs and statements.
66 McDowell, “Falsehood,” 130. The label “crude” applies to the approach to statements found in
Euthydemus 285–86 and 283e7–284c6, mentioned in Section 3.1 of this chapter; crude because it leaves
no room for a distinction between naming and saying—and hence none for false statement.
67 “So it seems to him that when we try to capture the falsity of ‘Theaetetus is in flight’ by saying that
it represents in flight, which is not (in relation to Theaetetus: given the mistake, the addition does not
help) as being, we must be talking nonsense” (McDowell, “Falsehood,” 133). Note the clause in
parentheses.
334   Lesley Brown

But this overlooks two salient features of the discussion of statement and of false
statement. First, it is Def A that is prominent when the ES pointedly moves the discus-
sion on to its final stage (260–64). At 260c3–4, he says, “because judging or saying things
that are not—that’s what falsehood in thought and statements is, surely,” and at 260d1–2,
he says that the sophist denied the existence of falsehood since “no one can judge or say
what is not.”68 These prominent descriptions of false saying and judging set the stage for
the final push. The more complex Def B does not get a further mention. Indeed, we may
hazard that Plato considered that Def B is not, at bottom, problematic but a rather
insightful definition of falsehood, provided, of course, that we add “about so-and-so.”
A solution to the “crude picture” is, contra McDowell, just what is needed, and it is
what we get.69
McDowell also overlooks how strongly the account of falsehood emphasizes the need
for a statement being about something. We saw how the ES stresses, apropos of his two
sample logoi, that they are about Theaetetus. Furthermore, after his first pass at the
account of true and false logos, 263b, discussed above, the ES re-emphasizes, first, that
the false one is a logos; second, that it is about something; and third, that it must be
“yours”—that is, it must be of or about Theaetetus.70 He then moves on to the Second
Final Formula, whose first words are “about you.” We can safely reject McDowell’s analy-
sis, then, and restore the view that Plato is concerned to combat the so-called crude
picture, and that a crucial move in doing so is to insist that a logos, whether true or false,
must be about something.71
Now one might think, with McDowell, that the second feature of the account must be
the more important, given the centrality of not being/is not in the architecture of the
Middle Part. But remember that the ES offered to vindicate not being, to clear away mis-
understandings that made talk of it contradictory. It should not therefore be one of his
aims to dispense with it entirely, and, indeed, the project of dispensing with “not,” in an

68 McDowell, “Falsehood,” 130 n.31, correctly queries Owen’s claim that back in 237b7–e7, the puzzle
(about legein to mē on) is “a version of the familiar paradox.” But he overlooks 260c3–4 and d1–3, where
the locution legein to mē on (or a variant) is used to designate false speaking.
69 A further objection to McDowell’s reading, on which it is the complex Def B, not the simple Def A,
which frames the problem about falsehood, is the following. On his reading, any occurrence of the
phrase “is not” was held to be problematic, until the “Eleatic error” of interpreting this as “has the
contrary of being” or as “utterly is not” is scotched (132). But in that case, even the equivalent formula
for true negative statement implied at 240e1, “judging that what is not is not,” would be suspect. But no
such aspersions are cast against it.
70 The use of the “possessive” pronoun sos 263c7 (and emos, 263a6) has puzzled commentators, and
some (including McDowell, “Falsehood,” 130, and Frede, “False,” 416) believe that this shows that Plato is
invoking the “old” concept of a logos as belonging to someone, by “putting that person into words.” Frede
writes: “Given that the language of ‘about’ is perfectly clear, and given that the language in terms of
possessive pronouns is neither ordinary nor natural, it is difficult not to see in it an allusion to the way of
thinking about statements underlying the antilogia argument.” But “your logos” can mean the logos that
describes you, just as “your picture” is the one that depicts you: that is, we have an “objective” use of the
pronoun, not a true possessive, so we need not find the usage puzzling. Cf. Szaif, Wahrheit, 464.
71 If the emendation to peri tou is accepted, then the second Formula starts with “about something,”
see note 60 of this chapter.
 The Sophist on Statements, Predication, and Falsehood   335

account either of negation or of falsehood, is bound to be a hopeless one, whose success

is at best illusory.72 Note further that in the neglected Second Formula for Falsehood,
the ES is happy to use both “different” and “is not,” when he describes a false statement as
both “saying what is different is the same” and “saying what is not is.” Note also, what
I remarked above, that the entire Second Formula begins prominently with “about you.”
If we consider the whole passage in which statement and false statement are discussed,
only a tiny portion of it—just a few lines—offer the paraphrases that dispense with “is
not” and rephrase the account in terms of the different. Although I believe we can inter-
pret Plato’s intentions here, by appealing to what I have called the “incompatibility range
interpretation,” I do not think it is, for him, the chief lesson he wants to convey. The chief
lesson is the one about the kind of interweaving a statement is, with functionally differ-
ent parts: this is what allows something both to be unambiguously a logos, about some-
one or something, and to say something false (something that is not) about the subject.
Both in the section devoted to the Late-learners’ problem and its solution, and in the
discussion of falsehood, Plato is concerned to disclose the nature of statements, particu-
larly predicative statements, and to stress that some parts of a logos have a function other
than that of naming. Furthermore, for this very reason, a logos itself is neither a name
nor a string of names. Plato’s new account, in emphasizing that a logos is a special “weav-
ing together” of terms with different roles, is of major importance. In their different
ways, both sections I have discussed make these key points and thereby enable some old
puzzles—ones that can be read both as eristic teasers and as revealing deeper philosophical
problems—to be finally put to rest.73

Bibliography
Ackrill, J. L. “Plato and the Copula: Sophist 251–59”; repr. in Plato 1: Metaphysics and
Epistemology, ed. G. Vlastos (Garden City, NY, 1971), 210–22.
Ademollo, F., The Cratylus of Plato (Cambridge, 2011).
Bostock, D. “Plato on ‘Is-Not’ (Sophist 254–9),” Oxford Studies in Ancient Philosophy 2 (1984),
89–119.
Brown, L., “Being in the Sophist: A Syntactical Enquiry,” in Plato 1, ed. G. Fine (Oxford, 1999),
455–78.
Brown, L. “Definition and Division in the Sophist,” in Ancient Theories of Definition, ed.
D. Charles (Oxford, 2012).
Brown, L. “Negation and Not-Being: Dark Matter in the Sophist” in Presocratics and Plato
Festschrift at Delphi in Honor of Charles Kahn, ed. R. Patterson, V. Karasmanis, and
A. Hermann (Las Vegas, NV, 2012).

72 Critics of interpretations that invoke incompatibility often object that that notion, in turn, needs
to be explicated with the help of negation and/or falsity. But the very same point can be made about
interpretations that rely on the simple notion of the different, or nonidentical.
73 I am very grateful to the editor, Gail Fine, for her helpful comments. Despite my disagreements
with his interpretations, I have benefited greatly from Paolo Crivelli’s writings on the Sophist and thank
him accordingly.
336   Lesley Brown

Brown, L. “Aporia in Plato’s Theaetetus and Sophist” in Aporia in Ancient Philosophy ed.
G. Karamanolis and V. Politis (Cambridge, 2018).
Burnyeat, M. “Plato on How Not to Speak about Not-Being,” in Le Style de la pensée, ed.
M. Canto and P. Pellegrin (Paris, 2002), 40–65.
Cornford, F. M. Plato’s Theory of Knowledge (repr. London, 1960).
Crivelli, P. “Allodoxia,” Archiv für Geschichte der Philosophie 80 (1998), 1–29.
Crivelli, P. Plato’s Account of Falsehood: A Study of the Sophist (Cambridge 2012).
Denyer, N. Language, Truth and Falsehood (Cambridge, 1991).
Ferejohn, M. “Plato and Aristotle on Negative Predication and Semantic Fragmentation,”
Archiv für Geschichte der Philosophie 71 (1981), 257–82.
Frede, M. Prädikation und Existenzaussage (Göttingen, 1967).
Frede, M. “The Literary Form of the Sophist,” in Form and Argument in Late Plato, ed. C. Gill
and M. M. McCabe (Oxford, 1996), 135–51.
Frede, M. “Plato’s Sophist on False Statements,” in Cambridge Companion to Plato, ed. R. Kraut
(Cambridge, 1990), 397–424.
Gill, M. L. Philosophos: Plato’s Missing Dialogue (Oxford 2012).
Gosling, J. Plato (London, 1973).
Hestir, B. Plato on the Metaphysical Foundation of Meaning and Truth (Cambridge, 2016).
Hoekstra, M., and F. Scheppers. “Onoma, rhēma et logos dans le Cratyle et le Sophiste de
Platon,” L’Antiquité Classique 72 (2003), 55–73.
Kahn, C. The Verb “Be” in Ancient Greek (Dordrecht, 1973).
Keyt, D. “Plato on Falsity,” in Exegesis and Argument, ed. E. Lee, A. Mourelatos, and R. Rorty
(Assen, 1973), 285–305.
Lewis, F. A. “Did Plato Discover the Estin of Identity?” California Studies in Classical Antiquity
8 (1975), 113–42.
Lockwood, M. “On Predicating Proper Names,” Philosophical Review 84/4 (1974), 471–498.
Mates, B. “Identity and Predication in Plato,” Phronesis 24/3 (1979), 211–29.
McDowell, J. “Falsehood and Not-Being in Plato’s Sophist,” in Language and Logos, ed.
M. Schofield and M. Nussbaum (Cambridge, 1982), 115–34.
Moravcsik, J. M. E. “Being and Meaning in the Sophist,” Acta Philosophica Fennica 14 (1962),
23–78.
Notomi, N. The Unity of Plato’s Sophist (Cambridge, 1999).
Owen, G. E. L. “Plato on Not-Being,” in Plato 1, ed. G. Fine (Oxford, 1999), 416–54.
Ryle, G. “Negation,” Proceedings of the Aristotelian Society (suppl. vol) IX (1929), 80–96.
Sedley, D. The Midwife of Platonism (Oxford, 2004).
Sedley, D. Plato’s Cratylus (Cambridge, 2003).
Sommers, F. “Do We Need Identity?” Journal of Philosophy 66/15 (1969), 499–504.
Szaif, J. Platons Begriff der Wahrheit (Munich, 1996).
Taylor, C. C. W. “Socrates the Sophist,” in Remembering Socrates, ed. L. Judson and
V. Karasmanis (Oxford, 2004), 157–68.
van Eck, J. “Falsity without Negative Predication,” Phronesis 40/1 (1995), 20–47.
van Eck, J. “Plato’s Logical Insights: Sophist 254d–257a,” Ancient Philosophy 20/1 (2000), 53–79.
van Eck, J. “Plato’s Theory of Negation and Falsity in Sophist 257 and 263. A New Defense of
the Oxford Interpretation,” Ancient Philosophy 34/2 (2014).
Vlastos, G. “An Ambiguity in the Sophist,” Platonic Studies (Princeton, NJ, 1981), 270–322.
 chapter 14

The Philebus

Constance C. Meinwald

The Philebus discusses the good human life and the claims of pleasure on the one hand
and a broad cognitive cluster containing understanding, intelligence, and right opinion
on the other in connection with that life. Plato includes extended treatment of meta-
physics and methodology: this is his typical supplement to the procedure of his own
Socratic dialogues, which considered human questions in isolation from other issues.
Some parts of our dialogue are intelligible locally as we read them, and some tricky bits
have benefited by treatment in the secondary literature. Yet the text as a whole is hard to
grasp. It is not clear why the discussion develops as it does and how all this is supposed
to work together. This means we do not understand the characteristic Platonic move of
addressing human questions with the aid of what he takes to be more fundamental
investigations. Moreover, we do not grasp the dialogue as an artistic success. Surely a
work studying good mixtures and thematizing the harmony (that is, the fitting together)
of unlike elements should itself fit together in some intelligible way. But Plato may well
have given the Philebus an obscure unity that we are challenged to find.1 Indeed, the
portions of the dialogue that are hardest to make anything of are introduced in the text
(remarks interspersed throughout 13c6–19b8; 23b5–92) as necessary for the discussions
of pleasure and cognition, which many readers approach directly, without bothering too
much with the apparatus. In this essay I start by discussing the characters and setting of
the dialogue and then provide an overview of the discussion (with a little more detail on
passages that I do not treat later than on ones that the overview serves merely to intro-
duce and locate). Then I give my interpretation of the metaphysics and method pas-
sages, and take Plato’s cue to use these as a basis from which to approach the discussions
of pleasure and reason.

1 Our dialogue fits an idea thematized in M. F. Burnyeat, The Theaetetus of Plato (Indianapolis,
Ind., 1990), xii–xiii and interspersed throughout: that in advanced work Plato sets us challenges that we
must do philosophy to meet.
2 Citations are to J. Burnet, Platonis Opera, vol. 2 (Oxford, 1901).
338   Constance C. Meinwald

1. Characters and Setting

The Philebus is a minimalist dialogue: it contains no action outside the philosophical

conversation, and Plato has not specified when or where that conversation occurs.
By contrast, in previous works3 Plato had often placed philosophical conversation in a
fictional world that had been built up to resemble the real world as it was within living
memory at the time of composition. He often specified settings: the homes of the
wealthy, the law courts, the lush countryside. Also usual was temporal location: after the
first victory of the tragic poet Agathon, at the time of the Great Panathenaea, in the days
leading up to the death of Socrates. The participants in a Platonic dialogue were typically
based on men known to history and often quite fully characterized as well; readers
will have been expected to recognize them. Yet in the Philebus the cast is reduced to a
minimum. There are apparently (16a4–5, 19c4–5, 19d6–e5) unnamed auditors who are
tacit throughout. Philebus himself has retired from official participation by the time the
dialogue starts: his role is so minimal that we cannot even tell if he is based on a real per-
son. He functions within the dialogue as the personification of the position he has
espoused4: holding that pleasure is the human good, he denies that the desirable human
life has any need of thought, and he embodies his position by his limited participation in
the inquiry. Plato thus shows that he is aware, even if Protarchus in his sometimes over-
stated formulations (e.g., at 21e3–4) is not, that the considerations advanced here do not
work on everyone.
Protarchus is taking over the hedonist position as we join the action: note the idiom
that he should pay attention to the position he is to defend lest what is said not be “in
accord with his mind/understanding” (kata noun) (11b1), a touch hinting that Protarchus
has understanding; he will prove reasonable and will be convinced by the considerations
Socrates brings forward in the course of their discussion. There is some mystery about
who this interlocutor might be, but we do have material for speculation. Our character
is “son of Callias” (19b5) and has heard Gorgias (58a7–8); thus he may be the offspring of
the Callias who was a well-known patron of the sophists (Apology 20a4–b9; Protagoras
311a1–2, 314b8–e2, 315c8–d4.) This famous Callias was a relative of Plato’s by marriage
and was at one time the wealthiest man in Greece. He was the father of two sons: we
know the name of only one; thus the other may have been Protarchus.
A son of “our” Callias would exhibit many features common among the youth of Plato’s
fictional world. In particular, the concern of the father for the best education money
could buy, along with the failure of the son despite that to amount to much (if we may

3 The Philebus is a member of the clearly established late group of dialogues. For discussion of
Platonic chronology, see C. Kahn, “On Platonic Chronology,” in J. Annas and C. Rowe, eds., New
Perspectives on Plato, Modern and Ancient (Washington, D.C., 2002), 93–127; C. Meinwald, Plato
(Abingdon and New York, 2016), 18–21.
4 Cf. L. G. Westerink, ed. and trans., Damascius Lectures on the Philebus (Amsterdam, 1959), 6–7:
Philebus represents the zôiôdes (animal type).
 The Philebus   339

assume that he actually failed to earn a place in history and isn’t just anonymous by
­accident), would resonate with the dialogue’s inquiry into the role of rational accom-
plishment in the good human life and what that accomplishment involves. Yet this
­historical person was too young to have conversed philosophically with Socrates.5 Does
this rule out the identification? It could be that Plato does mean him, despite the chron-
ological impossibility; alternatively, it could be that he invents a character about whom
he puts in virtually no details. Before we pursue these ideas, I would like to draw atten-
tion to the way in which the relationship of Protarchus with Socrates is represented.
Despite the fact that he is assigned the position Philebus originally held, Protarchus
seems to me not to be a natural opponent, and not so much to be converted6 in the
course of the dialogue by Socrates, as to have had a relationship with him all along. This
interlocutor is unusually dependent on his questioner: he asks Socrates to answer
questions in his stead (20a1–8 and 28b7–10). He is clingy in a childish way, reminding
Socrates “You agreed to be with us”7 for the purpose of settling the matter at hand;
Protarchus insists at 19d6–e4 on his role as the one who gets to say when Socrates may
be excused and refuses to let him go at 23b2–4.8 At the end of the dialogue, the notion
(in other works typically expressed by Socrates) that there are points that need further
discussion is expressed by Protarchus, as a reason here not (as has been usual) for
resuming discussion on another occasion but for not releasing Socrates now (67b11–13).
Though our window on the action closes, the idea seems to be that Socrates and
Protarchus are continuing their conversation.
These observations can be gathered in the service of a suggestion that would also help
answer the question why Plato makes Socrates the leading philosopher in this work
despite his not having this role in other dialogues of the late group. Scholars have already
pointed out that the overall topic and dialectical style of the Philebus hark back to the
interests of the Socrates of the early dialogues.9 But perhaps the present observations
concerning the literary elements may add something. Plato—obsessed throughout his
compositional career with the death of Socrates—now brings him back in his final
­starring role. In fact, Plato had long ago followed his Phaedo depiction of the death of

5 I have relied on D. Nails, The People of Plato (Indianapolis, Ind., 2002), 68–74, 257 for the historical
information in this and the previous paragraph, though Nails herself thinks Protarchus is not a son of
our Callias.
6 A term thematized by D. Frede in, for example, “Disintegration and Restoration: Pleasure and Pain
in the Philebus” [“Disintegration”], in R. Kraut, ed., The Cambridge Companion to Plato (Cambridge,
1992), 427. In the case of scholars such as Frede with overlapping publications on our dialogue, I favor
the accessible for the convenience of readers.
7 Note sunousian at 19c5. Literally “being with,” but used commonly for sexual intimacy, this is a
favorite way for Plato to refer to philosophical engagement—for example, at Lysis 223b3; Laches 201c2;
Protagoras 310a2, 335c3, 347e1; Gorgias 461b1; Symposium 172b7, 172c1; Theaetetus 150d4, 151a2–3;
Timaeus 17a5; and Sophist 217e1.
8 Compare the role of Callias in marshaling discussion at Prot. 317d5–e2, 335c8–d5, 338b2–3, and
362a1–3.
9 So D. Frede, “Disintegration,” 431–33. Note that the Socrates of our dialogue, when he speaks of
dialectic, has taken on board a conception of it that Plato has continued to develop/make explicit in
subsequent work.
340   Constance C. Meinwald

Socrates with works showing the character in the prime of his life. These were set
­fictively before the Phaedo, and touches such as the false security enjoyed by Cephalus
in the Republic underscored that this time was truly lost. (The death of his son,
Polemarchus,10 and the ruin of the family’s fortunes in the convulsions of war and revo-
lution were known to everyone by the time of composition.) Plato’s late period presents
another sort of death of Socrates. Not only does the Sophist allude to his actual death,11
but Socrates ceases to be the main figure in Plato’s works: after having been schooled by
the venerable Eleatic in the Parmenides, he is only a minor presence in the Sophist,
Statesman, Timaeus, and Critias, and does not figure at all in the Laws. The Philebus
restores to Socrates the starring role: we might almost think that it imagines how he lives
on in “Platonic heaven.”12 In this virtually bare representation of philosophical discus-
sion, unlocated in place and time, with a partner not available in real life, Socrates
embraces all of Plato’s recent metaphysical suggestions and more; in this fantasy he won’t
go if “we” don’t wish it.
If this suggestion is too fanciful, we can account for a subset of our collected observa-
tions with a thought that seems true to me in any case. The Philebus, like other late dia-
logues, generally leaves out entertaining elements that help draw people in because Plato
when composing them assumes advanced readers. We are no longer like listeners at
intro lectures but have, so to speak, taken ethics and been participating in the metaphys-
ics seminar. Thus we have in this latest of completed dialogues philosophical content
stripped down to its essentials.

2. The Course of the Discussion

As we have noted, when the curtain rises Protarchus is taking over the position
­originally held by Philebus. The theme question and the rival answers given to it are put
with some variation, compounded by translators’ choices, but the force of the position
Protarchus holds is clearly enough that pleasure alone makes human life good.
(Question and answers are first formally stated at 11b4–c3 and summarized at 60a7–b4;
cf. 11d4–6, 13a8, 13b7, 13e4–6, 19c4–d6.) Socrates, on the other hand, maintains that
understanding, intelligence, right opinion, and such are . . . we think he will say the
human good, but he pulls back and says more carefully “better than pleasure” (11b9).13

10 Made famous by his brother in “Against Eratosthenes” (Lysias XII).

11 By opening with the statement that this is the next day’s meeting agreed to at the end of the
Theaetetus, when Socrates was off to meet the indictment of Meletus.
12 The possibility that after death one has the opportunity to philosophize at leisure was articulated
by Socrates at Apology 41b5–c4.
13 Socrates foregrounds different members of his cluster, depending on context, and switches
between the infinitives noein and phronein (11b7) and the nouns nous and phronêsis (used more
pervasively, e.g., at 13e4; cf. 60b4). The idea, though, is uncontroversially that his cluster includes what
we might call “correct cognition” and the “cognitive faculties,” or “reason and its manifestations.”
 The Philebus   341

The motivation for this careful formulation will soon become clear: Socrates will argue
that our life requires a mixture of elements to be desirable and fit for a human being.
When one reads through the dialogue for the first time—or, indeed, for the tenth or
twentieth—one may well be puzzled by the way in which the discussion now starts to
skip around. Socrates, having claimed that pleasures are many and unlike each other
(as are forms of knowledge), announces a “One/Many Problem” (14c7–15c3) he consid-
ers significant and introduces (16b4–18d2) a method from “some Prometheus” that
helps to deal with it. Socrates says he has long been a devotee of this method, that it is the
mark of the dialectician, and that it is the source of all discoveries of technê. (As we will
see, technê in our dialogue is treated as (at least a kind of) knowledge, though it ranges
broadly from the sort of case we can designate in English as “expert skill/craft” through
the sort we might rather call an “art”—especially as in our usage “arts and sciences.”) The
Promethean Method is based on the fact that things have peras (limit or definition) and
apeiron (the unlimited or indefinite) in them, though Socrates does not explain what
this means. The method involves dividing subjects into their subkinds and knowing
how the subkinds combine with each other; Socrates says he and Protarchus should use
it on their candidates (18e8–19a2; cf. 19b2–4). Protarchus confesses his inability to do
this and asks first Philebus, who does not reply (19a3–b4), and then Socrates to do it for
him (19c1–e5, 20a1–8). But Socrates, instead of applying the method, suddenly recollects
an argument to show why neither of their candidates on its own can be the human good
(20b3–22c4).
For pleasure the key intuition is that a life of the greatest pleasures in which one did
not even know what one was experiencing, and had no awareness of one’s past or reck-
oning about the future (all of which are functions of the candidate of Socrates), is not
really desirable for a human being.14 Socrates calls this the life of a mollusk (the higher
animals might be thought to have some self-awareness or memory). Scholars are divided
about exactly why Protarchus here rejects the life devoid of any cognitive elements. He
could be recognizing that some kind of understanding, at least of one’s own experiences
and life trajectory, has value of its own that makes it necessary for a good human life.
Or he might, without thinking that any cognition has value in itself, be registering that
some distinctively human pleasures (enjoying memories of a childhood birthday party,
say) depend on our cognitive powers. And he might think that being deprived of these
pleasures cannot be compensated for by any amount of the sort of pleasures a mollusk
enjoys.15 In either case, this turns on its head a hedonist argument pattern. Hedonists
sometimes move from what they identify as the goal for all other observed animals
to conclude that the same thing is our natural goal as well. Here the thought is the

14 Note the resemblance between the setup here—that the human good must by itself be sufficient to
make a life desirable—with what we find in Aristotle (NE 1097b7-8, 1097b14–21).
15 For the first sort of view, see J. Cooper, “Plato’s Theory of Human Good in the Philebus” [“Human
Good”], Reason and Emotion (Princeton, N.J., 1999), 152 (originally published in Journal of Philosophy
74 [1977], 713–30) and T. Irwin, Plato’s Ethics (Oxford, 1995), 333–35; for the second sort of view, see
E. Fletcher, “The Divine Method and the Disunity of Pleasure in the Philebus” [“Disunity”], Journal of
the History of Philosophy 55 (2017), 192–93.
342   Constance C. Meinwald

opposite: since a human has different capacities from a mollusk, a life with only goods of
which a mollusk is capable is not suitable for a human.
Maintaining symmetry between his own candidate and that of Protarchus, Socrates
points out that a life of reason alone would also not be human; in this case, it might
rather be divine, but, after all, we are not gods.16 For us, then, a mixed life containing
both reason and pleasure must be the best, though this does not yet say whether all
types of cognition and all pleasures should be included. However, this victory of the
mixed life shows already both that neither the candidate of Protarchus nor that of
Socrates can be the human good, though that of Socrates may still be, as he had so care-
fully put it, better than pleasure. The rest of the dialogue is supposed (22c5–e3) to
explore each of the original candidates for the purpose of determining the relative
responsibility of each for the good life. The examination will turn out to prepare us to
make the mixture that is that life.
For his exploration of pleasure and reason, Socrates says he needs “new equipment,
though perhaps some will be the same” (23b6–9); this introduces a Four-fold Division of
things into peras (limit or definition), the apeiron (the unlimited or indefinite), what is
mixed from them, and the cause of the mixture. We will soon turn to understanding this
passage and its relation to the Promethean Method. But for now, the thing to note is the
way in which our topics follow one another. The introduction of the four fundamental
categories leads immediately to the placement of pleasure (27e5–7 with 31a8–10) in the
category of the apeiron, partly perhaps through misunderstanding on the part of
Philebus (evident from the reaction of Socrates at 28a1–4), but this is enough to go for-
ward with for now. As we will see, subsequent discussion will be applicable to this point.
Phronêsis (intelligence/wisdom), epistêmê (knowledge/science), and nous (under-
standing/mind), meanwhile, are placed in the category of the cause [of the mixture] in a
passage (28a4–31a317) that self-consciously takes off from the views of “all the wise”
(28c6–7). In Greek antiquity, strict regularity did not have the almost automatic associa-
tion with machinery it came to have later; the order observed in the heavens could
instead be thought to indicate the operation of a mind. Socrates and Protarchus accept
the tradition that mind (nous) is the ruler of the cosmos, and the mind in us is the ruler of
our body. Plato famously had made Socrates in the Phaedo (97b8–98c2) complain that
the philosopher of nature Anaxagoras, having raised expectations by the extremely
promising way in which he introduced this entity, then “did not make use of his mind.”
Plato may think to do better here: we will see what the balance of the dialogue
contributes.
Why, someone might ask, if the agenda is to find out which of the two candidates is
more responsible for the excellence of the mixture, does the placing of mind in the cate-
gory of the cause not settle the matter? Surely the cause of the mixture is responsible for

16 Cf. NE 1095b14–22 on pleasure as the goal of animals, and 1177a12–18 with 1177b26–1178a8 on
contemplation as divine.
17 This section tends to elude reconstruction as a formal argument; it may be better to take it as
expressing the tradition of the sages.
 The Philebus   343

it, and this is even clearer in Greek since aition lies behind both English terms. Here it is
helpful to draw on the critical discussion of Anaxagoras in the Phaedo. There Socrates
said that since the natural philosopher had made mind the source of motion in the cos-
mos, Socrates expected that Anaxagoras would explain the good that mind was seeking
to achieve in the unfolding of events, but that this hope was disappointed. So here, that
mind is the cause of one’s activity is all very well, but a satisfying explanation must go on
to tell more about what one has in mind in acting. The thing responsible for the desirabil-
ity of the mixture in this sense may at this point be anything.
To return to our synopsis, Socrates next says that in addition to placing each of the
candidates in its category, they must investigate in what each occurs and how it comes
about (31b2–4). Why they must investigate this he does not say here. Taking up the can-
didate of Protarchus, Socrates embarks on a long discussion (31b4–53c3), which starts
from the idea of the harmonious balance of a living thing’s constituents. When this is
disrupted, we have pain, and restoration/improvement of harmony is pleasure (31c2–d10).
At 43b1–c6 Plato will add a refinement: only felt disruption of a creature’s constitution is
pain; felt restoration is pleasure.18 We are treated to a lengthy discussion of varieties of
pleasure—though it does not amount to a proper structure of Platonic divisions, because
the varieties we get here both cut across and run into each other. The discussion as a
whole emphasizes the negative: ways in which the pleasures under discussion may be
called false, ways they are mixed with pain throughout their duration or involve pain as
a prerequisite, their unseemliness. Only in the final pages of the discussion do we get a
brief passage concerning special pleasures that have no association with pain and are
also called pure and true (50e5–53c2).
This is all capped by a powerful argument returning to the idea that pleasure cannot
be the good. We will return to it later in this essay; for now I note that it works by com-
bining the contrast between genesis and ousia (becoming and being) with the notion of
some things being for the sake of others (53c4–55a11).19 Next (55b1–c1) Socrates claims to
draw several absurd consequences from the claim that pleasure is the good. These
include that by whatever margin people have pleasure, they would stand out to that
degree for virtue (human excellence), while even the most outstanding persons, when
suffering, would be bad. Though the argument here is the merest sketch, Protarchus
goes along with Socrates.
With the discussion of pleasure over at long last, we get a fairly compact discussion
(55c4–59d9) of nous and epistêmê. Not surprisingly since he now wishes to learn what
within knowledge broadly understood is most pure, Socrates reviews a variety of types

18 For the quite prevalent interpretation of this formula as the general account of pleasure, see, e.g.,
D. Frede “Disintegration.” Diverging recent suggestions: T. Tuozzo, “The General Account of Pleasure
in the Philebus,” Journal of the History of Philosophy 34 (1996), 495-513; M. Evans, “Plato’s Anti-
Hedonism,” Proceedings of the Boston Area Colloquium of Ancient Philosophy 22 (2007), 121–45;
E. Fletcher, “Plato on Pure Pleasure and the Best Life,” Phronesis 59 (2014), 113–42 and “The Divine
Method and the Disunity of Pleasure in the Philebus,” Journal of the History of Philosophy 55 (2017),
179–208.
19 Again, the family resemblance to the setup in the NE (1094a1–22, 1097a15–1097b6) is striking.
344   Constance C. Meinwald

of cognition. The passage starts at the low end, distinguishing kinds of technê which it is
natural to speak of using a range of English terms, from “craft,” “expert skill,” and “art”
(including mathematical arts), before culminating with a discussion of dialectic as the
most pure knowledge.
Finally comes the production (59d10–64a6) of our mixture, which turns out to
­consist of every member of the cluster Socrates has advanced, from the purest and
most hegemonic to the most applied and most empirical, necessary “if we are even to find
our way home” (62b8–9). The cognitive cluster is supplemented only by the pure plea-
sures recently described in 50e5-52c2 and those that attend on virtue. This is followed
by reflections (64a7–65a5) on what makes a mixture good. The good is approached
under a triple aspect. We find, mentioned in varying order within the passage: (1) truth,
(2) measure and proportion, and (3) to kalon;20 the candidate of Socrates turns out to be
more closely related to these three than that of Protarchus is. Finally comes a ranking
(66a4–end) of elements responsible for the good life. Some from the cluster of Socrates
make the list in third place and others in fourth; the list continues through the pure plea-
sures (the only ones to make the list at all) in fifth place.21 Then, obeying an injunction
from Orpheus (66c8–10), Socrates and Protarchus cease their song.
The sequence of topics I have just recounted should raise some questions. We should
ask why Socrates says he and Protarchus should apply the Promethean Method to their
candidates and then veers off when asked to do so on behalf of his interlocutor. We
should ask why, if Plato means the Promethean Method and the Four-fold Division pas-
sages to work together, he presents them separately. We should ask whether the extended
discussions of the kinds of pleasure and later of the cognitive cluster represent the
Promethean Method’s technical division of kinds into subkinds: again, if they do, why
does Socrates not do the divisions earlier; and if they do not, then why introduce the
method at all? We should ask how the lengthy discussions of pleasure and later of the
various rational functions relate to the Four-fold Division: since the placing of pleasure in
the category of the apeiron and mind in that of the cause is done immediately on the
introduction of these categories, do the categories bear at all on the subsequent discus-
sions? We should ask how all this prepares for the final mixing and consequent ranking.
To put all these questions in one summary formulation: Why does Plato intercalate the
portions of the text dealing with pleasure and reason and those containing the meta-
physics and methodological equipment in the involved way he does?22
In what follows I do not take up passages in the order in which they occur; we have
already traced out this sequence. Rather, since the metaphysical and methodological

20 It is difficult to render to kalon in English: “the beautiful,” “the fine,” and “the good” are perennial
candidates. Whether or not we use the word “beautiful,” we must resist limitation to the narrowest
aesthetic sense the word tends to have in English today.
21 The brief but insightful overview of K. Vogt, “Why Pleasure Gains Fifth Rank: Against the
Anti-Hedonist Interpretation of the Philebus” [“Fifth Rank”], International Plato Studies 26 (2007),
250–55 and P. M. Lang, “The Ranking of the Goods at Philebus 661–67b,” Phronesis 55 (2010), 153–69
discuss the ranking.
22 For an alternative treatment of some of these issues, see now E. Fletcher, “Disunity.”
 The Philebus   345

equipment is meant to aid in the inquiry into pleasure and reason as contributors to the
best life, I proceed now by laying out how I understand the passages introducing this
equipment. With a detailed reading of these in hand, we then return to consider how
other parts of our dialogue relate to them.

3. Metaphysics and Methodology

I believe the Philebus should be understood in the context of Plato’s initiative to treat
forms23 in accordance with the process of “Platonic division,”24 which divides kinds
into subkinds by the use of differentiae and so produces genus-species25 trees. For us
today it may be easiest to catch onto the idea by thinking of the program of Linnaeus. We
can think of such a scheme as starting with a genus and producing its species by adding
differentiae, then adding differentiae to each species, and so on until lowest kinds are
reached. An account of each of the lower kinds is then available via genus and
differentia(e). Being an animal with a backbone is what it is to be a vertebrate. Indeed,
Plato in the Parmenides and Sophist marked out a special way of making assertions that
express such facts. The totality of such genus-species structures would map out the
underlying structure of reality; the totality of truths articulating the associated accounts
thus would express comprehensive deep understanding of that reality. While this way of
conceiving of forms becomes especially prominent in the late dialogues, it was hinted at
much earlier: Socrates suggested in the Euthyphro that one would have an account of
piety if one could say what part of justice it is (12d5–e4).26
The One/Many Problem of the Philebus arises naturally for someone who takes such
divisions and the entities involved in them seriously: What preserves the unity of a
genus if it is divided into many species?27 Socrates offers the Promethean Method in this

23 The same word, eidos, lies behind both many key assertions we put in terms of forms (e.g.,
Republic 476a5–6) and many central passages (e.g., Phaedrus 265e1–2) translated as concerning species.
24 On Platonic division, see the groundbreaking M. Frede, Prädikation und Existenzaussage
(Göttingen, 1967), or Meinwald, Plato, 263–66, 277–78, and 287–98. V. Harte, Plato on Parts and Wholes
(Oxford, 2002), 6, is agnostic about whether the kinds of the late dialogues are forms. M. Miller, “The
God-Given Way,” Proceedings of the Boston Area Colloquium in Ancient Philosophy 6 (1990), 329–59,
denies that the Philebus continues Plato’s tree program.
25 It is important to realize that “genus” and “species” in connection with Plato and Aristotle do not
refer to fixed levels in a classification tree; rather the pair can be used whenever one wishes to indicate
that one kind (the species) is a subkind of another (the genus).
26 The stylistic choices of scholars in capitalizing or not words such as “form” and “justice”
correspond to nothing in Plato’s Greek.
27 For my views on this passage, see C. Meinwald, “One/Many Problems,” Phronesis 41 (1996), 95–103.
Cf. F. Muniz and G. Rudebusch, “Plato ‘Philebus’ 15b: A Problem Solved,” Classical Quarterly 54 (2004),
394–405; R. Dancy, “The One, the Many, and the Forms: Philebus 15b1–8,” Ancient Philosophy 4 (1984),
160–93. A. Barker, “Plato’s Philebus: The Numbering of a Unity” [“Numbering”], Apeiron 19 (1996),
161–64, makes a suggestion similar to mine. I am glad to find that Barker’s essay and his “Text and Sense
at Philebus 56a” [“Text”], Classical Quarterly 37 (1987), 103–09, definitive concerning discussions of
music, point beyond music to an interpretation of the dialogue that is consonant with mine.
346   Constance C. Meinwald

connection, yet to some extent—the extent to which it uses an apparatus of division—it

would seem to repeat the basis for the problem. However, there is also material that
offers promise of a solution. Not only must someone with the technê associated with a
subject master all the divisions of a given domain down to infima species as well as ways
of combining the elements so produced. Our passage also emphasizes the opposite
direction, going up to grasp the one, that is, what unifies a field, illustrated with the
foundational work of the Egyptian god Theuth on letters/phonemes. This is in turn con-
nected with seeing that it is impossible to understand any of the elements in isolation,
and that a single art is set over all of them (grammar in the case of the letters).28
A striking addition the Promethean Method passage makes to previous descriptions
of division is that now, for the first time, the pair of terms peras and apeiron is introduced
in connection with it: the fact that things have peras and apeiron in them is said to be
in some way foundational, and we are told that after lowest kinds have been reached
we release them into the apeiron (though our passage underdetermines what exactly
this means). The new29 pair of terms is not really doing much in the Promethean
Method passage taken by itself, but they are placeholders for a great deal of foundational
information that will be introduced when we come to the Four-fold Division. (To preview
the results I obtain by this approach: each Promethean subkind is a member of the
Four-fold Division’s mixed class, and as such turns out to be a metaphysical combina-
tion of a desirable ratio (from the category of peras, i.e., limit) with an originally apeiron
pair of opposites.)
There is, in fact, disagreement in the secondary literature about whether peras and
apeiron pick out the same things in the Four-fold Division passage as they did in the
Promethean Method.30 Since Socrates himself makes the identifications (23c9–12), we
should work from that: otherwise we ignore a unifying element Plato has made explicit.
In effect, he is challenging us to come up with an interpretation of the two passages
that respects the constraint that peras and apeiron are the same in both, and the inter-
pretation I offer does this.31 This approach takes the facts that peras and apeiron in the
Four-fold Division combine to produce its mixed class and that the kinds studied by the
Promethean Method have peras and apeiron in them to mean that the Promethean
Method’s kinds are members of the Four-fold Division’s mixed class. The latter passage
includes as illustrations of this class music, health, fineness (kallos), strength, and “many
fine things in the soul”32 (26a4, 26b5–7). On this reading, the Four-fold Division is a

28 For detailed discussions see D. Frede, Plato Philebus (Indianapolis, Ind., 1993), xxv–xxx, 12 and
now E. Fletcher, “Disunity,” 189–90.
29 New, that is, in Platonic descriptions of division. I traced the relationship of the Philebus to earlier
philosophy in C. Meinwald, “Plato’s Pythagoreanism” [“Pythagoreanism”], Ancient Philosophy 22
(2002), 87–101.
30 The approach according to which they do not is associated with G. Striker, Peras und Apeiron
(Göttingen, 1970), esp. 80–81. For criticism, see J. C. B. Gosling, Plato Philebus (Oxford, 1975), ix, 186,
195–96.
31 For details, see C. Meinwald, “Prometheus’ Bounds,” in J. Gentzler, ed., Method in Ancient
Philosophy (Oxford, 1998), 165–80, and Meinwald, “Pythagoreanism.”
32 Cf. Republic 443c9–e2, 591d1–3, and 430e1–4.
 The Philebus   347

study of wide scope, giving us new insight not just into the subjects of the Promethean
Method but also into their constituents.
In the earlier passage, the apeiron can naturally be understood as exhibiting a blurred
condition in which kinds run together with no significant demarcations. Below the level
of scientifically distinguished species we must admit a wash of variety not so distin-
guishable: below the specific vowels is a continuum of sounds into which even Theuth
and Henry Higgins must release them; below the specific musical intervals are a blur of
indefinitely many other relations in which pairs of notes may stand to each other; below
the lowest division into kinds of cats there is still indefinite variation in softness of fur,
shape of eye, and so on—even at the level of types.33 The Four-fold Division’s treatment
of the apeiron is consistent with this but now focuses explicitly on pairs such as the hot
and the cold, the wet and the dry, the high and the low. Plato may well have thought that
what makes each of these pairs an apeiron is that its members, left to themselves, run
together. To see this we may apply a pattern of thought familiar from other dialogues.
The temperature of 40 degrees Fahrenheit is cold in the Ithaca summer yet warm in the
Chicago winter, so the hot and the cold run together. The lowest soprano sound would
be high for a bass, so the high and the low run together. And so on.
Yet the members of each such pair are capable of being distinguished and made
definite for the purposes of some art or science, and then being measured out and set
into good balances; the factor of peras is responsible for all of this organizing.34 Medicine
must specify what, for its art, the wet and the dry are, and it knows what ratios to impose
on them to produce health. Note the correctness of this traditional view in virtually all
eyes, from the Hippocratic corpus to the present day. Being dehydrated, for example, is
not a matter of the absolute amount of water in one’s body but of the proportion of water
to dry elements: a football player needs a greater volume of water than a ballet dancer,
yet in both cases the ratio between water and ash in the healthy body is the same. Music
is a star example in which, having specified what they took the high and the low to be,
theoreticians by Plato’s time had worked out which ratios should govern them to pro-
duce desirable intervals and, in turn, desirable combinations of intervals into modes.
Harmonious intervals were famously characterized by special ratios.
The factor of peras, which in the Promethean Method we deduced would be whatever
marked off kinds from each other to make them definite, is revealed in the Four-fold
Division to be proportion; Socrates first seems to make all ratios members of this cate-
gory, but then goes back and adds phraseology that gestures at the idea that some ratios

33 The suggestion that apeira in the Promethean Method are at the level of types is due to
J. M. Moravscik, “Forms, Nature, and the Good in the Philebus,” Phronesis 24 (1979), 81–104. The
overwhelming majority of scholars have assumed that the unlimited multitude of sensible particulars
participating in each form is in question.
34 I don’t think to de emmetron in 26a7 should be understood to pick out members of the mixed class
as J. Cooper seems to suppose (“Human Good,” 152), or translated as “what is measured” (Gosling).
Compare the renderings of the relevant stretch of Greek (italics added) by Diès (“pour y mettre à la fois
mesure et proportion”), Hackforth (“create measure and balance”), and D. Frede (“establishes moderation
and harmony . . .”). Note also the assignment of different places in the final ranking to measure and
proportion (66a6–b3), whose rationale must be that measure is prior to proportion.
348   Constance C. Meinwald

are better than others.35 Music cannot tolerate random pitches, or even pitches reflecting
ratios between any chance pairs of integers, but results when the high and the low have
the special proportions famously discovered by the Pythagoreans imposed on them.
This school took the lead in supposing the ratios that lie behind music to be fundamen-
tal throughout the cosmos; the program of the Philebus follows this lead. Thus while
discussions of division in other dialogues also urge us to look for accounts, the Philebus
envisages that these accounts take a very particular form.36 “Promethean” accounts as
I understand them will be mathematized in a Pythagorean way: subkinds of a genus will
be distinguished from each other by the characteristic proportions each shows between
an underlying pair of opposites, with ratios that are better for a mathematical reason
being responsible for the desirable qualities of the mixtures they determine.

4. Some Thoughts on Dialectic

We have seen that pleasure is placed in the category of the apeiron, and this might seem
at first to be inconsistent with my understanding of the Promethean Method passage.
After all, the method is supposed to be a response to a One/Many Problem that Socrates
introduced in connection with the variety of pleasures, and Socrates asks Protarchus to
apply the method to pleasure, which the interlocutor rightly understands to mean that
he should give its species and subspecies. Yet how can this be, if the method only applies
to members of the Four-fold Division’s mixed class, while pleasure is apeiron? I think
that considering this will help us with one of the questions I indicated earlier, the first
one to do with the apparent jumping around of topics, especially if we now avoid the
trap of making the whole thing into a string of assertions by Socrates, regarded as Plato’s
spokesman. The piece represents a dialectical exchange, and we should respect the spe-
cial roles that Socrates and Protarchus have in connection with the examination of the
position the youth is supposed to maintain. Thus it is not that Plato points out the vari-
eties of pleasures, says they instantiate the One/Many Problem, gives the Promethean
Method for dealing with that, proposes using the method on pleasure, and then veers
away without doing so.

35 Compare the criticism at Republic 531c1–4 of those who fail to inquire into which numbers are
concordant and why. A. Barker, “Ptolemy’s Pythagoreans, Archytas, and Plato’s Conception of
Mathematics” [“Mathematics”], Phronesis 39 (1994), 113–35, identifies someone who isolated a
mathematical criterion according to which some ratios are better than others; the Philebus seems to
presume some such thing.
36 On the legacy of the Philebus in accounts such as those distinguishing subspecies of birds by talon
length, understood in terms of the long and the short, see J. Lennox, “Kinds, Forms of Kinds, and the
More and the Less in Aristotle’s Biology,” in A. Gotthelf and J. Lennox, eds., Philosophical Issues in
Aristotle’s Biology (Cambridge, 1987), 339–59. For the standard scheme for understanding metrical
forms as invoked at Republic 400a3–c5, see J. Adam, The Republic of Plato (Cambridge, 1902; 2nd ed.
1963): the dactyl and spondee belong to the kind where the ratio between rise and fall (the two parts of
a foot) is 1:1; other metrical forms are characterized by 2:1 and 3:2.
 The Philebus   349

We can retell the sequence with greater dialectical sensitivity as follows: Protarchus
and Socrates each have a candidate, and Protarchus in particular maintains that pleasure
is the good. Socrates supposes that the human good will be treatable by the method,
which applies to all that is scientifically intelligible. Then he asks Protarchus to give the
Promethean treatment of pleasure; Socrates need not suppose himself that this is pos-
sible but requires dialectically that if pleasure indeed holds the position Protarchus
assigns to it, then it should be treatable in this way. The inability of Protarchus to carry
out the task thus might indicate not so much personal inadequacy as the fact that the
task cannot be carried out. And the awareness of Socrates that a true genus-species
treatment of pleasure is not possible would explain why he does not do the job for his
companion here but proceeds, instead, to introduce the idea that neither pleasure nor
cognition by itself can be the good.
I also think that embracing the idea that pleasure as an apeiron cannot be treated by
the Promethean Method has another advantage when it comes to the actual discussion
of varied pleasures. It makes the difficulty (which will manifest itself soon) of seeing how
to parcel that discussion into neat sections covering exclusive subtopics neither a com-
positional failure on Plato’s part nor an interpretative one on ours. It is rather a reflection
of the character of the pleasures. And of course the fact that pleasure cannot be treated
by the Promethean Method,37 though that is the way to understand everything that is a
subject of technê, is a significant step in Plato’s “degradation” of pleasure.38
On the question why Plato introduces the Four-fold Division and the Promethean
Method separately if he means them to work together, I suggest that this presentation
highlights what is being added in our dialogue. The Promethean Method passage, as far
as describing division goes, is fairly standard: it introduces the Pythagorean terms but
does not yet do anything with them. So far, Platonic division looks about as we already
thought. Then if we work out an understanding of peras and apeiron in the Four-fold
Division that is tailored to fit both passages, this adds foundational information not
­present in other dialogues. The whole notion that genera and species—our old friends,
the forms!—are made up of peras and apeiron is new and important; the use of a scheme
where the apeiron is an underlying blurred pair of opposites while peras is the factor of
proportion is also new.39 This proposal allows Plato to suggest a way of integrating
the mathematics he emphasized in the Republic (522b2–531d4) with his forms, thus

37 Fletcher in “Disunity” also holds that pleasure cannot receive a full treatment by the method; this
is because the discovery phase shows pleasure does not have the requisite unity.
38 This apt word is from D. Frede, “Disintegration,” 429.
39 It is attractive to think that the Philebus contains the scheme Aristotle had in mind when he
reported (Metaphysics 987b18–988a17) that Plato made the Forms from the One and the indefinite
dyad. Cf. K. Sayre, Plato’s Late Ontology (Princeton, N.J., 1983), 13 and throughout. To see why this
emerges on my interpretation, reflect that taking the apeiron as picking out blurred pairs of opposites
connects the Philebus scheme with the notion of the indefinite dyad. I take peras as amounting to good
or harmonious proportion, and there is evidence that Plato associated goodness, harmony, and
unification. See M. F. Burnyeat, “Platonism and Mathematics,” in A. Graeser, ed., Mathematics and
Metaphysics in Aristotle (Bern, 1987), 213–40; Barker, “Mathematics.”
350   Constance C. Meinwald

revealing their study to be not just a matter of mystical revelation but one that holds out
promise of an understanding at once articulable and mathematized, scientific and deep.

5. Pleasure

We turn now to the discussion of pleasure.40 As we have noted, the initial characterization
placed pleasure as apeiron. I mentioned before that this may not have been for the right
reason. But we can now return to consider for ourselves whether and why the idea is cor-
rect. On the interpretation I have developed, for the pair pleasure and pain to be apeiron
means that left to themselves they blur together and are not clearly distinguished, while
for the “subkinds” or “varieties” of pleasure to exhibit apeiria means that they run into
each other. While a situation analogous at this level of description holds also of opposite
apeiron pairs such as hot-cold and high-low, as well as of the blurred range of types (e.g.,
of pitch intervals) that lies below the level of the lowest species revealed in a successful
Promethean genus-species division, the dialogue will show us that there is a crucial
difference between such cases and that of pleasure. The success of Promethean practitio-
ners such as Theuth is based on (and in turn reveals) the fact that in a case like that of
music, it is possible to establish, for example, what the high and the low each are for the
purposes of the art, and to recognize desirable balances between them that mark off
the harmonically significant and desirable musical intervals. By contrast, though our
discussion does turn up some unmixed, pure pleasures, these aren’t combined either
with pains or with any opposite (newly demarcated) pleasures in favored ratios to gener-
ate desirable species. Indeed, no anatomization of genuine subkinds of pleasure emerges
in the discussion for us to understand before we release them into the apeiron. Let’s now
look at our text in a little more detail.
The investigation into how pleasure arises starts at 31b2 and introduces the ideas of
the harmony of a creature’s natural constitution, the pain when this is disrupted, and the
pleasure at restoration—note that this pleasure cannot even be approached without
mentioning first the pain that is its precondition. We enjoy eating, for example, only
while we are still hungry. The very account of pleasure this section introduces must
involve mention of the prerequisite pain. While this type of pleasure may still be a nec-
essary part of a good and healthy life, many parts of the coming discussion function to
throw a pejorative light on as many pleasures as possible and to make most of the unnec-
essary ones seem undesirable.
In the meantime, discussion of this joint pleasure of the body and soul together leads
into consideration of pleasures belonging to the soul alone—those of anticipation. Since
anticipations can be true or false, this in turn leads to consideration of the notion of false

40 For the context of the contemporary activity that may underlie what we find both here and in the
NE, see A. Diès, ed. and trans., Platon Philèbe, 2nd ed. (Paris, 1978; 1941 orig.), liii–lxx.
 The Philebus   351

pleasure (discussion starts at 36c6).41 This starts with consideration of false pleasure in
nonexistent facts (37e5–40d10). Of course, only some pleasures of anticipation are false
pleasures taken in nonexistent facts. Conversely, note that only some but not all plea-
sures false in this way are pleasures of anticipation: the nonexistent states of affairs in
question may also be in the present or the past.42
This discussion is succeeded by that of other situations in which pleasures are rightly
called “false.” Socrates discusses pleasure overestimated because of perceptual illusions
to do with the proximity of pain (41a7–42c3). Next he turns to supposed pleasure, which
is really only the cessation of pain (42c5–44a10). Finally comes discussion of unseemly
and intense pleasure whose precondition is the difficulty of addressing a painful and
urgent desire (44b6–50e2). Socrates makes the point that within the cluster of intense
and unseemly pleasures some are mixtures of pleasures and pains of the body, some
are mixtures of affections of the body with those of the soul, and some are mixtures of
affections of the soul alone (47c1–e3).
Plato has been writing throughout the dialogue so as to decrease the appeal of pleasure,
but in the final discussion of false pleasures—the unseemly and intense ones—he turns
the rhetoric up to maximum. First (46d9–e3) he details the grossly unpleasant case of a
putrid inner itch that is extremely hard to get rid of and so sets up an intense pleasure
that comes when rubbing and heat finally bring relief. And he follows this (47a3–b7)
by the case of sexual activity leading to orgasm, assimilating the sexual case to the dis-
gusting one. He then goes on to show how emotions—both in the theater and in the
“tragicomedy of life” (50b2–3)—essentially involve a mixture of pleasure and pain.
The discussion suggests that all emotions are pathological (47d5–50d2).
In all the discussions of false pleasures but the first (of pleasures taken in nonexistent
facts), the inextricability of pain and pleasure is evident from the start. In fact, Socrates
shows these pleasures too have that feature when he says that they illustrate the second
subpossibility of mixtures of pleasure and pain mentioned in the discussion of unseemly
and intense false pleasures (47c3–d3).43 This is another indication of the running
together of pleasure and pain. It is another indication as well of the way this discussion
fails to discover true Platonic divisions.
So we have many manifestations of the unsuitability of pleasure for properly techni-
cal Platonic divisions here: the treatment so far of pleasures features a pervasive running
together of pleasure and pain as well as a lack of true divisions in the “subtypes” under
examination. The “subtypes” overlap and their “distinguishing marks” cut across
each other.

41 See R. Hackforth, Plato’s Philebus (Cambridge, 1972; published under a different title in 1945),
69–98; D. Frede, “Disintegration,” 442–52; E. Fletcher, “Plato on Incorrect and Deceptive Pleasures,”
Archiv für Geschichte der Philosophie 100 (2018), 379–410.
42 D. Frede, “Rumplestiltskin’s Pleasures,” Phronesis 30 (1985), 151–80, illustrates the type of false
pleasure in question in this section with the well-known tale, and fills out the case that the relevant
nonexistent facts may be present or past as well as future. (Imagine a woman reveling in what she takes
to be her honeymoon with her true love, when in actuality she is with a conman.)
43 47c–d invokes previous discussion at 41c–d, which in turn refers back, apparently to 36b.
352   Constance C. Meinwald

At long last, Socrates and Protarchus turn to pure, true pleasures (50e5–53c2). These
are unmixed (50e6; cf. 51b6, 51e2–3, 52b6–7) with pain, and hence must correspond to
unfelt imbalances (unless they are bonuses that do not presuppose specific imbalances
as preconditions44). The examples we now find are pleasures to do with fine colors, fine
shapes, fine sounds, and also some to do with fragrances, and with learning. Socrates
introduces a distinction that makes the class of pure pleasures narrower than we might
originally have realized.
He says: “Well, what I mean is not quite obvious immediately; however, I must try to
explain it” (51b9–c1 tr. Hackforth). Consider the apparatus he uses in connection with
the fineness of shapes and sounds. These are somewhat laboriously described in terms
(at 51c1–d9) that need interpretation: Socrates is not talking of a pros ti (in relation to
something) fineness like that of animals. He wants us to think, by contrast, of something
straight, curved, and so on or, switching sense modalities, of smooth, clear (leias kai
lampras) sounds giving forth a pure phrase45—these are fine not relationally; rather they
are fine kath’ hauta (in accordance with themselves).
What is going on with this pros ti/kath’ hauta contrast? We should start from the basic
fact that the core function of pros in this construction, to signal relationality, works with
the context to determine what is meant. Our context has already introduced the idea
that some but not all pleasures correspond to felt needs/pains. And the contrast between,
for example, an animal on the one hand and pure geometrical figures or phrases could
go with this. For if we ask what it could mean to say an animal is fine in relation to some-
thing, one obvious answer (and one perhaps more to hand in antiquity than today, at
least for philosophers!) is that the animal is fine in relation to some need we have in
view: a fine ox does a good job around the farm. Similarly, I welcome the fine sound that
answers me from the horn of my comrades when I am hard-pressed in the forest. This
would contrast with cases—such as the proverbial stopping to smell the roses—where
something is fine without reference to any such consideration. These afford pleasures
“not like those from rubbing” (51d1)—that is, not conditioned on felt need and so not
mixed with necessary pains. Indeed, Plato’s first cluster of examples of pleasure in things
that are fine kath’ hauta (something straight, curved or the “smooth, clear” sound of a
pure phrase in music) are well chosen to make this point, since the most obvious ways of
finding these beautiful is not in relation to needs.
Thus, if we are guided in this way by the context in which this language is employed
here, we find that the relevant contrast is not that between the ideal and the instance or
that between the abstract and the representational. Nor does it equate purity with
monotony.46

44 If one wanted to make sense of this idea while maintaining the connection of all pleasure with
harmony, one could perhaps consider the possibility that an added note could improve an existing
(i.e. not discordant) state of affairs.
45 When rendering 51d7, melos is often taken as “note,” (see, for example, D. Frede, Plato Philebus,
and Diès, Platon Philèbe) but this is not a common meaning and, in fact, reflects—and then, in turn,
supports circularly—a certain interpretation, which I will reject.
46 I am suggesting we see the examples as applying the criterion, not as further limiting it (for the
latter, see D. Frede, “Disintegration,” 452–53, and D. Frede, Plato Philebus, liii). To equate purity and
 The Philebus   353

What about the circumstance that paintings from nature are mentioned together with
animals in the fine pros ti category?47 No doubt some people consider art fine when it
works as a means to satisfying their craving for status—pros ti as I have glossed that. Or
I could perhaps say that the production of life studies (one way of reading zôgraphêmata)
is subject to a need that is not a matter of the intrinsic harmony of the artwork—in this
case, it is representation of the object one is copying.48 There is actually an analogous issue
with animals too: the dialogue has accustomed us to the idea that animals themselves
can have a harmony—when they exhibit the natural constitution of their kind. So it is
possible to maintain that, while in the everyday case (“what most people would under-
stand as such . . . the beauty of a living creature etc.” 51c1–2 tr. Hackforth), the fineness
we recognize in animals is conditioned on our need and so pros ti, animals also have a
fineness with respect to their own balanced natural constitutions and this fineness
would be kath’ hauta.49
So the list of pure pleasures contains only certain special ones that are a proper subset
of pleasures to do with fine colors, fine shapes, and fine sounds, and with fragrances and
with learning. All of these are free from association with pain and free from unseemli-
ness. Yet none is a strong candidate for our ultimate end. The first four seem more means
to or enhancements of a good life than what it could be organized to achieve. What
about the “pleasures to do with learning” (tas peri ta mathêmata hêdonas, 51e7–52a1) in
view here? If we are persuaded by the overall tenor of the text that this too should be a
matter of improving natural harmony,50 then we should understand pleasures “to do
with” learning as those of the process of learning (i.e., of steps toward the acquisition of
knowledge, as distinct from its possession or exercise).
If all the pleasures Socrates has been discussing at such length are restorations/
improvements of the balance of our natural constitution, then we may suspect that the
real goal must be not these but what they lead up to: this balance or the consequent func-
tioning of the creature in question. This thought, which one could have had in outline
when reading the initial discussion of restoration back at 31c2–32b4, has now gained in
credibility through the development of so much detail in the cases considered: Plato has
managed to describe a myriad of pleasures in such a way that they do not seem very
attractive or possessed of ultimate value.

monotony would not only leave us with an unhappily restricted range of pleasures but would surely be
a problematic position for a program that enshrines harmony, which, after all, requires more than just
one note.
47 “. . . kallos . . . ê zôiôn ê tinôn zôgraphêmatôn” (51c1–3).
48 A third possibility might be that the zôgraphêmata that are beautiful pros ti pick up on the
productions of the artist within the soul introduced in the discussion of the first kind of false pleasure.
Here too a painful imbalance was part of the preconditions of the scenario.
49 On the point that Plato is here concerned with sensory pleasures that involve appreciating the
“fineness or good order of a sensory object” see C. Bobonich, Plato’s Utopia Recast (Oxford, 2002),
355–57; cf. E. Fletcher, “Plato on Pure Pleasure and the Best Life,” Phronesis 59 (2014), 124.
50 Remember that Plato can think that both wisdom and indeed virtue generally are needed for the
soul to be in its naturally harmonious condition.
354   Constance C. Meinwald

Thus we are now ready both intellectually and emotionally for the argument about
genesis and ousia at 53c4–55a3.51 Becoming in general is for the sake of being (e.g., ship-
building is for the sake of ships, and not vice versa). And the good must obviously be
that for the sake of which other things are. We have over and over seen pleasure as a
process of restoration (or improvement) of a creature’s natural harmony and thus as
becoming. So it is for the sake of something else and pleasure cannot be our goal.52 The
ousia at which it aims will be the sort of harmonious constitution studied in the
Promethean Method.

6. The Cluster of Socrates

Finally we turn from examination of the candidate of Philebus to that of the cluster
preferred by Socrates. We already know from the categorization after the Four-fold
Division that this cluster belongs in the category of the cause. Notice that cause in the
Four-fold Division had been tantamount to Aristotle’s efficient cause: we were told at
26e6–7 that “the maker” and “the cause” vary only verbally. Before, we wondered if Plato
would be able to develop his account more satisfactorily than he felt Anaxagoras had
done. This penultimate section of the Philebus will turn out to develop the Anaxagorean
tradition to the point of great similarity with what we are familiar with in Aristotle. The
Physics tells us that while in a sense we may pick out a carpenter as the efficient cause of a
house, or a sculptor as that of a statue, more precisely it is the art of each (and ultimately
the form in question) that is the real efficient cause (195b21–27, 202a9–12). In the
Philebus, analysis of what has been identified as the cause gives pride of place to bodies
of knowledge that grasp relevant Platonic forms.
Is the discussion of the cluster of Socrates itself a Promethean division? Considering
that the cluster has been put in the category of the cause rather than the mixed class, the
answer, strictly speaking, should be no. But the forms of cognition will be distinguisha-
ble in a way, as the varieties of pleasure were not, because of their close relation to their
objects, themselves, in turn, clearly distinguishable in some way (even if not, in all cases,
by the Promethean Method). And the objects of the higher cognitive forms, in fact,
do admit Promethean treatment; the forms of knowledge concerned with them will
automatically show an isomorphic structure.

51 I am glad to see recognition growing that this argument is intended to have a key role in the
dialogue, for example in M. Evans, “Plato’s Anti-Hedonism,” Proceedings of the Boston Area Colloquium of
Ancient Philosophy 22 (2007), 121–45. Evans also makes the point (126) that the position of the argument
right after the discussion of pure pleasures indicates that Plato meant it to apply to them as well.
52 If we are inclined to think that some sections of our text describe pleasures that are not
restorations at all, even of unfelt and/or psychic imbalances, then an additional epicycle is needed. But
since the most common pleasures are restorations and so subject to the genesis-argument, we can still
conclude that it cannot be that pleasure in general (or pleasure qua pleasure) is our goal.
 The Philebus   355

The discussion of the cluster of Socrates takes place in two phases: the first starts with
crafts/arts of manual skill (55d5–56c6). These are graded depending on the extent to
which their practice incorporates measurement. Carpentry (“Measure twice, cut
once!”) is emblematic of the more accurate kind of craft. Music, on the other hand, does
not—at the moment of performance—admit of measuring the quantities involved, even
though they structure it. As Andrew Barker has pointed out,53 this discussion is per-
fectly consistent with the treatment of music in the Promethean Method and Four-fold
Division even if it takes a little thought to see why: those treatments concerned music
theory, this one is about actual performance. If we follow Barker in this, then we can
actually locate the music theory of the Promethean Method among the sciences to come
in the second phase of the discussion.
In the meantime, we find that measure, weighing, and arithmetic are hegemonic with
respect to the manual arts, since as we saw, the more the arts admit measure (etc.), the
more accurate they are: without the mathematical component, nothing is left but estimates
born of practice. Now it turns out (56d1–57e5) that arithmetic and the other hegemonic
parts of the first group afford a further distinction, between applied and pure. To see
this, we need to keep in mind that for the Greeks there is not a unique series of natural
numbers. Rather, a number is a plurality of units, so that any pair of objects is a two, any
trio a three. Thus, addition can be done with sensibles, or with ideal nonsensible units.
By the lights of the Philebus, we have less pure arithmetic when the units are oxen or
armies, and we have pure arithmetic when we treat ideal, nonsensible units. This is the
most accurate apprehension mentioned to this point.
Now the second phase (57e6–59d9) of the discussion comes when Socrates intro-
duces dialectic as a contender for the most accurate knowledge. Dialectic, from the
Republic on, has had the task of saying of each thing what it is—that is, giving accounts
(532a6–7, 533b1–3, 534b3–4). As we saw earlier in this essay, Platonic division developed
as a science of genera and species (see, e.g., Phaedrus 265c8–266c1, 277b5–8). That pro-
cedure offers a type of account that is in view in our dialogue’s Promethean Method
(explicitly associated with dialectic at 17a4). Thus it is natural to locate the technical
understanding54 we were told resulted from the method here. Now we find that this type
of knowledge is most pure—it perfectly and clearly grasps its wholly stable objects—and
most deserves the name of nous and phronêsis. Coming down from this peak, there is a
less pure, less accurate type of apprehension (called doxa, or opinion), as, for instance,
when someone studies not the eternal but subjects such as how this cosmos came into
being. While 57e6–59d9 contains no mention of the role of any mathematics, if we do
connect it with the Promethean Method understood in turn in light of the Four-fold
Division, then mathematical harmonics will in fact have a hegemonic role to play in this
second phase of the discussion of the candidate of Socrates, a role corresponding to that
of measuring and arithmetic in the first phase concerning crafts/arts of manual skill.

53 Barker, “Numbering,” 158–61, relying on Barker, “Text.”

54 The method was associated with technê at 16c2–3.
356   Constance C. Meinwald

7. Sketch of a Finale

We are now ready for our job as mixologists of the good life. Note that Socrates is not
overambitious: he will do this in a sketch (61a4). In discussion we previewed earlier,
Socrates and Protarchus started the mixing at 61b11 and agreed to include the whole
cluster of Socrates, from its purest members to the most applied and empirical. Socrates
goes on to imagine (63b2–4) the pleasures being asked if they are willing to dwell with all
phronêsis. They (more enlightened than their champion, Philebus!) say that they are;
their generous welcome confirms the start Protarchus had made. When the cognitive
cluster is asked the corresponding question, they are more careful, asking (suspiciously?)
“What kind of pleasures?” (63c8). It develops that they are happy with pure pleasures
and embrace those that go with health and all virtue (pleasures such as those of eating
appropriate food when hungry). They vehemently reject the rest, focusing on the way in
which such pleasures are inimical to the development and retention of intellectual
achievement.
As we noted earlier, what makes this mixture and indeed all mixtures good is now
apprehended under a threefold aspect: (1) truth, (2) measure and proportion, and
(3) fineness. We can now perhaps understand this along the following lines. A good
mixture must be truly mixed. Proportion as we have learned at great length from the
Four-fold Division is key to good results, and in turn presupposes measure. And pro-
portion itself is directly akin to fineness.
To determine which of our two contenders is closer to what makes life good, then, we
must evaluate each one three times. That the candidate of Socrates is closer to truth is
obvious: the cognitive capacities aim at truth, while our dialogue has stressed the multi-
farious connections between various pleasures and falsity. If we had left to kalon as
beauty, we might have thought Protarchus would have a chance to claim that humans
characteristically take pleasure in beauty. In fact, he admits readily that pleasures are often
hidden at night, and he seems to think this is because they can be shameful (aischron
in Greek is an opposite to kalon). Finally, let us consider measure and proportion. My
interpretation allows us to see clearly why the candidate of Socrates comes out strongly.
For on this reading, understanding/knowledge is the science the Promethean Method
amounts to, grasping Platonic forms in terms of accounts specifying proportions
governing underlying opposite constituents. Thus the bodies of knowledge themselves
will be structured by the factor of proportion and more particularly by the desirable
ratios discovered by the synthesis of mathematics and philosophy that Plato wants
harmonics to be.
The final ranking of elements in terms of relative responsibility for the good life gives
measure first place, followed by proportion in second. (Measure is clearly prior to pro-
portion in the strict sense since the latter holds between measured quantities.) Third
place goes to nous and phronêsis; fourth come the bodies of knowledge and arts/expert
 The Philebus   357

skills (epistêmai and technai) 55 and right opinion.56 Finally, the recently discussed pure
pleasures take fifth place.57 A reader who felt comfortable with the internal dialectic of
the Philebus could go on to ask how all this compares with the corresponding discus-
sions in the Socratic dialogues, in the Republic, and in the works of other philosophers.
But now, if not in accordance with an injunction of Orpheus, then perhaps more in the
manner of musical chairs, this essay must come to a stop.58

Bibliography
Adam, J. The Republic of Plato (Cambridge, 1902; 2nd ed. 1963).
Barker, A. “Plato’s Philebus: The Numbering of a Unity,” Apeiron 19 (1996), 143–64.
Baker, A. “Ptolemy’s Pythagoreans, Archytas, and Plato’s Conception of Mathematics,”
Phronesis 39 (1994), 113–35.
Baker, A. “Text and Sense at Philebus 56a,” Classical Quarterly 37 (1987), 103–09.
Bobonich, C. Plato’s Utopia Recast: His Later Ethics and Politics (Oxford, 2002).
Burnet, J. Platonis Opera, vol. 2 (Oxford, 1901).
Burnyeat, M. F. “Platonism and Mathematics,” in A. Graeser, ed., Mathematics and Metaphysics
in Aristotle (Bern, 1987), 213–40.
Burnyeat, M. F. The Theaetetus of Plato (Indianapolis, Ind., 1990).
Cooper, J. “Plato’s Theory of Human Good in the Philebus,” Reason and Emotion (Princeton,
N.J., 1999); originally published in Journal of Philosophy 74 (1977), 713–30.
Dancy, R. “The One, the Many, and the Forms: Philebus 15b1–8,” Ancient Philosophy 4 (1984),
160–93.
Diès, A., ed. and trans. Platon Philèbe, 2nd ed. (Paris, 1978; originally published 1941).
Evans, M. “Plato’s Anti-Hedonism,” Proceedings of the Boston Area Colloquium of Ancient
Philosophy 22 (2007), 121–45.
Fletcher, E. “The Divine Method and the Disunity of Pleasure in the Philebus,” Journal of the
History of Philosophy 55 (2017), 179–208.

55 I construe the distinction of third place from fourth as registering something like the priority of a
faculty (which “nous and phronêsis” could be picking out here) to the results of its exercise.
Alternatively, it is possible (Hackforth, Plato’s Philebus ad loc) to maintain that “nous and phronêsis”
here has a narrow sense restricted to apprehension of the highest reality (introduced in 59d4–5).
56 Thus the Philebus holds back from the Stoic extreme of trying to extirpate opinion.
57 As Vogt has pointed out (“Fifth Rank”, 251) “To gain fifth rank with such competitors is not to
come in last and accept a lowly status. To gain fifth rank among such competitors is to be praised.”
58 I am grateful to Gail Fine for inviting me to contribute to this volume, and for her work as editor.
Thanks also to members of an audience at Cornell and Rachana Kamtekar for putting the questions
addressed here, and to Patricia Curd, Emily Fletcher, Rory Hanlon, Jan van Ophuijsen, Sandra
Peterson, Marya Schechtman, and the University of Chicago workshop in Ancient Greek and Roman
Philosophy for entertaining my proposed answers at various stages of work. For the second edition,
I have incorporated references to additional secondary literature and adjusted some vocabulary
and exposition in the interest of clarity for the Greekless reader, while keeping the main points of the
essay the same.
358   Constance C. Meinwald

Fletcher, E. “Plato on Incorrect and Deceptive Pleasures,” Archiv für Geschichte der Philosophie
100 (2018), 379–410.
Fletcher, E. “Plato on Pure Pleasure and the Best Life,” Phronesis 59 (2014), 113–42.
Frede, D. “Disintegration and Restoration: Pleasure and Pain in the Philebus,” in R. Kraut, ed.,
The Cambridge Companion to Plato (Cambridge, 1992), 425–63.
Frede, D. Platon: Philebos (Göttingen, 1997).
Frede, D. Plato Philebus (Indianapolis, 1993).
Frede, D. “Rumplestiltskin’s Pleasures,” Phronesis 30 (1985), 151–80.
Frede, M. “Plato’s Sophist on False Statements,” in R. Kraut, ed., The Cambridge Companion to
Plato (Cambridge, 1992), 397–424.
Frede, M. Prädikation und Existenzaussage (Göttingen, 1967).
Gosling, J. C. B. Plato Philebus (Oxford, 1975).
Hackforth, R. Plato’s Philebus (Cambridge, 1972; published under a different title in 1945).
Harte, V. Plato on Parts and Wholes (Oxford, 2002).
Kahn, C. “On Platonic Chronology,” in J. Annas and C. Rowe, eds., New Perspectives on Plato,
Modern and Ancient (Washington, D.C., 2002), 93–127.
Kraut, R., ed. The Cambridge Companion to Plato (Cambridge, 1992).
Lang, P. M. “The Ranking of the Goods at Philebus 661–67b,” Phronesis 55 (2010), 153–69.
Lennox, J. “Kinds, Forms of Kinds, and the More and the Less in Aristotle’s Biology,” in
A. Gotthelf and J. Lennox, eds., Philosophical Issues in Aristotle’s Biology (Cambridge 1987),
339–59.
Meinwald, C. “Good-Bye to the Third Man,” in R. Kraut, ed., The Cambridge Companion to
Plato (Cambridge, 1992), 365–96.
Meinwald, C. “One/Many Problems,” Phronesis 41 (1996), 95–103.
Meinwald, C. Plato (Abingdon and New York, 2016).
Meinwald, C. “Plato’s Pythagoreanism,” Ancient Philosophy 22 (2002), 87–101.
Meinwald, C. “Prometheus’ Bounds,” in J. Gentzler, ed., Method in Ancient Philosophy (Oxford,
1998), 165–80.
Miller, M. “The God-Given Way,” Proceedings of the Boston Area Colloquium in Ancient
Philosophy 6 (1990), 329–59.
Moravscik, J. M. “Forms, Nature, and the Good in the Philebus,” Phronesis 24 (1979), 81–104.
Muniz, F. and Rudebusch, G. “Plato ‘Philebus’ 15b: A Problem Solved,” Classical Quarterly 54
(2004), 394–405.
Nails, D. The People of Plato (Indianapolis, Ind., 2002).
Sayre, K. Plato’s Late Ontology (Princeton, N.J., 1983).
Striker, G. Peras und Apeiron (Göttingen, 1970).
Tuozzo, T. “The General Account of Pleasure in the Philebus,” Journal of the History of
Philosophy 34 (1996), 495–513.
Vogt, K. “Why Pleasure Gains Fifth Rank: Against the Anti-Hedonist Interpretation of the
Philebus,” International Plato Studies 26 (2007), 250–55.
Westerink, L. G., ed. and trans. Damascius Lectures on the Philebus (Amsterdam, 1959).
 chapter 15

Pl ato’s L aws

Susan Sauvé Meyer

The Laws is Plato’s longest dialogue and is generally taken to be his last.1 Three elderly
men—an unnamed Athenian, a Spartan named Megillus, and a Cretan named Clinias—
undertake a leisurely discourse on “constitutions (politeiai) and laws (nomoi)” (625a6–7).
While they discuss some norms, institutions, and forms of governance that are not a
product of written law, their focus is on the practice of legislation. Over the first three
books they assess the law codes of Crete, Sparta, Athens, Persia, and the failed states of
Argos and Messene. Book 4 begins the project of formulating laws for a new city to be
founded on Crete, and gradually over the rest of the work more than a hundred statutes
to regulate the private and public lives of its citizens are articulated, and the city’s social
structure and forms of governance are defined. The discussion, however, continues to be
oriented to theoretical issues, with persistent attention to questions of legislative and
political theory,2 moral psychology, anthropology, and theology.3
In its political and economic structure, the proposed “city of the Magnesians” (860e6)
has much in common with Athens,4 and it is strikingly different from the ideal city
defined in the Republic. The latter is governed by expert philosophers who are guarded
from the temptation to corruption by being forbidden private property and private
families. Magnesia, by contrast, is governed by property-owning citizens with private
families who serve in the assembly, on the council, or as one of the 37 law guardians
(a body with both executive and judicial functions). The institutional safeguards against

1 Aristotle writes that the Laws is later than the Republic (Politics 1264b26). The biographer Diogenes
Laertius 3.37 cites reports that the Laws was still “in the wax” when Plato died. Stylometric analysis
(see Kahn 2002) puts the Laws in the same general group as other dialogues considered late. For an
assessment of this and other evidence, see Bartels 2017, pp. 25–26 as well as chapter 3 by Irwin in this
volume.
2 On the political theory, see also chapter 24 by Bobonich in this volume.
3 The theology in book X is discussed in chapter 26 by David Sedley in this volume, and by
Mayhew 2008.
4 As Glenn Morrow 1960 has shown, the preponderance of the legislation is modeled on Athenian
law of the fifth century, the so-called “ancestral constitution” attributed in Plato’s day to Solon and
Cleisthenes.
360   susan sauvé meyer

political corruption are the division of political authority among dozens of high offices
and a rigorous system of formal public scrutiny of the conduct of officeholders. While
the Republic invokes a city ruled by experts, the Laws argues forcefully for the rule of law.
The only vestige in the Magnesian constitution of a political role for philosophers is the
so-called Nocturnal Council, an investigative body introduced late in the work as essen-
tial to the salvation of the city (961a–c, 968a–c), but not accorded any specific political
function beyond that of the law guardians, from whom its membership is drawn. Its task
is to investigate abstract questions about virtue and theology (963c–968a).5

1. The Right Way to Legislate

One of the most memorable doctrines of the legislative theory propounded in the Laws
is that a city’s laws should persuade its citizens, not simply command and coerce them.
To that end, statutes are to have persuasive preludes, whose goal is to make the citizens
amenable to the laws (718e–723d). Most of the statutes formulated for Magnesia are given
such preludes, either individually, or en bloc.6 In addition, the Athenian insists, the body
of statutes that comprise a law code must have a central focus. The legislator must have a
single target (skopos) in view, and every practice, law, and institution mandated in the
law code must be directed toward this end. This teleological conception of legislative
practice informs the discussion right from the very beginning, when the Athenian asks
Clinias to identify the principle on which Cretan law codes are organized (625c6–8).
He reiterates at regular intervals over the rest of the work the assumption that proper
legislators will “have in view” a single goal (using the locution apoblepein pros or a variant),
and that the institutions and practices they mandate will likewise have that goal “in
view” or be “for the sake of ” (charin, heneka) it.7
That goal, all the interlocutors agree, is virtue (630c, 630d–631a, 688a, 705d, 963a),
although they diverge at first over whether to construe virtue narrowly as courage,
or more broadly as “justice, moderation and wisdom combined together along with
courage” (630a7–b2). The Athenian calls this combination “virtue in its entirety”
(sumpasēs aretēs, 630b3) and argues that it is toward this entirety that legislators must
direct their efforts (630c1–6). “Courage on its own” (630b2), by contrast, “comes fourth

5 On the Nocturnal Council and the extent to which it approximates the philosopher-rulers of the
Republic see Morrow 1960, pp 500–15; Schofield 1997; Bobonich 2002, pp 391–408; Laks 1990; Brisson
and Pradeau 2007, pp 153–66; and Schöpsdau 2011, 575–85.
6 There is a debate about whether the preludes are to deliver rational persuasion akin to
philosophical argument, or more rhetorical or emotional appeals. See references in n49 in Bobonich’s
chapter 24 in this volume and the entries in Schofield 2016.
7 The legislator has the goal in view: 628a6–10, 628d6-7, 630c1–4, d5–6, 688a1–b1, 693b5–c6,
962d1–5, 962d6–e9, 963a1–b4. The mandated practices have the goal in view: 632e6, 705d3–706a4,
714b8–c1; heneka or charin substituted for apoblepein: 628c6–7, 628d8-e5, 631a3–4, 688a5.
 Plato’s Laws   361

in rank and honor” (630c8–d), being only “a part of virtue, and the least important
(phaulotaton) one at that” (630e1–2).
To illustrate the teleological structure of a properly formulated law code, the Athenian
presents a magisterial portrait of a law code structured around virtue, which we may call
the “grand hierarchy”:

The grand hierarchy: You should have said: ‘Stranger, it is for good reason that
Cretan laws are held in such high repute among the Greeks. They are correct laws
and bring happiness to those who live by them, since they provide them with every
good. Now goods are twofold—some of them human, others divine—and the former
depend on the divine. A city that receives the greater ones acquires the lesser as well;
otherwise, it is bereft of both.
Chief among the lesser goods is health, second beauty, and third is strength for
running and other physical activities. Fourth is wealth that is not blind but clear-
sighted, which comes from following wisdom.
Wisdom (phronesis) itself is first and leader of the divine goods. Second is a moderate
disposition of the soul along with intelligence (nous).8 Third would be justice, which
arises when these are mixed with courage, and fourth would be courage.The latter
goods are naturally ordered above the former, and the lawgiver must order them
thus as well. Next, he must encourage the citizens themselves to believe that his
other orders have these goods in view (eis tauta blepousas), with the human goods
looking toward the divine, and the divine goods in their entirety looking toward
their leader, intelligence. (631b3–d6)9

In the lines that follow (631d6–632c2), the Athenian elaborates upon the encouragement
that the legislator will deliver to the citizens. In every domain of private and public life,
in times of good fortune and bad, the laws will indicate the behavior that is moderate
and just, and they will employ praise and censure to train citizens’ “pains and pleasures,
their desires, and the general intensity of their passions” (631e4–632a1). Presumably, it is
the persuasive preludes to the statutes that will deliver this praise and criticism. The gen-
eral idea is that citizens’ emotional responses should align with the values that structure
the laws by which they live. The Athenian concludes this portrait by noting that the cap-
stone of the law code will be the appointment of guardians to oversee it:

Some [guardians] will operate with wisdom (phronesis), others with true opinion—
so that intelligence (nous) will bind all [the law code] together and it will follow
moderation and justice rather than wealth and ambition. (632c2–7)

8 Reading meta nou with Eusebius and Theodoretus, and most modern editors, instead of the
manuscripts’ meta noun; Bobonich 2002 p 124 translates the latter.
9 Translations are my own. Those from books I and II are from Meyer 2015 with occasional
modifications (as in the present quotation, where I follow Schofield and Griffith 2016 in rendering
631d2–4). The Greek text used is the Budé edition by des Places 1951 (books I–VI) and Diès 1956
(books VII–XII).
362   susan sauvé meyer

A law code that is directed by wisdom and that subordinates wealth and ambition (human
goods) to moderation and justice (divine goods) will be one whose project of delivering
“every good” to its citizens (631b6) will be informed and constrained by the hierarchy of
divine and human goods articulated at 631b–d.
Our project in the rest of this chapter will be to understand this hierarchy of goods
and to identify the ways in which it informs the legislative theory and practice in the
Laws as a whole. Limitations of space make it impossible to do justice to other important
aspects of the dialogue.10

2. The Hierarchy of Goods

According to the grand hierarchy at 631b–d, one class of goods (human goods) is ranked
below the other class (divine goods), and within each class there is also an ordering:

A Divine Goods:
1. wisdom (phronesis or nous)11
2. a moderate disposition of soul along with wisdom
3. justice (a blend of moderation and wisdom with courage)
4. courage
B Human Goods:
1. health
2. beauty
3. strength and speed
4. wealth

Although the Athenian offers no explanation for labeling the higher class of goods
“divine,” he does say that the divine goods all “look to” wisdom (631d5–6), and he will
later invoke wisdom as the “portion of immortality within us” (713e8–714a2).

10 For a brief overview of the dialogue as a whole, see the introduction to Schofield and Griffith
2016. An excellent book-length introductory treatment (in French) is Brisson and Pradeau 2007. For
an overview of the dialogue’s political philosophy see Laks 2000 and Bobonich and Meadows 2013.
Bobonich 2002 is a sustained examination of the interplay among ethics, psychology, and politics in the
dialogue. The most comprehensive treatment of the dialogue in English remains Morrow 1960. For a
guide to scholarship on the Laws, see Schofield 2016, an annotated online bibliography.
11 Here at 631b6–7, the leading divine good is first called phronesis (“wisdom”) and then nous
(“intelligence”), a substitution repeated at 632c5–6; in the listing of the four canonical virtues at the end
of book XII, the leading virtue is first identified as nous (963a8), but then as phronesis (963c12, 964b6,
965d2). At 688b2–3 the “leading virtue” is listed as “phronēsis and nous and doxa.” Other alternates for
nous and phronesis include logos (963e5) and sophia (689d4–e1, 691a7). The expression noun echein or
noun kektasthai is used as an alternate for the predicate phronimos (“wise”) at 687e7–9, 688b7, and
963e6–7; cf. 913a6.
 Plato’s Laws   363

The Athenian opens the grand hierarchy by saying that the human goods “depend”
(ērtētai) on the divine (631b7), and he follows up by explaining that a city12 provided
with the divine goods will also possess the human ones, and will not possess the latter
unless it possesses the former (631b8–c1). Exactly why this is so he does not explain. His
subsequent remark that the legislator will rank the human goods below the divine
(631d2) suggests that the dependence of the human goods on the divine in any particular
city will be a result of legislative practice. But this cannot be the whole story, since the
Athenian indicates that the ranking is natural: “the latter goods are naturally ordered
(tetaktai phusei) above the former” (631d1–2).
We find an elucidation of the natural subordination of human to divine goods in
book II, when the Athenian specifies the values that the laws should inculcate in
­citizens (660e–661c). State-sanctioned musical education must teach citizens that wealth,
strength, and other bodily advantages are good only to a person who is “moderate and
just” (660e3) and possesses “virtue as a whole” (661c4). That is, the goodness of the
human goods depends on their compresence with divine goods; thus Christopher
Bobonich has dubbed the former “dependent goods.”13 Related theses about the condi-
tional value of human goods are reiterated over the course of the dialogue. In book III
the Spartan law code is commended for adhering to the principle that people should not
be honored for wealth, beauty, and strength unless they also possess virtue (696b2–4).
The grand prelude to the law code at the beginning of book V—which ranks god first,
the soul second, and the body last (727b, 728c–d)—encourages citizens to honor bodily
health, beauty, and strength only in those with the virtue of moderation (728d–e).
Finally, in the penal code of book IX, incorrect praise of wealth—which reflects the mis-
taken view that wealth is the highest, rather than the lowest, ranked good (870a7–b1)—is
to be corrected by promulgating in cities the doctrine that wealth is worth pursuing only
in conjunction with justice and moderation (870b7–c1).
Now let us consider the order of goods within each class. Is it simply an expository
order, or does it convey a ranking? It would be a mistake to suppose that the order is
simply expository. Even within the text of the grand hierarchy, wisdom is clearly
accorded priority among the divine goods, which “look toward” it as their “leader”
(631d6). Moreover, in the simplified version of the hierarchy articulated in book IX, we
find the fourth human good ranked below the other three. There we are told that wealth
is “for the sake of ” the body, and thus ranked below excellence of body (870b2–6); simi-
larly, in book III, the first rank goes to goods of the soul, the second to “what is beautiful
and good in the case of the body”—the latter encompassing the first three human goods
in the grand hierarchy (health, beauty, and strength)—and the third to goods of wealth
and property (697b2–c2). If we compare these two versions of the hierarchy with the

12 Some editors delete polis (“city”) at 631b8 as a gloss. In that case the Athenian’s claim would be
that a person who possesses the higher goods will possess the lower ones as well, etc.—a thesis whose
first conjunct is similar to Socrates’ claim in the Apology 30b2–4.
13 See Bobonich 2002, pp 123–215 and Meyer 2015 pp 108–10, 256–59.
364   susan sauvé meyer

ranking in the grand prelude to the law code in book V, we can rank the divine and
human goods as follows:

Grand Hierarchy Simplified Hierarchy The Grand Prelude

(631b–d) (697b, 870a–b) (727b, 728c–d)

DIVINE wisdom God

moderation, justice, courage excellence of soul Soul
HUMAN health, beauty, strength excellence of body Body
wealth excellence of property

On this schema, the first divine good, wisdom, is ranked above all the other goods,
and the fourth human good, wealth, is ranked below all the other goods.
The fourth divine good, courage, is similarly ranked below the other three. The
Athenian makes this clear in the conversation that introduces the grand hierarchy
(630a–631a). He has singled out courage for discussion at this point in the conversation
because his interlocutors take courage to be the virtue aimed at by the legislator, whereas
he insists that a proper legislator must aim at “virtue in its entirety” (sumpasēs aretēs,
630b3; pasan aretēn 630e2–3). He explains what he means by “virtue in its entirety” by
comparing the fierce warriors celebrated by the Spartan poet Tyrtaeus—those “embold-
ened by the sight of bloody slaughter” and who “lay waste the enemy, assailing him at
close quarters” (629e1–3)—to citizens who excel at the project of keeping the polis free
from internal faction (629e9–630b3b). He will later characterize the latter as “knowing
how to rule and be ruled by justice” (643e6), and as “capable participants in the affairs of a
city and its towns” as opposed to an army camp (667a1–2). Here he claims such citizens
are “better by far” than the courageous warriors lauded by Tyrtaeus, to the degree that
“justice and moderation and wisdom combined together along with courage is better
than courage on its own” (630a7–b2). These well rounded citizens have “virtue in its
entirety” (sumpasēs aretēs, 630b3) and it is toward this, the “height of virtue” (tēn megistēn
aretēn, 630c3–4), that legislators must direct their efforts (630c1–4). This target is prop-
erly called “complete justice” (dikaiosunēn telean, 630c5–6). In comparison to it, “the vir-
tue that Tyrtaeus singles out for praise” is properly “ranked fourth in honour” (630c6–d1),
being virtue’s “least significant (phaulotaton)” (630e2) and “smallest” (631a5) part.
In stating that the entirety of virtue is properly called justice (630c5–6), the Athenian
indicates that his description of that entirety at 630a7–b2 should be read as “Justice, that is
(kai) moderation and wisdom combined together along with courage.” 14 This is exactly

14 This conception of justice (as encompassing wisdom and moderation and courage) would explain
why in the theodicy in book X the Athenian claims that the gods possess “all of virtue” (pasa aretē
900d1–2) and explains by attributing to them moderation, wisdom, and courage (900d5–e2), with
no mention of justice. Bartels 2017 pp 33–36 notes that, in contrast to the Republic, there are few
occurrences of the noun “justice” (dikaiosunē) in the Laws; however, the incidence of the cognate
adjectives and verbs is still very high. The Athenian continues to be interested in persons and behavior
that can be characterized as just (dikaion) or unjust (adikon).
 Plato’s Laws   365

how he will describe justice in the grand hierarchy, a page later. There, he lists justice
as the third divine good, and characterizes it as the result of blending the higher two
divine goods with courage (631c5–d1). That is, it is a blend of wisdom, moderation,
and courage. In ranking courage fourth after this blend, the grand hierarchy also replicates
the introductory passage’s contention that courage ranks fourth in relation to justice
(630c4–d1). Since the fourth-place status of courage in this introductory passage is
clearly normative—it is the “least significant” (630e2) or “smallest” (631a5) part of virtue—
we may suppose that the same is true of the grand hierarchy that it introduces. Courage
is listed fourth among the divine goods (631c8–d1) because it is the least important of
those goods.
To understand why courage is ranked fourth in this category, let us consider why it is
inferior to the blend of wisdom and moderation with courage that is ranked third. In
effect, this is to ask why one part of virtue, courage, is inferior to what he earlier referred
to as “the entirety of virtue” (630b3; 630e2–3). Now, if that entirety were simply a con-
junction of four virtues—wisdom, moderation, justice, and courage—the inferiority of
courage to the entirety would be straightforward and fairly uninteresting. Courage
would be inferior to the whole of virtue for the same reason that one meal a day is infe-
rior to three meals a day, and one hour’s sleep inferior to a full night’s sleep. But, as we
have seen, virtue in its entirety is not simply a conjunction of wisdom, moderation, jus-
tice, and courage, but a single entity, justice, that is blended from wisdom, moderation,
and courage.
Once we recognize that fourth-ranked courage is an ingredient in third-ranked justice,
it is no longer a trivial matter to establish that courage is inferior to justice. Compare the
case of olive oil, bread, and wine. These three comestibles together make a better meal
than bread on its own, if one dips the bread in the oil and then eats it while drinking the
wine. If, however, the oil and bread are ground together and mixed with the wine, any of
the ingredients on their own would be better fare than such a mixture. Of course, some
ingredients are always better fare in a mixture than on their own: baking soda, salt, and
hot sauce, for example. Whether a given ingredient is better or worse than a mixture of
which it is an ingredient will depend on the nature of the ingredient in its unblended
state, and on the nature of the mixture into which it is blended. The upshot for our
inquiry is that to understand why the Athenian ranks courage below justice, we need to
know how he conceives of courage when it is not a component of justice, how it is trans-
formed when combined with the rest of virtue, and what sort of mixture justice is.

3. Courage

In the grand hierarchy at 631b–d, the fourth ranked divine good is listed simply as courage.
However, in the conversation that introduces the grand hierarchy and where the
Athenian first assigns courage its fourth-place rank (630c8), he specifies it more precisely
as “courage on its own” (630b2), and he illustrates what he has in mind by invoking the
fierce warriors celebrated by Tytraeus, who can be “rash, unjust, arrogant (hubristai),
366   susan sauvé meyer

and exceedingly stupid (aphronestatoi)” (630b6–7). Courage on its own, it seems, is

courage that need not involve any of the rest of virtue.
This conception of courage persists over the course of the dialogue. In book II, for
example, the Athenian reiterates the possibility that a warrior endowed with Tyrtaean
courage can be unjust (661a2). In book III he invokes the dismal prospect of a fellow
citizen who is “a very courageous person [but] lacks moderation and is unrestrained”
(696b10–11). And at the end of book XII, when he returns to the notion of fourfold
virtue—wisdom, moderation, justice, and courage, with wisdom the leader of the rest
(963a6–9)—he again indicates that one can have courage without wisdom, as in the case
of small children and nonhuman animals:

Ask me why, although we call both [courage and wisdom] with a single name,
“virtue”, we turn around and speak of them as two things, courage on the one
hand and wisdom on the other. The reason, I will tell you, is that the one, courage,
concerns fear. Beasts partake of it, and so do the characters of very young children,
since there can be a soul that is naturally courageous without reason (logos), while
there never has been nor will there be a soul that is wise and intelligent without
reason. (963e1–8)

On this persistent conception of courage, it is a trait of fearlessness or boldness that can

be undiscriminating (as in the case of children and animals), or misdirected and unjust
(as in the case of the savage mercenaries). It is the conception of courage invoked in
dialogues such as in Euthydemus 281c and Meno 88a–b, as well as in Statesman 306a–b,
where it is a trait that can be employed well or badly, with or without knowledge. Let us
refer to courage so conceived as raw courage.
Raw courage may be contrasted with a refined variety, on which courage either is or
requires wisdom. Such a conception of courage is defended in other dialogues of Plato,15
and readers of the Laws tend to construe the courage that is ranked fourth among the
divine goods as the refined variety.16 However, this is unwarranted. Outside the divine
hierarchy, when the Athenian invokes courage, he either has in mind the raw variety, in
the passages we have considered, or he gives no indication that he is thinking of a refined
variety.17 The only instance of refined courage in the dialogue is a component of the

15 Protagoras: courage is wisdom 360d; cf. Laches 194d–198b; Phaedo: genuine courage requires
phronesis 69a–c; Republic: courage involves obedience to the verdicts of reason (429c–430b; 442b–c).
16 For example, Bobonich 2002 allows that in the Laws, Plato “sometimes” uses “courage” for a trait
that can exist in isolation from the rest of virtue (p. 289), but he denies that courage so conceived is a
divine good (p. 125).
17 Irwin 1995 proposes that the courage invoked at 6335d–6 is “a type of bravery that requires
wisdom” (p. 348). However, the contrast in that text is not between courage that does and does not
involve wisdom, but between the courage that resists only pains, and the ambidextrous courage (634a)
that resists both pleasures and pains. Wisdom is not mentioned, and the context gives no reason to
suppose it is implied. Bobonich 2002 finds in the Laws the doctrine that “genuine courage . . . cannot
exist apart from the other virtues, especially a correct conception of the ultimate end” (p. 117), but does
not indicate where it is to be found. He describes 633c8–d3 as articulating a conception of courage that
involves “appropriate” resistance to pleasures and pains (p. 288), but the modifier is absent from the text.
 Plato’s Laws   367

third divine good: a state in which courage (along with moderation) is blended with
wisdom (631c7–8). The courage that is ranked fourth, however, is not the refined variety
but the raw ingredient that is refined when blended with wisdom and moderation. It is
raw courage.
Once we recognize that the fourth divine good is raw courage, it is easy to see why it is
ranked below justice in the grand hierarchy, since justice is there construed as a mixture
in which raw courage is refined by being blended with wisdom and moderation.
Courage on its own (raw courage) is not something we value in our fellow citizens, but
we do value it when it is informed by wisdom and tempered by moderation—that is,
when its ferocity is directed at the right objective, on the right occasions, and with the
right intensity. In other words, its value is contingent on its being informed by wisdom.
Thus, this divine good “looks to” wisdom in the same way as the human goods do;
its goodness is contingent on its being combined with wisdom.18 In the culinary meta-
phor, courage is comparable to salt, which on its own is not palatable, but in the right
quantity, with a mix of other ingredients, is a valuable component of a dish.
We may find it puzzling that the Athenian should classify raw courage as a divine
good, or as a virtue, especially given his negative portrayal of it at 630b–c and 696b.
However, he is unequivocal about classifying it as a “virtue” across the dialogue (630c,
632e, 791c, 900d, 963c–e, 965d) and, as we have just seen, his introduction to the grand
hierarchy (630a–631a) indicates that raw courage is precisely what he has in mind as the
courage that ranks fourth among the divine goods.19 We will return to the question of
why he accords “divine” status to such a dubious trait in Section 5 of this chapter. But first
let us consider moderation, one of the other ingredients of justice.

4. Moderation

Recall that when justice is ranked third among the divine goods, it is specified with ref-
erence to the first and second divine goods. “Third would be justice, which arises when
these are combined with courage” (631c7–8). “These” (toutōn, c7) must refer back to the
“intelligence” (nous) and the “moderate disposition of the soul” invoked in the preced-
ing clause. The former is the leading divine good, wisdom (phronesis, 631c5–6),20 and the
latter is either the second divine good, or—on an alternative reading of the text21—it is

He notes that “Plato might reasonably deny that any psychic quality that does not benefit its possessor
is a virtue” in order to defend the thesis that wisdom, moderation, justice, and courage are reciprocally
entailing (290); however, this fails to establish that Plato makes such a denial in the Laws. On the
version of the reciprocity of virtue to be found in the Laws, see §6.
18 Here I agree with Bobonich 2002, p. 125.
19 Irwin 1995, p 347 agrees that the courage invoked in the introductory passage (630a–631a) and
in book XII is raw courage, but he takes no explicit stand on the fourth ranked courage in the grand
hierarchy (631b–d).
20 On the alternation between “nous” and “phronesis” see note 11.
21 The reading of Eusebius and Theodoretus at 631c7; see note 8.
368   susan sauvé meyer

combined with wisdom to yield the second divine good. On either alternative, what
gets combined with courage and wisdom to yield justice is “a moderate (sophrōn) dispo-
sition of the soul.”
While the Athenian never uses this expression again in the Laws, he does regularly
invoke “moderation” (sōphrosunē). It occurs on his lists of the four forms of virtue (630b,
964b), or as included in “all of virtue” (900d–e; cf. 936b), and in book III he insists, on
behalf of the legislator, that it is essential for citizens (696b–c), rulers (692a, 694c–696a,
709e–710b), and city (693b–c) alike to possess it. In line with his parody at 660d–661a of
the Spartan poet Tyrtaeus’s refusal to praise a rich or well-born person who lacks courage
(629a–b), he credits the Spartan law code for adhering to the principle that:

No person should get higher honour (timē) in a city because he is especially rich,
any more than he should for being swift or handsome or strong, if he lacks virtue,
nor should he if moderation is absent from his virtue. (696b2–5)

In other words, the human goods, health and strength, are not to be honored unless they
are accompanied by virtue, but not just by any kind of virtue; moderation is required.
In the same contexts in which he insists on the importance of having moderation
(696b6–c10), the Athenian also insists that, in isolation from the rest of virtue, it is of no
significance. On its own, it merits neither praise nor censure (696d3–8). It is only “an
adjunct” (prosthēma, 696d9) to what merits honor. On its own, “it is not worth discuss-
ing, but should be passed over in silence” (696d9–e1).
This disparagement of moderation is puzzling, since the same could be said of raw
courage, which the Athenian nonetheless has classified as a divine good in the grand
hierarchy back in book I. Nonetheless, it is not an isolated comment. The Athenian
makes the same point in book IV when he claims that the ruler of a city must have “that
quality which in our earlier discussion we said must accompany the parts of virtue
if there is going to be any benefit from their presence” (709e8–710a1), and then reiterates
the thesis that moderation on its own has no value. But before he repeats the disparaging
thesis, be distinguishes two kinds of moderation:

CL: I think, Megillus, that moderation is what the [Athenian] is saying must
accompany them. Am I right?
ATH: Yes, Clinias, but it is the ordinary variety, not what one might speak of in an
elevated discourse when insisting that having moderation (to sōphronein) is also
being wise (phronēsin. . . einai). I have in mind the quality that naturally develops
even in children and beasts—some of whom lack self control in the face of pleasures,
while others are self controlled. It is worth nothing to speak of, we said, when it is
separated from the other so-called goods. (710a3–b2)

The Athenian here contrasts ordinary (dēmodē, 710a5) moderation with a different
kind of moderation that involves wisdom. This contrast is sometimes identified with the
distinction in the Phaedo between “popular (dēmotikē) and political” virtue on the one
 Plato’s Laws   369

hand, and genuine, “philosophical” virtue on the other (82a11–b1; cf. 68b8–c3, 69a5–c2).22
However, the “popular” virtue denigrated in the Phaedo involves being motivated by
money and honor,23 whereas here in the Laws, the Athenian indicates it is the sort of
trait that can belong to children and nonhuman animals, who are unlikely to have such
a motivation.
In attributing ordinary moderation to children and nonhuman animals, the Athenian
invokes the same kind of examples he uses in book XII to distinguish courage from
wisdom (963e1–8). Since raw courage is what he has in mind there, it is most likely that
the ordinary moderation he has in mind here at 710a5 is raw moderation (analogous to
raw courage): a trait of restraint in the face of opportunities for indulgence, but that need
not be exercised intelligently. Such a trait is often mentioned in conjunction with raw
courage in other dialogues.24 In the Statesman, for example, we are told that it can lead to
seriously mistaken behavior—for example, appeasement when an aggressive response
is called for (Stsm. 307e). In the Laws, by contrast, the Athenian expresses no concern
about cases in which moderation may be deployed unintelligently—a marked contrast
with his treatment of raw courage, where he repeatedly invokes the example of the
Tyrtaean warriors (630b, 667a). In any case, it is presumably because ordinary modera-
tion need not involve wisdom that the Athenian classifies it as barely worthy of mention
(696e1, 710b1). This is an apt description of raw moderation.
What about the other kind of moderation, “what one might speak of in an elevated
discourse when insisting that having moderation is also being wise” (710a5–6)? A good
candidate for this elevated discourse occurs early in book III, where the interlocutors
discuss the failed states of Argos and Messene and the Athenian resists the diagnosis
that it was a lack of courage or military prowess on the part of their leaders that caused
them to fail. On the contrary, he argues, these regimes perished because their leaders
failed to have wisdom (689a–e). It was the “greatest ignorance (amathia)” that destroyed
these regimes, he claims (688e4), and he explains he is not talking about the “ignorance
of tradesmen” (dēmiourgoi) (689c2). His earlier remarks at 644a2–5 suggest that the latter
is narrow vocational education that leaves a “mechanical worker” (banausos) unlearned
in the affairs of state. The “ignorance” he has in mind as the greatest, he explains, is not a
lack of learning on the part of the leaders, but rather a lack of concord between their
judgments and their feelings:

When a person thinks something is beautiful or good but fails to love it—he hates it
instead, and loves and welcomes things he thinks are wicked and wrong—I call this
lack of agreement (diaphōnia) of his pleasure and pain with his reasoned opinion
[kata logon doxa] the ultimate and greatest ignorance (amathia). . . . When the soul
opposes itself to knowledge or beliefs or calculation [logos], which are its natural

22 Thus England ad loc. 23 See the discussion by Bobonich 2002, pp 14–21.

24 Both raw courage and raw moderation are invoked at Meno 88ab, Euthydemus 281c, and
Statesman 306a–b. They are arguably the two familial traits invoked at Laws 681b, as well as the natural
qualifications of the guardians at Republic 374e–376d, and among the natural aptitudes of a young
tyrant at Laws 710c.
370   susan sauvé meyer

rulers, I call this stupidity (anoia). The same goes for a city, when the multitude does
not obey the rulers and laws, and for an individual man, when the beautiful judgments
[logoi] in his soul amount to nothing, as he acts completely against them. I would
posit these cases of ignorance to be the most discordant either in a city or in an indi-
vidual citizen, rather than those of tradesmen, and I hope, Strangers, that you grasp
what I mean. (689a5–c3)

He proceeds to draw the corollary claim that phronesis requires “agreement” (sumphōnia)
between judgments and feelings:

How could there be wisdom (phronēsis) of even the smallest degree where there is
no agreement (sumphōnia)? There is no way. So the finest and greatest kind of agree-
ment is properly called the greatest wisdom (sophia), and it is in accordance with
this that the person living in accordance with reason (kata logon) conducts his life.
He who lacks it will invariably be the ruin of his family, and he will turn out to be no
saviour to his city either, indeed quite the opposite, since he is completely ignorant
(amathainōn) in that domain. (689d4–e1)

Later in book III, the Athenian will refer back to the present passage when he invokes
“the wise person we just mentioned, the one whose pleasures and pains agree with and
follow his correct judgments” (696c8–10). He invokes that admittedly recondite concep-
tion of wisdom to support the thesis that wisdom “does not arise without moderation”
(696c5–8). That earlier conversation in book III (689a–e) is therefore a good candidate
for the “elevated discourse” in which wisdom and moderation are yoked together that
is invoked in book IV (710a5–6).
Against this proposal, it might be objected that the elevated discourse, as paraphrased
at 710a5–6, requires moderation to be wisdom, while the burden of the earlier treatment
of wisdom is a different thesis, that wisdom implies moderation. However, the Athenian
also explains, in the course of book III, that the expressions “wisdom” and “moderation”
are different ways of characterizing the same thing:

You must take into account, when we say that [the legislator] must have in view
moderation, or wisdom, or friendship, that it isn’t a different target but the same one;
and don’t let it bother you if we employ other expressions of this sort as well.
(693c2–6)

He is speaking here of a city’s moderation and wisdom (where friendship is the civic
analogue of the concord (sumphōnia) required for wisdom in the personal case). He has,
however, indicated that his remarks about phronesis apply equally to a city and to an
individual person (687e, 689b). So it is not much of a stretch to describe the position he
presents here in book III as one on which “having moderation is also being wise” (710a5).
So construed, the treatment of wisdom and moderation in book III is the “elevated dis-
course” invoked in book IV, and the elevated discourse is the inquiry into legislation
embarked upon by the interlocutors.
 Plato’s Laws   371

With this understanding of the difference between ordinary moderation and the
moderation invoked in the elevated discourse, let us return to the list of divine goods in
the grand hierarchy. What kind of moderation is invoked when the second and third
divine goods are specified with reference to “a moderate disposition of the soul” (631c7)?
Ordinary moderation, we have seen, is raw moderation, a trait of self-restraint that need
not be accompanied by any other part of virtue. In isolation from the rest of virtue, the
Athenian has stressed, it is of insignificant value. So it cannot be what he has in mind as
the second divine good. It is, however, an excellent candidate for the “moderate disposi-
tion of the soul” that is blended with wisdom and courage to make justice, the third
divine good. It follows that if the “moderate disposition of the soul” is ordinary modera-
tion then that “moderate disposition” cannot be the second divine good. This gives us
reason to prefer Eusebius’s reading of 631c7: “second is a moderate disposition of the
soul along with wisdom (meta nou)”—a phrase that perfectly captures the moderation
invoked in the elevated discourse.25 We may conclude that the moderation of the ele-
vated discourse is the second divine good, while the ordinary kind of moderation is an
ingredient in the blend that constitutes justice, the third divine good.

5. Why Rank Courage as

a Divine Good?

Let us return to the puzzles raised previously about the Athenian’s very different
treatments of courage and moderation. We have seen that in books III and IV he dis-
tinguishes ordinary or raw moderation from a refined variety that involves wisdom, and
accords only trifling value to raw moderation. It is puzzling that he nowhere makes a
corresponding distinction in the case of courage; he never claims that courage on its
own is of trifling value. Indeed, in the grand hierarchy of book I, he lists it as a divine good.
If he is so attuned to the trifling value of moderation on its own (ordinary moderation),
then why does he not draw the same conclusion about courage on its own? And if cour-
age on its own is a divine good, why isn’t moderation on its own26 accorded a similar
status (perhaps as a fifth divine good, or tied for fourth)?27
It is tempting to solve these puzzles by interpreting the fourth divine good not as raw
courage, but as refined courage, analogous to the refined moderation that is the second
divine good. However, as we saw in Section 3 of this chapter, the evidence from the rest
of the dialogue weighs heavily against this interpretation. Nowhere else in the dialogue

25 The phrase is nicely parallel to such locutions as τὸ σωφρονεῖν νοῦν τε κεκτῆσθαι (900d7) “being
moderate and having wisdom”; and σωφροσύνη μετὰ φρονήσεως (906b2): “moderation along with
wisdom.”
26 Recall (from §4) that the second divine good is not moderation on its own, but refined
moderation.
27 Irwin 1995 p. 348 raises similar puzzles.
372   susan sauvé meyer

does the Athenian invoke a refined conception of courage. More importantly, his
introduction to the grand hierarchy (630a–631a) indicates that the courage ranked
fourth in that hierarchy is “courage on its own”—that is, raw courage. The point he is
making in the grand hierarchy by ranking courage below justice is that raw courage is far
less desirable that the entirety of virtue. He cannot make this point unless the courage
ranked fourth is raw courage.
Indeed, if we switch our focus from the fact that the Athenian lists courage as a divine
good to the fact that he ranks it fourth in that category—meaning that it is “the least
significant” (630e2) and “smallest” (631a5) part of virtue—we might recognize here a
disparagement of courage analogous to his later denigration of moderation. Against the
background assumption that citizens are expected to have—that goods of the soul rank
above those of the body (697b, 870a-b)—and the Athenian’s stipulation that goods are
either human or divine (631b)—it is unremarkable that the Athenian would place cour-
age in the class of divine goods. What is remarkable and emphatic is his insistence that
courage, so prized by his interlocutors, ranks last in its class. Still, for the Athenian to
include raw courage on the list of the divine goods is to accord it an honorific status that
he fails to accord to raw moderation. This is to treat raw courage with what we might see
as undue charity when we compare it with his uncompromisingly severe assessment of
raw moderation in books III and IV.
An unduly generous treatment of courage, however, is a persistent feature of book 1.
When the Athenian argues, against Clinias, that a legislator must aim at cultivating not
just courage but “complete justice” in the citizens (629a–631a), he hastens to assure Clinias
and Megillus that he is not criticizing the Cretan and Spartan law codes (630d; cf. 628e).
Since these law codes are divinely inspired, they must be correct (624a1–5, 631b3–4); thus,
it must be the interlocutors’ fault if they cannot identify the ways in which the laws of
Crete and Sparta promote the full range of virtue (632d–e).28 To remedy this purported
shortcoming in their understanding of those law codes, he proposes to take up one kind
of virtue at a time, seeking to identify, in each case, how those law codes promote it
(632d8–e4). They will start by completing the discussion of courage initiated by Cllinias,
and use that discussion as a model for their discussion of the other kinds of virtue (632e).
The ensuing discussion, taking its cue from Megillus (633b), treats courage as a matter
of enduring or resisting pains and fears, a conception of courage that the Athenian
invites his interlocutors to expand to include resistance to pleasures and desires (633d1–3),
a domain he will later identify as moderation (635e–636a). When Megillus and Clinias
prove unable to identify practices in their home states that cultivate “resistance to
pleasures” (634b7–c4), the Athenian proposes that such resistance is cultivated by the
drinking party (sumposion)—a venerable cultural institution in Athens that is banned in
Sparta and on Crete (637b6–638b9). Just as the dangers of combat give one experience in
resisting pains and fears, drinking parties provide the opportunity to battle against the

28 This polite fiction is abandoned in book II, where the Athenian offers direct criticism of the
military orientation of Crete and Sparta (666e–667a).
 Plato’s Laws   373

pleasures and desires that are unleashed when drink loosens a person’s self-control
(647c7–649d2).
The full defense of this educational proposal will not be complete until the end of
book II, and will turn out to have less to do with sumposia in particular than with musical
education in general. What is important for our present purposes is, first of all, the central
role played by courage in the discussion that launches and defends this educational pro-
posal. Even though the Athenian has insisted that courage ranks last among the forms of
virtue, it is clearly central to his interlocutors’ conception of virtue. Rather than challenge
the importance his interlocutors attach to courage, he exploits it, encouraging them to
take moderation seriously by getting them to construe it as a kind of courage.
Indeed, the conception of courage as a fight against pains, fears, pleasures, and desires
invokes an ideal that the Athenian has specifically repudiated. So construed, courage
would be an instance of the “victory over self ” or “self mastery” (to kreittō heautou) cele-
brated by Clinias early in book I as the “greatest victory of all” (626e). The Athenian
draws this point to his interlocutors’ attention at 633d. What he omits to point out, how-
ever, is that he has subjected that ideal to a devastating critique (627c–628d): when
­battling against oneself it is better to win than to lose, but it is better by far to be free
from internal conflict. Victory over oneself “is not an excellent condition but a necessity.
To think otherwise is like supposing that a disease-ridden body is performing at its best
after being flushed out by a purgative” (628c9–d3). This withering critique of self-mastery
notwithstanding, Clinias and Megillus are evidently still attached to the ideal, so when
Megillus invokes it at 633b–c, the Athenian accepts it as a paradigm for courage. Indeed,
when he introduces the famous figure of the divine puppets (644c–645a) to analyze the
inner struggle involved in such self-mastery (644b6–c2; 645b2–3),29 the Athenian
describes himself as having endorsed the ideal of self-mastery (“Now, we previously
agreed that those who are able to rule themselves are good, while those who are unable
to do so are bad” 644b6–8).30 While the Athenian has gone along with his interlocutors
when they have continued to rely on that ideal, and indeed has exploited that reliance, it
is surely an overstatement to characterize himself as having agreed to it.
Evidently, courage is central to Megillus’s and Clinias’s conception of virtue, and self-
mastery is central to their conception of courage. Although the Athenian has registered his
disagreement on both counts, he declines to reiterate his objections when his interlocutors
persist in these views. We may suppose that his generous inclusion of courage among the
divine goods is part of the same pattern of deference to his interlocutors. In this context, it
is significant that the Athenian articulates the grand hierarchy not in his own voice, but as
what Clinias “should have said” (631b3–4) about the laws of his own city.31

29 Those who struggle for self-mastery are like puppets pulled in opposing directions by the
golden cord of reasoning (logismos) and the tougher strings (“hard as iron”) of pleasure, pain, fear, and
daring. I discuss the large literature on the puppets in Meyer 2015, pp 178–87; more recent discussions
include Schofield 2016a, and Bartels 2017, pp 86–92.
30 I discuss this surprising claim in Meyer 2018.
31 632d1–4 indicates that what Clinias should have said includes all of 631b–632c.
374   susan sauvé meyer

6. Justice as the Entirety

of Virtue

We have now identified the two ingredients that, when blended with wisdom, yield
justice, according to the recipe in the grand hierarchy of 631b–d. These two ingredients
are raw courage and raw moderation; that is, they are courage and moderation on the ordi-
nary construal, dispositional tendencies toward fearlessness and self-restraint respectively.
They are the courage and moderation invoked in the Statesman and classified as “parts”
of virtue that are “opposed” to each other (306a–307c), where the Stranger clearly has in
mind the aggressive and restrained tendencies that we have labeled raw courage and
raw moderation (307e–308b). He states that the job of the statesman is to “weave” these
two tendencies together in citizens, as the warp and woof of character (308c–309b). The
weaving will be accomplished by means of education, primarily, as well as by eugenic
breeding (309c–311a). In Laws as well, the Athenian notes that these two natural tenden-
cies tend to run in families (681b) and he recommends intermarriage between the two
types, so as to produce offspring with a mixture of both traits (773c–d).32
In the Statesman, we are told, the appropriate mixture of courage and moderation will
be informed by stable true beliefs about what is admirable (kalon), just, and good (309c).
In book II of the Laws, we are told that it is the laws themselves (through the medium of
education) that are the source of citizens’ normative judgments (659d–661d). And in the
grand hierarchy of book I, the Athenian indicates that the phronesis that structures a
proper law code will also shape the emotional responses of citizens (631e–632a). Thus, it
is reasonable to identify the “woven” virtue of the Statesman with the mixture of moder-
ation, courage, and wisdom that is called “justice” in the grand hierarchy. So construed,
justice—that is, virtue in its entirety—would be a mixture in which raw courage and raw
moderation are informed and refined by the wisdom expressed in the laws.33
Virtue in its entirety, thus construed, would be a highly unified state, which befits its
status as a blend, rather than a compound of independently existing virtues. The raw
courage and moderation that are ingredients in the blend are transformed into refined
versions of themselves as a result of being mixed with wisdom. They are not simply com-
bined with wisdom, but informed by it. Their refined state depends on the wisdom with
which they are combined. Indeed, we find out in book III that the dependence is recip-
rocal. The Athenian indicates that he takes wisdom to require moderation (696b–c), on

32 The conception of virtue as a mixture that balances two opposing traits also appears in the
Republic, where the goal of the guardians’ education is to make them fierce against enemies, but gentle
and loyal to fellow citizens (Rep. 374e–376d; cf. Laws is 731b3–4).
33 In some places, the Athenian contrasts wisdom (phronesis) with true belief (632c–d), and in
others, he uses “true belief ” as an alternative that can play the same role as phronesis (653a, 688b).
Presumably the laws and the sanctioned musical education communicate true beliefs to the citizenry at
large, while those with higher cognitive competence will be responsible for selecting educational works
whose teachings are correct (670d–e).
 Plato’s Laws   375

the grounds that wisdom requires “pleasures and pains [that] agree with and follow one’s
correct judgments” (696c8–10; cf. 689a–e). Since such “pleasures and pains” include the
domains of both courage and moderation, the same argument would show that wisdom
involves courage.34 Thus none of the components of virtue as a whole—wisdom, refined
moderation, and refined courage—can exist without the others.
Exactly how to characterize the relationship between that integrated unity (virtue as a
whole) and the four types of virtue—wisdom, moderation, justice, and courage—is a
topic the Athenian is reluctant to address. He regularly refers to courage as a part (meros
or morion) of virtue (630e, 631a, 696a, 791c), but shies away from using that designation
for all of the four. In book I, after mentioning courage, he invokes “the rest of virtue’s
parts—or whatever else we should call them” (633a8–9). His preferred term for referring
to the four is “kinds” (eidē). At the end of the Laws, the Athenian returns to the question
of the relation between virtue as a whole and its four forms (eidē). While he takes it to be
relatively easy to characterize the four different forms, and thus within the ordinary
competence of a well-educated citizen, he takes the question about how to characterize
the one thing to be more difficult. Exactly how to characterize the unity of virtue in the
face of this fourfold multiplicity is a question that the Nocturnal Council will investigate
(963a–965e).35

7. The Hierarchy of Goods Revisited

Let us take stock. Our focus has been on the grand hierarchy of divine and human goods
that, according to the Athenian, will structure a properly formulated law code. We have
seen that not only are the human goods ranked below the divine goods, there are also
distinctions of rank within each class of goods. Our focus has been on the fourth place
status of courage among the divine goods (which turns out to be analogous to the fourth
place status of wealth among the human goods). Courage is ranked fourth in its class
because it is inferior to the entirety of virtue (named justice), which occupies third place
on the list, and the same is true of wealth, which ranks fourth among the human goods.
We may conclude that the distinction between the goods listed third and fourth in
each class marks a difference in value. It is a difference in rank, not just in expository
order. However, the same cannot be said regarding the order of the first three divine

34 The Athenian does not draw this conclusion explicitly, perhaps because his interlocutors need to
be convinced of the need for moderation, not of courage. In any case, since the beginning of book II, he
has deployed a new model for virtue in its entirety. Instead of moderation and courage blended with
wisdom (as on the grand hierarchy), virtue in its entirety is analyzed into wisdom (phronesis) and
paideia (“pleasures and pains, desires . . . ” trained to agree with correct judgment” 653b–c). Paideia
so construed captures the domain of both moderation and courage without marking a distinction
between them. On the collapse of the distinction between courage and moderation, see also Bobonich
2002, pp. 290, 547–48.
35 As Verity Harte (2002) has shown, abstract questions about parts and wholes are characteristic of
the metaphysical inquiry in Plato’s late dialogues.
376   susan sauvé meyer

goods. None of them is inferior to those listed ahead of it—at least, not in the way that
courage is inferior to justice, or that the human goods are inferior to the divine (the
inferiority of a “dependent good” to the good on which its value depends). Indeed,
the reciprocal dependence (noted in Section 6 of this chapter) between justice and its
components—refined courage and refined moderation—entails that the second divine
good (refined moderation) is a component of justice (the third divine good). In fact, the
first three divine goods comprise an integrated unity, and the same is arguably true of
the first three human goods: health, beauty, and strength (assuming beauty involves
something like the glow of good health). Thus, within both the divine and the human
goods, the first three goods comprise an interdependent unity, while the fourth is onto-
logically independent of them, but dependent on them for its goodness.
What, then, of wisdom’s status as “first and leader of the divine goods” (631c6)? In
stating that the divine goods all “look toward” wisdom (631d5–6; cf. 963a8–9), the
Athenian does assign it a kind of priority. But what kind of priority is it? It can’t simply
be the priority of unconditional to dependent goods. Wisdom stands in this relation
to courage, the fourth divine good, but not to the second and third divine goods, whose
goodness is not conditional on wisdom (since they involve wisdom). Nor can wisdom’s
priority be ontological, since we have seen that the Athenian takes it to require modera-
tion (696c).
Wisdom’s priority is expressed, if not elucidated, in the metaphor of leadership. In
casting it as “‘first and leader of the divine goods’ (631c6), and—in a version of the cos-
mological thesis attributed to Anaxagoras at Phaedo 97c1–2, —‘Nous, leader of all’
(963a8–9)— the Athenian assigns it a kind of causal priority, more prosaically expressed
in the characterization of the wise person’s ‘pleasures and pains’ as ‘agreeing with and
following’ his correct judgments” (696c8–10). The grand hierarchy of 631b–d casts
wisdom in this leadership role on a civic scale: the wisdom encapsulated and encoded in
the laws will shape the values of the citizenry. Educational effects, however, do not
exhaust its causal priority. Sitting atop of the grand hierarchy of goods, where it func-
tions as the target at which the legislator aims and the good for whose sake all elements
of private and public life in the polis are organized, wisdom has teleological priority in
the well-run polis. It is the final cause or ultimate good—for the city, and for the indi-
viduals within the city.

8. Aristotle’s Debt to the Laws

While the Athenian does little to elucidate the teleological priority of wisdom, it is worth
noting that elucidating this priority is a project to which Plato’s pupil Aristotle devoted
considerable attention. In his ethical writings, Aristotle offers his own version of a teleo-
logical hierarchy with wisdom at the top. It is plausible to read the basic structure of this
ethical theory as a version of the Athenian’s doctrine that human goods “look to” or are
“for the sake of ” the divine, and that divine goods stand in this same relation to wisdom.
 Plato’s Laws   377

Aristotle posits as central to the enterprise of ethics an ultimate skopos or telos because
of which (di’ ho) or for the sake of which (hou heneka, hou charin) we choose or pursue
anything else. With that telos specified as excellent activity according to reason, all
goods other than virtue (“human goods”) will be pursued “for the sake of ” virtue.
Within the domain of the virtues (“divine goods”), the virtues of character are both
dependent upon phronesis and required for it (EN 6.12–13); and despite the symmetry of
that dependence, Aristotle counts the relation as one in which phronesis is the leader, while
the virtue that concerns “pleasures and pains” (EN 2.3, 5; EE 2.2–3)—another echo of the
Laws36—is informed by the judgments of phronesis (EN 1107a1–2; cf. EE 1249a21–b4).37
With Aristotle’s ethical theory sketched at this level of generality, we can recognize it
as both conforming to and developing the basic teleological framework articulated in
the Laws.38

Bibliography
Bartels, M. (2017). Plato’s Pragmatic Project: A Reading of Plato’s Laws. Stuttgart: Franz Steiner
Verlag.
Bobonich, C. (2002). Plato’s Utopia Recast: His Later Ethics and Politics. New York: Oxford
University Press.
Bobonich, C., and K. Meadows. (2013). “Plato on Utopia.” In The Stanford Encyclopedia of
philosophy. Edited by Edward N. Zalta. Stanford, CA: Stanford Univ. Press.
Brisson, L. and J. F. Pradeau. (2007). Les Lois de Platon. Paris: Presses Universitaires de France.
Donini, P. L. (2014). Abitudine e Saggezza: Aristotele dall’«Etica eudemica» all’ «Etica nicoma-
chea». Alessandria: Edizioni dell’Orso.
Harte, V. (2002). Plato on Parts and Wholes: The Metaphysics of Structure. Oxford: Oxford
University Press.
Irwin, T. H. (1995). Plato’s Ethics. Oxford: Oxford University Press.
Kahn, C. H. (2002). “On Platonic Chronology.” in J. Annas and C. Rowe, eds. New Perspectives
on Plato, Modern and Ancient. Washington, DC: Center for Hellenic Studies, 93–127.
Laks, A. (1990). “Legislation and Demiurgy: On the Relationship between Plato’s Republic and
Laws,” Classical Antiquity 9: 209–29.
Laks, A. (2000). “The Laws” in C. Rowe and M. Schofield, eds., The Cambridge History of
Greek and Roman Political Thought. Cambridge: Cambridge University Press, 258–92.
Mayhew, R. (2008). Plato: Laws 10, translated with introduction and commentary. Oxford:
Clarendon Press.

36 Aristotle’s account of virtue follows the basic bipartite division we find at the opening of book II
of the Laws, between reason on the one hand, and “pleasures and pains” on the other (653a–c). The
two types of motivations are modeled in gold and iron cords in the figure of the divine puppets at
644b–645c.
37 I agree with Donini 2014 that Aristotle’s debt to the Laws is evident in the Nicomachean Ethics,
although I think the same can be said for the Eudemian Ethics. I thank Giulia Bonasio for help on this
point.
38 I would like to thank Brian Reese and Harold Parker for helpful discussion of this material, and
Gail Fine, Terence Irwin, and Malcolm Schofield for generous criticism of earlier drafts.
378   susan sauvé meyer

Meyer, S. S. (2006). “Plato on the Law” in H. Benson, ed., A Companion to Plato. Oxford and
Malden MA: Basil Blackwell.
Meyer, S. S. (2015). Plato: Laws 1 & 2, translated with introduction and commentary. Oxford:
Clarendon Press.
Meyer, S. S. (2018). “Virtue and Self Mastery in Plato’s Laws,” in D. Brink, C. Shields, and
S. Meyer (eds.), Virtue, Happiness, and Knowledge. Oxford: Oxford University Press.
Morrow G. (1960). Plato’s Cretan City: A Historical Interpretation of the Laws. Princeton, NJ:
Princeton University Press. Reissued 1993 with a new foreword by Charles Kahn.
Places, E. des and A. Diès. (1951–1956). Platon: Oeuvres Complètes XI–XII. Les Lois, texte
établit et traduit, 4 vols. Paris: Belles Lettres.
Schofield, M. (ed.) and T. Griffith (trans). (2016). Plato: The Laws. Cambridge Texts in the
History of Political Thought. Cambridge: Cambridge University Press.
Schofield, M. (1997). “The Disappearance of the Philosopher King.” Proceedings of the Boston
Area Colloquium in Ancient Philosophy 13: 213–41.
Schofield, M. (2016). Plato’s Laws. Oxford Bibliographies Online. http://www.oxfordbibliogra-
phies.com/abstract/document/obo-9780195389661/obo-9780195389661-0214.xml?rskey=4
RA6iz&result=3&q=schofield#firstMatch
Schofield, M. (2016a). “Plato’s Marionette.” Rhizomata 4 (2): 128–53.
Schöpsdau, K. (1994–2011). Platon: Nomoi (Gesetze). Ubersetzung und Kommentar. 3 Vols.
Göttingen: Vandenhoek und Ruprecht.
 chapter 16

The Epistemol ogy

a n d M eta ph ysics
of Socr ate s

Gareth B. Matthews

The portrayal of Socrates in the early dialogues of Plato is the most vivid picture we
have of any ancient philosopher. Prominent in that depiction is the story Socrates tells
us in the Apology about his reaction to being told that, according to the oracle at Delphi,
no one was wiser than he (21a). Socrates responded to this news, he says, by trying to
find someone of whom he could say, “This man is wiser than I, but you said I was [wiser]”
(21d).1 At least part of his test for determining whether a given candidate was wiser than
he was to determine if the candidate knew something “noble and good” (kalon kagathon).
Socrates himself claimed to know nothing of that sort (21d).
Another part of his test was to see whether candidates thought they knew things that,
in fact, they did not know. At Apology 22c–d, Socrates says of the craftspersons that they
knew many noble things [or fine or beautiful things, polla kalla], and so were, in that
respect, wiser than he. But in the end, he concluded that they were not really wiser
because they mistakenly thought themselves to be wise in other things. “This error of
theirs,” he explains, “overshadowed the wisdom they had” (22d–e). His own wisdom, he
had already said, lay in his not thinking he knew what he did not know (21d). Apparently,
no one excelled him in that!
On the face of it, the project of determining whether someone knows something
noble and good is, at least in part, an epistemological one. So is the project of determining
whether one thinks one knows something one does not really know. Both projects pre-
suppose an understanding of what is to count as knowledge.
Some commentators claim that Socrates, as he is portrayed in the early dialogues of
Plato, has no epistemology at all, or at least no epistemological theory. Thus Gregory
Vlastos, the dean of Socrates scholars in the twentieth century, writes: “In fidelity to our

1 All translations of Plato are taken from Cooper, (ed.)Plato: Complete Works.
380   Gareth B. Matthews

texts no epistemological theory at all can be ascribed to Socrates.”2 Vlastos is equally

dismissive of the suggestion that Socrates has a metaphysics. Socrates, he writes, “is as
innocent of epistemology as of metaphysics.” It is part of my purpose in this chapter to
show how it can be appropriate to speak of the epistemology and the metaphysics of
Socrates, but another part is to make clear why someone might also have doubts or res-
ervations about whether Socrates even has an epistemology or a metaphysics at all.
An even more fundamental issue dogs the project of this chapter: namely, the problem
of how we can know anything about the philosophy of the historical Socrates. We can call
this “the problem of the historical Socrates.” As I have already said, Plato has certainly
given us an extremely vivid picture of Socrates in the early dialogues.3 But one may well
ask how we can know that the picture Plato has given us is historically accurate.
The answer is surely that we cannot know. As Charles Kahn so vividly puts the point,
“The Socrates of the dialogues is an ambiguous figure, at once Plato’s historical master
and his literary puppet.”4 Still, the picture of Socrates in the early dialogues is a memorably
coherent portrait of a philosopher whose views deserve careful interpretation, analysis,
and criticism. It is the epistemology and metaphysics, if any, of the figure that emerges
from that portrait that I discuss here. I leave aside the issue of whether we have good
reason to think that that figure is, in fact, the historical Socrates.
In discussing the epistemology of Socrates, I focus especially on these three topics:
Socratic Ignorance, the Priority of Definitional Knowledge, and the Problem of the
Elenchus.

1. Socratic Ignorance

Socrates makes his claim of ignorance, or, more accurately and pedantically, his dis-
claimer of knowledge, in various forms. Here, from the Apology, is the example5 of that
disclaimer already mentioned previously:

T1. I am wiser than this man; it is likely that neither of us knows anything worthwhile
[or: noble and good, kalon kagathon], but he thinks he knows something when he
does not, whereas when I do not know, neither do I think I know. (21d)

We can distinguish at least two claims of ignorance in this passage:

C1. I do not know anything worthwhile (or: noble and good, kalon kagathon).
C2. I do not know, nor do I think I know, things that others think they know.

2 Vlastos, Socrates, Ironist and Moral Philosopher, 15.

3 I follow Vlastos in counting these as the early dialogues of Plato: Apology, Charmides, Crito,
Euthyphro, Gorgias, Hippias Minor, Ion, Laches, Protagoras, Republic I (Vlastos, Socrates, Ironist and
Moral Philosopher, 46–47).
4 Charles H. Khan, “Did Plato Write Socratic Dialogues?,” in Benson (ed.), Essays on the Philosophy
of Socrates, 35–52.
5 Other relevant passages include these: Apology 29b, Charmides 165b–c, Laches 200e, Gorgias 509a,
and Hippias Minor 372b–e.
 The Epistemology and Metaphysics of Socrates   381

For another statement of Socratic ignorance, we can look to this passage from the dia-
logue, Charmides, where Socrates is speaking:

T2. ‘‘But Critias,’’ I replied, ‘‘you are talking to me as though I professed to know the
answers to my own questions and as though I could agree with you if I really wished.
This is not the case—rather, because of my own ignorance, I am continually investi-
gating in your company whatever is put forward. However, if I think it over, I am
willing to say whether I agree or not. Just wait while I consider.’’ (165b–c)

In T2, Socrates seems to be making the following claim:

C3. I do not profess to know the answers to my own questions.

If we put these three claims, C1, C2, and C3, together, we get the picture that Socrates
claims ignorance of (1) things noble and good, (2) that others think they know, and (3) that
are the subject of his investigations when he questions his fellow Athenians. Although
it is not immediately clear how extensive this knowledge disclaimer is, it seems to fall
short of being a claim not to know anything at all. Indeed, a few pages later in the Apology
Socrates says this about his practice of questioning his fellow Athenians:

T3. I know well enough [oida schedon] that this very conduct makes me unpopular.
(24a)

And five pages later, he makes this morally significant claim:

T4. I do know [oida], however, that it is wicked and shameful to do wrong, to disobey
one’s superior, be he god or man. (29b)

Finally, the confident and well-targeted way in which he questions his interlocutors at
least suggests that Socrates must know quite a bit about what he is inquiring into, even
though he does not know what he most wants to know.

2. Complete Ignorance?

There is, however, at least one passage in a Platonic dialogue in which, astoundingly,
Socrates claims total ignorance of what he is clearly most interested in—namely, virtue.
What I have in mind is this passage from near the beginning of the dialogue, Meno:

T5. Socrates, ‘‘I myself, Meno, am as poor as my fellow citizens in this matter, and I
blame myself for my complete ignorance [literally, not knowing at all, to parapan]
about virtue. If I do not know what something is, how could I know what qualities
it possesses? Or do you think that someone who does not know at all [to parapan]
who Meno is could know whether he is good-looking or rich or well-born, or the
opposite of these? Do you think that it is possible?’’ (71b)
382   Gareth B. Matthews

Socrates adds that he has never met anyone who did know what virtue is (71c). What T5
commits Socrates to seems to be nothing less than this:

C4. Neither I nor, as it seems to me, anyone I have met, knows at all what virtue is.

The fact that Socrates here claims not to know at all what virtue is, and repeats the “at all”
[to parapan], should not be taken lightly. This is, in fact, the only dialogue in which Socrates
claims not to know at all what F-ness is. It is natural to conclude that Socrates is here dis-
claiming knowledge that, for example, bravery is a virtue or that any token person or action
is virtuous. Can this claim of complete ignorance of what virtue is be taken seriously?
Before we draw any rash conclusions from T5, we should note that the Meno seems to
be a transitional dialogue. The first part of the dialogue is fairly typical of the early dia-
logues. But we soon come to the Paradox of Inquiry, which threatens the very rationality
of Socratic inquiry:

T6. How will you look for [virtue], Socrates, when you do not know at all [to para-
pan] what it is? How will you aim to search for something you do not know at all?
If you should meet with it, how will you know that this is the thing that you did
not know? (Meno 80d)

A few lines later, Socrates introduces, apparently in response to the Paradox of Inquiry, a
grand epistemological and metaphysical hypothesis:

T7. As the soul is immortal, has been born often and has seen all things here and in
the underworld, there is nothing which it has not learned; so it is in no way surpris-
ing that it can recollect things it knew before, both about virtue and other things. As
the whole of nature is akin, and the soul has learned everything, nothing prevents a
man, after recalling one thing only—a process men call learning—discovering every-
thing else for himself, if he is brave and does not tire of the search, for searching and
learning are, as a whole, recollection. (Meno 81c–d)

T7 seems to entail the denial of T5. The idea it expresses is that, instead of Socrates being
completely ignorant of virtue, “there is nothing which the soul has not learned.”6 In the
context, the implication of this claim seems to be that “the soul,” and so, presumably, the
soul of Socrates, as well as the souls of his interlocutors, have already learned, among
other things, what virtue is. The knowledge that “the soul” has may, at any given time, be
only latent, until it is made manifest through recollection.7 However, since “the whole of

6 Exactly how this passage is to be understood is subject to debate. Indeed, the question of how to
understand the doctrine of recollection is far from settled. Vlastos, “Anamnesis in the Meno,” is an
important early article on the subject. Fine, “Inquiry in the Meno,” is an important later contribution,
which includes references to other interpretations. And Scott, Plato’s Meno, see especially pp. 96–97, is
the most recent contribution. ed.: See Judson, Chapter 7 of this volume.
7 “Recollection” may be taken “thinly” or “thickly.” On a very thin reading, “what Plato means by
‘recollection’ in the Meno is any enlargement of our knowledge which results from the perception of logical
relationships” (Vlastos, Studies in Greek Philosophy, vol. 2, 155–56). I suspect we should go for a reading
at least somewhat thicker than that.
 The Epistemology and Metaphysics of Socrates   383

nature is akin,” it seems that one who can manage to recollect a single bit of relevant
knowledge may be able to regain manifest knowledge of what he latently knows about
other things—provided “he is brave and does not tire of the search.”
If we now read T5 in light of T7, we see that we should probably understand Socrates’
claim in T5 to be in “complete ignorance” about what virtue is to be only this: he does not
know at all manifestly what virtue is, even though he knows latently everything there is
to know.
Let’s return now to claims C1, C2, and C3. Putting them together, and adding a refer-
ence to the distinction we get in the Meno between manifest and latent knowledge, we
get something like this as the claim of Socratic ignorance:
C5. Socrates claims that (i) he does not know (manifestly) anything noble and good,
where (ii) knowing something noble and good would bring with it the ability to
answer satisfactorily Socrates’ ‘‘What is F-ness?’’ questions and (iii) his interlocutors
(at least initially) think they know how to do that.

3. Examples

In the early dialogues, Socrates is unwilling to accept an example of a thing that is F, or

even a list of examples of F-things, as a satisfactory answer to his “What is F-ness?” ques-
tion.8 Thus, when he asks Euthyphro what piety is, Socrates refuses to accept the answer,
“Doing what I am doing now.” Here is part of the exchange:

T8.
Socrates: For now, try to tell me more clearly what I was asking just now, for,
my friend, you did not teach me adequately when I asked you what
the pious was, but you told me that what you are doing now, pros-
ecuting your father for murder, is pious.
Euthyphro: And I told the truth, Socrates.
Socrates: Perhaps. You agree, however, that there are many other pious actions.
Euthyphro: There are.
Socrates: Bear in mind then that I did not bid you tell me one or two of the
many pious actions but that form itself [auto to eidos] that makes all
pious actions pious, for you agreed that all impious actions are
impious and all pious actions pious through one form, or don’t you
remember?
Euthyphro: I do.
Socrates: Tell me then what this form itself is, so that I may look upon it, and
using it as a model, say that any action of yours or another’s that is of
that kind is pious, and if it is not that it is not. (Euthyphro 6c–e)

8 There is scholarly debate about whether the examples in question are types of action or action
tokens. See Nehamas, “Confusing Universals and Particulars,” and Benson, “Misunderstanding the
‘What Is F-ness?’ Question.”
384   Gareth B. Matthews

So what would Socrates require before he would recognize that he had been “taught
adequately”? That is, what would Euthyphro have to do to tell Socrates “what the form of
piety itself is,” so that Socrates could use the response as a “model” or pattern (paradeigma)
to determine which actions are pious? From the evidence of the early dialogues, it seems
that what Socrates requires as an adequate answer to his “What is F-ness?” question is
some privileged set of necessary and sufficient conditions for something to count as an
F. The set of necessary and sufficient conditions cannot just be features that happen to
belong to all and only people or things that are F; they must reveal that by which x is F—that
is, what makes x F. To put the point another way, they must reveal the nature (ousia) of
what it is to be F (Euthyphro 11a). Let’s call a set of necessary and sufficient conditions that
reveals the nature of what it is to be F an “explanatory set of necessary and sufficient con-
ditions.” My suggestion, then, is that the things “noble and good” mentioned in T1 that
Socrates is most interested in knowing, and that he thinks neither he nor his interlocu-
tors know, are an explanatory set of necessary and sufficient conditions for an action or a
person to count as being brave, pious, temperate, wise, just, and, more generally, virtuous.
In the commentary literature, knowing an explanatory set of necessary and sufficient
conditions for something to be F is called definitional knowledge of F-ness. I use that
terminology, however, with the warning that the definitional knowledge under discussion
is not the meaning of a lexical item, say, a word in Greek. It is, rather, an account of what
it is that makes something or someone brave, or temperate, or virtuous, or whatever. Put
another way, definitional knowledge of virtue, or one of the individual virtues, is having
an account of a moral kind that reveals its nature.
Socrates’ recognition of his own lack of definitional knowledge seems to be at least
one thing that motivates the “What is F-ness?” questions he asks his interlocutors. He
seems to want to learn whether his interlocutors know what F-ness is, when he himself
does not know. Moreover, he seems to want to learn for himself what F-ness is through
questioning his interlocutors.
We are now in position to state what I shall call the “moderate interpretation” of the
claim of Socratic ignorance:

Socratic Ignorance (Moderate Interpretation): Socrates claims that (i) he does not
know (at least not manifestly) anything noble and good, where knowing something
noble and good would be (ii) having definitional knowledge of virtue, or one of the
virtues, such as piety or bravery, and (iii) his interlocutors (at least initially) think they
have satisfactory knowledge of virtue itself, or whichever one of the virtues Socrates
asks them about.

4. Senses of “Know”

The claim of Socratic Ignorance, even on this moderate interpretation, expresses a

significant disclaimer. Some commentators have thought the Socratic disclaimer to
be far less radical than this. Prominent among the efforts to understand the Socratic
 The Epistemology and Metaphysics of Socrates   385

disclaimer in a less radical way than this is the interpretation offered by Vlastos.
Vlastos writes:

In Plato’s earliest dialogues, when Socrates says he has no knowledge, does he or

does he not mean what he says? The standard view has been that he does not. What
can be said for this interpretation is well said in Gulley, 1968: Socrates’ profession of
ignorance is ‘‘an expedient to encourage his interlocutor to seek out the truth, to
make him think that he is joining with Socrates in a voyage of discovery’’ (p. 69).
More recently the opposite interpretation has a clear-headed advocate. Terence
Irwin in his Plato’s Moral Theory9 holds that when Socrates disclaims knowledge he
should be taken at his word: he has renounced knowledge and is content to claim no
more than true belief.10

After considering these two alternatives—(1) Socrates is being insincere as a way of

drawing his interlocutor into discussion, and (2) Socrates really means that he knows
nothing—Vlastos proposes a way to go between the horns of the dilemma. He distin-
guishes two senses of “know” or, better, two senses of the Greek verbs we translate as
“know.” In what we can call the “strong sense” of the verbs for “know,” which Vlastos
marks with a subscript as “knowC,” we know all and only what we are infallibly certain
of. In the weak sense, by contrast, which Vlastos marks with a subscript as “knowE,” we
can know whatever has survived elenctic examination.
Vlastos then explains how we are to understand the apparent clash between Socratic
claims of knowledge and Socratic disavowals of knowledge this way:

When he says he knows something he is referring to knowledgeE; when he says he is

not aware of knowing anything—absolutely anything, ‘‘great or small’’ . . . —he refers
to knowledgeC; when he says he has no knowledge of a particular topic he may mean
either that in this case, as in all others, he has no knowledgeC and does not look for
any or that what he lacks on that topic is knowledgeE, which, with good luck, he might
still reach by further searching.11

So now we have this interpretation of the claim of Socratic Ignorance:

Socratic Ignorance (two-senses interpretation): Socrates knowsC (i.e., is infallibly

certain of) very little, even though Socrates knowsE (i.e., has found to survive elenctic
examination) quite a bit.

The proposal is ingenious. But it faces several difficulties. First, and most obviously,
Socrates never says anything in the early dialogues about using one or more of his verbs

9 The reference is to Irwin, Plato’s Moral Theory, 140. In Irwin, Plato’s Ethics, we find this: “If
[Socrates] says he has knowledge, but in a context in which he has made it plain that he claims only
human wisdom, his apparent claim to knowledge may simply amount to a claim that he recognizes
that his convictions do not really count as knowledge” (28).
10 Vlastos, Socratic Studies, 39. 11 Ibid., 58.
386   Gareth B. Matthews

for “to know” in different senses. Nor does he say anything of this form: “In one sense
I know this, but in another sense I do not.”
There is a second difficulty. It turns on the specific senses that Vlastos assigns to the
relevant Greek verbs for “know.” Here is the problem. In T1, Socrates says of himself and
his interlocutor, “It is likely that neither of us knows [eidenai] anything worthwhile
[kalon kagathon].” How are we to use Vlastos’s disambiguation proposal to clarify the
meaning of Socrates’ verb here for “to know,” eidenai? We have these two options:

1. It is likely that neither of us knowsC anything worthwhile.

2. It is likely that neither of us knowsE anything worthwhile.

According to Vlastos’s proposal, Socrates will certainly think (1) is true, on the
grounds that there are so few things that anyone could be infallibly certain of. But that
conclusion seems irrelevant to the practice of interrogating his interlocutors. After all,
the elenchus is not aimed at determining whether there is something that he is infallibly
certain of, such as the fact that he exists. It is, rather, aimed at determining whether
the interlocutor has any beliefs that will survive elenctic examination: that is, whether the
interlocutor knowsE anything. So let’s turn to (2).
The trouble with (2) is that, according to Vlastos, (2) will be false, or at least Socrates
will consider it false. For according to Vlastos, the set of moral beliefs Socrates holds at
any given time12 has survived elenctic examination. Those beliefs are then, presumably,
things that Socrates knowsE. So, on this reading, Socrates will consider his own statement
false. What all this means, I think, is that the Vlastosian proposal fails to make room for
a plausible reading of T1.13

5. Limited Knowledge

Hugh Benson accepts the Socratic claims of ignorance as the implication of a very strong
and restrictive conception of what it is to know something. Indeed, as Benson remarks
at the end of his book on the Socratic conception, knowledge is “very difficult to obtain.
Indeed, we should be surprised to discover that anyone has it.”14
Benson must, of course, explain away the passages in which Socrates claims to have
knowledge, such as T3 and T4. Here is T3 in its broader context:

12 After Socrates has engaged in a number of elenctic examinations, anyway.

13 There might, of course, be other two-sense interpretations that would avoid this difficulty. A
related interpretative move is to distinguish, not senses of Socrates’ words for “know,” but rather kinds
of knowledge. Woodruff, “Plato’s Early Epistemology,” distinguishes expert and non-expert knowledge,
where expert knowledge must include definitional knowledge. Other interpreters (e.g., Taylor, Socrates,
50ff.) follow Woodruff. Brickhouse and Smith, Plato’s Socrates, distinguish “knowing how something is”
from a lesser sort of knowledge (38–45). These latter lines of interpretation are more congenial to the
view of Socratic Ignorance I present here. On the importance of keeping views about senses distinct
from views about kinds, see Matthews, “Senses and Kinds.”
14 Benson, “Misunderstanding the ‘What Is F-ness’ Question,” 221.
 The Epistemology and Metaphysics of Socrates   387

T3*. To fear death, gentlemen, is no other than to think oneself wise when one is not,
to think one knows what one does not know. No one knows whether death may not
be the greatest of all blessings for a man, and men fear it as if they knew that it is the
greatest of evils. And surely it is the most blameworthy ignorance to believe that one
knows what one does not know. It is perhaps on this point and in this respect, gen-
tlemen, that I differ from the majority of men, and if I were to claim that I am wiser
than anyone in anything, it would be in this, that, as I have no adequate knowledge
of things in the underworld, so I do not think I have. I do know (oida), however, that
it is wicked and shameful to do wrong, to disobey one’s superior, be he god or man.
(Apology 29b1–7; emphasis added)

This passage includes a qualified claim of ignorance (“I have no adequate knowledge of
things in the underworld”), as well as a clear claim to knowledge of an important moral
truth. Socrates does not explain here why he thinks his knowledge of the “underworld”
is inadequate. It is plausible to assume that he thinks it is inadequate because, as yet, he
has had no experience of the underworld. By contrast, he claims to know “that it is wicked
and shameful to do wrong, to disobey one’s superior, be he god or man.”
On the moderate interpretation of Socratic Ignorance I advanced previously, Socrates
can consistently claim to know that it is wicked and shameful to disobey one’s superior
and also claim not to know anything worthwhile [kalon kathagon], if, for example, he
realizes that he lacks definitional knowledge of what it is to be wicked or shameful.
Indeed, on the moderate interpretation of Socratic Ignorance, Socrates can consistently
(1) claim to know certain moral facts, but (2) deny that he has definitional knowledge of
the elements of those moral facts (e.g., what it is to be shameful, wicked, etc.), and (3)
express puzzlement over how it is he can know such moral facts without having the
appropriate definitional knowledge. This is, in fact, the situation of many analytic philos-
ophers today. They will claim to know that 7 is a number but at the same time admit that
they have no definitional knowledge of what it is to be a number. Furthermore, they may
also be puzzled over how they can know that 7 is a number without having the appropri-
ate definitional knowledge. They may be puzzled because they find plausible a principle
that Socrates seems to accept.

6. The Priority of
Definitional Knowledge

According to a principle many commentators15 have thought Socrates is committed to,

one could not know that 7 is a number without having definitional knowledge of number.
Peter Geach once called this principle “the Socratic fallacy” and claimed it to be a morally

15 But certainly not all. Among the dissenters are Nehemas, “Socratic Intellectualism”; Beversluis,
“Does Socrates Commit the Socratic Fallacy?”; and Brickhouse and Smith, Plato’s Socrates, especially
pp. 45–55.
388   Gareth B. Matthews

pernicious fallacy.16 Hugh Benson, in a more recent formulation, calls it the “Priority of
Definitional Knowledge” and states it as the conjunction of (P) and (D):

(P) If A fails to know what F-ness is, then A fails to know, for any x, that x is F.
(D) If A fails to know what F-ness is, then A fails to know, for any G, that F-ness is G.17

According to (P), if Socrates and his interlocutor fail to know—that is, to have definitional
knowledge—of what, say, beauty is, they will fail to know of any given instance of beauty,
say, Helen of Troy, or the Parthenon, that Helen is beautiful or the Parthenon is beauti-
ful. According to (D), if Socrates or one of his interlocutors fails to have definitional
knowledge of bravery, they will not even know that bravery is a virtue.
(P) and (D), both individually and jointly, threaten to undermine the reasonableness
of thinking one might ever come to know what F-ness is through a Socratic elenchus. To
see that this is so, consider, for example, the way the elenchus proceeds in the dialogue,
Laches. When Laches proposes that courage is standing fast in the face of an advancing
enemy, Socrates asks whether one couldn’t be brave in retreat. (191a) Laches, like anyone
who reads this dialogue, recognizes immediately that a soldier might be brave in retreat
and that such a soldier would be a counterexample to the suggested analysis of courage.
But, according to (P), one could not know that this example is a genuine counterexample
unless one already knew what courage is—that is, had definitional knowledge of bravery.
But, if one had that, there would be no point in conducting the elenchus.
Again, when Charmides suggests that temperance might just be a certain quietness,
Socrates gains his agreement that temperance is one of the noble, or admirable, things
(tôn kalôn) (Charmides 159bc). Socrates then points out that quietness is sometimes noble,
sometimes not, and concludes that temperance cannot be just a certain quietness (160bc).
But if they do not actually know that temperance is a noble thing, they will not know
that it is not a certain quietness.
Perhaps, however, it is not the aim of the Socratic elenchus to provide knowledge of
what F-ness is. Perhaps the aim is only to determine what F-ness is. Or perhaps it is not
even that. Before we can make a reasonable judgment concerning the aim of the Socratic
elenchus, we need to consider what the form of an elenchus is and then what it would be
reasonable to take the aim of inquiry in that form to be.

7. The Elenchus

Vlastos describes the steps that make up the “standard” Socratic elenchus this way:

1. The interlocutor asserts a thesis [p], which Socrates considers false and targets for
refutation.

16 Geach, “Plato’s Euthyphro.” 17 Benson, Socratic Wisdom, 113.

 The Epistemology and Metaphysics of Socrates   389

2. Socrates secures agreement to further premises, say q and r (each of which may
stand for a conjunct of propositions). The argument is ad hoc: Socrates argues
from q and r, but not to them.
3. Socrates then argues, and the interlocutor agrees, that q and r entail not-p.
4. Thereupon Socrates claims that not-p has been proved true, p false.18

A quarter of a century before Vlastos wrote “The Socratic Elenchus,” he had main-
tained that “Socrates never meant to go beyond (3) in his elenctic arguments—that their
object was simply to reveal to his interlocutors muddles and inconsistencies within
themselves.”19 In “The Socratic Elenchus,” however, Vlastos argues that Socrates makes
two powerful assumptions that enable him to assert, with justification, that not-p—that,
for example, what his interlocutor had said piety (or bravery, or whatever) is—is simply
false. The powerful assumptions Vlastos attributes to Socrates are these:

A. Anyone who ever has a false moral belief will always have at the same time true
beliefs entailing the negation of that false belief.20
B. The set of moral beliefs held by Socrates at any given time is consistent.21

These two assumptions, Vlastos maintains, entail this:

C. The set of moral beliefs held by Socrates at any given time is true.22

Almost all commentators have declined to follow Vlastos in attributing these particu-
lar assumptions to Socrates.23 Nevertheless, the problem Vlastos uncovered, and tried
to solve, has dominated Socratic scholarship for the past two decades.

8. The Problem of the Elenchus

Vlastos states the problem this way: “how is it that Socrates claims to have proved a the-
sis false when, in point of logic, all he has proved in any given argument is that the thesis
is inconsistent with the conjunction of agreed-upon premisses [sic]?”24
I call this the “Weaker Version” of the problem and state it this way:

The Problem of the Elenchus (Weaker Version): How can the elenchus establish any-
thing more than inconsistency between a thesis about what F-ness is and certain
premises that have been agreed to by the interlocutor?

18 Vlastos, “Socratic Elenchus,” 39. 19 Ibid., 45.

20 Ibid., 52. 21 Ibid., 55. 22 Ibid.
23 For example, Kraut, “Comments on Gregory Vlastos”; Brickhouse and Smith, “Vlastos on the
Elenchus”; and Polansky, “Professor Vlastos’s Analysis of Socratic Elenchus.”
24 Vlastos, “Socratic Elenchus,” 49.
390   Gareth B. Matthews

In his article, Vlastos makes heavy use of the dialogue Gorgias for clues to the
­methodology of the elenchus. One of the passages he quotes is this one:

T9. For I think that we should all be contentiously eager to come to know [to eidenai]
what is true and what is false about the things we assert; for it is a common good for
all that this should be made evident. (505e)

This passage suggests that it is knowledge of what is true and what is false that the elenchus
aims to establish. It presents us with a stronger version of the Problem of the Elenchus:

The Problem of the Elenchus (Stronger Version): How can the elenchus be used to
achieve knowledge of what F-ness is?

9. Responding to the Problem

of the Elenchus

Vlastos does not mark off the stronger version of the Problem of the Elenchus from the
weaker version. As we have already noted, his general response to the Problem of the
Elenchus is to say that Socrates relies on assumptions A, B, and C. If, indeed, Socrates
does rely on these assumptions, then he does have available a response to the weaker ver-
sion of the Problem of the Elenchus. That is, given these assumptions, Socrates can claim
to have established truths about the virtues through his use of the elenchus. Moreover,
given Vlastos’s two-sense construal of Socratic Ignorance, Socrates can be said to know
whatever has survived elenctic examination. And so Vlastos also has a response to the
stronger version of the Problem of the Elenchus.
Suppose, however, we find it implausible to attribute assumptions A, B, and C to
Socrates. What response would be available to us to resolve the Problem of the Elenchus?
Terence Irwin suggests that Socrates makes other assumptions, what he calls the
“Guiding Principles of Socratic Inquiry.”25 According to Irwin, “If [Socrates] has good
reasons for believing these principles, then he has good reason to accept the conclusions
of elenctic inquiry.”26
Among the five principles Irwin supposes that Socrates’ interlocutors agree to is one
that begins this way:

[The interlocutor] agrees that a virtuous action must always be ‘‘fine’’ (kalon), ‘‘good’’
(agathon), and ‘‘beneficial’’ (ophelimon). If an action is shameful or harmful, the
interlocutor agrees that it cannot be virtuous, and that a state of an agent producing
such an action cannot be a virtue.27

25 Irwin, Plato’s Ethics, 48–49. 26 Ibid., 48. 27 Ibid.

 The Epistemology and Metaphysics of Socrates   391

Irwin adds:

These are guiding principles of the elenchos, not simply Socrates’ own beliefs. For he
assumes—and his assumption is proved right in the dialogues we have examined—
that the interlocutor will accept these principles and be influenced by them in
answering Socrates’ questions.28

It is Irwin’s view that Socrates does not attempt to argue for these guiding principles of the
elenchus until we get to the Gorgias. In that dialogue, Irwin tells us, Socrates claims to
demonstrate truths, even though he still does not “know how things are.” Irwin unpacks
that paradox this way: “By showing how anyone disagreeing with him becomes ridiculous,
[Socrates] has shown enough to prove his own position, but not enough to justify him in
claiming to know how things are.”29 In this way, Irwin offers a response to the weaker
version of the Problem of the Elenchus while rejecting the stronger version on the grounds
that Socrates does not suppose he has acquired any knowledge through the elenchus.30

10. Rejecting the Problem of

the Elenchus Altogether

Some commentators reject both the stronger version and the weaker version of the
Problem of the Elenchus. The clearest way to do that is to argue that Socrates never claims
to establish any positive conclusions through his use of the elenchus. Hugh Benson is
one commentator who takes this position. Benson reconstructs eight elenchi from early
dialogues, each of which shows only that the premises of that elenchus are inconsistent.
After discussing whether the Gorgias alters the picture significantly, Benson comes to
this conclusion:

I have shown, then, that ‘‘the problem of the elenchus’’ need not arise in Plato’s
early dialogues. A careful examination of the elenchoi employed in the Euthyphro,
Laches, and Charmides, three paradigmatic elenctic dialogues, requires only that
Socrates understands each of these elenchoi as establishing the inconsistency of
the interlocutor’s beliefs. The passages of the Gorgias that have been claimed to
suggest otherwise do not in fact require an alternative interpretation of Socrates’
understanding.31

28 Ibid., 49. 29 Ibid., 122.

30 Perhaps, however, this is not Irwin’s own view about whether one can acquire knowledge through
an elenchus. In his earlier work, Plato’s Moral Theory, he writes: “Socrates should not believe, then, that
the elenchus can never in principle meet his demand for definition and knowledge. Naturally the
definitions it finds will rely on the agreement between Socrates and the interlocutor; one interlocutor
might be wrong to agree, but answers can be confirmed by other people’s agreement. This all looks a
feasible programme for Socrates; when he and his interlocutors, after repeated tests, are satisfied with
an account, he can claim knowledge derived from the elenchus” (68–69).
31 Benson, Socratic Wisdom, 92.
392   Gareth B. Matthews

There are lots of good things to be said in favor of Benson’s interpretation. But there is
also something odd about it. As Socrates tells his story in the Apology, he set out to dis-
cover whether others are wiser than he is (21be). What would be a reasonable way of
doing that? Trapping someone in a contradiction to show that the interlocutor actually
has inconsistent beliefs might be a reasonable way to show that the interlocutor is not
wise at all, and so is not wiser than Socrates, assuming, of course, that Socrates cannot be
trapped in such contradictions. But setting out to determine whether people are wise by
simply determining whether they have consistent beliefs seems to be a hopeless way to
determine whether they are wiser than Socrates. It is quite conceivable that a very simple-
minded person with very few beliefs might actually have consistent beliefs. But finding,
after a thorough examination, no inconsistency in a simpleton’s belief set would hardly
tend to show that the simpleton is wise, let alone wiser than Socrates.
Socrates, moreover, often presents himself as deeply interested in finding out for him-
self, through the examination of an interlocutor, what courage is, or holiness, or virtue,
or friendship, or whatever the item up for discussion might be. In fact, Benson agrees
that Socrates himself wants to find out the nature of courage and virtue and holiness and
friendship and all the other things he discusses with his interlocutors.32 But he thinks
Socrates wants to find out what these important things are by finding someone who
knows what they are who can teach him.
This suggestion is also rather troubling. Testing putative experts on, say, bravery, for
the consistency of their beliefs about bravery seems a very unpromising way of determin-
ing whether they know what bravery is. For all we know, there are many distinct belief
sets about bravery that are internally consistent, yet incompatible with each other and
therefore not all correct. Thus trying to find out for oneself what bravery is by seeking to
identify a teacher who has merely consistent beliefs about bravery is hardly a promising
strategy. In the absence of solid textual evidence, we should be reluctant to attribute it to
Socrates.

11. Testing Beliefs for Consistency

Is it plausible to suppose, as Benson does, that testing the interlocutor’s beliefs for consist-
ency is the sole, or even the chief, aim of the Socratic elenchus? Only rarely does Socrates
himself suggest that he may be testing his interlocutor’s beliefs for their consistency. In
one such instance, Nicias had proposed that “courage is knowledge of the grounds of
fear and hope” (Laches 196cd). Socrates mentions, as a possible counterexample to
Nicias’s proposal, a ferocious beast, the fabled “Crommyon sow” (196). Nicias responds
by rejecting the counterexample. He says, simply, “By no means . . . do I call courageous
wild beasts or anything else that, for lack of understanding, does not fear what should be

32 “I do not deny that these passages indicated that Socrates aims to uncover truths and acquire
knowledge and to encourage his interlocutors to do the same as well. (How could one?)” (Benson,
“Problems with Socratic Method,” 107).
 The Epistemology and Metaphysics of Socrates   393

feared” (197a). In this case, one could certainly say that Socrates is testing his interlocutor
for the consistency of his beliefs. Moreover, Nicias passes the test!
Yet it is quite unusual in the early dialogues for an interlocutor to reject a counterex-
ample that Socrates has suggested. Moreover, Socrates typically claims outright that
he is asking what F-ness is in the hope of finding out for himself what it is. Thus when
Euthyphro, near the beginning of the dialogue named after him, claims to “know accu-
rately” (akribôs eideiên) what piety is (5a), Socrates offers to become Euthyphro’s pupil.
The clear implication, whether ironical or not, is that he might come to know what
Euthyphro knows. This thought is repeated later in the dialogue, for example, at 14bc.
Similarly, the Laches unfolds as an unsuccessful search for what bravery is, not as a
determination that the beliefs of Nicias and Laches are inconsistent. Thus when Socrates
and Nicias consider whether bravery might be analyzed as knowledge of what it is be
feared and hoped for (196d), Socrates points out that such a conclusion would be in con-
flict with a rather obvious assumption they had made earlier: namely, that courage is
only a part of virtue:

T10.
Socrates: Then the thing you are now talking about, Nicias, would not be a part
of virtue but rather virtue entire.
Nicias: So it seems.
Socrates: And we have certainly stated that courage is one of the parts of virtue.
Nicias: Yes, we have.
Socrates: Then what we are saying now does not appear to hold good.
Nicias: Apparently not.
Socrates: Then we have not discovered, Nicias, what courage is. (Laches 199e)

One might have expected at this point a confirmation that the most an elenchus can
establish is inconsistency of belief. The relevantly inconsistent beliefs would be these:

B1. Courage is only a part of virtue.

B2. Courage is the whole of virtue.

But Socrates does not say, “So, Nicias, you will have to give up at least one of your beliefs.”
Instead, he says, “Then we have not discovered [êurêkamen] what courage is.” That is,
they have not come to know what courage is.

12. Rejecting the Priority of

Definitional Knowledge

If we suppose Socrates is firmly committed to the Priority of Definitional Knowledge,

and we are unhappy about making the Socratic elenchus presuppose either the assump-
tions Vlastos makes or those Irwin proposes, we may see no alternative to limiting the
394   Gareth B. Matthews

aim of the elenchus to establishing inconsistency in the interlocutor’s beliefs, as Benson

does. But do we have to suppose that Socrates is committed to the Priority of Definitional
Knowledge?
Near the end of the Hippias Major, there is a long passage that seems to commit
Socrates to the Priority principle. Here is part of it. (Socrates is speaking.)

T11. I hear every insult from that man (among others around here) who has always
been refuting me . . . .’’Look,’’ he’ll say, ‘‘How will you know whose speech—or any
other action—is finely presented or not, when you are ignorant of the fine? And
when you’re in a state like that, do you think it’s better for you to live than die?’’
(304c1–e1)

The project in this dialogue is to investigate what to kalon is: that is, what the beautiful,
or the fine, or the noble, is. Socrates and Hippias have not been able to come up with a
successful analysis. Socrates says in T8 that he expects to be ridiculed by a neighbor when
he talks about a fine speech, since he cannot give a satisfactory analysis of what it is for
something to be fine. But before we immediately conclude that Socrates is committed to
(P) and (D), we should see how the speech that begins with T11 ends:

T12. That’s what I get, as I said. Insults and blame from you, insults from him. But I
suppose it is necessary to bear all that. It wouldn’t be strange if it were good for me.
I actually think, Hippias, that associating with both of you has done me good. The
proverb says, ‘‘What’s fine is hard’’—I think I know that. (304e3–9)

On a natural, and I think correct, reading of T12, Socrates here says that he thinks he
knows that what is fine is hard (more literally, that “noble things are difficult”—chalepa
ta kala). But if he did know about fine (or noble) things that they are difficult, he would
know something, indeed, something very important, about the fine, or the noble (to
kalon) without, as the dialogue amply demonstrates, having definitional knowledge of
the noble or the fine. So, just after he seems to have accepted the Priority of Definitional
Knowledge, he seems to flout it. What is going on?

13. An Aporetic Reading

It is important, I think, to recognize that a standard way in which Socrates expresses (P)
and (D) is by asking a question. “How can I know that x is F,” he asks, “if I don’t know what
F-ness is?” “How can I know that F-ness is G,” he asks, “if I don’t know what F-ness is?”33
It is, indeed, natural to suppose that, if I really do know that the bird on the tree in
front of my window is a robin, I must know the identifying features of robins and be able

33 In addition to T11 and T12, see, e.g., Charmides 176a–b and Laches 190b–c.
 The Epistemology and Metaphysics of Socrates   395

to “tick them off ” as I look at the bird in the tree to be sure that it has all these identifying
features. Suppose, however, that someone asks me something more basic, say, whether a
robin is a bird, or whether “7” is a number. Surely I know that a robin is a bird and that
“7” is a number. Yet it is overwhelmingly unlikely that I can specify an explanatory set of
necessary and sufficient conditions for the truth of “x is a number” or even for “y is a
bird.” But how can I know those basic things without being able to “tick off ” the right
necessary and sufficient conditions? It is puzzling.
If we read the passages in which Socrates is taken to commit himself to (P) or (D) as
expressions of puzzlement or perplexity over how he can know that a speech is fine
when he can’t informatively define “fineness”—or, in general, how he can know that x is
F without having definitional knowledge of F-ness—then we can understand such
passages, as I think we should, as motivating the elenchus without immediately under-
mining its chance of success in yielding knowledge. I suggest we read those passages this
way: that is, give them an “aporetic reading.”
Giving an aporetic reading to the passages in which Socrates asks how we can know
that x is F or F-ness is G without having definitional knowledge of F-ness makes this part
of the epistemology of Socrates relevant to recent discussions of natural kinds and natu-
ral-kind terms. Both Saul Kripke34 and Hilary Putnam35 have argued that we do not, in
general, recognize instances of gold or elm trees by applying the necessary and sufficient
conditions for something to be gold or an elm tree. Their conclusion has seemed radical
to many readers because of the plausibility of the very picture of instance recognition we
find in T8.
If we do give the passages that suggest (P) and (D) an aporetic reading, we should not
be surprised to find Socrates himself making claims that violate (P) or (D), or both. In
addition to T9, we might consider another passage from the Apology. The one I have in
mind comes after the jury at the trial has just ruled that Socrates is guilty as charged. To
understand this passage, we must understand a certain feature of the Athenian system of
justice. According to that system, if the defendant in court is found guilty, the party who
brought charges proposes a penalty and the party found guilty proposes an alternative
penalty. The jury then has to accept one of the two proposals; it cannot come up with some
completely different penalty. Meletus, Socrates’ accuser, proposes death, and Socrates
needs to propose an alternative. Should he propose some sort of prison term? The jury
might accept that rather than the death penalty. This is part of what Socrates says:

T13. Since I am convinced that I wrong no one, I am not likely to wrong myself, to
say that I deserve some evil and to make some such assessment against myself. What
should I fear? That I should suffer the penalty Meletus has assessed against me, of
which I say I do not know whether it is good or bad? Am I then to choose in prefer-
ence to this something that I know very well [eu oida] to be an evil and assess the
penalty of that? Imprisonment? Why should I live in prison, always subjected to the
ruling magistrates the Eleven? (Apology 37b2–c2)

34 Kripke, Naming and Necessity. 35 Putnam, “Meaning of ‘Meaning,’ ” 215–71.

396   Gareth B. Matthews

At the end of his trial Socrates gives reasons for explaining his uncertainty about
whether death is good or evil. But he thinks he knows very well, as he says here in T13,
that imprisonment would be an evil. Does he think he has definitional knowledge of
evil? We have no early dialogue in which he tries to get an interlocutor to say what evil is.
But one can be quite confident that he would not think he could say, in the rigorous way
he demands, what it is for something to count as being evil. On my aporetic reading of the
Principle of Definitional Knowledge, he might well have asked, “How can I know that
imprisonment would be an evil, if I cannot say what it is for something to be an evil?”
But he would not have been forced, on pain of inconsistency, to take back what he had
just said about knowing that imprisonment would be an evil on the ground that he did
not have Definitional Knowledge of evil.

14. The Meno Response to the

Problem of the Elenchus

I stated earlier that the Meno is a transitional dialogue and so should not be assumed to
be a reliable guide to the views of the figure of Socrates we have focused our attention on
here. One innovation in this dialogue is an analysis of knowledge as true belief that has
been “tied down” by giving “an account of the reason why” (aitias logismô) (98a). This
analysis of knowledge seems to be at least an ancestor of the familiar modern suggestion
that knowledge is justified true belief.36 It is clearly an epistemological theory. It could be
argued to be implicit in the early dialogues we have focused on, but it is not explicitly
stated in any of them. They offer no explicit analysis of knowledge.
On one plausible interpretation of the Meno, it also offers, among other things,
Plato’s own reflections on the Socratic elenchus and on what assumptions are needed to
justify thinking that it can produce positive conclusions, and perhaps even knowledge.
If one reads the Meno this way, then the fact that Socrates begins the dialogue by insist-
ing in T5, not just that he doesn’t know what virtue is, which would be a standard profes-
sion of Socratic Ignorance, but that he doesn’t know at all (to parapan) what virtue is, is
significant. Meno then sets up the Paradox of Inquiry in T6 so that it is not knowing
at all what virtue is that would make inquiry impossible. As commentators have been
quick to point out, we are usually in the position of knowing something, but not enough,
about the subject of our inquiry. But then the Paradox of Inquiry does not apply to us, at
least if it is stated in the strong form in which Meno first presents it. The application to
the Socratic elenchus, in particular, is this: We can’t reasonably expect the elenchus
to carry us from having no knowledge at all about what virtue is to having definitional
knowledge of what it is.

36 For a full and careful discussion of this passage and its relation to recent analyses of knowledge,
see Fine, “Knowledge and True Belief in the Meno.” ed.: See Judson, Chapter 7 of this volume.
 The Epistemology and Metaphysics of Socrates   397

According to the Doctrine of Recollection as presented in T7, however, each of us does

have some knowledge of what virtue is, though our knowledge may be purely latent. A
reasonable task for the elenchus is then to help us make our latent knowledge manifest.
Socrates’ interrogation of the slave boy is meant to show us that this can happen through
asking questions rather than by making assertions. Socrates gives this explanation of
how it can happen:

T14. So the man who does not know has within himself true opinions about the
things that he does not know? [Meno agrees.] These opinions have now just been
stirred up like a dream, but if he were repeatedly asked these same questions in various
ways, you know that in the end his knowledge about these things [epistêsetai peri
toutôn] would be as accurate as anyone’s. (85cd)

If we apply this suggestion to the Socratic elenchus, the idea is not that the elenchus
itself establishes positive conclusions through some logically coercive chain of rea-
soning, let alone produces knowledge in interlocutors by having them internalize
such reasoning. Rather, the idea is that repeated elenctic questioning about F-ness,
asking “these same questions in various ways,” may eventually lead an interlocutor
not simply to “recollect” true beliefs about F-ness but, finally, to gain knowledge of what
F-ness is.
On this interpretation of the Meno, then, we are meant to conclude that the Problem
of the Elenchus, both in its weaker and in its stronger forms, can be resolved positively.
However, it will not be resolved by seeing how a bit of stand-alone elenctic reasoning
establishes what F-ness is. Rather, it will be resolved by seeing that repeated elenctic
questioning can eventually lead an interlocutor to recollect true beliefs about F-ness and
finally gain manifest knowledge of what F-ness is.

15. Metaphysics

Plato’s grandest metaphysical theory is his theory of Forms. Does Socrates also have a
theory of Forms? Gail Fine, in her monumental study of Aristotle’s On Ideas, maintains
that he does. She writes:

As I interpret Socrates—an interpretation that basically agrees with, but goes beyond
Aristotle’s—he introduces forms for epistemological and metaphysical, but not for
semantic reasons. Further, Socratic forms are universals in the sense that they are
explanatory properties. The fact that they are self-predicative paradigms does not
jeopardize their status as explanatory properties; on the contrary, they are self-
predicative paradigms because they are explanatory properties.37

37 Fine, On Ideas, 53–54.

398   Gareth B. Matthews

When Fine refers to Aristotle’s interpretation, she seems to have in mind especially
remarks Aristotle makes in his Metaphysics M3, which she translates this way:

T15. Now Socrates was concerned with the moral virtues, and he was the first to seek
universal definitions in connection with them. . . . It was reasonable for Socrates to try
to find what a thing is, because he was seeking to argue deductively, and the starting-
point of deductions is what a thing is. . . . For there are two things one might fairly
ascribe to Socrates—inductive arguments and universal definitions, both of which
are concerned with the starting-point of knowledge. But Socrates did not make uni-
versals or definitions (horismous) separate (chōrista), but they (the Platonists) sepa-
rated them, and they called these sorts of beings ‘‘ideas.’’ (1178b12–32)38

Fine points out that in the Laches, Socrates maintains that “speed is some one thing, the
same in running, playing the lyre, speaking, learning, and so on.”39 Similarly, as Socrates
says in the Euthyphro, piety is one thing in all its instances. As pointed out previously,
Socrates seeks definitional knowledge of piety—what is called a “real definition” of piety,
rather than just a nominal definition.40
According to Fine, “Socrates offers an epistemological argument for the existence of
forms: the possibility of knowledge requires explanation, and this, in turn, requires the
existence of forms—real properties and kinds.”41 What Fine seems to have in mind here
is this passage from the Euthyphro, which we have already discussed:

T3. Tell me then what this form itself is, so that I may look upon it, and using it as a
model [or template or pattern, paradeigma], say that any action of yours or another’s
that is of that kind is pious, and if it is not that it is not. (6e3–6)

In this way, the metaphysics of Socrates seems to be driven by his epistemology.

I have said that, before the Meno, Socrates does not offer an analysis of knowledge.
Although he has an epistemology, we might be reluctant to credit him with an epistemo-
logical theory. Is the situation similar with regard to his metaphysics? In particular, does
Socrates really have a theory of Forms?
According to Aristotle in T12, Socrates, unlike Plato, did not consider his universal
Forms to be separate. Reginald Allen has challenged Aristotle’s assertion. With special
reference to what Socrates has to say about the form of piety in the Euthyphro, Allen
mounts this argument:

Ontologically, the priority of Forms is implied by the fact that they are essences and
causes by which things are what they are; their existence is a condition for the exist-
ence of their instances. That priority implies existential independence. If Euthyphro’s
action in prosecuting his father is holy [that is, pious], its existence as holy depends
upon the existence of the Form of holiness, by which it is holy; it would be merely

38 Ibid., 48. 39 Ibid.

40 Cf. Penner, “The Unity of Virtue,” and Irwin, Plato’s Ethics, 25–27. 41 Fine, On Ideas, 50.
 The Epistemology and Metaphysics of Socrates   399

queer to think that the Form of holiness depends for its existence on Euthyphro’s
action in prosecuting his father being holy.42

Allen’s reasoning may be persuasive to many of us who have read the middle dialogues
of Plato. But it does seem to go beyond anything made entirely explicit in the Euthyphro
or in any of the other early dialogues. Perhaps we could say there is a theory of Forms
“immanent” in the philosophy of Socrates.

16. Soul-Body Dualism

In the Apology Socrates characterizes his mission this way:

T16. For I go around doing nothing but persuading both young and old among you
not to care for your body or your wealth in preference to or as strongly as for the
best possible state of your soul. (30ab).

But did Socrates think of the soul as an entity distinct and separable from the body?
According to a speech he gives near the end of the Gorgias, he did. He says to Callicles:

T17. Death, I think, is actually nothing but the separation of two things from each
other, the soul and the body. (524b)

Socrates goes on in the Gorgias to describe the separated soul this way:

T18. All that’s in the soul is evident after it has been stripped naked of the body, both
things that are natural to it and things that have happened to it, things that the per-
son came to have in his soul as a result of his pursuit of each objective. (524d)

In the Apology, however, Socrates insists,

T19. To fear death . . . is no other than to think oneself wise when one is not, to think
one knows what one does not know. (29a)

And in his final speech at his trial he offers these two alternatives:

T20. Either the dead are nothing and have no perception of anything, or it is, as we
are told, a change and a relocating for the soul from here to another place. If it is a
complete lack of perception, like a dreamless sleep, then death could be a great
advantage. . . . If death is like this I say it is an advantage. If, on the other hand, death

42 Allen, Plato’s “Euthyphro,” 136.

400   Gareth B. Matthews

is a change from here to another place, and what we are told is true and all who have
died are there, what greater blessing could there be, gentlemen of the jury? (40ce)

Thus, in the Apology, Socrates remains explicitly agnostic about whether the soul
survives death. Still, whether Socrates believed in an afterlife, as he is given to say in
the Gorgias, or remained agnostic on this issue, he does seem to have conceived the
soul as a thing distinct from the body—something that might conceivably survive
bodily death.
On the other hand, Socrates never gives any arguments for soul-body dualism. In the
middle dialogue Phaedo, Socrates offers quite explicit arguments for not only the forms
but also the soul and its immortality. But the Socrates of the Phaedo is not our concern
here.43 Vlastos summarizes the situation this way:

The queries, ‘‘Is the soul material or immaterial, mortal or immortal? Will it be
annihilated when the body rots?’’ are never on his elenctic agenda. The first question
he never addresses at all. He does allude to the second at the close of the Apology but
only to suggest that it is rationally undecidible: both options—total annihilation
or survival in Hades—are left open. In the Crito he reveals his faith in the soul’s
survival. In the Gorgias he declares it. Nowhere does he try to prove it in the earlier
dialogues.44

So what can we say about the metaphysics of Socrates in Plato’s early dialogues? Socrates
seems to have believed in Forms, which Aristotle called “universals” and which we
might today call “properties.” He thought of these Forms as distinct from, if not separate
from, the concrete particulars that have them. He was at least inclined to think that we
cannot know that something has one of these Forms unless we can specify the nature of
that Form. He believed that one’s soul is something distinct from one’s body. It is some-
thing that may survive one’s body.

BIBLIOGRAPHY
Allen, R. E. Plato’s “Euthyphro” and the Earlier Theory of Forms (London, 1970).
Benson, Hugh H. “Misunderstanding the ‘What Is F-ness?’ Question,” Archiv f ür Geschichte
der Philosophie 72 (1990), 125–42.
Benson, Hugh H. Socratic Wisdom: The Model of Knowledge in Plato’s Early Dialogues
(New York, 2000).
Benson, Hugh H. “Problems with Socratic Method,” in Gary Alan Scott, ed., Does Socrates
Have a Method? (University Park, Penn., 2002).
Benson, Hugh H. ed., Essays on the Philosophy of Socrates (New York, 1992).
Beversluis, John. “Does Socrates Commit the Socratic Fallacy?,” American Philosophical
Quarterly 24 (1987), 211–23; reprinted in Benson, Essays on the Philosophy of Socrates,
107–22.

43 Ed. On the Phaedo, see Castagnoli, Chapter 8 of this volume.

44 Vlastos, Socrates, Ironist and Moral Philosopher, 55.
 The Epistemology and Metaphysics of Socrates   401

Brickhouse, Thomas C., and Nicholas D. Smith. Plato’s Socrates (New York, 1994).
Brickhouse, Thomas C., and Nicholas D. Smith. “Vlastos on the Elenchus,” Oxford Studies in
Ancient Philosophy 2 (1984), 185–95.
Cooper, John M., ed. Plato: Complete Works (Indianapolis, Ind., 1997).
Fine, Gail. “Inquiry in the Meno,” in Richard Kraut, ed., The Cambridge Companion to Plato
(Cambridge, 1992), 200–26.
Fine, Gail. On Ideas: Aristotle’s Criticism of Plato’s Theory of Forms (Oxford, 1993).
Fine, Gail. “Knowledge and True Belief in the Meno,” Oxford Studies in Ancient Philosophy 27
(2004), 41–81.
Fine, Gail. Plato on Knowledge and Forms (Oxford, 2003).
Geach, P. T. “Plato’s Euthyphro: An Analysis and Commentary,” Monist 50 (1966), 369–82.
Gulley, Norman. The Philosophy of Socrates (London, 1968).
Irwin, Terence. Plato’s Ethics (Oxford, 1995).
Irwin, Terence. Plato’s Moral Theory (Oxford, 1977).
Irwin, Terence. “Socratic Puzzles,” Oxford Studies in Ancient Philosophy 10 (1990), 24–66.
Kahn, Charles H. “Did Plato Write Socratic Dialogues?” in H. H. Benson, ed., Essays on the
Philosophy of Socrates (New York, 1992), 35–52.
Kraut, Richard. “Comments on Gregory Vlastos, ‘The Socratic Elenchus,’ ” Oxford Studies in
Ancient Philosophy 1 (1983), 59–70.
Kripke, Saul. Naming and Necessity (Cambridge, Mass.: Harvard University Press, 1980).
Lesher, James H. “Socrates’ Disavowal of Knowledge,” Journal of the History of Philosophy 25
(1987), 275–88.
Matthews, Gareth B. “Senses and Kinds,” Journal of Philosophy 69 (1972), 44–57.
Matthews, Gareth B. Socratic Perplexity and the Nature of Philosophy (Oxford, 1999).
Matthews, Gareth B., and Thomas A. Blackson. “Causes in the Phaedo,” Synthese 79 (1989),
581–91.
McPherran, Mark L. The Religion of Socrates (University Park, Penn., 1996).
Nehamas, Alexander. “Confusing Universals and Particulars in Plato’s Early Dialogues,”
Review of Metaphysics 29 (1975), 287–306.
Nehamas, Alexander. “Socratic Intellectualism,” Boston Area Colloquium in Ancient Philosophy
2 (1986), 275–316.
Penner, Terry. “The Unity of Virtue,” Philosophical Review 82 (1973), 35–68; reprinted in
Benson, Essays on the Philosophy of Socrates, 162–84.
Polansky, Ronald M. “Professor Vlastos’s Analysis of Socratic Elenchus,” Oxford Studies in
Ancient Philosophy 3 (1985), 247–59.
Prior, William J. “Socrates Metaphysician,” Oxford Studies in Ancient Philosophy 27 (2004),
1–14.
Putnam, Hilary. “The Meaning of ‘Meaning,’ ” in Mind, Language and Reality, Philosophical
Papers, vol. 2 (Cambridge, 1975, 215–71).
Reeve, C. D. C. Socrates in the Apology (Indianapolis, Ind., 1989).
Scott, Dominic. Plato’s Meno (Cambridge, 2006).
Taylor, C. C. W. Socrates, A Very Short Introduction (Oxford, 1998).
Vlastos, Gregory [1965]. “Anamnesis in the Meno,” Dialogue 4 (1965), 143–67; reprinted in
Gregory Vlastos, Studies in Greek Philosophy, vol. 2 (Princeton, N.J., 1995), 147–55.
Vlastos, Gregory. “Is the ‘Socratic Fallacy’ Socratic?,” Ancient Philosophy 10 (1990), 1–16.
Vlastos, Gregory. “The Socratic Elenchus,” Oxford Studies in Ancient Philosophy 1 (1983),
27–58.
402   Gareth B. Matthews

Vlastos, Gregory. Socrates, Ironist and Moral Philosopher (Ithaca, N.Y., 1991).
Vlastos, Gregory. Socratic Studies, ed., Myles Burnyeat (Cambridge, 1994).
Vlastos, Gregory. Studies in Greek Philosophy, 2 vols. (Princeton, N.J., 1993, 1995).
Woodruff, Paul. “Plato’s Early Epistemology,” in S. Everson, ed., Epistemology (Cambridge,
1990), 60–84; reprinted in Benson, Essays on the Philosophy of Socrates, 86–106.
 chapter 17

Socr atic Ethics a n d

Mor a l Psychol ogy

Daniel Devereux

1. Socrates and Plato

Our knowledge of Socrates’ ethics and moral psychology is based chiefly on Plato’s
dialogues. Socrates did not write any philosophical works; but since he is the main
speaker in most of Plato’s dialogues, we have plenty of information to work with. In a
sense, Plato gives us too much information: many scholars agree that, in some dialogues,
the dramatic figure “Socrates” represents the views and style of discussion of the histori-
cal Socrates, while in others “Socrates” speaks for Plato, setting out and arguing for views
that are not always in agreement with those of the historical Socrates. If Plato is our main
source of knowledge about Socrates, how can we tell when Plato’s “Socrates” speaks for
Socrates and when he speaks for Plato? As we shall see, Plato’s student, Aristotle, pro-
vides an important clue.
Among the 30 or so dialogues traditionally attributed to Plato,1 there is a group of 11
or 12 that share certain features setting them apart from the rest. In these dialogues, which
are considerably shorter than the others, Socrates always has the role of questioner. The
questions he discusses are mostly about specific virtues and how they are related to each
other: for example, piety is discussed in the Euthyphro, courage in the Laches, temper-
ance in the Charmides, and justice and temperance in the Gorgias. Socrates and his
interlocutors never reach satisfactory answers to the questions discussed, and because
the dialogues end in puzzlement, or (in Greek) “aporia,” they are often called “aporetic.”
Nevertheless, there are clear indications that Socrates favors an “intellectualist” view of
the virtues, according to which they consist in a kind of knowledge. The other, longer,
dialogues are generally more didactic: Socrates (or, in some cases, a different main

1 A few of these dialogues are not accepted as genuine by most scholars. For a comprehensive
discussion of the nature and composition of Plato’s works, see Irwin, chapter 3, in this volume.
404   Daniel Devereux

speaker) lays out and argues for ambitious theories, and the range of subjects discussed
is greatly expanded, encompassing the full range of topics that define the main areas of
philosophy as we know it today. And in these dialogues, Socrates argues for a more bal-
anced view of the virtues, a view according to which such virtues as courage, temperance,
and justice involve both intellectual and emotional elements.
Aristotle, in several brief passages, draws contrasts between the views and interests
of Socrates and Plato, contrasts that correspond closely to the differences just noted
between the two groups of dialogues. For instance, he says that Socrates “asked ques-
tions and did not answer” because he claimed not to have knowledge, and that he differed
from Plato in focusing his inquiries exclusively on the virtues (Sophistical Refutations
183b7–8; Metaphysics 987a29–b7; cf. Apology 38a1–7). He also reports that Socrates
viewed the virtues as purely intellectual qualities—as different forms of knowledge—
whereas Plato took account of the emotional and appetitive aspects of the psyche in
his treatment of such virtues as courage, temperance, and justice (Nicomachean Ethics
1144b28–30; Eudemian Ethics 1216b3–10; Magna Moralia 1182a15–30). These contrasts
between Socrates and Plato match the differences between the shorter, aporetic dia-
logues, such as the Protagoras, Laches, and Charmides, and the longer didactic works
such as the Republic, Phaedo, and Symposium. Although Aristotle was not yet born at the
time of Socrates’ death, he spent some 20 years in Plato’s school, the Academy, and he
undoubtedly came in contact with many people besides Plato who had firsthand knowl-
edge of Socrates. If we accept Aristotle’s reports as trustworthy,2 they provide strong evi-
dence that the shorter, aporetic dialogues present views of the historical Socrates, while
the longer dialogues put forward the different conception of the virtues developed by
Plato—in conscious disagreement with the intellectualist view of his mentor.
Many scholars agree that the shorter dialogues were intended to portray the charac-
teristic views and arguments of the historical Socrates; for this reason these dialogues
are also called “Socratic.” In fact, a number of scholars believe that these dialogues con-
tain a Socratic “theory” of the virtues: a unified, systematic account of their nature and
value. According to this view, one of Plato’s intentions in writing these dialogues was to
set out this systematic account of the virtues, and to defend it with arguments used by
Socrates. The claim is not that Plato gives us verbatim reports of Socratic discussions:
the discussions may well be fictional, but the views and arguments they contain derive
from Socrates. As we shall see, there are problems with this view: later in this essay
I suggest that while these dialogues are “Socratic” in the sense that they focus on the views
and style of discussion of the historical Socrates, they were not intended to give us a uni-
fied Socratic theory of the virtues—for the good reason that Socrates in all likelihood
did not have a unified theory of the virtues.

2 Doubts have been raised about the trustworthiness of Aristotle’s reports: see, e.g., C. Kahn, Plato
and the Socratic Dialogue (Cambridge, 1996), 79–87; L.-A. Dorion, “The Rise and Fall of the Socratic
Problem,” in D. R. Morrison (ed.), The Cambridge Companion to Socrates, (Cambridge, 2010), 1–23. For
further discussion of this issue, see Irwin, chapter 3, in this volume.
 socratic Ethics and Moral Psychology   405

2. Socrates and the Sophists

A prominent theme in the Socratic dialogues is Socrates’ opposition to the so-called

sophists. The sophists were a varied lot with different interests and claims to fame, who
shared certain characteristics that justified their common designation.3 They all claimed
to be experts (sophistai) in political affairs, public speaking, and debating, and they
offered to teach these skills, for a fee, to young men who were anxious to achieve success
in the political arena. They advertised themselves as “teachers of virtue,” and they regarded
virtue as an art or skill—a technê in Greek (Gorgias 519c3–d1; Protagoras 319a3–b1,
320b4–5; Hippias Major 282b1–e8, 283c3–5). The claim that virtue is an art or skill clearly
served their purpose, for if it’s a skill it ought to be teachable in the way that arts and
skills generally are.
Many Athenians were suspicious of the sophists. They regarded them as subverters of
traditional morals, and they were skeptical of the claim that virtue could be acquired
through the sort of teaching offered by the sophists (Protagoras 357e4–8; cf. Meno
90e10–91c5 and Gorgias 519c2–520e11). In view of Socrates’ opposition to the sophists, it
is ironic that he was considered to be one of the clan by many of his contemporaries. He
makes this clear in his Apology when he says that Athenian public opinion is biased
against him because of slanderous reports that he engages in the same activities as the
sophists (18a7–20c3). Actually, it is not so surprising that Socrates was lumped together
with the sophists by the general public. His interests were more or less defined by the
sophists’ claims and activities. For example, he questioned whether virtue could be taught,
and inquired into the nature of virtue. His discussions typically involved refuting his
interlocutor, and refutation was part of the sophists’ stock in trade. Although Socrates
denied having any special expertise or “wisdom,” he reports that his success in refuting
his interlocutors gave others the impression that he had superior knowledge of the mat-
ters under discussion (Apology 22e7–23a5). Moreover, some of his claims (his “Socratic
paradoxes”) were strikingly at variance with common opinion, and this may have led
many to think that he, like the sophists, was undermining traditional moral beliefs.
Socrates also seems to have accepted the sophists’ view that virtue is an art or skill. It is
true that Socrates never claimed to teach virtue, as the sophists did, and he never asked
for payment from those who associated with him. Nevertheless, to a casual observer,
he might well have seemed a fellow traveler, if not a bona fide member of the group
(cf. Sophist 226b1–231b8).
Socrates’ opposition to the sophists was focused mainly on their claim to teach virtue.
Although he seems to have accepted their view that virtue is an art or skill (a technê), he
argues in both the Protagoras and Meno that virtue is not teachable. He realizes that his
position is puzzling: if virtue is an art or skill, then it ought to be teachable; and if it is not

3 For background on the sophists, see R. Barney, “The Sophistic Movement,” in M. L. Gill and
P. Pellegrin (eds.), A Companion to Ancient Philosophy (Malden, Mass., 2006), 77–97; see also Schofield,
chapter 2, and Johansen, chapter 12 in this volume.
406   Daniel Devereux

teachable, this would seem to cast doubt on the claim that it is an art or skill (Meno
94b4–8; Protagoras 319c7–d7, 361a3–b7). Socrates finds himself in the following quandary
at the end of the Protagoras:4

(1) If X is an art or skill, then it is teachable.

(2) Virtue is a kind of knowledge—an art or skill.
(3) Virtue is not teachable.

It is possible that Socrates was genuinely perplexed and thought he had good reasons
for maintaining each of these three propositions. It is also possible that he had doubts
about one or more of them. Let us take a brief look at a couple of his arguments to see if
there is a way out of the quandary.
One of Socrates’ main arguments against the teachability of virtue (in both the
Protagoras and Meno) is that those who possess virtue are unable to pass it on to others,
in the way that skilled craftsmen can pass on their expertise to apprentices (Meno
90b3–e8). He cites the examples of Pericles and Themistocles, famous Athenian statesmen,
who had every reason to pass on their political skill and excellence to their sons, but
were apparently unable to do so (Protagoras 319d7–320b5; Meno 93a5–94e2). If these
great leaders were unable to teach their virtue to others, surely it’s because virtue is not
teachable.5
The main target of Socrates’ argument is the sophists’ claim to teach virtue. The “virtue”
he has in view is the ability to achieve success in politics—that is, to become a powerful
and influential leader like Pericles or Themistocles—and this ability is just what the soph-
ists promised to teach their students (Protagoras 318e4–319a7; Meno 91a1–b5). However,
there are good reasons to doubt that Socrates considered this ability to constitute genu-
ine virtue. In another dialogue, the Gorgias, he contends that Pericles and Themistocles
did not exercise good political leadership—they maintained their power by pandering
to the people and did nothing to improve them (515c4–517a6). He also suggests in several
places that someone who possessed genuine virtue would be able to explain its nature,
and he claims that he has never met anyone who was able to do this (Meno 71b1–c4; cf.
Charmides 158d8–159a4 and Apology 29e3–30a3). This is a pretty clear indication that he
did not regard the political leaders admired by his fellow citizens as examples of genuine
virtue. Socrates apparently regards genuine virtue as an extremely rare commodity—
something he has never encountered in his many years of searching.

4 See, e.g., R. Kraut, Socrates and the State (Princeton, N.J., 1984), 285–86.
5 In the Meno, Socrates argues more generally that, since there are no teachers of virtue and no
learners, it would seem that virtue is not teachable (89d3–96c10). He first points out that those who
are regarded as most virtuous, leaders such as Pericles and Themistocles, were unable to pass on their
excellence to others; he then considers those who claim to be teachers of virtue, the sophists, and
notes that their claim is highly disputed, which is not the case in other recognized arts and skills.
(Of course, if to be a teacher of virtue one must first possess it, and if virtue is an extremely rare
commodity (see later on in this chapter), then virtue might be teachable even if there are not, at
present, any teachers of it.)
 socratic Ethics and Moral Psychology   407

Socrates’ argument for the unteachability of virtue is ad hominem in the sense that it
relies on a conception of virtue accepted by his opponents, the sophists, but not by him.
He says to them in effect: “Let us suppose for the sake of argument that virtue is the sort
of thing you say it is, that is, the abilities that enabled leaders like Pericles to gain power
and influence in their communities and to secure the admiration of their fellow citi-
zens. I contend that this ‘virtue’ is not teachable, for those who possess it are unable to
pass it on to others.” If the argument is ad hominem in this way, we may conclude that
Socrates did not see it as a real threat to the view that genuine virtue is an art or skill,
and therefore teachable. We have found an escape route from the quandary at the end
of the Protagoras.

(1) If X is an art or skill, then it is teachable.

(2) (Genuine) virtue is an art or skill.
(3) (Sophistical) virtue is not teachable.

If the virtue that the sophists claim to teach is not teachable, it cannot be an art or skill
(Protagoras 319b3–d7, esp. c1–8; cf. Meno 94b4–6). But Socrates believes there is a genu-
ine art of politics, an art quite different from what the sophists profess to teach.6
We have seen that Socrates’ arguments against the teachability of virtue are not
intended to undermine the view that (genuine) virtue is an art or skill. There are some
indications in other Socratic dialogues that Socrates may have had doubts about this
view, however. For instance, in the Hippias Minor, he argues that an art is a “capacity for
opposites”: that is, the knowledge involved in an art or skill can be used for evil as well as
good purposes—medical skill can be used to spread disease as well as to cure it. Perhaps
someone who uses her medical skill to harm others for personal gain should not be called
a doctor, but she can be just as skillful as the doctor who cures; it’s precisely the skill that
enables her to harm others in the way that she does. Now if a virtue such as justice were a
skill and a capacity for opposites, it could be used for evil purposes as well as good: a just
person could use her justice to harm an innocent person (365d6–369a2, 375d7–376b6).
But this seems perverse: someone who possessed the virtue of justice could not manifest
that justice by harming an innocent person for gain. If a skill is a capacity for opposites, it
seems that justice cannot be a skill.
Another way of expressing this contrast between virtue and skill is that certain aims
or motives are essential to virtue but not to skill. In order to determine whether someone
has a skill, we can simply observe her at work: if she performs in a skillful manner and
produces the desired results, it is safe to conclude that she possesses the skill. But if some-
one consistently acts justly when called upon to do so, it doesn’t follow that he possesses
the virtue of justice: he might act justly out of fear of being punished, or for appearance,

6 For example, Gorgias 464b2–c3, 521d6–8; Meno 99e4–100b. Socrates’ claim about the sophists
is parallel to his claim about the rhetoricians in the Gorgias: while there may be a genuine art of
rhetoric (504d5–e4, 517a4–6), this art is quite different from the pseudo-art that Gorgias and other
contemporary rhetoricians profess to teach.
408   Daniel Devereux

or in order to further his career; acting justly for these reasons falls short of what is
expected of a just person.7
Socrates also argues in the Hippias Minor that in the case of an art or skill, one who errs
voluntarily is better or more skillful than one who does so involuntarily (372c8–376b6).
For instance, a mathematician might deliberately make a mistake in a proof in order to
see if his students are paying close attention; such a deliberate mistake would not count
against his knowledge or skill. But if someone made the same mistake without realizing
it (“involuntarily”), it would indicate a gap in his knowledge. If justice were an art or skill,
then someone who erred voluntarily—that is, performed an unjust act voluntarily—
would be better or more just than one who did something unjust involuntarily. But here,
in contrast with arts and skills, voluntarily performing an unjust act would seem to
count against being a just person.8
Socrates admits to being perplexed by these arguments: he is drawn to the view that
virtue is an art or skill; at the same time he sees that it leads to a consequence that he
regards as clearly unacceptable (376b8–c6; cf. 372d3–e1). Some have suggested that
Socrates points to a way out of the dilemma when he says, in the concluding lines of his
argument, “Then the one who goes wrong voluntarily and does base and unjust things,
Hippias—if there is such a person—would be none other than the good man” (376b4–6).9
The suggestion is that since Socrates argues elsewhere (as we shall see) that no one
voluntarily acts unjustly, he would regard the unacceptable consequence of the argument
as an impossibility. And yet Socrates seems genuinely perplexed at the end of his argu-
ment, and, on reflection, it seems that he has good reason to be. For if voluntary unjust
action is impossible, this would be a consequence of our psychological makeup, not of
the nature of justice itself: for if justice were an art or skill, there would be nothing in its
nature that would rule out someone displaying justice by voluntarily committing some
terrible injustice. But this seems clearly unacceptable: deliberately and willingly com-
mitting injustice (for example, for personal gain) is simply incompatible with our under-
standing of what it is to be a just person.
To sum up, Socrates is inclined to agree with the sophists’ view that virtue should be
understood as an art or skill; indeed, he relies on it as a working hypothesis in many of
his arguments. But he also has doubts about the sophists’ view because of certain strik-
ing disanalogies between the virtues and arts or skills. If virtue is not an art or skill, it is
no longer clear that it is teachable, at least in the way that arts and skills are: Socrates’
doubts about virtue being an art or skill would carry over to its teachability. However, he
has no doubts about whether the “virtue” that the sophists claim to teach is teachable; he

7 Aristotle makes a similar distinction between virtue and skill; see Nicomachean Ethics II 4,
1105a26–b5.
8 Aristotle cites this contrast as a reason for denying that virtue is an art or skill; see Nicomachean
Ethics VI 5, 1140b21–25. Another contrast noted by Aristotle is that to have a virtue is to have achieved a
standard of excellence in whatever the virtue is concerned with, but this is not the case with an art or
skill—one can possess a skill without having achieved a standard of excellence; see Nicomachean Ethics
VI 5, 1140b21–25; cf. VI 7, 1141a9–12. This contrast is perhaps alluded to in the Protagoras at 327c1–d4.
9 See, e.g., T. H. Irwin, Plato’s Ethics (New York, 1995), 69.
 socratic Ethics and Moral Psychology   409

believes that this pseudo-virtue—the ability to achieve status and power in the political
arena—is not teachable since those who have it are unable to pass it on to others.

3. Socrates’ Denial of Akrasia

If virtue is an art or skill that involves knowing what is truly good and evil, the virtuous
person should be able to “size up” a situation and determine what sort of action is called
for (Laches 199d4–e1). But will this knowledge guarantee performance? Isn’t it possible
to know what one ought to do in a given situation, but fail to do it because of a strong
desire to do something else instead? Don’t we sometimes act against our better judgment
because of weakness of will? Socrates believes that a virtuous person will consistently act
in a virtuous manner. If so, he must hold that knowledge of the good guarantees perfor-
mance: if you know what is good, you will always do it. Socrates’ conception of virtue thus
seems to rule out weakness of will (akrasia in Greek). But how can he deny what seems
to be a common, everyday occurrence?
Socrates takes up this question in the Protagoras, and tries to show that the common
understanding of akrasia is mistaken.10 The sorts of cases he considers are familiar
ones—for example, we are tempted to do something that promises immediate gratifica-
tion, but we realize that this pleasure is not worth the undesirable consequences that
would inevitably follow. Socrates recognizes that most people would say that we some-
times “give in” to such temptations even though we know we shouldn’t (352b3–c2); but
he suggests that this is a misdescription of what actually happens; in these situations we
only appear to be acting against our knowledge: we are actually acting in ignorance.
Socrates’ celebrated argument against akrasia has been picked apart and analyzed in
great detail by scholars over the past half century. I will not be able to go through the
argument in detail;11 instead, I focus on Socrates’ overall strategy, and attempt to clarify
how he understands apparent cases of akrasia. His argument has three parts: he first sets
out three claims that make up the “Common View,” he then points out that this Common
View leads to an absurd consequence, and finally he argues for an alternative descrip-
tion of apparent cases of akrasia that avoids the problem with the Common View.

10 “Acrasia,” like our expression “weakness of will,” covers more than just acting contrary to
knowledge: e.g., we might believe, mistakenly, that a certain kind of good-tasting drink is bad for one’s
health, and occasionally give in to a temptation to have some; this would be an instance of acrasia, as
well as of weakness of will.
11 For some detailed discussions, see G. Santas, Socrates, Philosophy in Plato’s Early Dialogues
(London, 1979), 195–217; C. C. W. Taylor, Plato’s Protagoras [Protagoras] (Oxford, 1991), 170–200;
T. Penner, “Knowledge vs. True Belief in the Socratic Psychology of Action,” Apeiron 29 (1996),
199–230, and “Socrates on the Strength of Knowledge: Protagoras 351b–357e” [“Strength”], Archiv für
Geschichte der Philosophie 79 (1997), 117–49; R. Singpurwalla, “Reasoning with the Irrational: Moral
Psychology in the Protagoras” [“Reasoning”], Ancient Philosophy 26 (2006), 243–58; T. Brickhouse
and N. Smith, Socratic Moral Psychology [“Psychology”] (Cambridge, 2010), 70–88; D. T. Devereux,
“Comments on Socratic Moral Psychology,” Analytic Philosophy 53 (2012), 216–23.
410   Daniel Devereux

As we will see, he does not give a proof of his claim that akrasia does not occur; rather, he
tries to show that the usual way of understanding apparent cases of akrasia is problem-
atic, and that there is another, less problematic, way that is compatible with his view that
knowledge of the good guarantees performance.
At the beginning of the argument, Socrates contrasts his own view of the “power” of
knowledge with the Common View.

Come now, Protagoras, and reveal how you think about this: How do you stand in
regard to knowledge? Do you take the same view of it as most people? Most people
do not regard it as something strong and controlling and ruling—they don’t think of
it in this way at all. They think that often knowledge is present in a man, but that
something else rather than this knowledge is in control: sometimes anger, some-
times lust, sometimes pleasure or pain, often fear—they think of knowledge as no
more than a slave, dragged around by everything else. Is this also your view of
knowledge, or do you take it to be something noble—the sort of thing that rules a
man; and if someone knows what’s good and what’s bad, he would not be overpow-
ered by anything so as to act contrary to what knowledge commands—wisdom is
powerful enough to come to his aid. (352a8–c7)

According to the Common View, people sometimes find themselves in the grip of
psychological conflict: their knowledge tells them to do one thing, but they are led by
passion or desire to do something different. Socrates focuses on the case of doing some-
thing you know to be bad because you are “overcome by pleasure”; he argues that when
most people call an action or experience “good” or “bad,” they consider only the pleas-
ure or pain that it brings: an action or experience is “good” if it brings pleasure or takes
away pain, and “bad” if it brings pain or takes away pleasure (353c4–354e2).12 Thus the
Common View has three elements:

(1) People sometimes do things they know to be bad.

(2) They do these bad things because they are “overcome by pleasure.”
(3) What is good = what is pleasant; what is bad = what is painful.

Socrates believes that there is something incoherent or absurd in the combination of

these three claims. For according to (3), we should be able to substitute “good” for “pleas-
ure,” and vice versa; but if we substitute “good” for “pleasure” in (1) and (2) we get:

(4) People sometimes do things they know to be bad because they are overcome by
good. (355a5–d3)

12 Does Socrates agree with the hedonism that he attributes to the Common View? For an
affirmative answer, see Taylor, Protagoras, 162–70, and Irwin, Plato’s Ethics, 82; for dissent, see D. J. Zeyl,
“Socrates and Hedonism,” Phronesis 25 (1980), 250–69, and J. Wilburn, “Plato’s Protagoras the
Hedonist,” Classical Philology 111 (2016), 224–44.
 socratic Ethics and Moral Psychology   411

But how could good be the cause of bad? Socrates thinks this is “laughable” or absurd.
The Common View’s explanation of the alleged cases of acting contrary to knowledge,
combined with its equation of the good with pleasure leads to an absurd consequence.
This does not mean that defenders of the Common View must give up their belief that
people sometimes act contrary to their knowledge; they can look for another possible
explanation, or they can give up the equation of good with pleasure. But Socrates has
shown that the Common View as it stands is problematic, and he thinks he has another
way of describing these cases that is not problematic.
At the very beginning of his argument, Socrates explains what he hopes to show.
According to the Common View, there is a certain experience that people describe as
acting against our knowledge because of “being overcome by pleasure”—an apparent
case of akrasia. Let us call this experience an “apparently akratic experience.” Where the
Common View goes wrong is not in claiming that there is such an experience; Socrates
accepts the existence of the experience, but not the Common View’s description of it.

Come, then, [Protagoras] and let us together attempt to persuade these people and
explain to them what this experience is which they call “being overcome by pleasure,”
which is the cause of their not doing what they know to be best. For perhaps if we
told them that they are not speaking correctly and that what they say is false, they
would ask us: “But Protagoras and Socrates, if this experience is not being overcome
by pleasure, what can it possibly be—what do you say it is? Tell us.” (352e5–353a6)

After he has pointed out the absurd consequence of the Common View, Socrates proceeds
to build a case for a revised description of the experience; at the end of the argument he
formulates his revised description as follows.

At that point you asked us: “But Protagoras and Socrates, if this experience is not
being overcome by pleasure, what can it possibly be—what do you say it is? Tell us.”
If we had then said straightaway that it is ignorance, you might have laughed in our
faces; but if you laugh at us now you will be laughing at yourselves. For you agreed
with us that those who err in their choice of pleasures and pains do so through lack
of knowledge. . . . And you must know that the erring act done without knowledge is
done through ignorance. So this is what “being overcome by pleasure” is—ignorance
in the highest degree. (357c6–e2)

What exactly does Socrates mean by saying that “ignorance” is a more accurate descrip-
tion of the experience? He is clearly denying that in an apparently akratic experience the
agent knows that the action chosen is a bad one: he says that the person who chooses
wrongly does so under the influence of a false opinion or belief (358c1–5). He is suggest-
ing, then, that the agent believes that the act chosen is good rather than bad: that is, the
agent chooses the present pleasure in the belief that it is the best option available.
Advocates of the Common View might wonder how this is a more accurate description
of the experience that they have in mind. Recall that a crucial aspect of an apparently
412   Daniel Devereux

akratic experience is psychological conflict: the agent’s knowledge dictates doing X, but
he chooses to do Y, instead, because he is overcome by a strong desire for immediate
gratification. According to Socrates’ revised description, the agent’s desire to do Y is in
accord with his (false) belief that Y is the right thing to do in the situation—there is no
longer any conflict. Instead of a more accurate description of an apparently akratic expe-
rience, Socrates seems to have given a description of a different sort of experience. Is he
simply denying the existence of psychological conflict in apparent cases of akrasia?
If we look at Socrates’ argument leading up to his revised description, it becomes clear
that his description does not eliminate psychological conflict. Central to the argument
is an ingenious analogy between our judgments regarding pleasure and pain and our
perceptual judgments about things that we see and hear. Socrates points out that
when we are looking at objects close up or far away we may make mistakes in judging
their size; similarly, when pleasures or pains are near at hand or in the future we may
misjudge them and think that they are greater or smaller than they really are (356c4–357b3).
He suggests that just as there is an art of measuring the size of objects, which saves us
from being taken in by misleading appearances, so there is an art of measuring pleasures
and pains that provides a similar protection.
Socrates’ analogy implies that pleasures and pains present misleading appearances
when they are near at hand or off in the future; when they are in the middle range, we are
less prone to misjudge them. Suppose we are considering the possibility of going to a
movie with friends, and we decide, correctly, that it would be foolish to go since we have
an important exam the next morning. But when the possible pleasure is there for the
taking—say our friends stop by on their way to the movie—we might waver and then
decide that it would be all right to join them; the proximity of the pleasure (and perhaps
the experience of anticipatory pleasure) leads us to overestimate its value in relation to
its cost (or underestimate its cost in relation to its value). The next day, after a disastrous
performance on the exam, we regret our ill-considered decision. Socrates seems to have
just such experiences in view in the following passage:

The power of appearance makes us wander back and forth, frequently changing our
minds about the same things and regretting our actions and choices of things large
and small; the art of measurement, on the other hand, would render the appearance
ineffective, and by showing us the truth would bring peace to the soul abiding in
that truth, and would save our lives. (356d4–e2)

Socrates suggests that the agent in apparent cases of akrasia changes his mind; he makes
the wrong choice under the influence of a false belief (358c3–5), but he later “has regrets”:
in retrospect, he sees that he made the wrong choice. The false belief or “ignorance” is
only a temporary, passing, condition (this is also implied by equating the ignorance with
an experience—something that happens to one; see 357c6–d3). It’s not that the agent has
a standing belief that going for immediate gratification is the right thing to do: in cooler
moments of reflection, before and after the choice, the agent’s judgment is clearheaded;
but at the moment of choice when the pleasure is near at hand (356a5–7; cf. 353c9–d4),
 socratic Ethics and Moral Psychology   413

the “power of appearance” clouds his judgment and leads him to believe that taking the
present pleasure is the right choice.13
The distinction between knowledge and belief—in particular, between knowledge
and true belief—is an important element in Socrates’ account of apparent cases of akrasia.
The person with knowledge can be counted on to make the right judgments about
pleasures and pains and to act accordingly. At the opposite extreme is the vicious,
self-indulgent person whose judgment has been permanently corrupted, and who has
standing false beliefs about the relative values of various pleasures and pains. Such a per-
son will not typically waver back and forth, but will consistently act in accordance with
his standing beliefs. In the middle is the person who wavers—the person who has true
beliefs but not knowledge, and whose judgment can be turned around by the “power of
appearance.” This is the person who has the apparently akratic experience: he seems to
know that X is the right choice, but chooses Y, instead, because of a strong desire for
immediate gratification. According to Socrates’ account, knowledge is not vulnerable to
the power of appearance, but belief is; in apparent cases of akrasia the agent’s true belief
about the relative values of the pleasures and pains (or goods and evils) in a course of
action is overcome by the misleading appearance of the present pleasure—that is, the
agent “changes his mind” at the time of choosing and (mis)judges that Y is the right
choice. Afterwards, he sees his mistake and “regrets” his choice.
This way of understanding Socrates’ account of apparent cases of akrasia preserves
the element of psychological conflict: the agent’s desire for immediate gratification
contends with and “overcomes” his belief that he should resist.14 As we have seen, the
account involves a contrast between the stability of knowledge, which can withstand
the “power of appearance,” and the instability of true belief, which can be temporarily
suppressed and replaced by a false belief that supports the desire for immediate gratifi-
cation. Socrates’ description of apparent cases of akrasia is similar to what is sometimes
called “rationalization”: fabricating a justification for some action that goes against our
better judgment—if, under the influence of passion or desire, we manage to persuade

13 What does Socrates mean by “the power of appearance”? Does he hold that the distorted
appearance is a “purely intellectual” mistake that gives rise to a desire for the immediate gratification?
Or is it the other way around—that the desire aroused by the anticipated pleasure causes the distorted
appearance? Or is the desire equivalent to a belief about the value of the present pleasure based
on the distorted appearance? For the first view, see T. Penner, “Strength,” 117–49; for the second,
D. T. Devereux, “Socrates’ Kantian Conception of Virtue,” Journal of the History of Philosophy, 33 (1995),
381–408, and T. Brickhouse and N. Smith, Socratic Moral Psychology (Cambridge, 2010), 78–88; for the
third, R. Singpurwalla, “Reasoning,” 249–54.
14 The conflict ceases at the moment of choice, if not before; when the agent chooses the acratic
action, his judgment accords with his desire; thus Socrates denies the possibility of what has been
called “synchronic belief acrasia” (see Penner, “Strength”). In the Republic, Plato apparently allows that
conflict can occur even at the moment of choice—that is, he allows for the possibility of synchronic
belief acrasia: see the story of Leontius in Book IV, 439d6–440a4. But it is not clear that Plato regards
the true beliefs instilled in the auxiliaries as subject to being overcome in this way—such beliefs have
staying power because of the way they are inculcated: see 429c7–430b9, and especially the distinction
between two types of “right opinion” (orthê doxa) at 430b2–9.
414   Daniel Devereux

ourselves that it is a good thing to do, we will have (temporarily) acquired a false belief
supporting the action.
Does Socrates provide a convincing account of what happens in apparent cases of
akrasia? In view of the familiar phenomenon of rationalization, we might agree that
sometimes what seems like “acting against knowledge” is more accurately described as
acting under the influence of a temporary false belief that rationalizes our action. But we
might also question whether it is only true opinion that gets pushed aside or suppressed
in rationalization: why can’t the same thing happen to knowledge? Furthermore, some
might argue that our experience testifies to the fact that we sometimes succumb to temp-
tation without any rationalization—we “know full well” that we are making a bad choice
and that we will come to regret it. Socrates might reply that knowledge must have a stronger
hold on us—if we had genuine knowledge of the good, it would exert such a powerful
influence on us that we would never act against it.15 And it is worth noting that philoso-
phers continue to find the phenomenon of akrasia puzzling, and many are sympathetic
toward Socrates’ view. But it also seems clear that he gives no argument for his claim that
knowledge, or “the art of measurement,” cannot be overcome by passion or desire; as
I mentioned earlier, Socrates does not give a proof of his claim that akrasia does not occur.
Socrates’ thesis that knowledge of the good guarantees performance of right action is
closely related to another paradoxical claim, which he argues for in the Protagoras: the
doctrine of the “Unity of the Virtues.” If each of the virtues requires knowledge of good
and evil, and if this knowledge guarantees that one will always perform the virtuous
action in the appropriate situation, then one couldn’t have one virtue without having all
the rest (see Laches 199d4–e1). Some scholars believe that what Socrates means by the
Unity of the Virtues is that, while each virtue has its own distinct essence, they are all
inseparably linked to each other through the knowledge of good and evil. Others argue
that he makes a stronger claim: that the virtues are identical with this knowledge, and
that there is just one essence and one definition for all the virtues.16 I will not pursue this
dispute, but it is worth noting that whether unity is understood as identity or insepara-
bility, the basis for the doctrine is the claim that knowledge of the good guarantees per-
formance of right action.

15 Cf. Aristotle, Nicomachean Ethics VII 2, 1145b23–24: “Socrates thought it would be astonishing
(deinon) if knowledge, being present in a man, could be overpowered by something else”; see also
Magna Moralia II 6, 1200b34–37.
16 Socrates seems to argue for identity in the Protagoras, but in the Laches he appears to claim that
the virtues are distinct parts of a whole. For a defense of the identity interpretation for both dialogues,
see T. Penner, “The Unity of Virtue,” Philosophical Review 82 (1973), 35–68, reprinted in G. Fine (ed.),
Plato 2 (New York, 1999), 78–104; for the inseparability interpretation, see G. Vlastos, “The Unity of
the Virtues in the Protagoras,” Review of Metaphysics 25 (1972), 415–58, reprinted with additional
notes in Vlastos, Platonic Studies, 2nd ed. (Princeton, 1981), 221–69, 418–23; for a “fence-straddling”
interpretation, see M. Ferejohn, “The Unity of Virtue and the Objects of Socratic Inquiry,” Journal of
the History of Philosophy 20 (1982), 1–21; and for the view that the two dialogues present conflicting
views, see D. T. Devereux, “The Unity of the Virtues in Plato’s Protagoras and Laches,” Philosophical
Review 101 (1992), 765–89.
 socratic Ethics and Moral Psychology   415

4. Desire and the Good

As we have seen, Socrates does not deny that there is an experience of psychological
conflict in apparent cases of akrasia. On one side is a desire for immediate gratification;
on the other is the correct belief that the gratification is not worth the undesirable conse-
quences that will inevitably follow. As the conflict plays out, Socrates suggests that the
agent “rationalizes” the choice of the present pleasure by temporarily suppressing
the judgment that it is a bad choice. In giving this account, Socrates seems to accept the
Common View’s assumption that we can desire things we recognize to be bad (bad for
us, or bad in some other way). However, in both the Gorgias and the Meno he claims that
we only desire what is good—no one desires what is bad.17 In the Gorgias Socrates makes
this claim in attempting to show that tyrants do not have great power because they do
not do what they want to do. It will be useful to survey the main steps of his argument.

1. Having great power is doing whatever one wants to do.

2. We only want what is good for us, not what is bad or indifferent.
3. Doing unjust acts is bad for us.
4. Tyrants do many unjust acts, thinking they are best for them.
5. Thus, by (3) and (4), tyrants do what is bad for themselves.
6. Thus, by (2) and (5), tyrants do not do what they want to do.
7. Thus, by (1) and (6), tyrants do not have great power.18

The most questionable parts of the argument are steps (2) and (3): What exactly is
meant by the claim that “we only want what is good for us,” and what is its basis?; and why
does Socrates think doing unjust acts is always bad for us? We return to the latter ques-
tion toward the end of our discussion.
It seems that (2) would not leave room for the psychological conflict we noticed in
Socrates’ discussion of akrasia in the Protagoras: if we recognize that a tempting present
pleasure would be bad for us, then according to (2) it cannot be something that we want or
desire. In fact, what Socrates means by (2) is that, whether or not we recognize it, if some-
thing is bad for us we cannot be said to want it. This is clearly implied by his claim that
when the tyrant, for example, does away with a potential rival, he does what seems best

17 In both contexts, it is clear that Socrates’ claim is that we only desire what is good for us. This
would seem to commit him to Psychological Egoism—that is, to the view that all of our actions are
motivated by self-interest. However, as we shall see, Socrates’ claim is not about what motivates us to
act; thus it is not clear that he is committed to Psychological Egoism. It is also worth noting that many
of his statements about his own motives do not seem to fit comfortably with such a doctrine (e.g.,
Apology 31a9–c3, Euthyphro 3d5–9).
18 Steps (1)–(4) are argued for at 467c5–481b5. The inferences in steps (5), (6), and (7) are not
explicitly drawn by Socrates, perhaps because Callicles breaks in at 481b5. But it’s clear that these
steps follow from the preceding, and that Socrates’ intention from the outset is to argue for (7);
see 466e13–467a10, 509e2–7.
416   Daniel Devereux

to him but not what he wants (since it is bad for him). Socrates’ claim in (2) is not that we
only want what we take to be good, but that we only want what is in fact good for us.
The assertion that all of our desires are directed toward what is actually good clearly
conflicts with Socrates’ view in the Protagoras that we sometimes have desires for things
that are bad for us; moreover, the claim seems clearly false—it seems obvious that we do
have desires for things that are bad for us. Socrates claims, in effect, that if we don’t know
what is good for us, then we also don’t know what we want. Suppose that we believe that
X would be a good thing for us to do, but actually it would be bad; according to Socrates,
we might believe that we want to do X, but in fact we don’t since it would be bad for us.
We sometimes say things that seem to fit with Socrates’ claim: for example, we might say
to someone who is very thirsty, “You don’t want that sugary stuff you’re about to drink.
What you want is something that will satisfy your thirst: this glass of cool water is what
you really want.” Similarly, Socrates might say that the tyrant wants what will make him
happy; being unjust will make him miserable in the long run, whereas being just will
make him happy; therefore he (really) wants to be just—he only thinks that he wants to
act unjustly.
Although we sometimes speak this way, it still seems extremely implausible to say
that the tyrant doesn’t want to do X even though he thinks it’s best and freely chooses to
do it. Sometimes we think that X would be the best thing to do in a situation, but we
don’t really want to do it. But this doesn’t seem to be true of the tyrant when he decides to
get rid of a hated rival: he has no hesitation, no misgivings. Socrates’ claim about desire,
as noted, also seems to clash with the Protagoras’ view that we can desire things that are
bad for us because of the lure of immediate gratification. What is even more puzzling is
that the claim conflicts with some of Socrates’ statements later in the Gorgias. One of the
important themes in his discussion with Callicles is the value of self-discipline and the
need to resist certain desires. He mentions at one point, for example, that doctors often
tell their sick patients they must refrain from satisfying their desires for the things they
are used to eating and drinking (505a6–10). In another place he says that we should only
satisfy those desires that make us better, not those that make us worse (503c6–d3, 505b1–12;
cf. 517b2–c2). In these passages he takes it for granted that not all of our desires are
directed toward what is actually good for us. There appears to be a clear inconsistency
between Socrates’ earlier claim that we only desire what is good for us, and his later
insistence on the need to curb and restrain certain desires.
For a reader of the Greek text, there is an interesting difference between the earlier
and later passages, which may help resolve the apparent inconsistency. Throughout the
earlier discussion in which Socrates claims that we only want what is good for us, he uses
a verb that is usually translated as “wish” or “want”; in his later discussion of desires that
need to be curbed, he uses a different verb that is usually translated as “desire.”19 Although

19 The two verbs are boulesthai (= “to want” or “to wish”) and epithumein (= “to desire”); in his
discussion with Callicles, Socrates more often uses the noun, epithumia (= “desire”) instead of the verb.
(In the later discussion, Socrates does use boulesthai once in referring to appetites, in speaking of the
healthy person’s appetites that do not need to be curbed (505a6–10); these are appetitive desires that are
at the same time desires for what is actually good.)
 socratic Ethics and Moral Psychology   417

these verbs are often used interchangeably (and are so used in parts of the Gorgias that
are not concerned with claims about desire), they do have somewhat different meanings.
In another Socratic dialogue, the Charmides, Socrates differentiates between the two by
pointing out that the object of wanting is the good, while the object of desiring is what is
pleasant—which may or may not be good (167e1–5). The notion that desire (epithumia)
is directed toward things that are pleasant but may or may not be good seems to fit the
examples of “desires” in the discussion with Callicles: Socrates refers to bodily appetites
such as hunger and thirst—appetitive desires whose satisfaction is pleasant (496c6–e9).
Given this difference between the two verbs, it is probably no coincidence that in the
Gorgias “wanting” (boulesthai) is used in the argument for the claim that all desires are
for the good, and “desiring” (epithumein) is used in the discussion of appetitive desires
that are aimed at pleasure and need to be restrained: in these contexts it seems likely that
Socrates is using the verbs in their special senses. If so, the apparent inconsistency is
resolved: if his statements are about two distinct types of desire, “wanting” and “appeti-
tive desire,” there is no inconsistency. If, alternatively, the claims are about a single notion
of desire, then we are faced with an obvious inconsistency.
Let us suppose, charitably, that Socrates is not guilty of an obvious inconsistency.
We might then sum up his view as follows. There are two types of desire, one of which
(“wanting”) is always directed toward what is actually good, while the other (“appetitive
desire”) is always directed toward what is pleasant. Some appetitive desires need to be
resisted since satisfying them would be bad for us. Thus we can only want what is good
for us, but we can desire what is bad. The tyrant doesn’t want to do something unjust
since it would be bad for him, but he may desire to do it if he believes it will bring him
pleasure (or fend off future pain). This way of understanding Socrates’ claims eliminates
the apparent inconsistency in the Gorgias, and also leaves room for the sort of psycho-
logical conflict presupposed in the Protagoras—a conflict between our appetitive desire
for immediate gratification and our judgment that it isn’t worth the cost. But we are still
left with a puzzle: What is Socrates’ reason for distinguishing between the two types of
desire in the way that he does? The distinction seems somewhat artificial, and the claim
that there is a type of desire that is directed only at what is actually good is particularly
problematic.20 Does Socrates make this claim simply in order to support his paradoxical
thesis that tyrants do not have great power?
Later in the discussion with Callicles, it becomes clear that the problematic claim
about wanting the good is also related to Socrates’ well-known paradox that “all unjust
action is involuntary.”

Why don’t you at least answer me this, Callicles—Do you or don’t you think Polus and
I were correct in being compelled to agree in our previous discussion that no one acts
unjustly wanting to do so, but all unjust action is done involuntarily? (509e2–7)

20 Aristotle also connects “wanting” (boulesthai) with the good, but in a more plausible way: he
argues that for each individual the object of wanting is what appears good rather than what is actually
good, but also that what appears good to the good person is actually good; see Nicomachean Ethics III 4,
1113a15–31.
418   Daniel Devereux

Actually, Socrates and Polus did not agree that “no one acts unjustly wanting to do so,”
or that “all unjust action is done involuntarily.” However, these claims do seem to follow
from two of their earlier agreements: (1) if a tyrant does X, which is actually bad for him,
he cannot be said to want to do X; and (2) a tyrant’s unjust actions are bad for him in that
they bring about a corrupt, unhealthy condition of the soul. From these it follows that
tyrants (or anyone else, for that matter) cannot want to do the unjust actions that they do;
and if they don’t want to do them, they do them involuntarily. Since it is the “not want-
ing” that makes the actions involuntary, Socrates must establish that no one wants to act
unjustly, in order to support his claim that all unjust action is done involuntarily.21
Socrates’ controversial thesis about wanting the good serves two important purposes:
it allows him to argue that tyrants do not have great power, and it provides a basis for his
paradox that all unjust or immoral action is done involuntarily. But it seems likely that
the thesis is also attractive to Socrates in its own right, as expressing an important truth
about our relationship to the good.22 The thesis implies that we have an innate, natural
attraction to what is actually good. When we mistakenly choose something that only
appears to be good we are still pursuing what is actually good (Meno 77d7–e4; cf. Republic
505d5–9). Socrates would perhaps say that when we attain the object of our misguided
desire, we are not satisfied—we have a sense that this is not what we really wanted
(cf. Gorgias 493d5–494a3).
Plato speaks of our innate orientation toward the good in the following passage in
book VI of the Republic.

Every soul pursues the good and does whatever it does for its sake, having an intui-
tion of its existence but unable to form an adequate grasp of what it is or to acquire
the sort of stable beliefs about it that it has about other things, thus missing out on
whatever benefit other things may provide. (505d11–e4)

Since this passage comes from the Republic, a later “Platonic” dialogue, we cannot assume
that it expresses a view that Socrates would agree with. However, it does seem to accord
with Socrates’ thesis that what everyone wants is the good, and it captures the idea that
even when our choices are misguided we are still pursuing what is actually good. But we
should also remember that Socrates recognizes another type of desire, appetitive desire,
which can be directed toward what is pleasant but bad. It is this type of desire that over-
comes our better judgment in apparent cases of akrasia.

21 One might argue that what makes the tyrant’s unjust action involuntary is the fact that it is done
in ignorance. But if the tyrant would still have wanted to do the act after finding out its true character,
I think it would not be right to describe it as involuntary (akôn or akousion); cf. Aristotle’s distinction
between the “not voluntary” and the “involuntary”: Nicomachean Ethics III 1, 1110b18–24. For a
different view of how the controversial claim about wanting relates to the paradox, see K. McTighe,
“Socrates on Desire for the Good and the Involuntariness of Wrongdoing,” Phronesis 29 (1984), 193–236,
reprinted in H. Benson (ed.), Essays on the Philosophy of Socrates (New York, 1992), 263–97.
22 For development of this suggestion, see R. Kamtekar, “Socrates on the Attribution of Conative
Attitudes,” Archiv für Geschichte der Philosophie 88 (2006), 127–62.
 socratic Ethics and Moral Psychology   419

5. Wisdom, Virtue, and Happiness

Socrates maintains, as we’ve seen, that those who know what is truly good will be virtu-
ous and will always act rightly. He also holds that everyone wants the good. But what
exactly is the good? In a well-known passage in the Apology, he chastises his fellow
Athenians for caring too much for such things as wealth and status, and too little for the
things that really matter: wisdom, truth, and the perfection of the soul (29d7–30b4;
cf. 36c3–d1). This suggests that he regards virtue and wisdom as the highest goods. But
Socrates also boasts that he confers “the greatest benefit” on his fellow citizens by making
them truly happy (36c3–d10; cf. 38a1–7, 41b7–c4). So happiness, too, is a paramount good.
How does Socrates understand the relationship between wisdom and virtue, on the
one hand, and happiness, on the other? In several places, he claims that they are inextri-
cably linked: one cannot be happy without virtue, and if one has virtue one cannot fail
to be happy (Gorgias 470e4–11, 507b8–c5, 508a8–b3; Crito 48b4–6). Is virtue a good
because it contributes to happiness, or does it have some value in itself, apart from its
relationship to happiness (as some modern moral philosophers have argued)? There is
general agreement that Socrates subscribes to a form of “eudaimonism”—that is, the view
that happiness (eudaimonia in Greek) is “the good,” or the highest good, and that every
other good, including virtue, derives its goodness from its relationship to happiness.
Scholars disagree, however, about the way in which virtue is related to happiness, some
holding that it “produces” happiness as a separate product, others that it contributes to
happiness by being a part of it.23
According to the first view, the value of virtue would be purely “instrumental,” and
this seems to fit well with the idea that virtue is an art or skill, similar to carpentry, medi-
cine, or shipbuilding; for these arts have a product separate from the art itself, and the art
derives its value from the product. But there are passages, as we shall see, that are hard to
square with a purely instrumental understanding of the value of virtue: in these passages
Socrates seems to regard virtue as having value in its own right, and not simply as some-
thing that produces happiness as a distinct product. Socrates’ views on these questions
are obviously fundamental to his ethics, so we must try to clarify how he understands
the nature of happiness and its relationship to virtue and the good.
Socrates takes up the question of the nature of happiness in the Euthydemus. He
points out to his young interlocutor, Cleinias, that there is no need to ask whether every-
one wants to “do well” and be happy—the answer is obvious; what is not clear is how one

23 For Socrates’ commitment to eudaimonism, see C. D. C. Reeve, Socrates in the Apology

(Indianapolis, Ind., 1989), 126–30; G. Vlastos, Socrates, Ironist and Moral Philosopher [Socrates] (Ithaca,
N.Y., 1991), 200–32; Irwin, Plato’s Ethics, 52–53; T. Brickhouse and N. Smith, The Philosophy of Socrates
[Socrates] (Boulder, Colo., 2000), 127–29; C. Bobonich, “Socrates and Eudaimonia,” in D. R. Morrison
(ed.), The Cambridge Companion to Socrates (Cambridge, 2011), 293–332; N. Reshotko, “Socratic
Eudaimonism,” in J. Bussanich and N. Smith (eds.), The Bloomsbury Companion to Socrates (London,
2013), 156–84. For the view that virtue produces happiness as a separate product, see Irwin, Plato’s
Ethics, 72–76; for a defense of virtue as a part of happiness, see Vlastos, Socrates, 209–32.
420   Daniel Devereux

goes about “doing well” and being happy. Socrates begins by citing commonly held views,
and then proceeds to “correct” them in certain ways (just as he does in the discussion of
akrasia in the Protagoras). Most people would agree that to be happy is to possess an
abundance of goods such as wealth, health, power, and status, as well as the virtues of
justice, temperance, courage, and wisdom (278e3–c2). But is possession of these goods
a guarantee of happiness? It seems that one also needs a certain amount of luck or good
fortune; for even if someone had all of these goods she might be deprived of happiness
by tragic misfortunes. But Socrates manages to persuade Cleinias that there is no need
to add good fortune to the list of goods, since wisdom, like any art or skill, guarantees
success in one’s activities—it ensures “doing well”—so if one is wise one will have no
need of good fortune (279c4–280b3).24
But Socrates is still not satisfied. He points out that even if we have all of these goods
we will not be happy unless we benefit from them, and this means that we must use them
(280b5–d7). Moreover, we must use them rightly—that is, wisely—for if goods are used
unwisely they will result in more harm than good. Happiness and “doing well” will there-
fore consist in the wise use of such goods as wealth, health, power, and so on. Wisdom
turns out to be the key to happiness (280d7–281b4).25 In fact, Socrates goes on to argue
that, strictly speaking, wisdom is the only thing that is good. For if a good is something
that never fails to be beneficial, wisdom seems to be the only thing that qualifies as a
good. Someone who lacks wisdom would not be better off with wealth or power, or any
of the other things on the list of “goods,” for if these things are used unwisely (as they can
be), they turn out to be harmful rather than beneficial. To underline the fact that the other
“goods” apart from wisdom are not genuine goods, let us call them “assets.” Happiness,
then, will consist in the wise use of assets: one must have some assets to “use wisely,”26
but the most important factor is wisdom.

24 But suppose an expert ship captain is blown off course and shipwrecked by an unforeseen storm:
has he “done well”? “Doing well” for Socrates apparently consists in the skillful exercise of one’s craft,
or of wisdom; we may “do well” even if we fail to achieve the goal of our skillful actions. Socrates’
discussion of the relationship between “doing well” and good fortune has interesting connections with
Stoic views; on the relationship between the Euthydemus and Stoic ethics, see J. Annas, “Virtue as the
Use of Other Goods,” in T. H. Irwin and M. C. Nussbaum (eds.), Virtue, Love and Form: Essays in
Memory of Gregory Vlastos (Edmonton, 1993), 53–66.
25 There is a puzzle about Socrates’ claim that possession of goods is not sufficient for happiness. If
we possessed wisdom and a sufficient supply of assets, it seems clear that we would use those assets
wisely—after all, it would be foolish not to. Hence, contrary to Socrates’ claim, it seems that the
possession of wisdom and these assets would be a guarantee of doing well and living happily. Perhaps
Socrates wants to emphasize that, while doing well is guaranteed by the possession of these goods,
it consists in the activity of using them. The claim that happiness consists in activity is standard in
Socratic dialogues: see, e.g., Charmides 171e7–172a5, 173c7–d5; Crito 48b2–8; Gorgias 507b8–c5; Republic
I, 353d11–354a2.
26 See, e.g., 280c3–d7: the craftsman cannot “do well” unless he has tools and materials to work
with. The necessity of having assets in addition to wisdom/virtue seems to conflict with the claim
noted previously that wisdom/virtue is sufficient for happiness. Perhaps in making the claim for the
sufficiency of virtue, Socrates presupposes that the virtuous person has at least some assets, for
example, health (see Crito 47e4–6; Gorgias 505a2–4, 512a2–5; cf. Brickhouse and Smith, Socrates,
139–49).
 socratic Ethics and Moral Psychology   421

In the course of his discussion with Cleinias, Socrates revises or corrects the common
view of happiness as “the possession of many goods” in two ways. First, he drastically
prunes the list of goods: only wisdom qualifies as a genuine good, since it is the only
thing we are always better off having. Second, he points out that happiness does not lie in
the possession of wisdom and other commonly recognized “goods” but in the wise use of
these “goods” in activity. We should also note that Socrates treats wisdom and virtue as
equivalent (282d8–283b3; cf. 278d1–3). Wisdom/virtue is distinct from happiness in that
it is a “possession” that can be used in activity, whereas happiness is not a possession that
can be used—it is the use itself of wisdom/virtue (and various assets) in activity.
In the second stage of the argument, Socrates and Cleinias attempt to clarify the
nature of wisdom/virtue (288d–292e5). It was earlier agreed that wisdom is the only gen-
uine good, and that it is beneficial insofar as it guarantees the correct use of “assets.”
Socrates now points out that among the various arts or skills there is generally a separa-
tion of production and use: some arts are concerned with production but not use (it is
not part of flute-making to know how to play the flute), while others are concerned with
use but not production (the art of flute-playing is not concerned with how flutes are made)
(289b7–d7). Given the preceding argument, one would expect wisdom to be classed
as one of the arts concerned with use. But Socrates surprisingly claims that wisdom
is concerned with both production and use: it has its own characteristic product, and
knows how to use that product (289b4–6). He and Cleinias then set off on a search for
an art that combines production and use. After considering and eliminating various
possible candidates, they decide that the “political” or “royal” art is most likely to be the
one they’re looking for (289b4–291c2).27
But their attempt to identify the product of this art leads to an impasse. Socrates
points out that if the political art is beneficial, its product must be something good; but
the earlier argument showed that the only genuine good is wisdom. So if the political art
produces a genuine good, it must reproduce itself: the function of the political expert
would be to “produce” others who are wise and good in the same way that he is (292b1–d6).
But if these “products” of the political art are good in the same way, then they will
be good insofar as they make others good, and these will be good insofar as they make
others good, . . . and so on. If the question is “How is wisdom beneficial?” the answer,
“It is beneficial insofar as it produces itself,” is of no help at all; as Socrates puts it, the
search for the valuable product of the political art has landed them in a “labyrinth” with
no way out (291b7, 292d8–e5). The rest of the dialogue provides no hints as to how this
impasse or aporia might be avoided.
It might seem as if Socrates has overlooked an obvious solution. At the beginning of
the second stage, he says that wisdom is the art that “provides and produces happiness”

27 The view that wisdom is equivalent to the political art might be seen to follow from Socratic
premises: (1) if wisdom is an art, and if it is characteristic of an art that its possessor can transmit it to
others, then someone who is wise will be able to make others wise; (2) if wisdom and virtue are
equivalent, then someone who is wise will be able to make others virtuous; (3) the function of the
political art is to make others virtuous; (4) hence someone who is wise will have the ability to carry out
the function of the political art: that is, they will possess the political art.
422   Daniel Devereux

(291b4–7). And in the first stage of the argument it was agreed that wisdom guarantees
correct use of assets, thereby ensuring that we will “do well” and be happy. Isn’t it clear
that happiness is the good product provided by wisdom or the political art? Perhaps this
is the solution that Plato had in mind, but there are a couple of reasons for skepticism.
First, let us recall that the product they are looking for is something that the political art
knows how to use. If happiness is the product, then the political art must know how to
use happiness. But is happiness something that can be used? If happiness consists in the
wise use of assets, can we make sense of this wise use of assets itself being something that
can be used?
A second reason for doubting that happiness provides the solution to Socrates’ aporia
has to do with the distinction between the exercise of an art or skill and the goal at which
it aims. In a typical art or skill such as carpentry, its “exercise” will be the various activi-
ties involved in the production of, say, a table. Socrates says that “doing well” for the car-
penter is a matter of using appropriate tools and materials in a skillful manner (280c3–d1,
281a2–6). The product aimed at is not the activity of production but the result—for exam-
ple, a table. Even in the case of an art such as flute-playing that is concerned with the use
of something produced by another art, there is a distinction between the use and the
result aimed at—between the flutist’s playing of the flute and the melody produced. So if
wisdom is an art, there should be a product or result, which it aims at in its use of assets.
The very expression “wise use of assets” implies a goal or product: when we use some-
thing it is always with a view to some end or goal. Since happiness is identified with
the use of assets, it cannot be the product of the political art. If wisdom is equated
with the political art, and happiness consists in the exercise of that art, there must be
something else that is its product or goal.28
If we turn to the Gorgias, keeping in mind the aporia of the Euthydemus, we find
clear suggestions pointing to a solution. In the Gorgias Socrates claims that the goal of
any genuine art is a certain virtue or excellence in the thing to which the art is applied
(503d6–504e4, 506c5–e4). Medicine and “gymnastic” (physical training) are concerned
with the body, and their aim is to bring about and maintain the excellence or “virtue” of
the body—health and strength. Socrates divides the political art into two parts corre-
sponding to gymnastic and medicine: the art of legislation, and the art concerned with
the administration of justice (464b2–c5, 520b2–3). The political art tends to the soul: its
aim is to bring about and maintain the soul’s excellence or virtue. In his discussion with
Callicles, Socrates introduces a very general and simplified concept of virtue: the virtue
or excellence of a thing has to do with how its parts are related to each other; a thing will
be an excellent specimen of its kind—it will possess virtue—if its parts are ordered in
the appropriate way for a thing of its kind (506c5–e4, 503d5–504d3). Thus the virtue of
the body, health, consists in the various parts of the body being appropriately ordered
(504b7–c9).

28 According to this scheme, the wise person does not aim at happiness in exercising his or her
wisdom, but happiness is nonetheless something “produced” by wisdom (291b4–7, 292b6–c1); we might
call it a “by-product” of wisdom.
 socratic Ethics and Moral Psychology   423

The virtue of the soul is analogous to physical health: it consists in the parts of the soul
being ordered in the appropriate way. But what are the “parts” of the soul? Socrates does
not take up this question directly, but he indicates what he has in mind in a couple of
brief remarks: in one passage he mentions “that [part] of the soul in which appetitive
desires are found” (493a1–b3), and in another he refers to a part that should exercise
rule over these desires—presumably the rational, thinking part of the soul (491d7–e1).
The virtue of the soul, then, and in particular justice and temperance (504d1–e4), will be
the condition in which the appetitive and the rational parts of the soul are properly
ordered: that is, the condition in which the rational part rules the appetitive. Just as the
goal of medicine and gymnastic is to bring about and maintain the virtue of the body,
health, so the goal of the political art is to bring about and maintain the virtue of the
soul: that is, the proper order of its parts. (Let us call this “psychic order.”)
In the Gorgias, in contrast with the Euthydemus, Socrates does not treat virtue as
equivalent to wisdom or the political art. Virtue as psychic order is the product of the
political art, as health is the product of the art of medicine. The analogy between virtue
and health clearly implies that virtue is distinct from the political art—it is not identical
with a form of knowledge, but consists in a relationship between the rational and appeti-
tive parts of the soul. The analogy suggests that, just as one doesn’t need to acquire the
art of medicine to be healthy, so one doesn’t need to acquire the political art to possess
virtue; the good political leader will work to bring about virtue (especially justice and
temperance) in the souls of his fellow citizens, but this will not necessitate teaching them
the political art (504d1–e4, 515a1–d1, 517b2–c2).29
The treatment of the political art in the Gorgias provides what is missing in the
Euthydemus: a product distinct from the art itself, and also distinct from happiness. But
is the product of the political art, virtue, a good in its own right? Recall that the aporia of
the Euthydemus demands that the product be a genuine good. Does the Gorgias recog-
nize a distinct kind of good embodied in virtue understood as psychic order? Socrates
seems to give an affirmative answer in the following passage.

Socrates: Listen, then, as I take up the discussion from the beginning. Is the pleasant
the same as the good?—It is not, as Callicles and I have agreed.—Should the pleas-
ant be done for the sake of the good, or the good for the sake of the pleasant?—The
pleasant for the sake of the good.—And pleasant is that by which, when present, we
experience pleasure, and good is that through which, when it is present in us, we are
good?—That’s right.—But surely we are good, both we and everything else that’s
good, when some virtue has come to be present in us?—Yes, this seems to me neces-
sary, Callicles.—But the way in which the virtue of each thing comes to be present
in it, whether it’s the virtue of an artifact, a body, a soul, or of any living thing—the

29 See T. H. Irwin, Plato, Gorgias (Oxford, 1979), 214. At 507a5–c5, Socrates enumerates the virtues
that make one a good person: temperance, justice, piety, and courage; the striking absence of wisdom
may perhaps be explained by the fact that Socrates here has in view the virtue that a good political
leader works to instill in the populace at large (504d5–e4, 515b8–c3, 517b2–c2). Being just and acting
justly may require a knowledge of which actions are just and which unjust (509d7–e2), but this would
be only a part of the knowledge that a good political leader uses to instill virtue in his fellow citizens.
424   Daniel Devereux

best way it comes to be present is not at random but through the order, correctness,
and art that has been bestowed on each of them. Isn’t this right?—I for one would
say so.—So it is through ordering that the virtue of each thing is ordered and
organized?—I would agree.—Each thing, then, is rendered good by the presence in
it of a certain organization, an organization that is appropriate for that thing?—
I think so. (506c5–e4)

Socrates here treats virtue as a source of goodness: it is the presence of the appropriate
virtue in a thing that makes that thing good (cf. Charmides 161a8–9, Gorgias 497d8–e3,
Meno 87d8–e1). As a source or principle of goodness—as a “good-maker”—virtue itself
must be good; and it would seem to be a good in its own right.
We have seen that the Gorgias provides a possible solution to the aporia of the
Euthydemus by identifying a distinct product of the political art, a product that is a gen-
uine good in its own right.30 We might say that the Gorgias “corrects” the Euthydemus
by introducing a conception of virtue that is distinct from wisdom or the political art—
virtue understood as psychic order. Wisdom is a virtue; but there is another kind of
virtue, psychic order, which is exemplified most clearly in justice and temperance
(504d1–e4, 506e1–507a2).
We began this section by noting that Socrates in the Apology considers both happi-
ness and wisdom/virtue as the most important or highest goods, but he doesn’t indicate
how these are related to each other. Our examination of the Euthydemus and Gorgias has
provided some clarification. Let us sum up our results.

1. Happiness or “doing well” is not a possession that can be used; it consists in use—
the use of assets guided by wisdom.
2. Wisdom is a possession that can be used, but it differs from other things that can
be used in that it can only be used well or rightly whereas they can also be used
badly or wrongly.
3. Wisdom, understood as the political art, not only guides the use of assets (including
the products of other arts: Euthydemus 291c4–9, Gorgias 517c7–518a7), but also
has the function of producing virtue in human souls.
4. The virtue produced by the political art consists in the proper order of the parts of
the soul—the rational part ruling the appetitive part—and it is distinct from the
political art in the way that health is distinct from the art of medicine.
5. Virtue as psychic order is not like an asset that may or may not be beneficial; it is
like wisdom—it is always beneficial, it always contributes to our happiness.
6. Virtue as psychic order is also a good-making characteristic: it “makes” those who
possess it good specimens of their kind—that is, good human beings. As a “good-
maker,” it must be good in its own right.

30 What about the demand that the political art know how to use its product? The product of the
political art is virtuous citizens. We may surmise that a wise statesman will know how to “use” these
products in the sense that he knows how to direct their activities for the overall good of the civic
community.
 socratic Ethics and Moral Psychology   425

Happiness is also a good in its own right, of course, but not a “good-maker” in the way
that virtue is; happiness is a good insofar as it is a kind of activity and life that fulfills the
desires of a person whose soul has psychic order (Gorgias 492c4–494a5, 504e6–505b1;
cf. Symposium 204e1–205a3). We noticed earlier that Socrates speaks of happiness as
something “produced” by wisdom or virtue (produced as a sort of by-product: see note
28 of this chapter). Does this mean that happiness is related instrumentally to its “pro-
ducers”? This depends on how one understands the relationship between a state or condi-
tion of the soul (wisdom or virtue) and its exercise or expression in activity (happiness).
The relationship seems closer than that between a typical art and its product, but not
quite the same as that between whole and part: perhaps the relationship is similar to
that between an art such as dancing and its “product”: the product simply is the exercise
of the art.31

6. Socrates and Plato Revisited

Some of the claims made in Section 5 of this chapter are controversial. For example, the
claim that Socrates views virtue, understood as psychic order, as having value in its own
right apart from its contribution to happiness, would be disputed by those who believe
that Socrates subscribes to a strong form of eudaimonism: the view that happiness is “the
good,” or the highest good, and that every other good, including virtue, derives its good-
ness from its contribution to happiness. These scholars might argue that, although there
are passages in the Gorgias that link the goodness of virtue to the notion of “proper order,”
the value of this proper order is, for Socrates, tied to its contribution to happiness.32
Another controversial aspect of our interpretation is the claim that Socrates (in the
Gorgias) distinguishes between wisdom and virtue—that is, virtue understood as psychic
order—and the further suggestion that these are separable in the sense that one can have
virtue as psychic order without possessing wisdom or the political art. The latter view

31 Cf. Aristotle, Eudemian Ethics II 1, 1219a11–18.

32 According to a number of recent scholars, Socrates regards happiness as the single ultimate end of
action—that every deliberate action is done for the sake of happiness (what is often called
“psychological eudaimonism”; see note 23 of this chapter). But even if this were true, it would not rule
out the possibility that virtue as psychic order has value in its own right, apart from its contribution to
happiness: virtue may be choiceworthy for its own sake as well as for the sake of happiness, while
happiness is choiceworthy only for its own sake. It is questionable, however, whether Socrates does hold
that happiness is the single ultimate end of all choices and actions. At Gorgias 499e7–500a3, he
characterizes “the good” as “that for the sake of which everything that we do should be done”; then at
506c5–e4 he picks up this line of thought (see c5), and argues that “the good” is virtue understood as
proper order. Nowhere does he characterize happiness as “that for the sake of which everything is done
(or should be done).” Instead, he says that virtue (or wisdom) is necessary for one who “wants to be
happy” or “is going to be happy” (507c–-d6; Euthydemus 280d4–6; cf. Republic 358a1–3, 427d1–7,
498b8–c4); in other words, we should act for the sake of virtue, and only by acting in this way will we
be able to attain happiness (see note 28 of this chapter).
426   Daniel Devereux

would imply a rejection of the Unity of the Virtues, a doctrine that is generally believed
to be a standard fixture of the Socratic dialogues.
While it seems clear that a conception of virtue as psychic order is introduced in the
final section of the Gorgias, some have argued that this is a Platonic innovation and
should not be attributed to Socrates.33 This is linked to the view that the Gorgias is a “tran-
sitional” dialogue, that is, a dialogue that shares many features with the other Socratic
dialogues, but also includes some Platonic elements that are not in harmony with
Socrates’ views.34 In fact, the Gorgias contains both the “Platonic” conception of justice as
psychic order (504c5–d3, 507e6–508a4; cf. 525a3–6), and the “Socratic” view that it is an
art or skill (460a5–c6)—and we are given no hints as to how we are supposed to fit these
views together. But this is nothing new. We noticed, for instance, that Socrates assumes
in many arguments that virtue is an art or skill, but in the Hippias Minor he raises diffi-
culties for this view. He also seems to argue for inconsistent views about how the virtues
form a unity (see note 16 of this chapter).
If the conception of virtue as psychic order appeared only in the Gorgias and not in
any of the other Socratic dialogues, it would be plausible to regard it as a Platonic inno-
vation. But the analogy between virtue and health (and vice and sickness) that underlies
the conception of virtue as psychic order appears in other Socratic dialogues. In the Crito,
for instance, Socrates suggests that just as “healthful” things promote the good condi-
tion of the body while “diseaseful” things tend to cause its destruction, so just actions
promote the good condition of the soul (virtue) while unjust actions corrupt and destroy
it (47c8–48a4).35 And when he claims in the Apology that his accusers are doing more
harm to themselves than to him by their unjust actions, and that through their unjust
actions they acquire vice and wickedness, he seems to have the same view in mind
(30c7–d6, 39a6–b6; cf. Gorgias 479b3–c4, 480a6–b2). The health analogy and the associ-
ated conception of virtue as psychic order are not “Platonic” innovations: they already
appear, at least in rudimentary form, in the Socratic dialogues. Perhaps the development
of the health analogy in the Gorgias is “Platonic,” but if so, Plato probably thought of
himself as developing a Socratic idea.
The intellectualist conception of the virtues as forms, or a single form, of knowledge is
clearly dominant in the Socratic dialogues. But there is another conception, virtue as
psychic order, which appears in several of these dialogues and is developed in some
detail in the Gorgias. These different ways of understanding virtue do not fit together to
form a unified conception. The presence of the two incompatible conceptions of virtue

33 See, e.g., W.H. Thompson, The Gorgias of Plato (New York, 1973), viii–x.
34 See A. Gomez-Lobo, The Foundations of Socratic Ethics (Indianapolis, Ind., 1994), 109–11;
J. M. Cooper, “Socrates and Plato in Plato’s Gorgias,” Reason and Emotion: Essays on Ancient Moral
Psychology and Ethical Theory (Princeton, N.J.,1999), 29–75; and Irwin, Plato, Gorgias, 7–8.
35 For “healthful” (hygieinon) and “diseaseful” (nosôdes) as the things that produce health or disease,
see also Gorgias 504c5–9; Republic 444c8–d10. The notion that a virtue such as justice can be acquired
by performing just actions goes hand in hand with the analogy between virtue and health, and does not
fit comfortably with the intellectualist conception of virtue as a kind of knowledge (cf. Euthydemus
283a1–4; Gorgias 507c9–d1, 527c6–e5; Republic 444c1–d1, 518d9–519a1).
 socratic Ethics and Moral Psychology   427

in the Socratic dialogues is just one of several inconsistencies and ambivalences we have
noticed. The lesson to be drawn is that Plato is not setting out a systematic, unified,
“Socratic” theory of the virtues in these dialogues; rather, he is exploring and developing
the provocative claims and ideas of his mentor—claims and ideas that are not always
consistent with each other. An implication of this way of understanding Plato’s project
in the Socratic dialogues is that the break between “Socratic” and “Platonic” may not be
as sharp as modern scholars tend to believe; for example, he may have viewed his elabo-
rate account of the virtues in terms of psychic order in the Republic as a departure from
Socrates’ intellectualism, but at the same time a development of another aspect of his
mentor’s conception of virtue.36

Bibliography
Annas, J., Platonic Ethics, Old and New (Ithaca, N.Y.: Cornell University Press, 1999).
Annas, J. “Virtue as the Use of Other Goods,” in T. H. Irwin and M. C. Nussbaum (eds.)
Virtue, Love and Form: Essays in Memory of Gregory Vlastos (Edmonton: Academic Printing
and Publishing, 1993), 53–66.
Barney, R., “The Sophistic Movement,” in M. L. Gill & P. Pellegrin (eds.), A Companion to
Ancient Philosophy (Malden, Mass.: Blackwell, 2006), 77–97.
Benson, H. (ed.), Essays on the Philosophy of Socrates (New York: Oxford University Press,
1992).
Bobonich, C., “Socrates and Eudaimonia,” in D. R. Morrison (ed.), The Cambridge Companion
to Socrates (Cambridge, 2011), 293–332.
Brickhouse, T., and Smith, N., The Philosophy of Socrates [Socrates] (Boulder, Colo.: Westview
Press, 2000).
Brickhouse, T., and Smith, N. Socratic Moral Psychology (Cambridge: Cambridge University
Press, 2010).
Cooper, J. M., “Socrates and Plato in Plato’s Gorgias,” in Cooper, Reason and Emotion
(Princeton, N.J.: Princeton University Press, 1999), 29–75.
Devereux, D. T., “The Unity of the Virtues in Plato’s Protagoras and Laches,” Philosophical
Review 101 (1992), 765–89.
Devereux, D. T., “Socrates’ Kantian Conception of Virtue,” Journal of the History of Philosophy
33 (1995), 381–408.
Devereux, D. T., “Comments on Socratic Moral Psychology,” Analytic Philosophy 53 (2012),
216–23.
Dorion, L.-A., “The Rise and Fall of the Socratic Problem,” in D. R. Morrison (ed.), The
Cambridge Companion to Socrates (Cambridge: Cambridge University Press, 2010), 1–23.
Ferejohn, M. (1982). “The Unity of Virtue and the Objects of Socratic Inquiry,” Journal of the
History of Philosophy 20 (1982), 1–21.
Ferejohn, M. “Socratic Thought-Experiments and the Unity of Virtue Paradox,” Phronesis 29
(1984), 105–22.
Gomez-Lobo, A., The Foundations of Socratic Ethics (Indianapolis, Ind.: Hackett, 1994).

36 My thanks to Tom Brickhouse, Gail Fine, and Rebecca Stangl for helpful comments on an earlier
version of this chapter.
428   Daniel Devereux

Irwin, T. H., Plato, Gorgias (Oxford: Clarendon Press, 1979).

Irwin, T. H. Plato’s Ethics (New York: Oxford University Press, 1995).
Kahn, C., Plato and the Socratic Dialogue (Cambridge: Cambridge University Press, 1996).
Kamtekar, R., “Socrates on the Attribution of Conative Attitudes,” Archiv für Geschichte der
Philosophie 88 (2006), 127–62.
Kraut, R., Socrates and the State (Princeton, N.J.: Princeton University Press, 1984).
McTighe, K., “Socrates on Desire for the Good and the Involuntariness of Wrongdoing,”
Phronesis 29 (1984), 193–236; reprinted in H. Benson (ed.), Essays on the Philosophy of
Socrates (New York: Oxford University Press, 1992), 263–97.
Penner, T., “The Unity of Virtue,” The Philosophical Review, 82 (1973), 35–68; reprinted in
G. Fine (ed.), Plato 2 (New York: Oxford University Press, 1999), 78–104.
Penner, T. “Socrates on the Strength of Knowledge: Protagoras 351b–357e,” Archiv für Geschichte
der Philosophie 79 (1997), 117–49.
Reeve, C. D. C., Socrates in the Apology (Indianapolis, Ind.: Hackett, 1989).
Reshotko, N., “Socratic Eudaimonism,” in J. Bussanich and N. Smith (eds.), The Bloomsbury
Companion to Socrates (London, 2013), 156–84.
Santas, G., Socrates, Philosophy in Plato’s Early Dialogues (London: Routledge & Kegan Paul,
1979).
Singpurwalla, R., “Reasoning with the Irrational: Moral Psychology in the Protagoras,” Ancient
Philosophy 26 (2006), 243–58.
Taylor, C. C. W., Plato’s Protagoras, 2nd ed. (Oxford: Oxford University Press, 1991).
Thompson, W. H., The Gorgias of Plato (New York: Arno Press, 1973).
Vlastos, G., “The Unity of the Virtues in the Protagoras,” Review of Metaphysics 25 (1972),
415–58; reprinted with additional notes in Vlastos, Platonic Studies, 2nd ed. (Princeton, N.J.:
Princeton University Press, 1981), 221–69, 418–23.
Vlastos, G., Socrates, Ironist and Moral Philosopher [Socrates] (Ithaca, N.Y.: Cornell University
Press, 1991).
Wilburn, J., “Plato’s Protagoras the Hedonist,” Classical Philology 111 (2016), 224–44.
Zeyl, D. J., “Socrates and Hedonism,” Phronesis 25 (1980), 250–69.
 chapter 18

Pl ato’s Epistemol ogy

C. C. W. Taylor

A convenient starting point for consideration of Plato’s treatment of knowledge is

Socrates’ notorious disavowal of knowledge; we can hope to arrive at a conception of
how knowledge is seen in the Socratic dialogues if we can see what it is that Socrates
claimed not to have. In antiquity, Socrates was widely supposed to have claimed that he
knew nothing or, in some versions, that he knew only this one thing, that he knew noth-
ing.1 But in Plato’s dialogues, Socrates never makes either claim. The nearest he comes to
the former is at Apol. 21d, where he describes himself as having shown that someone
who claimed to have wisdom (which, in context, amounts to wisdom in organizing his
life as a whole) did not, in fact, have the wisdom that he claimed. Socrates remarks that
while it is likely that neither he nor the person whom he has just exposed knows any-
thing fine and good, he (i.e., Socrates) is wiser in that he does not think that he knows
what he does not know (ha mē oida oude oiomai eidenai), while the other does think that
he knows what he does not know. Assuming that “is likely” is an intentional understate-
ment, this is naturally taken as asserting that Socrates knows nothing fine and good. But
in this same work Socrates claims to know that it would be wrong for him to abandon his
divine mission to improve the souls of his fellow citizens by philosophical criticism
(29b); that implies that he knows that carrying out his mission is fine and good, which
I take it he would count as a case of knowing something fine and good. Later (37b) he
claims to know that his being imprisoned would be something bad (presumably because
it would prevent him from fulfilling his divine mission). So either Plato depicts Socrates
as inconsistent within a single work, or the assertion that he knows nothing fine and
good is to be interpreted so as to be consistent with the knowledge claims just mentioned.
The context readily suggests such an interpretation. Socrates is explaining how he was
prompted to his mission by the pronouncement of the Delphic oracle that no one was
wiser than he. He was puzzled by this because he was conscious of not being wise about
anything, great or small (21b)―that is, of not being an expert in any subject matter―and
he tried to show that the oracle could not have meant what it seemed prima facie to

1 Cicero, Academica I, 16 and 45, II, 74.

430   C. C. W. Taylor

mean by seeking genuine experts, whether experts in the conduct of life as a whole, as
the sophists claimed to be (20a–c), or experts in particular areas, such as builders. The
result was that neither kind of expert proved wiser than Socrates: the former because
they had no expertise at all, the latter because they mistakenly believed that the techni-
cal expertise that they did possess extended to the conduct of life as a whole. I suggest,
then, that the assertion at 21d that neither Socrates nor the supposed expert knew any-
thing fine and good is to be interpreted as “neither knows anything fine and good in that
way,” that is in the way that the supposed expert had claimed: the possession of expertise
in how to live. Lacking such expertise, Socrates may still be able to know some particular
moral truths, such as that mentioned previously, though how he knows them is as yet
unexplained.2
So far, the texts warrant a distinction between the highest level of epistemic achieve-
ment, wisdom or expertise, which Socrates claims not to possess, and a lower level,
exemplified by knowledge of particular moral or evaluative truths, which he does
claim.3 It has been suggested4 that this amounts to the distinction between knowledge
and true belief. That distinction is certainly important in Plato’s epistemological thought
(see later on in this chapter), but it is not the distinction drawn in the Apology. In that
work, Socrates is made to claim particular moral knowledge without qualification, or
any other indication in the text that the verbs rendered “know” are not the most appro-
priate terms to use. The contrast between that knowledge and the wisdom that Socrates
disavows is never explicitly spelled out, but it is evidently connected with the fact that
the possessor of wisdom is thereby qualified to impart that wisdom to others, and
regularly does so, whereas Socrates insists that he does not have any wisdom to impart:
that is, that he does not teach anyone anything (19d–20c). What the expert is typically
qualified to teach is a systematic body of knowledge, whether theoretical, practical, or a
combination of the two; in the latter case the relative importance of the two aspects
depends on the nature of the expertise.
It is that kind of expertise that Socrates disavows; he is not expert in any specific
subject or in the kind of general expertise in running one’s life that the sophists claimed
to have. But he does not argue that expertise is impossible. He recognizes experts in
specific areas, and, as far as general expertise goes, he does not argue that there can be

2 For a fuller discussion see G. Fine, “Does Socrates Claim to Know That He Knows Nothing?”,
Oxford Studies in Ancient Philosophy 35, 2008, 49–88.
3 For a similar, though different, view, see G. Vlastos, “Socrates’ Disavowal of Knowledge,” Socratic
Studies, ed. M. Burnyeat (Cambridge, 1994), 39–66, reprinted in G. Fine, ed., Plato 1: Metaphysics and
Epistemology [Plato 1] (Oxford, 1999), 64–92. He claims that what Socrates disavows is certainty,
and that what he claims is a form of knowledge falling short of certainty, derived from successful
application of elenctic argument. My claim is that what Socrates disavows is systematic knowledge, and
that what he claims is unsystematic―that is, piecemeal knowledge. I see nothing in the texts to suggest
that wisdom requires certainty or that Socrates gives any general account of what grounds his
particular claims to knowledge, beyond the claim that he has arguments for them. I discuss the matter
more fully in Socrates (Oxford, 1998), 42–48.
4 By T. Irwin, Plato’s Moral Theory: The Early and Middle Dialogues (Oxford, 1977), 40–41, and
Irwin, Plato’s Ethics (Oxford, 1995), 28–29.
 Plato’s Epistemology   431

no such expertise, merely that those who claim to possess it, including sophists and
statesmen, fail to meet the ordinary standards for possessing expertise, notably the
ability to impart it to others. In various dialogues, we see Socrates in conversation with
self-styled experts in different areas (e.g., Euthyphro claims to be an expert on religious
matters (4e–5a), and Meno on virtue in general (81b)), and as the conversation progresses,
we find their claims to knowledge of the subject evaporating. A notable feature of these
discussions is that they reveal that Plato is using a specific conception of expert knowl-
edge. Central to any expertise is the knowledge of what that expertise is concerned with,
and that knowledge consists in the ability to specify those things. The requirements for
such a specification are exacting; it must apply to all and only the things in question, it
must reveal the feature or features in virtue of which things count as of that kind, and
that feature (or those features) must be the same in all cases. For example, the expert in
holiness must be able to specify a feature or set of features such that (1) all and only holy
things possess that feature, and (2) it is in virtue of possessing it that those things count
as holy (6d–e).
The ability to give that kind of specification is primary in a number of ways. In the
Euthyphro, it serves as a template for the solution of disputed cases; anything that satis-
fies the specification of holiness is holy, and anything that does not satisfy it is not holy
(6e). The specification is thus explicitly said to be sufficient for resolving disputes, and
one may plausibly suppose that it is assumed to be necessary also. In the Meno, having
the specification of what virtue is is necessary for knowing further things about virtue,
specifically how it is to be acquired (71b). In at least one dialogue, the Hippias Major,
Socrates maintains that it is impossible to know whether anything is an instance of a
property (the example is that of beauty) unless one is able to specify what that property
is (304d–e), and that seems also to be the implication of the conclusion of the Lysis
(223b), where Socrates says that he and his young friends appear ridiculous in thinking
that they are friends, though they have proved unable to say what a friend is. It is clearly
implied that in that situation they do not know that they are friends and, perhaps, even
suggested that they are not entitled to believe that they are.5
The evidence surveyed so far has not suggested any general account of knowledge or
any concern with how knowledge is acquired or how it relates to other mental states
such as belief or activities such as perception or thought. What has emerged is the view
that a certain kind of knowledge is primary. This is, roughly, knowledge of what things
are, where things are conceived as universals of one sort or another, chiefly properties
(holiness, temperance) or states (virtue), but also kinds, not sharply differentiated from
properties and states. (I take it that the question “What is a friend?” may be expressed
without change of meaning as “What kind of thing is a friend?” or as “What is it to be
a friend?”―that is, “What is the property of being a friend?.”) Knowledge of “things”

5 For a defense of the thesis that Socrates maintains a strong form of the principle of the priority of
definition―“If A fails to know what F-ness is, then A fails to know anything about F-ness”―see
H. H. Benson, “The Priority of Definition and the Socratic Elenchus,” Oxford Studies in Ancient
Philosophy 8 (1990), 19–65, revised version in Benson, Socratic Wisdom (New York, 2000), ch. 6.
This article contains copious references to other literature on the topic.
432   C. C. W. Taylor

consists in the ability to specify them, as set out in the preceding paragraph. Apart from
the formal characteristics set out there, specifications take various forms, some approxi-
mating to conceptual definitions, as in “speed is the ability which achieves many things
in a short time” (Laches 192a–b), others to scientific accounts, such as “Color is an efflux
of shapes adapted to (the sense of) sight and (hence) perceptible” (Meno 76d). Plato never
makes any explicit theoretical discrimination between these types of specification, and
it is a moot point how far he was aware of the distinction.6
This kind of knowledge is primary in that it is presupposed by any other kinds of
knowledge in the respective area (e.g., knowledge that friendship is good presupposes
knowledge of what friendship is). As is clear from the citations so far, this paradigm
is found in a number of dialogues generally regarded as written early in Plato’s career, as
well as in the Meno. There we find it brought into connection with a number of ques-
tions about knowledge not raised in any of the other dialogues cited, including a puzzle
about how it is possible to acquire knowledge, and the question of how knowledge relates
to true belief. These questions require closer examination.

1. Meno

The dialogue opens abruptly with the question how virtue or excellence (aretē) is to be
acquired, that is: How is one to become an outstanding individual and thereby achieve
overall success in life? Socrates immediately turns the question to that of what virtue or
excellence is, in accordance with the primacy thesis elucidated previously, and various
attempts at specification are explored and rejected. When Socrates says (80d) that
though he and Meno do not know what excellence is, they should continue to try to find
out, Meno asks how it is possible to try to find out anything that you do not already
know. He poses two specific problems:

1. Of the many things that you do not know, which one will you set up as the object
of your inquiry?
2. Even if you were to happen upon what you were looking for, how will you know
that that is what you did not know? (80d6–8).

6 Of the extensive literature on Socratic definition, the following may be particularly mentioned:
T. Penner, “The Unity of Virtue,” Philosophical Review 82 (1973), 35–68, reprinted in H. H. Benson, ed.,
Essays on the Philosophy of Socrates (New York, 1992), 162–84; and in G. Fine, ed., Plato 2: Ethics,
Politics, Religion, and the Soul (Oxford, 1999), 78–104; G. Vlastos, “What Did Socrates Understand by
His “What Is F?” Question?,” Vlastos, Platonic Studies, 2nd ed. (Princeton, N.J., 1981), 410–17.
C. C. W. Taylor, “Socratic Ethics,” in B. S. Gower and M. C. Stokes, eds., Socratic Questions (London,
1992), 137–52; H. H. Benson, Socratic Wisdom (New York, 2000), ch. 5; D. Charles, “Definitions in the
Meno,” in V. Karasmanis, ed., Socrates 2400 Years since His Death (Athens, 2004), 357–66, revised
version, entitled “Types of Definition in the Meno” [“Types of Definition”], in L. Judson and
V. Karasmanis, eds., Remembering Socrates: Philosophical Essays (Oxford, 2006), 110–28.
 Plato’s Epistemology   433

The first problem makes the point that, in order to undertake any inquiry, one must
know what inquiry one is undertaking, and therefore understand the terms in which
that inquiry is identified. To undertake an inquiry is to ask a question, and one must
know what question it is that one is asking. (Knowing what question one is asking is of
course distinct from knowing the answer to that question.)
The second problem asks how you will know whether you have found what you
were looking for, with the implication that you will never know. If you did not know
the answer to the question in advance of inquiry, how will you recognize any answer
as the one you were looking for? In general, the answer to the second problem is that the
understanding of what question one is asking provides a specification of what counts as a
correct answer to it, and that one recognizes an answer as correct when one recognizes
that it satisfies that specification. Thus understanding the question “What is the cube
root of 27?” involves knowing that one has correctly answered it when one has found
a number n such that ((n × n) × n) = 27, and understanding the question “Who is the
murderer of Smith?” involves knowing that one has correctly answered it when one
has identified an individual of whom it is true that that individual murdered Smith. (The
question of how one knows that one has found the right number, or the right individual,
is not a question about how doing so is in principle possible, but about how one knows
that one has employed the right method of inquiry and employed it correctly.)
The general answer to the second problem―that the phrasing of the question provides
a specification of the correct answer―poses a particular problem in the special case
where the question is itself a request for a specification. Prior understanding of what a
cube root is specifies the correct answer to the question “What is the cube root of X?,”
but if one’s question is “What is X?” (e.g., “What is virtue?”), it is problematic what prior
understanding might be supposed to specify the correct answer. It is tempting to think
that that prior understanding could be nothing other than understanding of what virtue
is: that is, in the special case of the kind of knowledge that Plato regards as primary (see
previous discussion), Meno’s second problem is unanswerable. In fact it is not. The prior
understanding in question is the pre-theoretical understanding of the concept that is
presupposed by the ability to pose and to understand the request for a specification,
and the specification itself consists either in sharpening that pre-theoretical under-
standing via a conceptual definition or in providing a substantive account satisfying
the requirements indicated in that pre-theoretical understanding. (There is a residual
problem―namely, which type of specification is (1) sought and (2) appropriate for the
particular case.)7
In response, Socrates represents Meno as arguing that it is impossible to try to find
out anything; he poses the dilemma that either one already knows what one is trying to
find out, in which case one cannot try to find it out, or one does not already know it, in
which case one does not know what one is trying to find out (and hence cannot look for
it) (80e). This “captious argument,” (eristikos logos) as Socrates describes it, does not
do justice to the genuine problems that Meno has raised or to the insights about the

7 For example, Charles, “Types of Definition,” 118–19.

434   C. C. W. Taylor

presuppositions of inquiry that those problems reveal. The dilemma that Socrates
ascribes to Meno is solved simply by the distinction between knowing what question
you are asking and knowing the answer to that question. Contrary to his description,
however, Socrates does not treat Meno’s problem as a facile sophism but as a deep problem
whose solution involves an ambitious theory not only of the acquisition of knowledge
but also of the nature of the soul. The essence of the solution is that it is possible to find
out what you (by ordinary standards) do not know, provided that in a deeper sense you
do already know it. What we think of as discovery is in fact the recovery of knowledge
that the soul has previously possessed but that it has forgotten. The detailed exposition
in which that broad outline is spelled out raises a number of difficult questions about
precisely what cognitive resources it is that the soul has previously possessed, how it has
come to possess them, and how its mode of possession relates to the distinction between
knowledge and belief.
Socrates begins by simply stating the theory on the authority of priests, priestesses,
and poets; the human soul is immortal and undergoes many incarnations, in the course
of which it has “seen everything here and in Hades” and has thereby learned everything.
Hence it is not surprising that it should be able to recall what it previously knew about
virtue and other things (81c). This suggests a simple model of the revival of experiential
knowledge; knowledge is originally acquired by experience, whether of things in the
world, experienced, presumably, via the senses, or of things in Hades. What those things
might be, and what kind of experience apprehends them, we are not told. We are told,
however, how this theory allows for the process of arriving at knowledge via some
sequence of mental acts; since the whole of nature is akin, and one has learned everything,
recalling one thing allows one to find out everything else, provided one perseveres in one’s
search (81d). This sounds like a description of the acquisition of knowledge by inference
from things known by experience, but a key feature of inferential knowledge is that
previous experience of what one knows by inference is no part of the explanation of
one’s knowledge. One is not reminded of what one knows by inference.8
Socrates’ explanatory account of this kind of knowledge, in contrast, includes the
repetition of the claim that the soul has learned everything (d1), and he concludes (d4–5)
with the assertion that seeking and learning is just recollection (or, in other words, being
reminded (anamnēsis)). That is not, then, inferential knowledge. Rather, what is envis-
aged is something more like sequential revival of experiences via association of ideas.
Suppose that I have been previously acquainted with every member of a given family.
Since they all share a family resemblance, recalling what Robert Smith looks like serves

8 One might have inferential knowledge of something one had already experienced, provided that
the previous experience is not causally productive of one’s inferential knowledge. For example, if one
saw a dog running over a snow-covered lawn, but then forgot having done so, one might infer from the
tracks that a dog had run over the lawn. But if one remembered seeing the dog run over the lawn, one’s
knowledge that a dog ran over the lawn is not inferential, even if one did, in addition, infer from the
tracks that a dog ran over the lawn.
 Plato’s Epistemology   435

to remind me of what his brother Richard looks like, and that, in turn, of what their
cousin Winifred looks like, and so on. There is no inference here, merely serial reminding.
Challenged to show that this theory is true, Socrates conducts the famous experiment
with the slave, which he claims to be an instance of the process he has just described.
This, however, appears to be an instance of the acquisition of knowledge by inference.
The slave certainly works out the answer to the problem of doubling the square by infer-
ence, specifically inference from the premises that the diagonal of a given square bisects
it; that the square on the diagonal contains four triangles, each equal to half the area of
the given square; and that 4 × ½ = 2. Insofar as his reaching the correct solution is to be
explained by recollection, it is quite unnecessary to suppose that the slave is recollecting
the solution of that particular problem, a solution that, ex hypothesi, he had arrived at in
some previous existence. It suffices to suppose that he recollects the crucial properties of
the square and the diagonal, from which he now (for the first time) infers the solution.
But it is unclear whether that is how Plato sees the matter. It is possible that he does not
distinguish the acquisition of knowledge by inference from what I have termed serial
reminding and therefore intends what, in fact, is a description of the latter to apply to
both indifferently.
This raises the important question of what it is that one recollects. On the model of
serial reminding, one recollects literally everything that one finds out by any kind of
investigation, and everything that one recollects one has previously experienced. Even
if we restrict the application of the theory to a priori investigation (a restriction for
which there is no explicit textual warrant), it is still a vastly uneconomical theory and
one that depends on a quite obscure conception of experience. It is uneconomical in
supposing that every particular arithmetical or geometrical truth that anyone discovers
has previously been known by that person, and obscure in attributing that knowledge to
experience. To stick to the example of doubling the square, what would it be for the slave
to have “seen” that the area of the square on the diagonal of a given square is double the
area of the original square? Ex hypothesi, that would be to recognize that particular
truth without inference; but that leaves us quite in the dark how the slave knows that
truth (he “just knows,” it seems) and darker still how the process of thought that he
undertakes together with Socrates revives that particular item of immediate knowledge.
These difficulties are at least alleviated if we suppose that what is recollected is some
restricted set of items (elements, principles, or basic entities), knowledge of which pro-
vides the basis for inferential knowledge of further truths. The idea that there might be
such a thing as immediate apprehension of the properties of, say, the square or the even
seems not an obviously hopeless suggestion. It remains problematic how closely that
notion of immediate apprehension can be modeled on perception.
This question connects with the topic of the distinction between knowledge and true
belief. When the slave has reached the correct answer to the problem, Socrates says that
he has true beliefs about it (which are his own, not imposed on him by someone else) but
that he does not yet know the answer. His true beliefs have been stirred up as if in a
dream, and if he is subjected to repeated and varied questioning, he will eventually attain
436   C. C. W. Taylor

exact knowledge (85b–c). Yet, immediately, Socrates describes him as having recovered
his knowledge (my italics) from within himself, without anyone’s having taught it to him
(d3–4) and proceeds to argue that the knowledge that he now has9 he must always have
had, since he could not have acquired it during his present life (d9–13). That is clearly
incompatible with the suggestion that all that he now has is true belief and that he is yet
to acquire knowledge, which he will do as a result of subsequent questioning. True
belief, it appears, presupposes the permanent possession of knowledge (86a–b), and the
transition from true belief to knowledge is in fact the transition from a state of partial
recovery of the knowledge that we have always possessed to its full recovery.
The connection with the perceptual model of knowledge arises from Socrates’ argu-
ment (86a6–9) that since the slave’s true beliefs are always in his soul, both in its incarnate
and in its discarnate state, his soul is always in a state of “having learned” (ton aei chronon
memathēkuia estai hē psuchē autou.) If the soul is always in a state of having learned (i.e.,
having acquired knowledge), there was no time at which it did acquire that knowledge;
“always having learned” is thus equivalent to “never having learned, but always knowing.”
And since the knowledge we have was never acquired, but was always possessed, it follows
that it was not acquired by experience.
The distinction between knowledge and true belief reappears at the end of the dialogue,
when Socrates points out that true belief is as good a guide to action as knowledge (97a–b).
The crucial difference is one of stability; true beliefs are as useful as knowledge as
long as one retains them, but they are liable to be lost, “until one ties them down by
­reasoning about the cause” (aitia).10 That (i.e., tying down true beliefs by reasoning
concerning the cause) is recollection, and when true beliefs are thus tied down, they
become stable items of knowledge (98a). This brief passage contains a cluster of problems.
The first is how we are to understand “reasoning about the cause”; the cause of what?
Reasoning about the cause of one’s having a belief does not seem appropriate to turn

9 It may be, as G. Fine suggests (Plato on Knowledge and Forms [Knowledge], (Oxford, 2003), 5 and
69), that “now” at d9 refers not to the actual time of Socrates’ utterance but to the envisaged future
time at which the slave has achieved complete knowledge. But even if that is so, the argument still
requires that the slave has never acquired knowledge but has always possessed it.
10 Following on from G. Vlastos’s influential discussion of the meaning of aitia in Plato and
Aristotle, G. Fine suggests that aitia should be rendered “explanation” rather than “cause.” Fine explains
that she reserves the term “cause” for an event that is sufficient for bringing about change and points
out, correctly, that Plato’s aitiai―for example, the aitia of the correct solution of the geometrical
problem in the Meno―are not restricted to events. See Vlastos, “Reasons and Causes in the Phaedo,”
Philosophical Review 78 (1969), 291–325, reprinted in Vlastos, ed., Plato I: Metaphysics and Epistemology
(Garden City, N.Y., 1971), 132–66; and in Vlastos, Platonic Studies, 2nd ed. (Princeton, N.J., 1981), 58–75;
Fine, “Knowledge and True Belief in the Meno” [“Knowledge and True Belief ”], Oxford Studies in
Ancient Philosophy 27 (2004), 56.
In common with some other writers on Plato and Aristotle, I use “cause” more widely, to apply to
whatever answers the question “Why?” and hence as virtually interchangeable with “reason” and
“explanation.” For this usage, see, e.g., D. Bostock, Plato’s Phaedo [Phaedo] (Oxford, 1986), 135; for a
defense of the translation of aitia as “cause,” see D. Furley, “What Kind of Cause Is Aristotle’s Formal
Cause?,” in M. Frede and G. Striker, eds., Rationality in Greek Thought [Rationality] (Oxford, 1996),
60–62. Given the wider usage of “cause,” the issue of whether aitia should be rendered “cause” or
“explanation” is stylistic rather than substantial.
 Plato’s Epistemology   437

true belief into knowledge. On the other hand, reasoning about the cause of the belief ’s
being true makes good sense of the point about stability, as well as giving a good account
of the example of the slave. Someone who understands why a given belief is true―for
example, because it follows from the basic principles of the discipline to which it
belongs―is not liable to be persuaded by apparent counterarguments. And we can see
why frequent and varied questioning would be needed to give the slave the systematic
grasp of geometry that would enable him to see not just that this is the correct solution
of this problem but why it is.
It is problematic, however, whether reasoning about the cause is required for every
case of knowledge or only for some. In this passage, someone who has traveled the road
to Larisa is said to know it, which presumably implies that his or her true beliefs about
the road amount to knowledge about it. But acquaintance with the road is surely insuffi-
cient to give one understanding of why one’s beliefs about it are true; one may know that
the road passes a certain hill on the north side because one has been there but still not
understand why the road passes the hill to the north rather than to the south. Again,
the acquisition of this kind of knowledge seems to have nothing to do with reasoning;
experience of the actual road (together with memory of what one has experienced) is
sufficient for knowledge of it. It seems plausible, then, that perceptual knowledge is not
supposed to be explained by recollection, which is what we should expect, especially in
light of the assertion that knowledge that is recollected was not acquired at any time
but was always possessed. No one can believe that knowledge of the road to Larisa, or
knowledge that Socrates is now standing in front of me, has been in my soul as long as
my soul has been in existence.
Recollection, then, provides an explanation of a special kind of knowledge, which
contrasts with perceptual knowledge.11 It is characteristic of that kind of knowledge to
be grounded in an understanding of what makes the beliefs constitutive of that knowl-
edge true, an understanding that is reached via reasoned inquiry.12 The objects of that

11 Alternatively, only the kind of knowledge that is explained by recollection is knowledge, strictly
speaking. On that supposition, perceptual knowledge counts as knowledge in a reduced sense, perhaps
on the strength of some resemblance to knowledge properly so called (e.g., that it gives one’s true
beliefs the same degree of stability as “reasoning about the cause” does). On either view, knowledge
grounded in recollection is primary in the evaluative sense.
12 Some commentators (e.g., Fine, Knowledge, 5–6, 50, and “Knowledge and True Belief,” 61–67)
interpret the requirement that knowledge involves tying down true beliefs by reasoning concerning the
cause as amounting to the definition of knowledge as justified true belief (a view shared by E. L. Gettier
in his epoch-making article “Is Justified True Belief Knowledge?,” Analysis 23 (1963), 121–23). But the
requirement that one should have a grasp of what makes one’s beliefs true is a stronger requirement
than that one’s true belief should be justified. At the conclusion of his discussion with Socrates, the slave
is justified in his true belief that the square is doubled by constructing the square on the diagonal,
since he has constructed (or at least followed) a sound argument leading to that conclusion, but Socrates
insists that he does not yet know the conclusion. (In fact, some of Gettier’s counterexamples show
how one can be justified in having a true belief yet lack knowledge, since one does not grasp why
one’s belief is true. Someone who has excellent evidence that Jones owns a Ford may believe on the
strength of that evidence that either Jones owns a Ford or Brown is in Barcelona, and that might be
true despite the fact that Jones no longer owns a Ford, because Brown is, as it happens, in Barcelona.)
438   C. C. W. Taylor

understanding are most plausibly to be thought of as whatever function as principles or

elements of reasoning. At this point, we should recall Socrates’ methodological princi-
ple that knowledge of what things are is primary in the investigation of the properties of
those things. In the Meno, Plato seems to be moving toward a systematization of that
principle in an ambitious combination of epistemology and metaphysics.
On this theory, the transformation of true belief into knowledge via intellectual
inquiry is, in fact, the recovery of knowledge that the soul has always possessed. Souls
are created with a grasp of the basic principles of reality, including an understanding
of what are the primary things, and the task of systematic intellectual inquiry is to
reactivate that knowledge, which includes both formulating those principles, including
definitions of the basic things, and drawing consequences from them. This program is
admittedly hinted at, rather than made explicit, in the Meno itself. It is displayed more
explicitly in other dialogues, especially the Phaedo and the Republic.13

2. Phaedo

The thesis that the objects of recollection are basic principles, including basic entities,
which was suggested for the Meno by considerations of economy and fit with the text, is
explicitly confirmed by the text of the Phaedo (72e–77a). Forms (see ch. 8), which are not
explicitly mentioned in the Meno, are central to the epistemology and metaphysics of
the Phaedo, and they fill two gaps in the schematic theory sketched in the Meno, first as
objects of recollection and second as the causes, reasoning about which transforms true
belief into knowledge. Recollection of Forms is presupposed by the ability to give philo-
sophical accounts of properties such as equality; we are prompted to give such accounts
by experience of instances of them, and that experience prompts us to think of the prop-
erties as something over and above the instances themselves.14 The instances, then,

Fine responds by tightening the requirements for justification to include the grasp of what makes
one’s belief true as a necessary condition for justification. As she points out (“Knowledge and True
Belief,” 64, 78), the question whether the Meno provides an account of knowledge as justified true belief
then turns on the question (still disputed in contemporary discussions) of how demanding the
standard for justification is.
13 The most recent comprehensive discussion of the Meno is D. Scott, Plato’s Meno (Cambridge,
2006). For discussion see G. Fine, “Enquiry and Discovery: A Discussion of Dominic Scott, Plato’s
Meno”, Oxford Studies in Ancient Philosophy 32 (2007), 331–67 and L. Brown, “Review of Scott, Plato’s
Meno”, Philosophical Review 117 (2008), 468–71.
G. Fine, The Possibility of Inquiry (Oxford, 2014) presents in part I a meticulous examination of
Meno’s paradox and the theory of recollection in the Meno. Part II discusses the treatment of the
paradox by later philosophers from Aristotle to the Sceptics and Epicureans.
14 I accept the interpretation of recollection in the Phaedo proposed by D. Scott, Recollection and
Experience (Cambridge, 1995), part 1 (condensed version reprinted in Fine, Plato 1, 93–124), according
to which recollection explains the ability to give theoretical accounts of Forms. The more traditional
view that what recollection explains is ordinary concept formation is maintained by, among others,
J. L. Ackrill, “Anamnesis in the Phaedo: Remarks on 73c–75c,” in E. N. Lee, A. P. D. Mourelatos, and
 Plato’s Epistemology   439

remind us of something distinct from them, as a picture may remind us of its subject,
and what we are reminded of in each case is the appropriate Form. That confirms the
more economical interpretation proposed for the Meno. The explanation of the slave’s
solving the puzzle is his ultimate recollection not of that very solution but of the square
and the diagonal to which he was prompted by Socrates’ rough representation of them
and from which he was able to work out the solution itself.
For a case where what is sought is itself the account of some Form, the doctrine that
all nature is akin allows us to see how such an account may be worked out. Since Forms
are systematically connected with one another, recollection of the properties of one may
lead by inference to recollection of the properties of another. That Forms are the causes
of things is central to the Phaedo’s account of explanation, where the first stage in the
explanation of anything’s being F is that it shares in the Form of F. Admittedly, that is
only the first stage, which Socrates describes as a “safe and ignorant” answer (105c1); a
more subtle answer explains the thing’s being F via the presence of an entity or the
instantiation of a property such that anything in which that entity is present, or anything
instantiating the property, necessarily is F―for example, whatever contains fire is neces-
sarily hot (105c2–6). The explanatory entities and properties are of various types, and
while some may themselves be conceived as Forms, not all can be; fire and snow are
perceptible stuffs that come to be and cease to be, not changeless and eternal Forms.15
It is then obscure how the theory in which they play a central role is supposed to be
radically superior to the empirically based theories of the physicists of the sixth and
fifth centuries.
I have two suggestions on this point. First, Socrates’ principal objection to his prede-
cessors is their neglect of teleological explanation, and though the theory sketched in
the Phaedo is not explicitly teleological, it is unlikely that Plato had simply abandoned
that ideal. Second, the Timaeus, which is Plato’s fullest sketch of a theory of the physical
world,16 (1) depicts the world as teleologically designed, specifically as the best material

R. M. Rorty, eds., Exegesis and Argument: Studies in Greek Philosophy Presented to Gregory Vlastos
(Assen, 1973), 177–95, reprinted in Ackrill, Essays on Plato and Aristotle (Oxford, 1997), 13–32, and
Bostock, Phaedo, 66–69. Scott gives a lucid account of the controversy. See also A. Silverman, “Plato’s
Middle Period Metaphysics and Epistemology” [“Plato’s Middle Period”], The Stanford Encyclopedia of
Philosophy (Fall 2014 Edition), sect. 11.
15 For fuller discussion, see C. C. W. Taylor, ”Forms as Causes in the Phaedo,” Mind 78 (1969), 45–59;
Bostock, Phaedo, ch. 7; and Fine, Knowledge, ch. 14, “Forms as Causes: Plato and Aristotle.”
16 At Timaeus 27d–28a and 51d–52a, Plato asserts that there can be no knowledge, but only belief,
about the physical world since knowledge requires stable objects, whereas the physical world is in a
state of systematic instability. Hence the account that he proceeds to give of it is only a “likely story”
(eikota muthon, 29d2). But that should not be taken as an expression of epistemic despair about the
physical world. I take it that the teleologically grounded mathematical physics is intended to be,
given the pervasive instability of matter, the closest approximation to knowledge of which that
subject matter is capable. Mathematical and evaluative Forms are universal principles of
intelligibility, which, instantiated in various subject matters, make that subject matter knowable or
the nearest approach thereto. In “The Philosophical Economy of Plato’s Psychology: Rationality and
Common Concepts in the Timaeus,” in Frede and Striker, Rationality, 29–58, D. Frede gives a
persuasive account of how those universal principles permeate the flux of sensible phenomena in the
440   C. C. W. Taylor

approximation to the Form of the Living Being; and (2) gives a basic explanatory role
to the geometrical properties of the fundamental particles of matter. Physical stuffs
and their properties thus fit into an overarching theory that is both teleological and
mathematical. Ultimately, matter behaves the way it does because it instantiates mathe-
matical structure, and mathematical structure is as it is because that is the best way for
it to be. Mathematical and evaluative Forms are thus the ultimate aitiai, and it is by
reasoning about them that we achieve the systematic understanding of reality that
constitutes knowledge.

3. Republic

That systematic understanding is further elucidated in the two principal treatments of

knowledge in the Republic, in book V and in books VI–VII. In the former passage,
Socrates, in defending his claim that the ideal state can come into existence only if it is
ruled by philosophers (473c–e), argues that only the person with knowledge of the
Forms is entitled to be called a philosophos (lover of wisdom). The first argument for
this conclusion (475e–476d) presupposes the existence of the Forms, but the second
(476d–480a), which is designed to convince someone who does not antecedently accept
the existence of Forms, argues from the generally accepted premise that knowledge is a
grasp of what is to the conclusion that only the person who knows the Forms grasps
what is. The main rival, the person who lacks knowledge of the Forms, but is restricted
to acquaintance with sensible instances of them, is thereby confined to what is intermediate
between being and not-being, and hence to a grade of cognition lower than knowledge―
namely, belief.
Both the premise and the argument itself are highly problematic.
The premise is that knowledge is a grasp of what is. Someone who knows ( gignōskei)
knows something (rather than nothing), and what he knows is (476e7–477a1). The range
of uses of the Greek verb “to be” admits three interpretations of “what he knows is”:

Timaeus, allowing the achievement, if not of knowledge of phenomena at least of true and reliable
beliefs about them (37b4–8).
In different contexts, Plato expresses mutually inconsistent views about whether there is any
knowledge of the sensible world. It is recognized in the Meno (see earlier in this chapter, though note
the reservation expressed in note 11 of this chapter) and in the Theaetetus (see later in the chapter).
In the Republic, it is assumed that the philosopher-rulers know particular truths about good and bad
(520c). The Phaedo seems more optimistic about the possibility of knowledge of the sensible world than
the Timaeus; Socrates is dissatisfied with the theories of his predecessors, not because they were
attempting something impossible in principle but because they failed to give the right kind of
explanation of physical events, and he appears to envisage that the theory of Forms will ultimately
make good that deficiency. And even in the Timaeus, though knowledge is apparently impossible, true
and reliable belief is attainable (see previous discussion).
 Plato’s Epistemology   441

1. What he knows is true.

2. What he knows is something (e.g., is beautiful).
3. What he knows exists.17

All three are appropriate marks of knowledge: if someone knows that p, it must be
true that p, and if someone knows something (e.g., knows Socrates), then what he knows
must be something (e.g., be a man) and must exist. This elucidation indicates that the
three interpretations apply most readily to different types of knowledge: on the one hand,
(1) is a characteristic of propositional knowledge, where what is known is a proposition,
capable of truth and falsehood, and linguistically represented by a sentence, such as
“Plato knows that Socrates is wise.” Interpretations (2) and (3), on the other hand, apply
primarily to what is traditionally called “knowledge by acquaintance,” or familiarity with
some object. But (a) neither here nor elsewhere does Plato show a firm grasp of that
distinction, and (b) in this discussion, he is chiefly concerned with cases such as that of
“knowing beauty” (i.e., knowing what beauty is), where the distinction becomes blurred.
Someone who knows beauty may be conceived as being familiar with something that is
such and such and that exists, and as being ipso facto aware of the true proposition that
beauty is such and such. We should not expect this discussion to focus on truth as a
mark of knowledge, then, if such a focus is assumed to presuppose a sharp distinction of
truth from the other marks.
Plato’s thought is not well represented by interpretation (3), “what is known must
exist.” He goes on immediately (477a2–7) to describe things as being more or less, and
some things as “such as to be and not to be” and hence intermediate between “being
unqualifiedly” and “not being in any way.” The notion of degrees of existence is not only
unintelligible in itself; nothing suggests that Plato accepted it. “Degrees of being” in
479a–d fit readily with interpretation (2), being such and such, and also with (1), being true,
but not at all with (3), existence. It is best, then, to assume that Plato’s argument is to be
interpreted in terms of (1) and (2), to the exclusion of (3), while bearing in mind that
(1) and (2) are unlikely to be sharply distinguished from one another.
On this undifferentiated interpretation, the argument proceeds fairly smoothly. At
477a–b, Plato correlates degrees of being with degrees of knowability; what totally or
unqualifiedly is is totally or unqualifiedly knowable, what is not in any way is totally
unknowable. Anything that both is and is not is in between the totally knowable and the
totally unknowable and is the object of a mental state in between complete knowledge
and total ignorance or error, if there is such a state.
This argument may be understood either in terms of truth (Elucidation A) or in terms
of being something (Elucidation B). For the reasons given previously, I incline to think
that Plato does not distinguish between the two elucidations.

17 So G. Fine, “Knowledge and Belief in Republic V,” Achiv für Geschichte der Philosophie 60 (1978),
121–39, reprinted as Fine, Knowledge, ch. 3. (The passage cited occurs at 69 in the latter volume.) My
discussion of Rep. V owes a great deal to hers, though she distinguishes senses (1) and (2) more sharply
than I do.
442   C. C. W. Taylor

Elucidation A:
Any proposition that is totally or unqualifiedly true (e.g., “2 + 2 = 4”) is capable of
being known without qualification; any that is totally or unqualifiedly false (e.g.,
“2 + 2 = 5”) expresses nothing but ignorance or error. But any proposition that is
sometimes (in some contexts) true and sometimes false, (such as “Englishmen are
phlegmatic,”) is an expression neither of unqualified knowledge (since it is not true
without qualification that Englishmen are phlegmatic, it cannot be known without
qualification) nor of total ignorance or error.
Elucidation B:
Something that is totally or unqualifiedly such and such (e.g., “Cruelty is bad”) is
capable of being known as such without qualification; something that is not such and
such at all can be said to be such only in error (e.g., “Cruelty is good”). But suppose
we have something that is such and such qualifiedly, for example, “Swimming is
good for you.” Someone who accepts that without qualification does not have
knowledge but is not totally wrong, either.

Plato’s strategy is to try to show that the philotheamōn (the person who is acquainted only
with sensible instances of Forms) cannot escape from that situation. The only general
beliefs available to him are characterized as much by falsehood as by truth. That claim is
not distinguished by Plato from the claim that the objects of the philotheamōn’s general
beliefs are characterized by not being F (e.g., beautiful) as much as by being F.
Plato has already assumed that distinct capacities are directed onto (epi) distinct
objects; belief is a distinct capacity from knowledge (477b5), so belief is directed onto one
thing and knowledge onto another (b7–8). This assumption is spelled out at 477b11–478a4.
At 477c6–d5, Plato’s Principle of Differentiation of Capacities is stated: capacities are
differentiated by two factors, their object and their effect (i.e., what possession of the
capacity enables its possessor to do). Capacity A and capacity B are one and the same
capacity if they have the same object and the same effect, and they are distinct capacities
if they have different objects and different effects. It is apparently assumed that objects and
effects cannot vary independently of one another: it is impossible that the same object
should be subject to distinct effects or the same effect applied to distinct objects. Plato,
then, must assume that there is a necessary connection between objects and effects. That
result would be achieved if the object were itself specified in terms of the effect, as what
is susceptible of the effect―for example, the capacity to see has as its object the visible,
and the capacity to touch has its object the tangible. But that would merely yield the trivial
result that the concept of the knowable is distinct from the concept of the believable, which
is compatible with its being the case that the application of the two concepts is identical.
Plato is not aiming at the trivial connection between knowledge and the knowable but at
the connection, necessary but nontrivial, between knowledge and the character of what
is known―namely, that knowledge is of what is true and/or real, whereas belief lacks
those necessary connections with truth and reality.
He is correct to distinguish the two concepts in that way but wrong to try to derive
that distinction from his general Principle of Differentiation of Capacities. That general
principle is either trivial or false. Heating is a distinct effect from cooling, but there is no
 Plato’s Epistemology   443

nontrivial sense in which the object of the one effect, the heatable, is distinct from the
object of the other, the coolable.
By 478d, belief has been established as a capacity intermediate between knowledge
and error or ignorance. Objects intermediate between being and nonbeing will be
objects appropriate to that intermediate capacity (478d5–9). Arguments familiar from,
for example, Rep. I, 331c and Hippias Major 289a–c show that the instances of Forms,
which are all that the philotheamōn is familiar with, are so characterized. These instances
include both kinds (e.g., paying back what you borrowed) and particulars (e.g., a beautiful
woman). Beliefs about these things, such as “Justice is paying your debts” and “Helen is
beautiful” will be neither unqualifiedly false nor unqualifiedly true but sometimes true
and sometimes false; in Plato’s words, “the many beliefs (nomima) held by the many
about beauty and the rest roll about, as it were, between not being and being without
qualification” (479d3–5). So those who are familiar with nothing beyond the instances
of Forms, lacking any grasp of the Forms themselves, must recognize that they have
nothing more than belief, and therefore accept the title “lovers of belief ” (philodoxoi)
instead of that of philosophoi that they had attempted to usurp (479e–480a).
But the philotheamōn can have unqualifiedly true particular beliefs―for example,
“In these particular circumstances, paying back this particular debt was just.” As such
beliefs will simply be true, they will be epi tōi onti (literally, “onto what is”; that is, they
will latch onto what is so). So why will they not amount to knowledge?
To defend Plato against that objection, we must return to the Meno’s distinction of
knowledge from true belief, by the criterion that, in order to count as knowledge, true
beliefs have to be grounded in “reasoning about the cause”―that is, in a grasp of the
grounds of their truth. In order to have knowledge, the philotheamōn must understand
what makes his beliefs true, for example, why this particular repayment was unquali-
fiedly just. And in order to do that, he must have a systematic grasp of the standards that
govern the characterization of types and particular instances: that is to say, he must be
familiar with the Forms, as well as the instances. Knowledge, even of particular cases,
must be grounded in understanding of why things are as they are, and that understand-
ing requires knowledge of Forms. Hence the philotheamōn, who has no knowledge of
the Forms, lacks the understanding that is necessary for knowledge.
This suggestion does not claim to represent Plato’s actual argument but to reply to an
objection on the part of the philotheamōn. It has the advantage of assimilating Rep. V to the
Meno and thereby removing an apparent difficulty: that whereas in the Meno (and Tht.
201b) knowledge and belief can have the same objects, in Rep. V they have, by the Principle
of the Differentiation of Capacities, different objects. In fact, the two positions are compat-
ible; the thesis in the Meno and Theaetetus concerns particular items of knowledge, while
that of Rep. V concerns the objects of the capacities as such. Plato does not claim in Rep. V
that there can be no knowledge that is not knowledge of Forms (which would deprive phi-
losopher-rulers of knowledge of events in the sensible world). He does (implicitly) claim
that there can be no knowledge of anything that is not grounded in knowledge of Forms.18

18 However, this does imply that perceptual knowledge such as knowledge of the road to Larisa
either is not knowledge (or, at least, not knowledge strictly speaking) or is somehow grounded in
444   C. C. W. Taylor

Teleology and mathematics are central to the discussion in books VI–VII, whose
context is the description of the advanced education of the philosopher-rulers. The
“greatest subject” of their education is the Form of the Good, since their grasp of what is
beneficial in the political sphere depends on their understanding of goodness as such
(505a). Since Socrates does not know what goodness (506c) is, he cannot give a scientific
account of it, but he states his beliefs in the form of the famous images of the Sun
(506e–509d), the Divided Line (509d–511e), and the Cave (514a–517a). The central point
of the first of these is that just as the sun is both the ultimate generative force and the
primary source of illumination in the visible world, so the Form of the Good is primary,
both epistemologically and ontologically, in the intelligible world of the Forms (509a–b).
That is to say, the other Forms exist, and are what they are, because it is best that they
should be, and understanding what any Form is involves understanding why it is what it
is―that is, understanding how that is the best way for it to be.
This immediately raises the difficulty that teleological explanation requires that what
is actual is the best of a range of alternative possibilities, whereas the Forms exist, and are
what they are, necessarily. I suggest that we can best approach Plato’s meaning if we take
it that his starting point is the ordinary conception of goodness as consisting in order
and proportion, as illustrated in Gorgias 504a–b, where goodness in a range of things,
from a house to the soul, consists in order and arrangement of parts, whereas badness
consists in disorder. To be good, then, is to manifest rationally satisfactory order; so to
say that the intelligible Forms are as they are because that is best is to say that they are
what they are because that system is maximally intelligible.
It seems fairly clear that Plato believed that order and proportion were ultimately to
be understood mathematically. Hence the curriculum that is to lead the philosophers
to the systematic study of the Forms is mathematical, not merely because mathematics
leads the mind from reliance on the senses to abstract thought (524b) but because the
grasp of the basic principles common to the various mathematical sciences is useful in
leading to the search for the beautiful and the good (531c–d). To understand goodness is
to understand order, and fundamental to the understanding of order is the understanding
of its mathematical basis; hence the understanding of goodness is to be sought via the
basic principles of mathematics.19 Some confirmation of this suggestion is provided by
Aristotle’s evidence of Plato’s lecture on the Good, which was all to do with mathematics
and which culminated (on the most likely interpretation of the text) in identifying the
Good with Unity.20

knowledge of the Forms. Does Plato perhaps think that you cannot know the road unless you know
what a road is, and that knowing the latter is (or involves) knowing the Form of the Road?
19 For a fuller exposition and defense of this view, see M. Burnyeat, “Plato on Why Mathematics Is
Good for the Soul,” Proceedings of the British Academy 103 (2000), 1–81.
20 The evidence comes from the Elements of Harmony of Aristoxenus (a pupil of Aristotle), II.30–31:
This, as Aristotle was always saying, was the experience of most of those who heard Plato’s
lecture On the Good. Each of them attended on the assumption that he would hear about one of
the recognised human goods—such as wealth. health, strength, and in general some marvellous
happiness. When Plato’s lectures turned out to be about mathematics—numbers, geometry,
 Plato’s Epistemology   445

This suggestion is open to some obvious objections. According to the simile of the
Sun, the Form of the Good is epistemologically primary; according to the image of
the Divided Line, the principles of the mathematical sciences are themselves fully
intelligible only when they are derived from the “unhypothetical principle of everything”
(510c–511d), which, in context, must be the Form of the Good. Yet, on this suggestion,
the Form of the Good is itself elucidated as a fundamental principle of mathematics,
specifically the Form of Unity. The difficulty arising from the Divided Line is compara-
tively superficial; the principles of the individual mathematical sciences, when taken in
isolation from one another, have the status of mere hypotheses. Only when they are
derived from a single unhypothetical principle (i.e., when they are tied down by reasoning
concerning their cause) are they themselves known, and are hence the grounds of the
knowledge of what is derived from them.
The other difficulty is deeper. Ex hypothesi goodness was the basic explanatory
concept, but if understanding what goodness is requires that one explain what it is in
terms of other concepts, those concepts are now more basic than goodness. We have
here an instance of the classic problem of the hierarchical structure of knowledge.
If knowledge of X is founded on knowledge of Y, and that, in turn, on knowledge of Z,
then either we have an infinite regress of knowledge or we have some foundations of
knowledge, knowledge of which is grounded on nothing but themselves. Plato’s insistence
that we must be able to give an account of what we know seems to rule out self-evident
foundations of knowledge; he asserts the necessity of a logos in many passages,21 notably
in Rep. 534b–c, where the philosopher’s task is that of giving the logos of each of the Forms
and his ultimate aim is that of differentiating the Form of the Good (tōi logōi22) from the
other Forms. In contrast, the account at the end of book VI of the priority of the Good―
and the simile of the Sun, in particular―strongly suggest traditional foundationalism.
Just as the sun makes everything else visible by its own light, and is itself visible by that

astronomy—and to crown all about the thesis that the good is one, it seemed to them, I fancy,
something quite paradoxical, and so some people despised the whole thing, while others
criticised it. (Translation from J. Barnes, The Complete Works of Aristotle: The Revised Oxford
Translation, 2 vols. (Princeton, N.J., 1984), 2397.)
The crucial phrase, translated above “the good is one,” is agathon estin hen, which would most naturally
be rendered “there is one good” (presumably as opposed to many). The translation given above
assumes the emendation tagathon estin hen, “the good is one” (presumably, again, as opposed to many
different things). But it is hard to see why the thesis that there is a single supreme good should have
seemed so paradoxical as to provoke the reactions mentioned; it is clear from the context that the
disappointed audience was expecting to hear that one of the recognized goods was the (i.e., the
supreme) good. What was so outrageous must have been not the claim that the good was one as
opposed to many but the account of what it was, and that must have been such as to require the
mathematical build-up described. I propose that we should adopt the reading tagathon estin hen,
understanding that as “the Good is the One,” the article before hen being omitted (as is standard in
Greek) when an expression with the definite article is the complement of the verb einai (to be).
21 Rep. 510c, 531e, 533b–c; Phaedo 76b; Symp. 202a; Tht. 202c; Tim. 51e; Laws 966b, 967e.
22 The phrase tōi logōi may be translated either “by (its) definition” or “by reasoning.” The translation
does not affect the doctrine; Forms have to be distinguished from one another by reasoning (since
reason alone grasps them), but what reasoning does is to reach accounts of them that differentiate one
from another.
446   C. C. W. Taylor

same light, so the Good makes the other Forms intelligible and, we should expect from
the analogy, is itself intelligible in and of itself.
Holistic (alternatively, coherentist) pictures of knowledge offer an escape from this
dilemma.23 To give an account of a concept is not to explain it in terms of anything more
basic but to locate it in a coherent structure of concepts, and specifically to show the
explanatory role that each concept plays within that structure. The suggestion that Plato
identifies goodness with unity can be seen as instantiating that model. The basic explan-
atory role of goodness in a teleological scheme of explanation is adapted to unity, in that
goodness is order, harmony, symmetry, and so on, those features are understood math-
ematically, the mathematical understanding of them is grounded in basic mathematical
principles, and, given the holistic model, the account of those principles consists in
showing their contribution to the system as a whole.
This discussion of Rep. VI–VII is doubly speculative, first in suggesting that Plato
intends the nature of goodness to be understood mathematically, and then in raising the
possibility that the type of account of it that he intends is a holistic one. Both suggestions
are recommended by the extent to which they achieve plausibility; neither can claim
direct textual confirmation. The problem that the second suggestion attempts to meet is
also prominent in the Theaetetus, the only dialogue of Plato’s to be devoted to the topic
of knowledge. We shall therefore return to it in the context of that dialogue.

4. Theaetetus

The topic of the dialogue is the question “What is knowledge?” Three answers are proposed,
and examined in turn:

1. Knowledge is perception (aisthēsis).

2. Knowledge is true belief.
3. Knowledge is true belief with an account (logos).

Each answer is rejected, and the dialogue ends aporetically. The discussion of the first
suggestion, which is considerably longer than the other two combined, is largely devoted
to a complex and sophisticated treatment of two theses, which Socrates argues to be
logically connected with the proposed account of knowledge as perception―namely,
Protagoras’s thesis that things are as they appear to each individual, and a thesis derived
from Heraclitus that everything is in a state of total flux. I shall not discuss the treatment
of these theses (for which see chapter 17 of this volume) but shall confine myself to the
direct discussion of the proposed account of knowledge as perception. This is undertaken

23 As suggested by G. Fine, “Knowledge and Belief in Republic V–VII,” in S. Everson, ed.,

Companions to Ancient Thought 1: Epistemology (Cambridge, 1990), 85–115, reprinted as ch. 4 of Fine,
Knowledge, and in Fine, Plato 1, 215–46. See also A. Silverman “Plato’s Middle Period”, sect. 15.
 Plato’s Epistemology   447

in a brief section (184b–186a) whose central point is a distinction between, on the one
hand, properties apprehended by the individual bodily senses (colors by sight, acoustic
properties by hearing, flavors by taste, etc.) and on the other properties (beauty, ugliness,
goodness, badness, being, sameness and difference, likeness and unlikeness, number, etc.),
which are not specific to any individual sense.24
Socrates says (185a6–7) that the soul discerns or apprehends (episkopein) the sensible
properties “by means of the powers of the body,” and the non-sensible properties “itself
by means of itself ” (i.e., solely by its own power). The appplication of those concepts is
not the work of any individual sense, or of the senses collectively, but of the integrating
capacity of the mind, which unifies the data of the several senses into a single coherent
diachronic picture (186a10–b1), and which also applies evaluations, such as beautiful
and ugly, good and bad (a8). Perception, identified as the apprehension of the sensible
properties, is thus distinguished from judgment, which is the work of the mind, and
since being is one of the properties that belong to judgment, it is judgment, not percep-
tion, that grasps truth, since truth belongs to being (186b–c).25 And since knowledge
implies truth (c9–10), the conclusion is reached that “Knowledge is not in our experi-
ences, but in our reasoning about them; for it is here (i.e., in reasoning) that it is possible,
it seems, to attain being and truth, but it is impossible there (i.e., in experiences)” (d2–5).
At first, this seems very straightforward: knowledge is propositional, knowledge
that p. And propositions are the objects of judgment: judgment is always judgment that
p. Perception, by contrast, lacks propositional content. Hence, knowledge cannot be
perception. If that is the argument, it is a bad one, since the premise that perception
lacks propositional content is just false. Some perception at least has propositional
content: for example, I can see that the table has already been laid. Even if we restrict
ourselves to perception of the proper objects of the senses, one sees that the color sample
is green. Moreover, Socrates himself says just that, when he points out that, asked to
examine (skepsasthai) whether something is salty or not, one does so by taste (185b7–c2).
A great deal of perception, then, is perception that something is the case, and it is plausi-
ble that that was part of what Theaetetus had in mind in his original suggestion (151e1–3)
that “someone who knows something perceives what he knows, and as it now seems to
me, knowledge is nothing other than perception.” And certainly, when Socrates imme-
diately equates Theaetetus’ suggestion with Protagoras’s doctrine that things are as they
seem to each individual, that seeming has propositional content; the wind’s seeming
(feeling) cold to me is its seeming (feeling) to me that the wind is cold. So if Socrates’

24 On this section, see, in addition to the works cited later on in the chapter, M. Burnyeat, “Plato on
the Grammar of Perceiving,” Classical Quarterly, n.s. 26 (1976), 29–51, and J. M. Cooper, “Plato on
Sense Perception and Knowledge: Theaetetus 184 to 186,” Phronesis 15 (1970), 123–46, reprinted in Fine,
Plato 1, 355–76, and in Cooper, Knowledge, Nature and the Good: Essays on Ancient Philosophy
(Princeton, 2004), 43–64.
25 The crucial sentence is 186c7, “Is it possible for what cannot even attain to being to attain to
truth?,” to which Theaetetus answers, “It is impossible.” The connection between being and truth is
presumably made via the idiomatic use of to on and ta onta in the sense of “what is true”; the ordinary
Greek for “speak the truth” is to on (or ta onta) legein.
448   C. C. W. Taylor

final argument is the straightforward one set out previously, it assumes a conception
of perception that is not the one intended by the original proposal. In that argument,
propositional content is assigned exclusively to judgment, leaving perception to be
construed as contentless, that is, as the reception of raw data whose interpretation is the
work of a distinct faculty. But no one could conceivably maintain that knowledge is
perception thus conceived.
That may, however, be the argument. Socrates’ claim would then be that, strictly
speaking, perception is nothing more than the contentless reception of stimuli26 and talk
of perceiving that p is an illegitimate conflation of perception itself with judgment
consequent on perception. The proponents of the thesis that knowledge is perception
would then have misdescribed their own position. There are, however, some indications
in the text that Socrates’ argument may be different. First, in the statement of his
conclusion quoted previously (186d2–5), he says not that knowledge is in judgment
about our experiences (which would presumably be doxa) but that it is in reasoning
(sullogismos) about them. That suggests that knowledge is to be found not in the class
of conceptualized judgments about perception but in some more restricted class of
judgments arrived at by reasoning.27 That is supported by what immediately precedes
(186b–c). The soul perceives by touch the hardness of what is hard and the softness of
what is soft, but certain other properties it attempts itself to judge by examination and
comparison; these properties are “their being and what they are (or that they are)28 and
their opposition to one another and again the being of their opposition.”
By contrast with perception, which is innate in humans and animals, “reasonings (or
“calculations,” analogismata) about them with reference to their being and their utility”
are arrived at through a long and arduous process of education. It is hard to see that it
takes such a process to arrive at the judgment, concerning something hard, that it is
hard, but easier to see that it might take such a process to be able to understand what
hardness is, that hardness is not just different from but opposite to softness, and again
what oppositeness is. For these tasks, one needs not just experience of hardness and
softness but a theory of the nature of those properties and of the properties of those
properties (such as oppositeness). The references to evaluation may make the point that
evaluation, like understanding what things are, requires not just experience but theory;
one cannot determine whether something is good or bad, or beautiful or ugly, just by
experiencing it, but rather needs to understand the appropriate standards of evaluation.29

26 Which is presumably all that animals and infants experience (186b11–c1).

27 The reminiscence of the Meno’s “by reasoning concerning the cause” (aitias logismōi) is highly
suggestive.
28 The Greek is hoti eston, which may be translated either as “what they (both) are” or “that they
(both) are,” “both” referring to the experienced instances of hardness and softness. For the reasons
given in what immediately follows, I think that “what they are” is more likely. For the opposite view,
see D. Bostock, Plato’s Theaetetus [Theaetetus] (Oxford, 1988), 139–40. Given that translation “being”
(ousian) is best understood in the same way: as equivalent to “nature,” the kai connecting the two
expressions being epexegetic, “their nature, i.e. what they are.”
29 Experience is necessary, but not sufficient, for correct evaluation; to the untutored ear a piece by
Stockhausen may sound ugly, but it takes understanding of the genre and its conventions to determine
whether it is really beautiful or ugly.
 Plato’s Epistemology   449

It is, then, at least possible that the conclusion of the first main section of the dialogue
is that knowledge is not perception, not because knowledge is always propositional,
whereas perception lacks propositional content, but because knowledge is primarily
knowledge of what things are, whereas perception is never sufficient to reveal what
things are.30 That suggestion is not without its difficulties,31 but it is worth keeping in
mind when we turn to the remaining sections.
The second proposed definition is that knowledge is true belief. Since perception can-
not be knowledge, the latter must be found in the activity of the soul “by itself ” (see the
distinction previously described), which is said to be belief or judgment (187a), and
since there can be false belief, knowledge cannot be belief as such but must be true belief
(187b). This proposal is threatened by the claim that false belief is impossible (in which
case, knowledge would collapse into belief), and the bulk of this section (to 201a) is
devoted to discussion of how false belief is possible (for details, see chapter 17 of this
volume). The substantive suggestion is dealt with only briefly, being refuted (201a–c) by
the distinction between an eyewitness’s knowledge of some event, say an assault, and
the true beliefs that a member of the jury has about that event. The latter cannot have
knowledge of what “only the person who saw” can know (b7–8). Socrates describes
the jury as having only a short time to decide the matter and as being persuaded, but not
“taught” or “instructed” by the litigants (201a–b), which conveys the suggestion that
they are unfairly manipulated rather than being presented with evidence sufficient to
reach a proper verdict, but the insistence that only the eyewitness can know what
occurred clearly implies that testimony, however compelling, and however fairly pre-
sented, can never produce that knowledge. We have returned to the distinction between
the person who knows the road to Larisa from experience and the person who has true
second-hand beliefs about it. Knowledge by experience is admitted without qualification
as knowledge, and there is no suggestion that the eyewitness is better placed epistemi-
cally than the jury member because the former has some “reasoning about the cause” of
the event which the latter lacks.32
It is clear that the eyewitness’s knowledge is knowledge that such and such occurred.
But it is less clear how sharply Plato distinguishes that from knowledge of the event. Just
as the person who knows the road knows various things about it―for example, that it
passes to the north of such and such a hill―whereas the person who merely “has the
road in mind” merely believes those things, so the person who knows the event knows
various things about it, whereas the person who relies on testimony merely believes those
things. The suggestion that knowledge that and knowledge of things are not seen by Plato
as two distinct kinds of knowledge is supported by the fact that the dialogue passes
immediately to the final suggestion, that knowledge is true belief with an account (logos),
which is primarily a discussion of knowledge of things, in the sense of knowledge of
what things are. The basic idea is that knowing what something is is having a true belief

30 This suggestion was originally made by J. McDowell, Plato, Theaetetus (Oxford, 1973), 188–93.
31 See, for example, Bostock, Theaetetus 140–42.
32 On the jury passage, see M. Burnyeat and J. Barnes, “Socrates and the Jury,” Proceedings of the
Aristotelian Society, Supplementary Volume 54 (1980), 173–91 (Burnyeat) and 193–206 (Barnes).
450   C. C. W. Taylor

about that thing together with an account of it. The kinds of account discussed are the
enumeration of the elements of a thing, illustrated by the example of the analysis of a
syllable into its component letters, the statement of one’s true belief, and a description of
the thing sufficient to distinguish it from everything else.
The first suggestion is rejected on the strength of the regress difficulty discussed ear-
lier; if knowledge requires an enumeration of the elements of the things known, then the
elements themselves must be unknown. But it is impossible that unknown elements can
be the basis of knowledge of what they compose. On the contrary, the elements must be
better known than the things composed of them (206b), but that is impossible on this
compositional model. Stating one’s true belief is immediately rejected on the ground
that since everyone with a true belief is able to state it, this proposal merely restates
the previously rejected suggestion that knowledge is true belief (206d–e). Finally, the
suggestion that knowledge of something is true belief of or about that thing together
with a distinguishing mark of that thing is rejected on two grounds. First, true belief
about anything requires that one already possesses a distinguishing mark of it (otherwise
one’s belief would not be about it specifically); hence, once again, knowledge adds
nothing to true belief. Second, if one responds to the first objection by requiring knowledge
of the distinguishing mark, the proposed account of knowledge is circular (208c–210a).
The dialogue thus ends inconclusively.33
This outcome raises the question whether the aporia reflects genuine uncertainty on
Plato’s part, or whether his intention is to suggest some positive answer to the original
question. Specifically, is the reader to infer that, given some other sense of “account,”
knowledge will indeed prove to be true belief with an account? An obvious suggestion is
that we should revive the Meno’s proposal, defining knowledge as true belief with
reasoning concerning the cause. But that proposal fits derivative, rather than basic,
knowledge. One has the kind of knowledge defined in the Meno’s terms when one has
some true belief together with understanding of what makes that belief true. But that
understanding is itself a sort of knowledge, and application of the Meno’s formula to it
raises the dilemma that we have already encountered: either it, too, has to be accompanied
by understanding of something else that makes it true, which leads to a regress, or there
are some beliefs that are true in virtue of nothing other than their own truth―in other
words are self-evident. If there is a distinction between basic and derivative knowledge,
and if we assume that Plato is looking for an account of the former, then that would need
to be an account of self-evidence; but “true belief with reasoning concerning the cause”
cannot be an account of self-evidence. An alternative is the suggestion that we have
already encountered: that the understanding of what makes any belief true is provided
by the whole conceptual structure into which that belief fits. The regress is halted by the
abandonment of the distinction between basic and derivative knowledge; “reasoning
concerning the cause” would then have to be construed as “elucidation of the conceptual
scheme to which the belief belongs.”

33 On the final section of the dialogue, see G. Fine, “Knowledge and Logos in the Theaetetus,”
Philosophical Review 88 (1979), 336–97, reprinted in Fine, Knowledge, as ch. 10, and Bostock,
Theaetetus, ch. 6.
 Plato’s Epistemology   451

This certainly has some affinities with some things said about knowledge in the
Phaedrus, Sophist, Statesman, and Philebus, all plausibly regarded as later than the
Theaetetus. Of these, the Phaedrus and Philebus describe,34 and the Sophist and Statesman
exemplify,35 a method of specifying what things are by a process of constructing defini-
tions per genus et differentiam. A genus is collected together from many different things,
and then successively divided into species and subspecies until indivisible species are
reached. But while that method provides systematic knowledge of the various species
and of their connections with their higher genera and therefore with each other, it does
not avoid the problem of basic knowledge. For knowledge of being an X as being an F
that is G presupposes that we know what being an F and being G is. It will not do to say
that being an F is being a member of that genus that is constituted by the species, F that is
G, F that is H, and so on. For that assumes that we know which species constitute a unity,
and it seems that, for that, we have to have some way of identifying the genus independ-
ently of the species. We cannot identify the genus as the genus that is collected from
the different things we started from, for that requires that we know what different
things to collect. If we begin by collecting indivisible species, then how can we know
which species to collect? And if we begin from individuals, then how shall we know which
individuals to collect?
A system of classification cannot by itself be adequate to provide knowledge of reality
but rather has to be supplemented by means of fixing the application of the classificatory
terms, whether by observation, or by taking as primitive some pre-theoretical catego-
ries, or in some other way. The Sophist gives some intriguing hints in this direction in the
suggestion that one of the principal tasks of philosophy is working out the conceptual
interrelations of what it calls the “Greatest Kinds”―that is, some of the most general and
abstract concepts: namely, Being, Sameness, Difference, Change, and Stability (251c–261b).
It may have been Plato’s view that a full specification of these interrelations will amount
to an account of what each of these kinds is and thus to identifications of these highest
genera, which will then be divisible via the method of division. But that is speculation;
there is nothing in the text to connect the discussion of the Greatest Kinds with the
method of collection and division.

5. Conclusion

The overall picture of Plato’s views on knowledge is not particularly tidy. While some
themes remain constant from his earliest dialogues throughout, there are a number of
important points on which he does not appear to have reached a fixed position. The
following are constant themes.

34 Phaedr. 265d–266c; Phil. 16b–18d.

35 Soph. 219a–232a, 264b–268d; Statesm. 258b–268d, 279a–311c.
452   C. C. W. Taylor

I. Knowledge is systematic. Over specific areas, such as mathematics and morality,

and even conceivably for reality as a whole, items of knowledge are systematically
interconnected, and it is the task of inquiry in those areas to reveal those connections.
II. For any department of knowledge (and, conceivably, for reality as a whole), the
primary knowledge is knowledge of what things are. Given the metaphysical theory
of Forms as the basic things that there are, Forms are the primary objects of inquiry.
Changes in (or uncertainties about) that theory are reflected in corresponding
changes in Plato’s views (or uncertainties) about knowledge.
III. Knowledge of what things are is achieved a priori, by critical reflection.
Empirical investigation has at best a secondary role in the achievement of knowledge
(see later on in this section).

Those themes may be summed up as the doctrine that the aim of inquiry is to achieve
systematic understanding of the intelligible principles of reality. While that remained
Plato’s constant ideal for philosophy, his conception of how, and how completely, it might
be achieved seems to have fluctuated, in various ways.

a. The Theory of Recollection expresses the view that the soul has been in perma-
nent possession of a total grasp of the principles of reality and that the task of
critical inquiry is to recover that grasp. But that theory is found only in the Meno,
Phaedo, and Phaedrus, and even in those dialogues it appears in different versions.
b. The Republic presents the ambitious ideal of a single all-embracing system, on the
model of a mathematical axiomatic system, founded on a single fundamental
principle, the nature of goodness. It is plausible that goodness was itself conceived
mathematically. No other dialogue gives that universal role to any single principle.
c. There is no single view of the status―or, indeed, the existence―of empirical
knowledge. In the Timaeus, Plato denies that knowledge of the sensible world is
possible but allows that there can be reliable belief about it. Knowledge of the
sensible world is recognized in numerous dialogues, but there is no uniform view
how it is achieved. In Rep. V, knowledge of the sensible world appears to be admit-
ted, provided that it is grounded in knowledge of Forms, and the same view is
indicated by the thesis in the Meno that knowledge requires reasoning concerning
the cause of one’s true beliefs. But in the Meno and Theaetetus, we find instances
of knowledge acquired by direct perception, where it is not clear how, or whether,
knowledge of Forms is presupposed. Equally, it is not clear how, or whether, such
items of knowledge are systematically connected to others. It may be that such
knowledge is thought of as knowledge of a secondary kind or as not, strictly
speaking, knowledge, but no distinction of that kind is explicitly drawn.
d. Plato asserts repeatedly that in order to know what something is one must be able
to give an account or definition of that thing. He is clearly aware of the difficulty
that that requirement leads to an infinite regress of accounts, but his response to
that difficulty is disputed. On some views, he modified the requirement to the
 Plato’s Epistemology   453

extent of recognizing some things, perhaps including the Form of the Good,
which were self-intelligible. On others, he extended the notion of an account to
include the system in which such alleged primitives have their place, so that
knowledge of the primitive elements and knowledge of what is derived from
them is mutually self-supporting.
There are traces of such views in some of the later dialogues, but they are not
explicitly related to the regress problem.
e. Some of the later dialogues exhibit definitions in genus-species hierarchies. The
method raises a number of questions, including how these hierarchies are sup-
posed to apply to the sensible world, and how the method is supposed to account
for knowledge of the summa genera and the infimae species. It is possible that the
former kind of knowledge is somehow grounded in the kind of investigation of
the interrelation of basic formal concepts conducted in the Sophist, but there is no
explicit connection in the texts between these two kinds of investigation.

Bibliography
Ackrill, J. L. “Anamnesis in the Phaedo: Remarks on 73c–75c,” in E. N. Lee, A. P. D. Mourelatos,
and R. M. Rorty, eds., Exegesis and Argument: Studies in Greek Philosophy Presented to
Gregory Vlastos (Assen, 1973), 177–95; reprinted in Ackrill, Essays on Plato and Aristotle
(Oxford, 1997), 13–32.
Barnes, J. The Complete Works of Aristotle: The Revised Oxford Translation, 2 vols. (Princeton,
N.J., 1984).
Benson, H. H. “The Priority of Definition and the Socratic Elenchus,” Oxford Studies in
Ancient Philosophy 8 (1990), 19–65; revised version in Benson, Socratic Wisdom (New York
and Oxford, 2000), ch. 6.
Bostock, D. Plato’s Phaedo (Oxford, 1986).
Bostock, D. Plato’s Theaetetus (Oxford, 1988).
Brown, L. “Review of Scott, Plato’s Meno,” Philosophical Review 117 (2008), 468–71.
Burnyeat, M. “Plato on the Grammar of Perceiving,” Classical Quarterly, n.s. 26 (1976), 29–51.
Burnyeat, M. “Plato on Why Mathematics Is Good for the Soul,” Proceedings of the British
Academy 103 (2000), 1–81.
Burnyeat, M., and Barnes, J. “Socrates and the Jury,” Proceedings of the Aristotelian Society,
Supplementary Volume 54 (1980), 173–91 (Burnyeat) and 193–206 (Barnes).
Charles, D. “Definitions in the Meno,” in V. Karasmanis, ed., Socrates 2400 Years since His Death
(Athens, 2004), 357–66; revised version, entitled “Types of Definition in the Meno,” in L. Judson
and V. Karasmanis, eds., Remembering Socrates: Philosophical Essays (Oxford, 2006), 110–28.
Cooper, J. M. “Plato on Sense Perception and Knowledge: Theaetetus 184 to 186,” Phronesis 15
(1970), 123–46; reprinted in G. Fine, ed., Plato 1: Metaphysics and Epistemology (Oxford,
1999), 355–76, and in Cooper, Knowledge, Nature, and the Good: Essays on Ancient Philosophy
(Princeton, N.J., 2004), 43–64.
Fine, G. “Knowledge and Belief in Republic V,” Archiv für Geschichte der Philosophie 60 (1978),
121–39; reprinted in Fine, Plato on Knowledge and Forms (Oxford, 2003), ch. 3.
454   C. C. W. Taylor

Fine, G. “Knowledge and Belief in Republic V–VII,” in S. Everson, ed., Companions to Ancient
Thought 1: Epistemology (Cambridge, 1990), 85–115; reprinted in Fine, Plato on Knowledge
and Forms (Oxford, 2003), ch. 4, and in Fine, ed., Plato 1: Metaphysics and Epistemology
(Oxford, 1999), 215–46.
Fine, G. “Knowledge and Logos in the Theaetetus,” Philosophical Review 88 (1979), 336–97;
reprinted in Fine, Plato on Knowledge and Forms (Oxford, 2003), ch. 10.
Fine, G. “Knowledge and True Belief in the Meno,” Oxford Studies in Ancient Philosophy 27
(2004), 41–81.
Fine, G. “Enquiry and Discovery: A Discussion of Dominic Scott, Plato’s Meno,” Oxford Studies
in Ancient Philosophy 32 (2007), 331–67.
Fine, G. “Does Socrates Claim to Know That He Knows Nothing?,” Oxford Studies in Ancient
Philosophy 35 (2008), 49–88.
Fine, G. Plato on Knowledge and Forms (Oxford, 2003).
Fine, G., ed. Plato 1: Metaphysics and Epistemology (Oxford, 1999).
Fine, G. The Possibility of Inquiry. Meno’s Paradox from Socrates to Sextus (Oxford, 2014).
Frede, D. “The Philosophical Economy of Plato’s Psychology: Rationality and Common
Concepts in the Timaeus,” in M. Frede and G. Striker, eds., Rationality in Greek Thought
(Oxford, 1996), 29–58.
Furley, D. “What Kind of Cause Is Aristotle’s Formal Cause?,” in M. Frede and G. Striker, eds.,
Rationality in Greek Thought (Oxford, 1996), 59–79.
Gettier, E. L., “Is Justified True Belief Knowledge?,” Analysis 23 (1963), 121–23.
Irwin, T. Plato’s Ethics (Oxford, 1995).
Irwin, T. Plato’s Moral Theory: The Early and Middle Dialogues (Oxford, 1977).
McDowell, J. Plato, Theaetetus (Oxford, 1973).
Penner, T. “The Unity of Virtue,” Philosophical Review, 82 (1973), 35–68; reprinted in
H. H. Benson, ed., Essays in the Philosophy of Socrates (New York, 1992), 162–84, and in G. Fine,
ed., Plato 2: Ethics, Politics, Religion, and the Soul (Oxford, 1999), 78–104.
Scott, D. Plato’s Meno (Cambridge, 2006).
Scott, D. Recollection and Experience (Cambridge, 1995).
Silverman, A., “Plato’s Middle Period Metaphysics and Epistemology,” The Stanford
Encyclopedia of Philosophy (Fall 2014 Edition), Edward N. Zalta (ed.), https://plato.stanford.
edu/archives/fall2014/entries/plato-metaphysics/.
Taylor, C. C. W. “Forms as Causes in the Phaedo,” Mind 78 (1969), 45–59.
Taylor, C. C. W. Socrates (Oxford, 1998).
Taylor, C. C. W. “Socratic Ethics,” in B. S. Gower and M. C. Stokes, eds., Socratic Questions
(London, 1992), 137–52.
Vlastos, G. “Reasons and Causes in the Phaedo,” Philosophical Review 78 (1969), 291–325;
reprinted in Vlastos, ed., Plato I: Metaphysics and Epistemology (Garden City, N.Y., 1971),
132–66, and in Vlastos, Platonic Studies, 2nd ed. (Princeton, N.J., 1981), 58–75.
Vlastos, G. “Socrates’ Disavowal of Knowledge,” in Vlastos, Socratic Studies, ed. M. Burnyeat
(Cambridge, 1994), 39–66; reprinted in G. Fine, ed., Plato 1: Metaphysics and Epistemology
(Oxford, 1999), 64–92.
Vlastos, G. “What Did Socrates Understand by His ‘What Is F? ’ Question?,” Platonic Studies,
2nd ed. (Princeton, N.J.,1981), 410–17.
 chapter 19

Pl ato’s M eta ph ysics

Verity Harte

1. Identifying Our Topic

Any attempt to write about Plato’s metaphysics must be, to some extent, a work of
construction and runs the risk of artificial separation between topics that are, for Plato,
naturally related. Plato’s writings are not themselves shaped in reflection of modern
subdivisions of philosophical areas and the form in which they are shaped—the often
heavily and self-consciously crafted dialogue form—does not naturally invite separate
identification and treatment of the writings’ often tightly interwoven philosophical
threads. With the possible exception of the Parmenides, no work of Plato presents itself
as being as a whole on a topic that we could without distortion understand as metaphysics
narrowly construed, although it is fair to say that some works are more obviously meta-
physical in character than others. In what follows, therefore, readers should understand
that there is an engagement with the works of Plato from a perspective that, in certain
respects, may differ from his own.
“Metaphysics” is a heading under which a range of topics might be considered. In dis-
cussions of Plato’s metaphysics, what takes center stage is, typically, a certain feature of
Plato’s ontology: his commitment, at least in certain works, to the existence of a special
class of entities, once known in English as “Ideas,” these days more commonly referred
to as “Forms.” The present essay is no exception in this regard. This narrowing of the
subject has some justification. Forms are seen to play a central role in Platonic counter-
parts to many of the topics one might expect to find discussed in a modern course on
metaphysics (topics, for example, such as the nature of reality, the metaphysics of prop-
erties, and causal responsibility), while not all the topics one might find in such a course
(topics, for example, such as possible worlds or paradoxes of time travel) have obvious
counterparts in the work of Plato. There are, however, recognizably metaphysical topics,
Plato’s treatments of which would undoubtedly be valid and interesting objects of study
but which are not considered in any detail here. Examples include the metaphysics of
456   Verity Harte

composition, the nature of time and space, personal identity, and the existence and
nature of god(s).1 Omission of such topics is partly due to the limitations of space and
partly due to the desirability of having a relatively unitary focus.
This narrowing of the topic of Platonic metaphysics to Plato’s ontology itself has some
advantage as regards locating Plato’s metaphysical theorizing within his own immediate
tradition. For, unlike metaphysics as such, ontology—understood as the rational investi-
gation of what there is or of being—is a branch of study for which Plato could find obvious
precursors in his philosophical predecessors, perhaps most notably, the Eleatic philoso-
pher, Parmenides, in whose Way of Truth one finds an account of a subject identified
only as “being” (in Greek: to eon), which, as has often been noted, attributes to being
many of the characteristics that Plato would subsequently ascribe to Forms.2 In Plato’s
works, Forms themselves are identified most generally as “the beings” (in Greek: ta onta,
or at least in many places apparently equivalently: ousiai).3
Plato’s place in this tradition provides the overall focus of this essay. Like Parmenides,
and like Democritus, the atoms of whose atomic theory are also noticeably Parmenidean,
at least on common understandings of these two Presocratic thinkers,4 Plato is a philos-
opher for whom reality differs from the way in which it presents itself to us in perceptual
experience and must be rationally discovered. Plato is a realist, at least in one common
use of the term realist; he is committed to the existence of a world that is objective and
mind-independent.5 But he is a realist, we might say, of an essentially optimistic variety.
Given the existence of a world that is genuinely objective and independent of human
thinking, there are, we might think, no very good reasons to suppose that human think-
ing will have any means of access to the character of the world. Plato, like rationalist-
minded philosophers before and after him, believes that our most prominent apparent
sources of access to the world—our senses—are often radically mistaken about it.
Nevertheless, he nowhere doubts that knowledge—through rational inquiry—is possible.6
This metaphysical orientation underlies the central contrast in his metaphysical theorizing,
a contrast between what is intelligible and what is perceptible. It is this contrast and no
other, I argue, that shapes the contours of his ontology.

1 There has also, it’s fair to say, been rather less discussion of these topics in the literature on Plato
generally. However, on composition, see my discussion in Harte 2002; on time and space, see
Algra 1995, Owen 1966a, Sorabji 1983 and 1988; on personal identity, see Bostock 1999, Gallop 1982,
Gerson 2003, Gill 1996, McCabe 1994, ch. 9, and 2000; on the existence and nature of god(s), see
Menn 1995, Morgan 1992.
2 For a sophisticated treatment of Plato’s relations to Parmenides, see Palmer 1999.
3 See, e.g., Phaedo 65d13, 66a3.
4 For an introduction to Parmenides and Democritus, see Long 1999, chs. 6 and 9.
5 This, if anything is, is a point on which there is now broad consensus, although this has not always
been the case: see Natorp 2004.
6 Again, there is now broad consensus that Plato is not skeptical about the possibility of knowledge.
In antiquity, however, there was a long-standing tradition of skeptical readings of Plato, on the history
of which, see Brittain 2001.
 Plato’s Metaphysics   457

2. Is There a Theory of Forms?

And Does That Theory Develop over
the Course of Plato’s Writings?

Our focus is on Forms. But we must first consider what sort of evidence is available to us
about Plato’s views about Forms. In addition to talking about Forms, discussions of
Plato’s metaphysics commonly talk of Plato’s Theory of Forms. But not everyone agrees
that Plato has what should be described as a theory of Forms,7 and many people who are
content to talk in terms of a theory find that theory only in one or other subset of
Platonic works. Discussion of Plato’s Theory of Forms thus gets quickly caught up in
controversies regarding the development of Plato’s thought. Indeed, on one view, the
Theory of Forms, its development, and its subsequent rejection, is the central narrative
in this development, whose transitions are marked, first, by the introduction and elabo-
ration of a theorized account of Forms in central works of Plato’s so-called middle
­period—works such as the Symposium, Phaedo, and Republic, in particular—and, second,
by Plato’s signalled rejection of this account of Forms in the Parmenides.8 A few words
on these matters are in order, then, although my remarks are made with the intention of
setting such questions about development aside so far as is possible.
The answer to the question of whether there is a Theory of Forms will depend on one’s
criteria for theory. What does seem clear is that Forms are theoretical entities. By this,
I do not mean simply that they are not given in perception, nor are they among the data
of “commonsense,” although, at least prima facie, they are not. Rather, Forms are theo-
retical entities in the sense that they do some theoretical work. I give four (what seem
to be the) central examples. As I have already said, Forms have a role to play in Plato’s
theory of being or what there is:

1. Forms are (among the primary) beings.

2. Further, especially in the Phaedo (96–106), Forms are identified as having causal
responsibility9 for things other than Forms having some of the character they do;
the Form of beauty, for example, has causal responsibility for the beauty of any-
thing else that is beautiful. In this way, Forms are not only themselves beings, they
are causally responsible for at least certain other aspects of the character of the
world, as well.

7 For example, consider the doubts expressed in Annas 1981, ch. 9.

8 Contrast, for example, Ryle 1966 and Owen 1953 and 1966b with Kahn 2007; and see Peterson,
chapter 10 in this volume.
9 I choose “causal responsibility” as the least misleading translation of the Greek terms under
discussion in this passage of the Phaedo: the adjective aitios and the noun aitia. For discussion of the
terminology, see Frede 1980, and for the notion in Plato, see Sedley 1998.
458   Verity Harte

Given these roles in Plato’s theory of being, it comes as no surprise that Forms have
central roles to play in Plato’s theories about the ways in which we talk and think about
the world also.

3. In the case of language, it seems from several works that Forms play a special role
in relation to the language we use to describe the world; they are in some way
privileged bearers of the terms that we use to describe those aspects of things for
which they turn out to be causally responsible.10
4. In Plato’s theory of knowledge, Forms turn out to be objects of knowledge and of a
privileged sort.11

It is, of course, conceivable that Plato started out with some (independently moti-
vated) commitment to this favored sort of entity—the Form—and then sought out con-
texts in which to put it to theoretical work. More likely, however, is that Forms are
theoretical entities in the sense of being entities whose claim to existence is justified or
defended in light of the theoretical work they do. One might defend this view by appeal
to a passage of the Parmenides (130b1–e3) in which Socrates, invited to answer questions
about the range of Forms to which he is committed and finding himself uncertain, sug-
gests that the reason not to subscribe to a Form for such items as hair, mud, and dirt is
that these are things that are “just as we see them to be” (130d3–4). Socrates appears to
reason here in the following (reasonable) way: where there is no theoretical work for
Forms to do, there is no reason to posit them.12 In general, this understanding of the
theoretical status of Forms gains support from the fact that, within the Platonic corpus,
there are no clear examples of direct arguments for the existence of Forms.13
Given this understanding of Forms as theoretical entities, when it comes to possible
lines of development, one might expect that any developments in the conception of
Forms would be driven by developments in his views on questions associated with the
various theoretical roles that Forms play, developments in his views about the nature of
language or knowledge, for example. This makes the task of considering whether Plato’s
theorizing about Forms is something that develops over the course of his writings
considerably more complicated. In what follows, questions about development are left
outside the frame of this discussion, to the considerations elsewhere in this handbook of
the larger topics within which Forms have theoretical work to do.

10 See Phaedo 102b11, Republic X 596a7–9, Parmenides 130e5–131a2. Passages such as these have
sometimes led people to think that Platonic Forms are meanings; see Bostock 1986. See also Crivelli,
chapter 20 in this volume.
11 See Phaedo 73b–76e and Republic 476a–480a. I take no stand here on the controversial question
of whether, especially in this Republic passage, Forms are assumed to be the only objects of knowledge.
Contrast Annas 1981, ch. 8, and Fine 1978 and 1990; see also Taylor, chapter 18 in this volume.
12 For this understanding of his reasoning and its significance, see McCabe 1994, 78–81.
13 Arguments for the existence of Forms can be found in Aristotle’s On Ideas, together with his
criticisms of them. See Fine 1993.
 Plato’s Metaphysics   459

3. The Language of Forms

This still leaves open the question of where we should look for evidence of Plato’s
views about Forms. As far as use of language goes, the central terms used to identify
Forms in indisputably canonical accounts of Forms—in particular, the Greek terms idea
and eidos14—turn up in a wide range of works across the corpus that cut across candi-
date boundaries between developmental stages in Plato’s thought. In some places, these
go along with comparatively rich characterizations of the nature and role of the objects
picked out by these terms, in others not; in some places, these characterizations are
obviously similar, in others less obviously so.
For example, in Socratic dialogues of definition, such as the Euthyphro, for example—
works that on widely accepted chronologies of the order in which Plato’s works were
written were produced earlier rather than later—we find the language of Forms, includ-
ing hallmarks of what, as we have already indicated, are Forms’ central roles.15 In
Euthyphro 6d11, for example, Socrates indicates that he is looking for “that form (eidos)
in virtue of which all the pious things are pious,” using causal language comparable to
that found in the explicit theory of Forms as causally responsible set out in Phaedo
96–106. But the Socratic dialogues provide no real detail as to the ontological character
of Forms. Had an accident of survival left us in a position where these and only these
works of Plato survived, it would, I think, be something of a challenge to reconstruct
from them the theory of Forms of scholarly conception or, indeed, of the sort that could
warrant Aristotle’s much publicized objections.
This lack of detail might be taken to indicate that these works constitute an early stage
in the development of Plato’s theory, where later works develop or extend the account
of Forms so as to provide the metaphysical underpinnings for Socrates’ search for defi-
nitions.16 Alternatively, one might view it as a consequence of a presentational strategy
that takes one through an ordered sequence in which the picture does not develop but is
gradually filled out.17 The evidence does not seem to me clearly to decide between these
positions. Nor do we need to, for there are pragmatic reasons not to consider the evi-
dence about Forms from these Socratic dialogues. Precisely because they do not offer a
rich characterization of the ontological character of Forms, it is difficult to derive much
of our view about the nature of Forms from them. Note, however, that this is in reality

14 On the terminology, see Motte et al. 2003.

15 The dialogues of definition I have in mind are Euthyphro, Charmides, Laches, Lysis, and Hippias
Major; Republic I and Theaetetus share the general form but for different reasons do not naturally go
with this group—Republic I because it opens the Republic; Theaetetus because it is generally viewed as a
later return to the form and is much more elaborate. Not all the works that would typically be identified
as early works are dialogues of definition; to name just two examples, the Apology and Crito are not
works of this type.
16 On the idea of “Socratic” Forms in contrast to “Platonic” Forms and the possible relations
between them, contrast Vlastos 1991, Irwin 1999, and Penner 1987.
17 Kahn 1996.
460   Verity Harte

only a matter of degree. Even in the “canonical treatment” of Forms in the Phaedo, the
Republic, or Symposium, the characterizations of Forms are richer, but not rich; at the
least, they leave open many unanswered questions, a fact the Parmenides’ searching
“reprise” might be thought to acknowledge.
More difficult is the question of what other dialogues to include; the Timaeus,
Sophist, Politicus, and Philebus, for example, all have discussion involving the language
of Forms. These are dialogues generally held to be among the later group of Plato’s writ-
ings, postdating not only the discussions of Forms in the Phaedo, Symposium, and
Republic but also those in the critical examination of the Parmenides.18 Their discus-
sion is in certain respects like and in certain respects unlike the discussions of Forms in
these earlier works, so that it is unclear the extent to which it indicates a departure from
their view of Forms. These later works do not play a central role in the discussion here,
again for pragmatic reasons. Both they and the Parmenides, the nature and import of
whose treatment of Forms would be critical to any attempt to tackle the question of
where these later discussions fit within the context of Plato’s treatment of Forms, are
considered in detail elsewhere in this handbook.19 So far as is possible, however, I
attempt to remain neutral on the question of development related to the characteriza-
tions of Forms therein.
For better or worse, then, our (main, if not exclusive) focus is the somewhat, but by no
means fully rich characterizations of the ontological character of Forms in the canonical
discussions of the Phaedo, Symposium, and Republic (to which we might add also the
Phaedrus).

4. What Forms Are There?

One striking feature of discussion of Forms in both Phaedo and Republic, especially, is
that discussions of Forms are typically framed as though all participants in the conversa-
tion are already familiar with Forms and have some idea of what Forms there are. This is
somewhat surprising if Forms are here being introduced and theorized for the reader for
the first time, although it is possible that the mismatch between reader and participant is
precisely designed to draw the reader’s attention to the novel features of what is being
said. Whatever its intended purpose may be, one effect of the strategy is resulting unclar-
ity as to what the scope of the theory is intended to be.

18 In the case of the Timaeus, this dating has been disputed, however, by Owen 1953, precisely on
grounds related to questions concerning the developments in Plato’s attitude to Forms. Contrast
Cherniss 1957.
19 For the view that Plato abandons Forms in light of the Parmenides’ criticisms, see, for example,
Owen 1953 and 1966b. For an alternative, more “unitarian” approach to Plato’s treatment of Forms,
post-Parmenides, see now Kahn 2007.
 Plato’s Metaphysics   461

Consider, for example, the following passage that occurs early in the Phaedo and is its
first introduction of Forms:

What about the following, Simmias? Do we say that there is such a thing as the Just
itself, or not?—We do say so, by Zeus.—And the Beautiful, and the Good?—Of
course.—And have you ever seen any things of this sort with your eyes?—Not at
all, he said.—Or have you grasped them with any of your bodily senses? I’m speak-
ing of them all, for example, of Largeness, of Health, of Strength, in sum of the
being of all of the others that each happens to be.20 (Phaedo 65d4–e5)

Here, we have a list of examples of Forms:21 Just, Beautiful, Good, Largeness, Health, and
Strength. This list, in itself, is a rather odd assortment of items, including values, a size
property, and physical characteristics. It is completed by a generalization—“all the rest”—
whose scope is utterly opaque.
Somewhat better is the generalizing move that follows Socrates’ subsequent argu-
ment, in the Phaedo, to the effect that the soul can be shown to preexist its embodiment
on the grounds that, when embodied, it has cognitive abilities requiring that, prior to
embodiment, it had knowledge of Forms. The argument takes the Form of Equality as
example, but it applies, Socrates says, to Forms in general, whose range is indicated
as follows:

For our present argument is no more about the Equal than about the Beautiful itself,
the Good itself, the Just, the Pious, and, as I say, about all the things on which we put
as a seal this mark ‘‘what is,’’ and about which we ask and answer in our questions
and answers. (Phaedo 75c10–d5)

Socrates here ties the scope of Forms to the scope of Socratic questions and answers.
Socratic questions ask “What is F?” for some range of properties. The Form is identified
as “What is [F]”—that is, as the referent of the answer to this Socratic inquiry. In this
way, he fixes the scope of Forms. But this fixing is not very informative, since we are no
clearer on the intended scope of such Socratic questions and answers than on the scope
of Forms.
If the passages containing examples and generalizations of this sort are not helpful
in fixing the scope of Forms, we might turn to the arguments in which we find them. As
I have said, the Platonic corpus does not provide us with direct arguments for the exist-
ence of Forms, which we might use to establish their scope. Forms do play roles in a
number of arguments, however, and we might turn to these arguments to investigate
what range of Forms they could be used, indirectly, to establish. The results of such
investigation, however, turn out not to be straightforward.

20 Translations of Plato here and elsewhere are taken from or take as a starting point those in
Cooper 1997, although in some cases I have modified them more or less extensively.
21 The passage does not refer to the items mentioned as Forms, but it seems clear that this is what
they are.
462   Verity Harte

Commonly, Forms are introduced as pairs of opposites. In both Phaedo 102b–105b

and Republic 475e–476e, for example, we find as examples of Forms a series of pairs of
opposites: in the Phaedo, Largeness and Smallness, Hot and Cold, Odd and Even; in the
Republic, Beautiful and Ugly, Just and Unjust, Good and Bad.22 And it is a central feature
of Forms, in these passages, that a Form cannot be characterized by its own opposite,
something that isn’t the case for other, perceptible bearers of the same name as the Form
(this point is central to the Phaedo passage cited previously; see especially 102d–103c).23
And it is sometimes suggested that a passage of Republic VII makes explicit a restriction
of Forms to opposite properties.24
In Republic 523b ff., in preparation for the establishment of the educational curricu-
lum for the philosopher-rulers, Socrates contrasts two sorts of sense perception: one
sort does not summon the understanding to investigate, and one sort does exhort it to
investigate. Socrates illustrates this contrast by the example of looking at three fingers:
the smallest, ring, and middle fingers. Perception of a finger as a finger is an example of
the sort of perception that does not summon the understanding to investigate, precisely
because perception does not deliver up two opposing perceptions at the same time:
“sight doesn’t suggest to [the soul] that the finger is at the same time the opposite of a
finger” (523d5–6). In the case of perception of the finger as having certain opposing
properties, by contrast, as being, for example, large or small, thick or thin, hard or soft,
Socrates says that perception precisely reports that the very same thing that the sense
reports as large, thick, or hard, it says to be the opposite also (524a6–10 especially). And
perception of opposing properties like these is the sort of perception that, for this reason,
does summon the understanding to investigate.

Sight, however, saw the large and small, not as separate, but as mixed up together.
Isn’t that so?—Yes.—And, for the sake of clarity on this, understanding was compelled
to see in turn large and small, not mixed up, but distinguished, in the opposite way
from that.—True.—And isn’t it from these sort of cases that it first occurs to us to
ask what the large is and what the small.—Absolutely.—And thus we called one
intelligible, the other visible.—That’s right. (Republic 524c3–d1).

The investigation that is initiated by perceptions that summon is an investigation of the

sort that leads to the recognition and identification of Forms. Hence the contrast between
those properties perception of which summons and those properties perception of
which does not summon could be taken as an indication of a restriction upon the range
of Forms to that of the summoning properties: that is, to Forms that are opposites.25
However, if this passage is understood to imply such a restriction in the scope of Forms,

22 Are these contraries or contradictories? The examples suggest that opposing Forms are contraries,
not contradictories, but whether this distinction is observed throughout is unclear.
23 This feature of Forms is central to the contrast between Forms and their perceptible counterparts,
which I consider in detail later on in this chapter.
24 Annas 1981, ch. 9. 25 So Annas (ibid.).
 Plato’s Metaphysics   463

it is inconsistent both with examples of Forms we find elsewhere and with another
­passage of the Republic that has also been taken to indicate the scope of Forms.
First, the examples: even without going outside the works on which we are focusing
(and which are indisputably home to the canonical theory of Forms), it is easy to find at
least candidate examples of Forms that do not have opposites: in Republic X, Forms of
Couch and Table (596b1–2); in the Phaedo, Forms of Fire and Snow (103c13). And if we
were to consider works throughout the corpus, examples would come easier still. But
these latter examples will be moot, because of questions about development, and the
first group of examples can all be brought into doubt, if doubt is sought. Republic X is an
unusual context, and it is just not clear what we should make of this talk of a Form of
Couch and of Table, which plays a role in Socrates’ development of an elaborate analogy
between painting and poetry in the service of his notorious criticisms of mimetic art.26
And there is some external evidence (for what this is worth) that Plato did not, in fact,
believe in Forms of artifacts.27 As to the Phaedo’s examples, the passage does not provide
unequivocal evidence that Fire and Snow are themselves understood as Forms, as
opposed to being entities that stand in some necessary relation to a Form, which Form
conforms to the restriction of Forms to opposites.28
Turning from the examples to the other passage of the Republic that appears to speak
to the question of the scope of Forms, we find, at least on the face of it, a different result
from the book VII passage. In book X, immediately before the introduction of a Form of
Couch and of Table, Socrates offers what appears to be a procedure for generating
Forms, which is commonly translated along the following lines:

Do you want us to begin our examination, then, by adopting our usual procedure?
As you know, we customarily hypothesize a single form in connection with each of
the many things to which we apply the same name. (Republic 596a5–7)

Read in this way, the passage proposes a range of Forms far wider than that implied by
reading the scope off the distinction in Republic book VII. Indeed, the range would be
wide to the point of potential absurdity: Do we really want a Form for any general term,
no matter how unnatural, gerrymandered, or empty it might be? Again, however, the
evidence is not decisive, again because of the unusual context and also because the passage
need not be translated in this way. Smith proposed that the passage should be construed,
rather, as making the claim that we commonly assume, “[as a rule of procedure,] that the
Idea which corresponds to a group of particulars, each to each, is always one, in which

26 Annas (ibid.) remarks on the unusual context; for the salience of couches and tables to this
context, see Burnyeat 1999, 232–36.
27 For the evidence and discussion, see Fine 1993, ch. 6. This same external evidence would not
restrict Forms to opposites, however, since it would include Forms of natural kinds.
28 Phaedo 104d1–7 is the best evidence that Three, and so, arguably, by analogy, Fire, Snow, and so
on, are indeed Forms, but it is not indisputable. The Timaeus does provide unequivocal evidence as to
the existence of a Form of Fire (see especially 51b7–d3), but the Timaeus is an unusual work in many
respects and, as I have said, one whose dating has been controversial in light of views about the ways
in which Plato’s views about Forms develop.
464   Verity Harte

case we call the group of particulars by the same name as the [Form].”29 On this construal,
the passage does not carry any implication about the scope of Forms.
My view is that it is a mistake to seek to use either of the Republic passages considered
(book X or book VII) to settle the question of the scope of Forms, and not simply
because they appear to answer the question in ways that are inconsistent with each other.
For all practical purposes, the book X passage is unavailable for use to settle this question.
Its construal is vexed, and its context is such that it is hard to know what more general
use can be made of the points that are made therein. In the case of book VII, I think it
mistaken to view the passage either as making or implying a point about the scope of
Forms. Notice the care with which Socrates puts his claim, at 523d4–5: in the case of
those properties, such as being a finger, perception of which does not summon the
understanding, “the soul of the many is not compelled to question what a finger might
be” (emphasis mine). What could and should be questioned by the soul of the few
(the author of the Parts of Animals, for example) is another matter.
It thus does not follow from what Socrates says in book VII that properties such as
being a finger are ones whose content does not merit rational inquiry of the sort that
would discover and identify a Form. And it certainly does not follow—as Socrates does
not claim, anyway—that the distinction he draws between properties that summon and
those that do not corresponds to an ontological distinction between Forms and other
non-Formal properties. What follows is just what Socrates emphasizes and the sort of
point that the passage’s educational context requires: properties that summon are those
for which the fact that an understanding of them needs rational inquiry is conspicuous
or obvious in a way that it is not in other cases; such properties are thus well chosen for
use in the design of an educational curriculum that has as its object the turning of atten-
tion away from perception to reason.
Suppose, nevertheless, that we ask ourselves what this passage can tell us about the
intended scope of Forms. Its moral, I suggest, is the contrast from which we began:
between reason and perception. The scope of Forms is set by the limits to the unprob-
lematic deliverances of perception, if unproblematic deliverances there be. But this pas-
sage has not told us what these limits might be, and the limits may themselves be things
about which Plato has shifting conceptions, according as his views of the respective
contributions of perception and reason develop and change. This general claim may not
satisfy, inasmuch as it fails to deliver a determinate answer as to what Forms there are.
However, it has the merit of being consistent with the verdict arising from the one pas-
sage in which Plato explicitly raises, without settling, the question of the scope of Forms
for our consideration. This is the passage of the Parmenides mentioned previously that
gives indications of the sort of criteria that ought to be used to settle the question. Forms
are not needed in those cases where things are “just as we see them to be.” What cases
these are may be for us to discover.

29 Smith 1917; translation put together from pp. 70–71.

 Plato’s Metaphysics   465

5. How, In General, Are Forms

Characterized?

When Forms are characterized, it is, as often as not, as part of a contrast between the
characteristics attributed to Forms and the characteristics attributed to certain perceptible
counterparts to Forms. These perceptible counterparts are generally called “particulars,”
but I argue later on in this chapter that this label is importantly misleading. Typically,
Forms are identified as having features that their perceptible counterparts prominently
lack (such as unity and stability, for example) or as lacking features that their perceptible
counterparts prominently have (a susceptibility to qualification by conflicting pairs of
opposite qualities most notably among them). Questions about the characteristics of
Forms are thus bound up with questions about the differences between Forms and their
perceptible counterparts.
Consider, for example, the contrast drawn in the following passage from the Phaedo:

Let us then return to those same things with which we were dealing earlier. That
being of whose being we give an account in our questions and answers: is it always
in the same condition in the same respects or does it vary from one time to another?
Does the Equal itself, the Beautiful itself, each thing itself—that which is—ever
admit any change whatever? Or does each of these things that is, being of a uniform
character taken by itself, remain the same in the same respects and never in any way
admit any sort of change whatsoever?—Necessarily, said Cebes, it remains the same
in the same respects, Socrates.—But what about the many beautifuls, such as people
or horses or clothing or any other things of this sort, or about the equals, or about all
those sharing a name with those things? Do they remain the same, or, in complete
contrast to those others, do they, practically never in any way remain the same as
themselves or each other?—The latter is the case, said Cebes, they never remain the
same.—Then, is it the case that, whereas you could touch and see and perceive with
the other senses these latter, there is no way to grasp those that always remain the
same than by reasoning of the mind; rather, such things are invisible and not seen?—
You’re absolutely right. (78c10–79a5)

Socrates uses this contrast to establish that there are two sorts of being: one visible,
the other invisible (79a6–7). And this is the overarching contrast between Forms and
their counterparts: Forms are not perceptible, but intelligible; their counterparts are
perceptible. These two sorts of beings are further characterized in terms of their
respective stability or instability. Intelligible Forms are invariant; they do not change.
Their perceptible counterparts, by contrast, are in no way invariant but subject to
change. It is unclear quite how these two contrasts are meant to be related, but the
shape of the passage suggests that the receptivity to change of their perceptible coun-
terparts is intended to support the view that changeless Forms are intelligible as
opposed to perceptible.
466   Verity Harte

The passage raises a number of questions. First, how should we understand the terms
of the contrast here and elsewhere—the contrast between the “many beautifuls” and the
Form with which they share a name? This is a question I return to later. Second, how
should we understand the comparative instability of the perceptible counterparts to
Forms? Is the suggestion that Forms’ perceptible counterparts “never in any way remain
the same” meant to imply that they are, instead, subject to variation in every respect?
Plato has sometimes been regarded as taking such an extreme view of the condition of
perceptible things. However, if this extreme view were in question, it would be hard
to see why Cebes would immediately agree with this picture without any question. Still,
even if we do not suppose that Plato’s view is extreme in this way, we must still ask our-
selves what sort of change is at issue.30 This is linked to the third question, which is how
susceptibility to change of this sort (or these sorts) would support the view that (insus-
ceptible) Forms are intelligible as opposed to being perceptible.
The change to which the perceptible counterparts to Forms, unlike Forms, are subject
may include such unproblematic examples of change as coming into being or perishing,
growth or diminution, and so on. But it seems likely, also, to include a phenomenon that
we might not be immediately inclined to think of as an example of change. This is the
phenomenon generally known as “the compresence of opposites.” Certainly, when, in
the Symposium, Diotima seeks to explain to Socrates the Form of Beauty’s manner of
“always being,” she denies both that it is subject to ordinary sorts of changes and that it is
subject to the compresence of opposites; this is in implied contrast to its perceptible
counterparts.

First, it always is and neither comes to be nor passes away, neither waxes nor wanes.
Second, it is not beautiful this way and ugly that way, nor beautiful at one time and
ugly at another, nor beautiful in relation to one thing and ugly in relation to another,
nor is it beautiful here but ugly there, as it would be if it were beautiful for some
people and ugly for others. (Symposium 210e6–211a5)

At its most general, the compresence of opposites is a situation in which it would be true
to say of some subject both that it is F and that it is un-F (the opposite of F): that it is, for
example, both beautiful and ugly. This is among the things here denied of the Form of
Beauty. A simple example of the occurrence of the compresence of opposites in a percep-
tible counterpart to a Form might take the form of the following example from Phaedo
102b3–6: Simmias is both large and small (or, perhaps, both larger than and smaller
than—it is the comparative terms that Socrates himself uses at b5): large in comparison
with Socrates (larger than Socrates), small in comparison with Phaedo (smaller than
Phaedo). This example may be misleadingly simple. Whether it illustrates the only or
central form of example and the manner in which it might be expected to provide
­support for the intelligibility of Forms are matters I return to later.
The respective invulnerability and vulnerability to the compresence of opposites of
Forms and their perceptible counterparts is one candidate, and, in my view, the best

30 A now classic discussion of these questions is Irwin 1977. See also Irwin 1999.
 Plato’s Metaphysics   467

candidate explanation of what is meant by another broad contrast between them, which
has prominence in the Phaedo especially. This is the view that the perceptible counter-
parts to Forms are in some way deficient in comparison with the perfection of Forms.
Consider, for example, the following agreement between Socrates and Simmias, applied
to the Form of Equal and its perceptible counterparts:

Well, then, he said, do we experience something like this in the case of the equals
among sticks and the other equals we mentioned just now? Do they seem to us to
be equal in just the same way as what is Equal itself is? Is there some deficiency in
their being such as the Equal, or is there not?—A considerable deficiency, he said.
(Phaedo 74d4–8)

Equality here means geometrical (rather than, as it might be, political or social) equality,
that property in virtue of which things are of the same measurement in some dimension
of measurement. Socrates and Simmias agree that the perceptible counterparts of the
Form of Equal have some deficiency in respect of this property when compared with
the Form itself. This deficiency has been interpreted in one of two ways.31 On the
Approximation View, Socrates and Simmias agree that two sticks, for example, cannot
be exactly equal in any dimension of measurement; they may look equal, but, with suffi-
ciently accurate measuring equipment, we would find they are not. On the Compresence
of Opposites View, by contrast, Socrates and Simmias agree that equal sticks are both
equal and unequal (albeit in different respects); they may, for example, be equal in length
but not in weight; equal to each other but not to some third stick of different dimen-
sions.32 Notice that Plato cannot simultaneously maintain both of these views, for they
are inconsistent with each other. By Approximation, sensible equals are not in any
respect exactly equal; they merely approximate equality. By Compresence, in contrast,
sensible equals are, indeed, exactly equal, in some respect; they are also unequal in some
(other) respect.
I favor the Compresence of Opposites view of deficiency for the following reasons.
First, it seems to me that it would at the least be hugely controversial to claim that, as a
matter of fact, no two perceptible objects could have exactly the same measurements as
each other in some dimension of measurement. The very existence of one case of the
dimension in some perceptible object seems to prove the possibility of its occurring
twice. The claim at issue, it should be noted, is much stronger than the possibly trivial
claim that we are often fast and loose in our identification of things as being equal, and
that many things we identify as such turn out to fall short of equality upon closer exami-
nation. It is, however, the stronger claim that is needed for the Approximation View.
And it seems to me that we should avoid the attribution of controversial claims where
none are needed. Second, the Approximation View seems unable to deal with those
instances in which there are Forms for each of a pair of (binary) opposites. The Form of

31 For the contrast, see Nehamas 1975.

32 For Approximation, see, e.g., Ross 1951; for Compresence, see, e.g., Nehamas 1975 and
Irwin 1977, 1999.
468   Verity Harte

Equal is one example, if there is a Form of Unequal also.33 The problem for the
Approximation View is that whatever is only approximately equal seems to be some-
thing exactly unequal. The view cannot thus be simultaneously maintained for each of a
pair of (binary) opposites.
These first two reasons have been illustrated with reference to the Phaedo’s own exam-
ple, but both would appear to generalize across at least a wide range of candidate Forms.
It also seems important that both reasons do apply so readily to the very Form that
Socrates chooses as an example when making the point about the deficiency of sensibles
in comparison with Forms; the greater plausibility of one or other view with respect to
this very example should count in its favor. The final reason to favor the Compresence
View is that it seems to cohere much better with those passages in which there seems
undeniable interest in the compresence of opposites, both in the Phaedo and else-
where (Phaedo 102b3–6, mentioned above; more controversially, but, I think, plausibly,
in a vexed passage in the immediate context, 74b7–c3;34 and, for example, Republic V,
478e7–479d5), especially since, as we have seen, Plato cannot consistently maintain both
views of the status of Forms’ perceptible counterparts.
Notice, however, that we now find ourselves confronted once more by the question of
the scope of Forms—in particular, by the question of whether there are Forms only for
pairs of opposites. If perfection is a defining characteristic of Forms in contrast to their
perceptible counterparts, and if what perfection amounts to is an invulnerability to the
compresence of opposites with which their perceptible counterparts are afflicted, then it
looks as if there can be Forms only of opposites.35 As it is, however, there seems no clear
evidence for this restriction in the scope of Forms (which would have been easy enough
to state). And there is some evidence against such a restriction in scope, in candidate
examples of Forms that do not have opposites.
When one is faced with this question, there seem to be three different options for
keeping the scope of Forms broader than the focus on compresence of opposites
might be taken to suggest. First, one might decide that Plato thinks the phenomenon of
­compresence of opposites is found more broadly than we might think.36 Second, one
might deny that the contrast between perfect Forms and imperfect sensible counter-
parts is, in fact, a defining characteristic of Forms.37 In this way, we need not take imper-
fection, so understood, to constrain our understanding of the scope of Forms. But we
would still need to explain the prevalence of interest in the presence or absence of the

33 Perhaps this might be doubted, if one thinks that Formal pairs of opposites are contraries
(see note 22 of this chapter). The reference to “inequality” at Phaedo 74c2, identified using the abstract
noun anisotês, might be taken as evidence for a Form of Unequal.
34 For recent discussion of this vexed passage, see Sedley 2007.
35 This problem arises on the Compresence of Opposites View of the imperfection of Forms’
perceptible counterparts, but it is not clear that we would be in a much better position if we, instead,
adopted the Approximation View of their imperfection, for it seems at least less obvious what would be
meant by the claim that perceptibles approximate nonoppositional features such as humanity, for
example, than that they do so for oppositional features such as equality or beauty.
36 For indications of this sort of strategy, see Fine 1993, 100–01.
37 This strategy has recently been defended by Sedley 2006.
 Plato’s Metaphysics   469

compresence of opposites, whether or not it is part of a contrast between imperfect

­perceptibles and perfect Forms. Finally, then, in a manner similar to the point made
above in connection with Republic book VII, one might suggest that the compresence of
opposites is given attention as an especially conspicuous aspect of some broader phe-
nomenon that has the potential to apply to a broader range Forms, under which broader
phenomenon compresence may be subsumed.
How might this final strategy be cashed out? Without pretending to the sort of
detailed examination one would really need of this question, two possibilities suggest
themselves. One is to recall that Forms are contrasted to their perceptible counterparts
as being invulnerable to ordinary sorts of change, as well as to the compresence of oppo-
sites (as in the Symposium passage quoted previously). Suppose that Plato shares with
Aristotle the view that negative predicates such as “is not human” are true not only of
presently existing things that are not human but also of things that previously existed as
humans, but which no longer exist.38 Then, it is as true to say, today, that Socrates is not
human, as it would have been true to say of him that he was human, on the fateful day
recorded in the Phaedo.39 This is not a case of compresence of opposites. But it is a case
of something of which compresence of opposites might be construed as a more vivid
example.
A second, possibly related way in which to cash out the strategy would be to draw on
one final broad contrast associated with the difference between Forms and their percep-
tible counterparts: the contrast between being and becoming, a key, but unclear state-
ment of which is found in the Timaeus:

In my judgement, then, we must first make the following distinction: what is that
which is always, having no becoming, and what is that which becomes always, never
being? The former is such as to be grasped by thought with reason, being always in
the same condition, whereas the other is such as to be grasped by judgement with
unreasoning perception, becoming and ceasing to be, but never really being.
(27d5–28a4)

While the relation between the Timaeus and the discussions of Forms on which we have
been focusing has been left an open question, the contrast drawn here seems clearly in
some way related to the contrast drawn at Phaedo 78–79, quoted previously.40 Consider,
then, one persuasive interpretation of what Plato may mean by the contrast between
that which becomes and that which is, put forward by Michael Frede.41 Things that
become are things that, relative to some specific times, contexts, or relations take on the
character or marks of some formal feature, F, but not in virtue of having or being some

38 For this view in Aristotle, see De Interpretatione 3, 16b11–15 and Categories 10, 13b14–19.
39 I set aside the complications raised by questions about the possible humanity of Socrates’
putatively immortal soul.
40 Note that the dual contrast will be further complicated, as the Timaeus proceeds, by the
introduction of the receptacle. See, e.g., 50c7–d2, 51e6–52b5.
41 Frede 1988. See also response by Code 1988.
470   Verity Harte

nature that is F. Only Formal natures—that which is captured by a definition of “F”—are,

as opposed to become, F. Occurrence of the compresence of opposites, in respect to some
F, is one conspicuous, but not the only, indication that perceptibles fail to satisfy the
requirements on things that are (what is) F, and hence merely become F, at some times
and in some contexts or relations.

6. Where Are Forms?

The question “Where are Forms?” may seem an odd one, but it seems to me worth
considering, insofar as it will sharpen our understanding of the questions there are
about the relation between forms and their perceptible counterparts. Further, odd
though the question may be, it is one that, even in popular thought about Plato, as
found in a nonspecialist encylopedia, is commonly given an answer: Forms exist in
some “Platonic heaven.”42 This answer may be intended metaphorically, since Forms
are also (and with justification) commonly understood to be immaterial, nonspatially
extended objects of a sort that are not naturally thought of as having spatial location.
Nevertheless, the metaphor implies the existence of a location or quasi-location for
Forms, which is distinct from that of the location of ordinary material objects. Since
Forms are the objects of intellect and material objects are the objects of perception,
the metaphor often extends to talk of two quasi-spatially distinct “realms”: the sensible
realm and the intelligible realm.
Such talk, of course, reflects the sort of contrast between two sorts of being—the
­perceptible and the intelligible—on which we have focused thus far. And the “location” of
these two sorts of beings in two different “realms” undoubtedly reflects some of Plato’s
own choices of image and language. In the Republic’s analogy of the cave, for example, the
intellectual ascent involved in turning one’s attention from the perceptible to the intelligi-
ble is depicted as a journey out of a cave to an environment outside. And the Phaedrus
talks of the “place beyond heaven” (247c3) as the location of truth. But there are questions
as to what is the best way in which to understand this sort of language and image.
One direction in which we should be careful to avoid being led is that of talking as
though Plato is somehow committed to two different realities. Assuming that reality is
what there is—whatever that turns out to be—then it is hard to see that it makes any
sense to talk of two realities; Plato’s view, rather, should be understood as the view that
the deliverances of perception do not exhaust (and may in some way distort) the con-
tents of reality. There remain, however, two rather different ways to understand this
view. On one, the view is that there is in reality what we ordinarily think that there is (the
perceptibles), but that there is, in addition, an aspect of reality besides what is evident in
perceptual experience and that is in various ways metaphysically explanatory of what is

42 See, e.g., the entry on Wikipedia, the Free Encyclopedia: http://en.wikipedia.org/w/index.

php?title=Theory_of_Forms&oldid=73131620%Wikipedia.
 Plato’s Metaphysics   471

evident in experience. On another, the view is that the evidence derived from perceptual
experience in certain respects distorts our understanding of what there is in reality, and
that the reality discovered through rational inquiry corrects or replaces aspects of what
our experience suggests to us there is. On the first view, perceptibles and intelligibles
work in tandem, though in distinction from each other; on the second, perceptibles and
intelligibles are more like rivals.
The second view is more in line with the rationalist tradition antecedent to Plato, from
which I began. On this view, Plato (like Parmenides and Democritus before him) is best
understood as proposing that it is only by using our intellect, as opposed to our senses, that
we will come to understand what there really is in the (single) world around us, the world
with which we are, indeed, in contact through perception, but about which perception to a
greater or lesser extent misinforms us. And, we may note, in the case of Democritus, for
example, there is no parallel temptation to talk of a distinct “atomic realm.” Of course, the
prevalence of this temptation with regard to Plato may be a reflection of (another) respect
in which Plato differs from Democritus (one of many). But it may also be a hazard arising
from an overly literal interpretation of spatial imagery that is, in fact, designed to accentu-
ate the intelligibility—as opposed to the perceptibility—of Forms.
A second potential hazard of an overly literal reading of the talk of Forms as residing
in some “Platonic heaven” is the assumption that, if Forms are separate, as Aristotle sug-
gested, they are therefore not immanent, not in the things that have them. However, as
Fine has argued, these matters are far from clear.43 Even if Plato does assume that Forms
are in some sense separate—not literally spatially, but in the sense of being capable of
independent existence—it is not at all clear that he concludes from this that Forms are
not also in certain things. In the Phaedo, presence (parousia, 100d5) is among the candi-
date relations that Socrates canvases for the relation between participant and Form, and
he is prepared to license inferences of the following sort:

When you then say that Simmias is larger than Socrates but smaller than Phaedo, do
you not mean that there is in Simmias both Largeness and Smallness? (102b3–5)

It is disputed whether, in this passage, Socrates has in mind that the Forms, Largeness
and Smallness, are themselves in Simmias, or whether there are, in addition to Forms,
additional corresponding items, so-called Immanent Forms or Immanent Characters,
and it is one of these Immanent Forms or Characters that is, for example, the Largeness
in Simmias.44 On a credible reading of the passage, however, Forms do turn out to be
items that can be located.45 But they are not located off in some remote “Platonic

43 Fine 1984 and 1986. Contrast Devereux 1994.

44 Contrast Fine 1986 and Vlastos 1969.
45 Does the idea of their being located call into question their immaterial character? No: no more
than does the claim that the immaterial soul is to be found—at least some of the time—in a body.
472   Verity Harte

heaven”; they are located where everything else is, around here, sometimes in (at least
some of) the things we see.46

7. What, Metaphysically Speaking,

Are Forms? and What, For That
Matter, Are “Particulars,”
The Perceptible Counterparts
to Forms?

Are Forms universal in character, or are Forms particular? That is, are Forms repeatable
items—not only located, but multiply located in many spaces and times in the things
that have them in common—or are they unique and nonrepeatable in character? Both
views of the metaphysical character of Forms have been defended.47 On balance, there
seems to me reason to favor the view that Forms are universal in character. This is, in
part, because Forms appear to perform the central function that is typically adduced as
the reason for introducing a universal, the performance of which has some claim to be
constitutive of being a universal; Forms underlie genuine similarities in the character of
things by being (in some way) common to them. But it does not follow from this that the
theory of Forms is itself a theory of Universals. After all, it is not clear that their perform-
ing this function, if they do, constitutes the central reason for their introduction as
Forms, and performing this function is not the sole function of Forms.
If, in the theory of Forms, Plato were giving us a theory of Universals, then he would, in
all likelihood, be the first to do so (and, indeed, he is cited as such by Armstrong,48 for
example). One of the consequences of being the first to offer a theory of the existence of a
certain sort of metaphysical object is that the metaphysical terrain is not already carved
up in such a way that distinctions of the sort that might emerge from such a theory are
readily available to draw on. I argue that the distinction between universal and particular,
understood as the distinction between items that are repeatable and those that are not, is
not, in fact, central to the contours of Plato’s ontology as he conceives them, if, that is, he
would recognize the distinction at all. This fact may go some way toward explaining why

46 “Sometimes”: if Forms’ capacity for independent existence includes (or amounts to) the capacity
to exist even if no perceptibles participate in them, then Forms need not always be located in some
perceptible object(s). But it does not follow from this that they are—as well or instead—in some
alternative location, a Platonic heaven; it may be that in this case they exist without any specific
location(s).
47 For the view that Forms are universals, see, e.g., Fine 1993; for the view that Forms are particulars,
see, e.g., Geach 1956 (at least implicitly); yet another view is that Forms are best understood as
something like chemical elements, for which, see Denyer 1983.
48 D. M. Armstrong in his entry on “universals” in Kim and Sosa 1995, 502.
 Plato’s Metaphysics   473

these two, as it seems to us, fundamentally different metaphysical characters, universal

and particular, have both seemed feasible in characterizations of Forms.
There are two main reasons to suppose that the distinction between universal and
particular is not, in fact, central to Plato’s ontology (at least, not in his construction of
the theory of Forms). These two reasons complement each other. The first is that
Forms do not appear to be the only item in Plato’s ontology that are universal in char-
acter, so it would seem that, if he does think of Forms as being universal in character,
this cannot be what he takes to be especially distinctive of them. The second is that
when Plato constructs the “other” to Forms, he does so in a way that encompasses
both items that are particular and items that are universal. By the “other” to Forms,
I mean, of course, not merely whatever is different from Forms but the items that are
typically contrasted with Forms in arguments centrally involving features of Forms—
that is, in the type of argument that one might take to indicate Plato’s reasons for
­positing Forms (with the features proposed). Misleadingly, these “other” to Forms are
often referred to as “particulars”; Plato’s “particulars,” however, are not all metaphysically
particular, or so I argue.
I take the first reason first. In saying that Forms are not the only items in Plato’s ontol-
ogy that appear to be universal in character, I follow an interpretation according to
which, in talk of the perceptible counterparts to Forms, Plato at least sometimes refers
to perceptible universals. Consider, for example, the following central passage from
Republic book V:

(S) Now that these points have been established, I want to address a question to our
friend who doesn’t believe in the beautiful itself or any form of the beautiful itself
that remains the same in all respects but who does believe in the many beautiful
things—the lover of sights who won’t allow anyone to say that the beautiful is one or
the just or any of the others; and let me ask him this: of these many beautiful things,
friend, is there one which will not also appear ugly? Or, of the many just, one which
will not appear unjust? Or, of the many things that are holy, one that will not appear
unholy? (G) No, he said, rather they must appear in some way both beautiful and
ugly, and the same goes for the others you asked about. (S) What about the many
doubles? Do they appear less halves than doubles? (G) No. (S) And the many large
and small things, or light and heavy things, is any one of these any more whichever
of these we say it is than the opposite? (G) No, each will always be both. (S) Then is
each of the many any more whatever someone says it is than it is not? (478e7–479b10)

According to the interpretation I follow, the items referred to here as, for example,
“the many beautifuls” are universal perceptible properties such as “being brightly colored”
of the sort that might (erroneously, in Socrates’ view) be offered as a candidate expla-
nation of the beauty of some perceptible beautiful object (a lithograph by Miró, for
example).49 Such properties are universal, insofar as they are themselves repeatable

49 For this reading, see Gosling 1960 and compare Irwin 1977 and Fine 1993. For doubts, see
Silverman 2002, ch. 4.
474   Verity Harte

items. Many Miró lithographs, for example, have in common being brightly colored. But
they are clearly distinguished from Forms, which are nonperceptible universals.50
This brings us to the second reason for supposing that the distinction between
­universal and particular is not central to Plato’s ontology. For it is these very items—the
perceptible universals of, for example, Republic book V—that turn out to be included in
Plato’s construction of the “other” to Forms. They are included in, but, in my view, do not
exclusively constitute the other to Forms, which elsewhere seems to include things that
are metaphysically particular in character. Consider, for example, a portion of the
Phaedo passage quoted before:

But what about the many beautifuls, such as people or horses or clothing or any
other things of this sort, or about the equals, or about all those sharing a name with
those things? Do they remain the same, or, in complete contrast to those others, do
they, practically never in any way remain the same as themselves or each other?
(Phaedo 78d10–e4)

Here again, we have mention of “the many beautifuls.” On this occasion, however, the
expression would appear to refer to metaphysically particular items—people, horses,
clothing.51 Part of the difficulty is that Plato does not have explicit terminology with
which to mark the particular-universal distinction, a fact which itself is grist to my mill.
Further, as we have seen, he is prepared to use the very same expression—“the many
beautifuls,” for example—for items on both sides of this metaphysical divide. Neither
the lack of explicit terminology nor the indifferent use of terminology across the divi-
sion shows that Plato could not draw the distinction. But it does support my case that
the distinction, if he has it, is not central to his own conception of the contours of his
ontology, nor to where he puts the fault lines in his arguments about Forms.
Further, from the point of view of his theorizing, the heterogeneity apparent in Plato’s
construction of the “other” to Forms has certain advantages. I focus on two. The first
takes us back to some questions left outstanding in Section V of this chapter about the
compresence of opposites. It is clear from the Republic V passage already quoted that
perceptible universals can take a prominent role in arguments involving compresence
of opposites. I now argue that, even when not directly referred to, it is the perceptible
universals that do the lion’s share of the philosophical work involved in appeals to
compresence.
While sometimes more nuanced, claims about the occurrence of the phenomenon of
compresence of opposites are sometimes put as the claim that all “perceptibles” (of the
relevant sort) necessarily give rise to compresence of opposites. Our Republic V passage
has this tone.52 But what does this mean? Is it the claim (PC) that, for any particular per-
ceptible having some relevant feature, F, necessarily, that particular perceptible also has

50 Or: nonsensible universals. So, for example, Fine 1993. 51 But, contrast here, Irwin 1977.
52 The “sometimes” of Phaedo 74b8 may be an example of nuance.
 Plato’s Metaphysics   475

the opposing feature un-F? Or is it the claim (UC) that, for any perceptible type, a token
of which is F, for some relevant feature, necessarily that type has un-F tokens also?
Plato would certainly be well advised not to commit himself to (PC) as stated, which
seems an implausibly strong claim. Could there not, for example, be an action that was
just and that was not, in any respect, unjust?
We may make this point (and the force of (UC) more concrete) with an example
exploiting the Phaedo’s chosen Formal exemplar, Equality. Consider the following
(apparently reasonable) possibility that (PC) rules out. Imagine a world in which there
are exactly two objects that, as a matter of fact, are equal in every dimension. Ex
hypothesi, they are not unequal in any respect, contra (PC). However, just because
these equalities of length, weight, and so on involve specific lengths, weights, and so on,
then clearly there could be an object to which these objects were unequal, although, in
fact, there is not.53
But what does it mean to say that there could be an object to which the equals of this
world were unequal? One aspect of the possibility in question is a possible object that
does not, but could, exist in the actual world we’re considering. Call it U (for unequal).
Another aspect relates to the objects that do exist in the actual world considered in view
of the possibility of U. It is this aspect that matters as far as the actual equals are con-
cerned, and this is their possible inequality to U (a possibility realized in all those worlds
in which both they and U exist). Such possible inequality to U must have some basis
in some (actual) feature of the equal objects in every (relevant) world, including the
actual.54 But what this feature amounts to is just the claim that there is some type that
these equals instantiate, and that this type has equal and unequal tokens across the rele-
vant worlds.55 Once considered across worlds, then, it becomes easier to see that the pos-
sibility of compresence is grounded in (UC). But any actual occurrence is possible and
thus open to the same explanation. Both actual and possible occurrences of compres-
ence in perceptible particulars may thus be grounded in the occurrence of the phenom-
enon at the level of types.
A second advantage of the heterogeneity of Plato’s “other” to Forms is the effect it has
on our understanding of predicates, in particular as applied to their perceptible coun-
terparts and to Forms. Given the existence and pertinence of certain perceptible univer-
sals, metaphysically particular beautiful objects—such as a lithograph by Miró—turn
out often to be instances both of a perceptible universal (being brightly colored) and of a
Form (the Beautiful). They are not instances of the Form in virtue of being instances of
the perceptible universal (because the perceptible universal is vulnerable to compresence).

53 The advisability of stepping down to a modal claim about (particular) compresence is noted in
Kelsey 2000, 105 (where the thought is attributed to Sarah Broadie, n.26).
54 The domain of worlds must be fixed to those in which the equal objects (or their counterparts)
exist and where all relevant dimensions bearing on their equality in the actual world are constant.
55 The argument proceeds on the assumption that (at least in some cases) compresent opposite
properties are attributed on the basis of one and the same feature of the object(s) in question. The case
of Simmias, who, while remaining the same in height, can be viewed as large in relation to Socrates and
small in relation to Simmias is of this type.
476   Verity Harte

And this leaves open how we should understand the relation between the particular’s
instantiation of the perceptible universal (its bright color) and its instantiation of the
Form (beauty). This question approaches, albeit somewhat indirectly, one of the most
controversial features of Plato’s theory of Forms: self-predication.
Self-predication is the view that a Form can in some sense be predicated of itself: that
the Form of Beauty can have the predicate “beautiful” applied to it. Self-predication
might occur in certain specific cases without being a matter of theory. For example,
if every Form is one and if there is a Form, One, then this Form self-predicates.
The interesting question, however, is whether Forms self-predicate systematically and
as a matter of theory. And there are grounds for thinking they do. Consider, for example,
the following passage from the Phaedo:

Consider, then, he said, whether you share my opinion as to what follows, for I think
that, if there is anything beautiful besides the Beautiful itself, it is beautiful for
no other reason than that it shares in that Beautiful, and I say so with everything.
Do you agree to this sort of cause? (Phaedo 100c2–7)

Since this passage assumes that the Beautiful itself is beautiful—and goes on to make a
claim about what must be true about anything besides the Form that counts as beautiful—
we have here a pretty clear statement of self-predication in what looks to be a sample
case: the Form, the Beautiful itself, is beautiful. Further, it is sometimes thought that the
theory of causal responsibility that Socrates is here in the process of developing and
illustrating requires that a cause resemble its effect in the relevant causal respect.56 This
would provide a theoretical motivation for systematic self-predication. Notoriously,
self-predication plays some central role in the so-called Third Man Argument at
Parmenides 132a1–b2.57
If there are good grounds for supposing that Forms self-predicate, it is nevertheless
hard to deny the apparent absurdity of some pictures of how this would work. (No
doubt, this is one reason that, among Parmenides interpreters, self-predication is high
on the list of targets for attitudes to Forms to be repudiated or revised.)58 The apparent
absurdity is brought out nicely by Fine: self-predication would have the consequence
that “the form of White (if there is one) is coloured white; the form of dog (if there is
one) can scratch its ears.”59 And, lest we think absurdity occurs only in cases where it is
disputable whether there are Forms, consider two very clearly evidenced Forms, the
Large and the Small. Is the Large some massive object? And how small would the Form
of Small have to be?60

56 See Sedley 1998; for discussion of the causal principle itself, see Makin 1990.
57 This was originally brought out by Vlastos 1954 and has been the subject of much discussion; see,
among many others, Meinwald 1992, Peterson 1973, Sedley 1998.
58 This is the strategy of Meinwald 1992, for example.
59 Fine 1992, 25. And see discussion of self-predication in Peterson, chapter 10 of this volume.
60 This sort of picture is only encouraged—to its discredit—by the Approximation View of
imperfection, rejected in Section 5 of this chapter.
 Plato’s Metaphysics   477

The absurdity arises on what Fine describes as “Narrow Self-Predication,” the view
that “the Form of F is F in roughly the same way in which F sensibles are F.”61 This is
probably why attempts to rescue (or revise) self-predication have focused on identifying
some different way in which the Form “is F.” Without rejecting this strategy, I want to
suggest that we should have in mind a question about the way in which perceptible Fs
are F, Forms aside.
Think once again about my lithograph, a perceptible beautiful (metaphysically par-
ticular) object. It is an instance both of a perceptible universal (being brightly colored)
and of a Form (the Beautiful). But it is not an instance of the Form in virtue of being an
instance of the perceptible universal (because the perceptible universal is vulnerable to
compresence). Being brightly colored cannot be the explanation of my lithograph’s
beauty, because these same bright colors have ugly instances (such as the sweater
I bought, but never wear). Not only that: many cases of beauty will not be brightly
colored—in the case of beautiful souls or beautiful theories, for example, the beautiful
items in question will not be colored at all. But Plato is committed to the view that an
explanation of beauty must be capable of covering all cases.62 Bright coloration, then, is
at most coextensive with some cases of beauty. But this would appear to leave it an open
question how, if at all, the beauty of my lithograph relates to its being an instance of this
perceptible universal? The (salient) perceptible features of my lithograph could be either
(1) in no way constitutive of the beauty of my lithograph or at least (2) not constitutive of
it in any way that invites the drawing of the absurd parallel when it comes to consider-
ing the way of being beautiful that applies to the Form. Self-predication might be
defended from evident absurdity, that is, by supposing that the basis for the application
of predicates to perceptible particulars is already somewhat different from what we
might have been ordinarily inclined to think. Indeed, I take this to be one way to under-
stand the claim that Socrates makes at Phaedo 100c (quoted previously).
Finally, therefore, this raises a question about the perceptibility of the properties cor-
responding to Forms. In adjudicating between the two options regarding my lithograph
presented previously, we may be concerned about proving too much. It proves too
much, one might think, if the beauty of my lithograph turns out to be nonperceptible,
just like the Form. Or perhaps this is not too much. After all, if Forms are immanent, the
beauty of my lithograph is brought about by the presence of the nonperceptible Form of
Beauty within it. This issue has arisen, indirectly, more than once over the course of my
discussion. Take, for example, some particular beautiful human being. This is a metaphys-
ically particular object that I can directly perceive. In some sense, I can directly perceive
it as human and as beautiful. But it seems to me far from clear whether, on Plato’s view,
I can directly perceive its humanity or its beauty. While it may seem unsatisfactory for
me not to be able to answer this question, it does have the merit of being consistent

61 Fine 1992, 25.

62 At least he often appears so committed, although it is not clear how well this would work in the
case of the “more subtle” forms of explanation endorsed in Phaedo 105b ff., for while it may be the case
that, for example, fever, when present in a body, always makes it sick, it is far less clear that whenever a
body is sick, fever is present.
478   Verity Harte

with the emphasis of my overall theme: Plato as metaphysician for whom the fact that
Forms are intelligible rather than perceptible is the primary point of focus, and who,
in positing Forms, is concerned to argue that many aspects of the (single, local) world
that appears to us in perception are not in reality how they appear.63

Bibliography
Algra, K. (1995) Concepts of Space in Greek Thought (Leiden: Brill).
Annas, Julia (1981) An Introduction to Plato’s Republic (Oxford: Clarendon Press).
Bostock, D. (1986) Plato’s Phaedo (Oxford: Clarendon Press).
Bostock, D. (1999) “The Soul and Immortality in Plato’s Phaedo,” in Fine 1999b, 404–24.
Brittain, Charles (2001) Philo of Larissa (Oxford: Oxford University Press).
Burnyeat, M. F. (1999) “Culture and Society in Plato’s Republic,” in Tanner Lectures on Human
Values, vol. 20, ed. G. B. Peterson (Salt Lake City: University of Utah Press).
Cherniss, H. F. (1957) “The Relation of the Timaeus to Plato’s Later Dialogues,” American
Journal of Philology 78, 225–66.
Code, A. (1988) “Reply to Michael Frede’s ‘Being and Becoming in Plato,’ ” Oxford Studies in
Ancient Philosophy, supplementary volume, ed. Julia Annas and Robert H. Grimm, 53–60
(Oxford: Clarendon).
Cooper, J. M., ed. (1997) Plato: Complete Works (Indianapolis, Ind.: Hackett).
Denyer, N. C. (1983) “Plato’s Theory of Stuffs,” Philosophy 58, 315–27.
Devereux, D. T. (1994) “Separation and Immanence in Plato’s Theory of Forms,” Oxford Studies
in Ancient Philosophy 12, 63–90.
Everson, S., ed. (1990) Epistemology. Companions to Ancient Thought No. 1 (Cambridge:
Cambridge University Press).
Fine, Gail (1978) “Knowledge and Belief in Republic V,” Archiv fu̇r Geschichte der Philosophie
60, 121–39.
Fine, Gail (1984) “Separation,” Oxford Studies in Ancient Philosophy 2, 31–87.
Fine, Gail (1986) “Immanence,” Oxford Studies in Ancient Philosophy 4, 71–97.
Fine, Gail (1990) “Knowledge and Belief in Republic V–VII,” in Everson 1990, 85–115.
Fine, Gail (1992) “Aristotle’s Criticisms of Plato,” Oxford Studies in Ancient Philosophy, supple-
mentary volume ed. James C. Klagge and Nicholas D. Smith (Oxford: Clarendon Press),
13–41.
Fine, Gail (1993) On Ideas: Aristotle’s Criticism of Plato’s Theory of Forms (Oxford: Clarendon
Press).
Fine, Gail, ed. (1999a) Plato 1: Metaphysics and Epistemology. Oxford Readings in Philosophy
(Oxford: Oxford University Press).
Fine, Gail, ed. (1999b) Plato 2: Ethics, Politics, Religion and the Soul. Oxford Readings in
Philosophy (Oxford: Oxford University Press).
Frede, M. (1980) “The Original Notion of Cause,” in Schofield et al. 1980, 217–49.
Frede, M. (1988) “Being and Becoming in Plato,” Oxford Studies in Ancient Philosophy, supple-
mentary volume, ed. Julia Annas and Robert H. Grimm (Oxford: Clarendon Press), 37–52.

63 For helpful discussion of the issues and/or comments on drafts of this essay, I am grateful to the
editor, Gail Fine, and to Ursula Coope, Melissa Lane, M. M. McCabe, and Dominic Scott. The chapter
is unrevised from its original version in the first edition of the Handbook, though not for reason of my
thinking it the last or perfect word to be said on these matters by me or anyone else.
 Plato’s Metaphysics   479

Gallop, D. (1982) “Plato’s ‘Cyclical’ Argument Recycled,” Phronesis 27, 207–22.

Geach, P. T. (1956) “The Third Man Again,” Philosophical Review 65, 72–82.
Gerson, Lloyd P. (2003) Knowing Persons (Oxford: Oxford University Press).
Gill, Christopher (1996) Personality in Greek Epic, Tragedy, and Philosophy (Oxford: Clarendon
Press).
Gosling, J. C. B. (1960) “Republic V: Ta Polla Kala, etc.,” Phronesis 10, 151–61.
Harte, Verity (2002) Plato on Parts and Wholes: The Metaphysics of Structure (Oxford:
Clarendon Press).
Irwin, T. H. (1977) “Plato’s Heracleiteanism,” Philosophical Quarterly 27, 1–13.
Irwin, T. H. (1999) “The Theory of Forms,” in Fine 1999a, 143–70.
Kahn, C. H. (1996) Plato and the Socratic Dialogue: The Philosophical Use of a Literary Form
(Cambridge: Cambridge University Press).
Kahn, C. H. (2007) “Why Is the Sophist a Sequel to the Theaetetus?” Phronesis 52, 33–57.
Kelsey, S. (2000) “Recollection in the Phaedo,” Proceedings of the Boston Area Colloquium in
Ancient Philosophy 16, ed. J. J. Cleary and G. M. Gurtler (Leiden: Brill), 91–121.
Kim, J., and Sosa, E., eds. (1995) A Companion to Metaphysics (Oxford: Basil Blackwell).
Kraut, R., ed. (1992) The Cambridge Companion to Plato (Cambridge: Cambridge University
Press).
Long, A. A., ed. (1999) The Cambridge Companion to Early Greek Philosophy (Cambridge:
Cambridge University Press).
Makin, S. (1990) “An Ancient Principle about Causation,” Proceedings of the Aristotelian
Society 91, 135–52.
McCabe, Mary Margaret (1994) Plato’s Individuals (Princeton, N.J.: Princeton University
Press).
McCabe, Mary Margaret (2000) Plato and His Predecessors (Cambridge: Cambridge University
Press).
Meinwald, C. (1992) “Good-Bye to the Third Man,” in Kraut 1992, 365–96.
Menn, Stephen (1995) Plato on God as Nous (Carbondale: Southern Illinois University Press).
Morgan, M. L. (1992) “Plato and Greek Religion,” in Kraut 1992, 227–47.
Motte, A., Rutten, C., and Somville, P. (2003) Philosophie de la Forme: Eidos, Idea, Morphé
dans la philosophie Grecque des origines a` Aristote (Louvain-la-Neuve: Éditions Peeters).
Natorp, P. (2004) Plato’s Theory of Ideas: An Introduction to Idealism, ed. Vasilis Politis (Sankt
Augustin: Academia Verlag).
Nehamas, A. (1975) “Plato on the Imperfection of the Sensible World,” American Philosophical
Quarterly 12, 105–17.
Owen, G. E. L. (1953) “The Place of the Timaeus in Plato’s Dialogues,” Classical Quarterly
NS 3, 79–95.
Owen, G. E. L. (1966a) “Plato and Parmenides on the Timeless Present,” Monist 50, 317–40.
Owen, G. E. L. (1966b) “The Platonism of Aristotle,” Proceedings of the British Academy 51,
125–50.
Palmer, J. A. (1999) Plato’s Reception of Parmenides (Oxford: Clarendon Press).
Penner, T. (1987) The Ascent from Nominalism (Dordrecht: Reidel).
Peterson, S. (1973) “A Reasonable Self-Predication Premise for the Third Man Argument,”
Philosophical Review 82, 451–70.
Ross, D. (1951) Plato’s Theory of Ideas (Oxford: Clarendon Press).
Ryle, G. (1966) Plato’s Progress (Cambridge: Cambridge University Press).
Schofield, M., Burnyeat, M., and Barnes, J., eds. (1980) Doubt and Dogmatism: Studies in
Hellenistic Epistemology (Oxford: Clarendon Press).
480   Verity Harte

Scott, D., ed. (2007) Maieusis: Studies on Greek Philosophy in Honour of M. F. Burnyeat
(Oxford: Clarendon Press).
Sedley, David (1998) “Platonic Causes,” Phronesis 43, 114–32.
Sedley, David (2006) “Form-Particular Resemblance in Plato’s Phaedo,” Proceedings of the
Aristotelian Society 106, 309–25.
Sedley, David (2007) “Equal Sticks and Stones,” in Scott 2007, 68–86.
Silverman, Alan (2002) The Dialectic of Essence: A Study of Plato’s Metaphysics (Princeton, N.J.:
Princeton University Press).
Smith, J. A. (1917) “General Relative Clauses in Greek,” Classical Review 31, 69–71.
Sorabji, R. K. (1983) Time, Creation and the Continuum (London: Duckworth).
Sorabji, R. K. (1988) Matter, Space and Motion (London: Duckworth).
Vlastos, G. (1954) “The Third Man Argument in the Parmenides,” Philosophical Review 63,
319–49.
Vlastos, G. (1969) “Reasons and Causes in the Phaedo,” Philosophical Review 78, 291–325.
Vlastos, G. (1991) Socrates, Ironist and Moral Philosopher (Cambridge: Cambridge University
Press).
 chapter 20

Pl ato’s Phil osoph y

of L a nguage

Paolo Crivelli

1. Introduction

Ideas in and problems of philosophy of language surface frequently in Plato’s dialogues.

Some passages briefly formulate, or presuppose, views about names, signification,
truth, or falsehood; others are extended discussions of important themes of philosophy
of language.
It is impossible, within the limits of this essay, to follow all the leads. I shall focus on
three topics. The first is the linguistic dimension of the theory of Forms; the second is the
discussion of names in the Cratylus, Plato’s only dialogue almost completely dedicated
to linguistic themes; the third is the examination of semantic and ontological issues
in the Sophist, whose linguistic section (259d9–264b10) presents Plato’s most mature
reflections on statements, truth, and falsehood.1

1.1 The Linguistic Dimension of the Theory of Forms

Are there Forms corresponding to every predicative expression? The case for. (1) In the
Phaedo, in the context of his earliest extensive presentation of the theory of Forms, Plato
says that “each of the forms exists and the other things that partake of these derive
their names from these themselves” (102b1–2, cf. 78d10–e2; 103b5–c2; Prm. 130e5–131a2;
133c8–d5; Ti. 52a4–7; Arist. Metaph. Α6, 987b9–10). (2) In Republic x, Socrates announces

1 I have benefited greatly from remarks by the editor and Francesco Ademollo. The responsibility for
the remaining deficiencies is only mine.
482   Paolo Crivelli

that he and his interlocutors will begin the enquiry by their “usual method”: “We are
accustomed to assume that each form [eidos hekaston] corresponding to each set of
many things [hekasta ta polla] is one, in which case we apply the same name to the
many things [sc. as to the Form]” (596a6–8). (3) In the Meno, Socrates adduces the fact
that “you call these many things [sc. roundness and the other shapes] by a single name
[sc. ‘shape’]” (74d5–6) as a reason for believing that there is a single item that “occupies”
(74d8) all shapes and has “shape” as its name (74e11).
These texts induce some commentators to attribute to Plato the view that for every
predicative expression there is a corresponding Form.2 For the sake of precision, let me
pin down the terminology: a Form F corresponds to a predicative expression P just if 3
the range of F is identical with the extension of P; the range of a Form F is the set whose
elements are all and only the items that partake of F; and the extension of a predicative
expression P is the set whose elements are all and only the items of which P is true.4
Are there Forms corresponding to every predicative expression? The case against. Plato
seems to reject Forms that correspond to negative predicative expressions. For, in the
Politicus (262a3–263b12), he seems to deny the existence of a Form that corresponds to
“barbarian” (i.e., “non‑Greek human being”). Again, he indicates (287b10–c5) that in some
cases a genus cannot be divided into only two species, but must be divided into three
or more. This would be false if for every negative predicative expression there were a
corresponding Form: instead of dividing a genus G into the species S1, S2, and S3, one
would divide first G into S1 and G-but‑not-S1 and then the latter into S2 and S3.
Moreover, in the Parmenides (130b1–e3), the young Socrates, after endorsing the
existence of the Forms like, one, many, just, beautiful, good, “and everything of that sort,”
hesitates to admit the Forms man, fire, and water, and is disinclined to accept Forms of
undignified things such as hair, mud, and dirt. Parmenides comments that philosophy
has not yet gripped the young Socrates as it will later, when he will not despise the items
last mentioned. To be sure, one must be cautious in using the Parmenides as evidence
for reconstructing Plato’s own views. Nevertheless, it is not unreasonable to regard
our Parmenides passage as alluding to a development in Plato’s theory of Forms. Since
the position held by the young Socrates denies that for every predicative expression
there is a corresponding Form, it is tempting to infer that at some stage Plato did not
think that for every predicative expression there is a corresponding Form.5
Forms and predicative expressions. Let us then examine again the passages that appear to
provide evidence for crediting Plato with the view that for every predicative expression
there is a corresponding Form. A bit of reflection shows that the passages mentioned
under headings (1) and (2) of the penultimate subsection commit Plato to the claim that

2 Cf. D. Ross, Plato’s Theory of Ideas [Theory] (Oxford, 1951), 79.

3 I use “just if ” to mean “if and only if.”
4 Other passages that appear to bear witness to a linguistic dimension of the theory of forms (Prm.
135b5–c2; Phdr. 249b5–c3; 266b3–5) are vague: they do not show that according to Plato for every
predicative expression there is a corresponding Form.
5 Cf. D. Bostock, Plato’s Phaedo [Phaedo] (Oxford, 1986), 201–02.
 Plato’s Philosophy of Language   483

if many perceptible particulars partake of the same Form, then they bear its name, and
therefore to the claim that if many perceptible particulars partake of the same Form,
then they bear the same name as one another. What these passages do not commit Plato
to is the converse claim, that if many perceptible particulars bear the same name as one
another, then they partake of the same Form. Only this last claim would point toward
(without, however, entailing) the thesis that for every predicative expression there is a
corresponding Form. Let me stress that the rendering of the passage from Republic x
given previously is not the only possible one, and differs from those endorsed by most
translators and commentators. It has however been defended both on philological
grounds and with a view to the passage’s contribution to the argument for which it
provides a premise.6 The passage under (3) provides stronger support for the interpretation
in question: even if it mentions only one example (the name “shape” and the corre‑
sponding form), it paves the way for an understood generalization. How far does this
generalization go? Since a Form in the Meno is a single cause (sc. item that can be
mentioned in providing an explanation) of why all things partaking of it are in a certain
way (cf. 72c7–8),7 the generalization cannot go as far as to cover all predicative expres‑
sions. For, in the case of certain predicative expressions, there is no single cause of
why all things they are true of are in the way the predicative expressions say they are
(consider the predicative expression triat, an abbreviation of “triangle or cat”: there is no
single cause of why every triat is a triat).
So, the passages mentioned fail to prove that according to Plato for every predicative
expression there is a corresponding Form. In view of the considerations of the last
subsection, one ought to avoid crediting Plato with such a view.
Forms that correspond to basic predicative expressions? Plato might nevertheless hold a
less ambitious theory: that for every basic predicative expression there is a correspond‑
ing Form, whereas non‑basic predicative expressions have no corresponding Forms and
are to be analyzed by appealing to Forms that correspond to basic predicative expressions.
For instance, the theory could acknowledge the existence of the Forms Greek, triangle,
and cat, which correspond to the basic predicative expressions “Greek,” “triangle,” and
“cat,” but deny the existence of Forms that correspond to the non-basic predicative
expressions “barbarian” and “triat,” which are to be analyzed by appealing to the Forms
Greek, triangle, and cat.
Such a theory invites the question of what makes a predicative expression basic. One
possible answer is that whether a predicative expression is basic depends on whether
its extension is a genuine group. Intuitively, “Greek,” “triangle,” and “cat” satisfy this
condition, while “barbarian” and “triat” do not. Obviously, one would like to go beyond
intuitions, that is, to know what makes a set of items into a genuine group.

6 Cf. J.A. Smith, “General Relative Clauses in Greek,” Classical Review, 31 (1917), 69–71 at 70;
D. Sedley, “Plato and the One-over-Many Principle,” in R. Chiaradonna and G. Galluzzo (eds.),
Universals in Ancient Philosophy (Pisa, 2013), 113–37 at 122–32. A more standard rendering has been
recently supported by R. Sharma, “On Republic 596a,” Apeiron, 39 (2006), 27–32.
7 Cf. G. Fine, On Ideas: Aristotle’s Criticism of Plato’s Theory of Forms [Ideas] (Oxford, 1993), 48, 50.
484   Paolo Crivelli

There is evidence that Plato allows only Forms whose ranges are genuine groups
(Phdr. 265e1–3; Plt. 262a5–263b11). If Plato endorses the converse claim, that is that every
genuine group is the range of a Form, then he is committed to the view that for every
basic predicative expression, that is every predicative expression whose extension is a
genuine group, there is a corresponding Form. The evidence does not warrant crediting
Plato with the converse claim.
Forms as missing standards. What remains beyond doubt is that there is a linguistic
dimension to Plato’s theory of Forms: Plato does say that perceptible particulars derive
their names from the Forms they partake of. What does Plato precisely mean when he
says this?
A deflationary explanation is possible. The view it attributes to Plato is simply that a
Form and the perceptible particulars that partake of it share the same name.8
One might however feel that this explanation is too deflationary. Specifically, one
might complain that this explanation does not take into account Plato’s point that per‑
ceptible particulars derive their names from the Forms they partake of. An alternative,
more substantive explanation rests on attributing to Plato two assumptions.9 The first,
intuitively plausible but never formulated in the dialogues, is that the mastery of a predi‑
cative expression is acquired by confronting an unambiguous standard for it, that is,
something to which that predicative expression applies whereas its negative counterpart
does not. For instance, the mastery of the predicative expression “red” is acquired by
confronting an unambiguous standard for “red,” something to which “red” applies
whereas its negative counterpart “not red” does not. The second assumption, often made
in Plato’s early and middle dialogues, is that for some predicative expressions, including
those from the all-important spheres of ethics and aesthetics (e.g., “good,” “just,” and
“beautiful”), there are no unambiguous standards among perceptible particulars: for
instance, to any perceptible particular to which “beautiful” applies, “not beautiful” also
applies, in some different respect or context (cf. Hp.Ma. 289a8–289b3; Smp. 210e2–211b5;
R. v 478e7–479c5).10 How then is the mastery of these predicative expressions acquired?
According to some commentators, the Phaedo is committed to the view that in the case
of these predicative expressions too, mastery is acquired by confronting unambiguous
standards, which however are not perceptible particulars, but intelligible Forms. For
instance, the Form beautiful is something to which “beautiful” applies whereas its
negative counterpart “not beautiful” does not (in any respect or context). Forms can be
contemplated by the soul when it is disembodied. They are forgotten at birth, but they
leave latent memory-traces, which can be triggered when perceptible particulars are
encountered that partake of them. Such a triggering ignites a disposition to apply the

8 Cf. Fine, Ideas, 318, 326.

9 Cf. N. P. White, Plato on Knowledge and Reality (Indianapolis, 1976), 75–77; Bostock, Phaedo,
94–115, 194–96.
10 Some commentators (e.g. Fine, Ideas, 56; T.H. Irwin, “The Theory of Forms” [“Forms”], in G. Fine
(ed.), Plato 1: Metaphysics and Epistemology (Oxford, 1999), 143–70 at 156–65) take Plato’s view to be that
it is perceptible properties (rather than, or as well as, particulars) that instantiate certain predicative
expressions and their negative counterparts.
 Plato’s Philosophy of Language   485

predicative expression whose mastery had been acquired by confronting the form
whose latent memory-trace has been triggered. In this sense, perceptible particulars
derive their names from the Forms they partake of.
Let me pause for a few remarks. (1) Rather few Forms are required by the view that
Forms function as missing standards—specifically, only Forms that correspond to pred‑
icative expressions for which there are no unambiguous standards among perceptible
particulars. Obviously, other considerations might postulate the existence of more
Forms. (2) The view that Forms function as missing standards does not make Forms
into meanings. One should have no more inclination to regard the Form beautiful as the
meaning of “beautiful” than to regard the finger one was shown when learning to use
“finger” as the meaning of “finger.” (3) According to the view that Forms function as
missing standards, the soul learns to use certain predicative expressions in its disem‑
bodied existence. Did my disembodied soul learn to use “beautiful,” “beau,” or “schön”?
The beginnings of an answer can perhaps be gleaned from the Cratylus, where Plato
assumes that there are Forms of names (cf. later in this chapter). So, my soul became
perhaps acquainted with the association of the Form of the name “beautiful” with the Form
beautiful. Since not only “beautiful,” but also “beau” and “schön” partake of the Form of
“beautiful,” having grasped the Form-to-Form association enables me to apply “beautiful”
(or the translation of it in whatever language I happen to be speaking) to beautiful things.
(4) The textual evidence for crediting Plato with the view that Forms function as missing
standards is scant (for instance, as I pointed out earlier, the assumption that the mastery
of a predicative expression is acquired by confronting an unambiguous standard for it is
never formulated in the dialogues). Some commentators11 therefore refrain from attrib‑
uting this view to Plato.

2. Names in the Cratylus

Theme and structure. The Cratylus examines the problem of the correctness of names: if
a name is correctly given to something, what is the source of this correctness? Plato dis‑
cusses two contrasting solutions, associated with two of the speakers, Hermogenes and
Cratylus. The third speaker, Socrates, is called to adjudicate.
The Cratylus is an aporetic dialogue: Socrates discloses difficulties for the views
defended by both of his interlocutors, but reaches no positive conclusion. To be sure,
this does not exclude the possibility that the dialogue might contain some of Plato’s
positive views on the correctness of names and other themes of philosophy of language.
Nevertheless, one should be cautious in one’s attempt to unearth such positive views.
The problem addressed. There is something woolly about the problem of the correctness
of names. One reason is that names constitute a loose category: they comprise not only

11 E.g., Fine, Ideas, 59, 137–38; Irwin, “Forms,” 155–65.

486   Paolo Crivelli

proper nouns such as “Cratylus” (383b2–3), but also common nouns such as “man” (399c1),
adjectives such as “large” (433e8), participles such as “flowing” (421c5), and infinitives
such as “to grow” (414a8) (words of primarily syntactic function are excluded).
There is a further, deeper source of woolliness. What is it for a name to be correctly
given to something? Is it for a certain usage of a name to be established correctly (as
when the stipulation is made that “hydrogen” will be the name of a certain element, or
“splash” of events of a certain sort), or for a name whose usage has already been estab‑
lished to be employed correctly? In the second alternative, is it for a name whose usage
has already been established to be applied truly (as when someone applies “hydrogen” to
hydrogen, or “splash” to splashes), or for a name whose usage has already been estab‑
lished to be employed correctly to convey the intended point, independently of whether
this is true or false (as when someone happens to employ “hydrogen” to mean a certain
element, or “splash” to mean events of a certain sort)? In the second alternative, is it the
case that a name whose usage has already been established is employed correctly to
convey the intended point just if it is being employed in conformity with its previously
established usage? Or is it the case that a name whose usage has already been established
is employed correctly to convey the intended point just if its previously established
usage, in accordance with which it is being employed on this specific occasion, is
appropriate to the intended point? On the first of these last two alternatives, the use of
“public,” in accordance with British English, to express a point concerning private
schools of a certain sort, is correct; on the second alternative, it is (arguably) incorrect.
These alternatives are not explicitly mapped in the Cratylus. However, the problem of
the correctness of names is sometimes connected with the first kind of correctness,
sometimes with the last. In other words, sometimes it is the problem of what the source
is of the fact that a certain usage of a given name is correctly established; sometimes it
is the problem of what the source is of the fact that a certain name whose usage has
already been established is, in its being employed in accordance with this previously
established usage, adequate as an expression of the intended point. These two issues are
closely connected.12
Conventionalism versus naturalism. Hermogenes favors a conventionalist solution:
“the correctness of names is determined by [. . .] convention and agreement” (384d1–2),
“any name you give a thing is its correct name” (384d2–3), and “if you change its name
and give it another, the new one is as correct as the old” (384d3–5). Something like this
conventionalism is the position likely to be endorsed by most people, non-philosophers
and philosophers alike, and Hermogenes seems to lack an elaborate linguistic theory to
support it.13

12 Cf. N. Denyer, Language, Thought and Falsehood in Ancient Greek Philosophy [Falsehood]
(London and New York, 1991), 69–71.
13 Here I follow D. Sedley, Plato’s Cratylus [Cratylus] (Cambridge 2003), 51–54, against the widely
held view that Hermogenes’s conventionalism is a philosophically extreme position (cf. e.g.
B. Williams, “Cratylus’ Theory of Names and Its Refutation” [“Theory”], in M. Schofield and M. Craven
Nussbaum (eds.), Language and Logos: Studies in Ancient Greek Philosophy Presented to G.E.L. Owen
(Cambridge, 1982), 83–93 at 90).
 Plato’s Philosophy of Language   487

Cratylus defends a naturalist solution: “there is a correctness of name for each thing,
one it is endowed with by nature” (383a4–5), and “a thing’s name isn’t whatever people
agree to call it, [. . .] but there is a natural correctness of names, which is the same for
everyone, Greek or foreigner” (383a5–b2). Such a naturalism is counterintuitive. Cratylus
seems to have a linguistic theory to back it, but he is unwilling to expound it.
Socrates’ criticism of conventionalism. Cratylus’s naturalism and Hermogenes’s con‑
ventionalism are presented in the dialogue’s initial exchange (383a1–384e2). The next
section (385a1–391a1) contains an extended criticism of Hermogenes’s conventionalism.
After a preliminary skirmish (385a1–e3), Socrates inquires about Hermogenes’s view
on Protagorean relativism (385e4–386d2), according to which things are for each subject
in the way they appear to him or her. To paraphrase a famous example from the Theaetetus
(152b2–9): whenever a wind feels cold to a person, it is cold for that person, and whenever
it does not feel cold to a person, it is not cold for that person (different people can be
involved, or the same person at different times). Hermogenes rejects Protagorean rela‑
tivism (386a5–7). Socrates offers a brief argument for the rejection: were Protagorean
relativism correct, there would be no difference between experts and laymen (because
things would be for all subjects whatever ways they appear to them). An annihilation of
the difference between experts and laymen would also follow if Euthydemus were right
when he claims that “everything always has every attribute simultaneously” (386d4), a
position quickly dismissed.14
Having done away with Protagorean relativism, Socrates and Hermogenes conclude
that things are what they are “by nature” (386e4), that is, not relatively to subjects nor
dependently on what appears to them. From this they infer (386e6–9) that actions also
are what they are by nature (the grounds of this inference are unclear).15 This result is
taken to hold also for that aspect of actions which is the use of tools: it is by nature that
given actions are performed by using certain tools. For instance, what instrument
weaving is performed with is a matter that is not relative to subjects nor dependent on
what appears to them: it is a natural matter. And, however things may seem to you, by
nature weaving is not performed with a drill, but with a shuttle.
This applies also to speech-acts, since they too are actions. So, it is by nature that given
speech-acts are performed by using certain tools. Specifically, since “naming is a part of
speaking” (387c6), it is by nature that given acts of naming are performed by using cer‑
tain tools. But the tools of naming are names. So, it is by nature, that is, not relatively
to subjects nor dependently on what appears to them, that given acts of naming are
performed by using certain names.
So far, the argument has been rather abstract: it has concluded that naming is what it
is by nature, and that by nature given acts of naming are performed by using certain

14 Cf. Tht. 161b9–163a3; 166c9–167d5.

15 For different interpretations, and the difficulties they face, see J. L. Ackrill, “Language and Reality
in Plato’s Cratylus” [“Language”], in Ackrill, J. L., Essays on Plato and Aristotle (Oxford, 1994), 33–52 at
38; C. D. C. Reeve (trans.), Plato: Cratylus [Cratylus] (Indianapolis, Ind., and Cambridge, MA, 1999),
xv‒xvi; Sedley, Cratylus, 55–58.
488   Paolo Crivelli

names. But what is it that naming by nature is? Socrates and Hermogenes agree that by
using names “we instruct [didaskomen] one another and we separate [diakrinomen]
objects according to how they are” (388b10–11), so that “a name is a tool for giving
instruction and separating being” (388b13–c1). The concept of instruction introduces, in
one go, two fundamental features of language: communication and truth.16 As for the
concept of separating being, at first one might regard it as connected with reference:
we use names to “separate being” in that we isolate certain specific beings from others as
topics of discussion (as the name “snow” in the statement “Snow is white” isolates snow
from other beings as a topic of discussion). However, Plato later indicates that the art of
instruction involving the use of names coincides with dialectic (390c2–12), and that
certain names imitate not sounds, shapes, or colors, but the essence of their nominata
(423b9–424a1, cf. later in this chapter). These later hints suggest that the separation of
being is the taxonomic division of genera into subordinate species,17 and perhaps also
the analysis of kinds into their constituents (genera and differentiae) performed by defi‑
nitions. The ideas of reference, taxonomic division, and analysis are probably operating
jointly in Plato’s description of names as separating being.18
Since speech-acts of naming are what they are by nature, that is, not relatively to sub‑
jects nor dependently on what appears to them, it follows that people can be more or less
successful at naming. Most successful will be those who possess the relevant art. As with
other kinds of action, so with naming: craftsmen make the best use of the tools, that is,
names. Moreover, again in analogy with what happens with other kinds of action, in the
case of naming also the tools, that is, names, are produced by another art, the art of
the legislator (nomothetēs) who lays down the rules governing the use of names. The
idea of a legislator who lays down rules for names is strange for us, but was probably
well known in Plato’s time thanks to the many (now almost completely lost) Sophistic
discussions of the correctness of names: Plato could mention it without causing surprise
in his readers.19 Names are therefore at the crossroad of two arts: that of the user (who
gives instruction and separates being), and that of the producer (the legislator).20
Socrates and Hermogenes make two points concerning the legislator’s art. The first
(389a5–390a10, cf. R. x 596a10–b10, Ti. 28a6–b2, 28c2–29b2) is that like other tool-
producing craftsmen, so also the legislator produces names by looking at forms and
realizing them in certain materials, namely syllables and letters. Forms of names are

16 Cf. N. Kretzmann, “Plato on the Correctness of Names” [“Correctness”], American Philosophical

Quarterly, 8 (1971), 126–38 at 128; Ackrill, “Language,” 42.
17 Taxonomic division is mentioned at 424c5–d4.
18 Cf. Sedley, Cratylus, 59–61.
19 Cf. Chrm. 175b3–4; Sedley, Cratylus, 66–74. Some commentators (e.g., Kretzmann, “Correctness,”
128–29, and M. Schofield, “The Dénouement of the Cratylus” [“Dénouement”], in M. Schofield and
M. Craven Nussbaum (eds.), Language and Logos: Studies in Ancient Greek Philosophy Presented to
G.E.L. Owen (Cambridge, 1982), 61–81 at 66) regard the legislator as a mythical personification of an
accepted linguistic authority.
20 Cf. Euthd. 289b7–c4; R. x 601d1–3; T. Borsche, “Platon,” in P. Schmitter (ed.), Sprachtheorien der
abenländischen Antike, 2nd edn (Tübingen, 1996), 140–69 at 143.
 Plato’s Philosophy of Language   489

presupposed: both a generic form and specific forms (like the form of the name for dogs
and that of the name for horses). It is also presupposed that syllables and letters play the
role of materials. As blacksmiths can produce tools of the same kind by looking at the
same form and embodying it in different kinds of iron, so legislators can produce names
of the same kind by looking at the same form and embodying it in different syllables and
letters. As not every kind of material is apt for the blacksmith to produce (say) a drill, so
not all syllables and letters are apt to realize a certain name. Nevertheless, there is some
flexibility in the choice of syllables and letters: this explains away the obvious fact (which
at first blush tells against Cratylus’s naturalism) that distinct linguistic communities
have equally adequate names that sound differently.
The second point concerning the legislator’s art (390b1–d8, cf. R. x 601d1–e3) also
relies on an analogy with other tool-producing arts: the person in the best position to
know whether the tool has been properly produced is the skilled user. In the case of
names, the skilled user is the person who is able to ask and answer questions, that is, the
dialectician (implicitly identified with the person who gives instruction and separates
being). The dialectician will therefore supervise the legislator’s production of names.
In the end, Hermogenes gives up his conventionalism, and Socrates concludes that
Cratylus’s naturalism is correct (390d9–391b3). This seems too sweeping a conclusion.
For the result established by the argument is rather abstract: it requires only that certain
general aspects of naming should be by nature, and it says nothing with regard to partic‑
ular aspects of the process. As far as we have been told, convention could still play an
essential and abundant part in these particular aspects. Plato perhaps intends his read‑
ers to see that Hermogenes and Socrates are hasty in reaching their conclusion.
Etymology and imitation. Having established that names have a natural correctness,
Socrates and Hermogenes proceed to “inquire what correctness of names is” (391b4–5).
First (391c10–421c2) they examine many derivative names, that is, names that can be
analyzed by etymological techniques whereby they are brought back to further names
out of which they are composed. Then (421c3–427d3) they investigate the correctness of
primary names, that is, the basic names that are not composed out of further names.
They begin this investigation by assuming (422c7–10) that the correctness of names is
the same for all, that is, for primary as for derivative names. By reflecting on what their
examination of derivative names has shown, they agree that the correctness of derivative
names consists in being “fit for revealing what each being is like” (422d2–3). They then
infer (422d5–7, cf. 393d3–4) that the correctness of primary names also consists in the
capacity to reveal what their nominata are like. However, in the case of primary names,
this revelatory capacity cannot be based on being composed out of further names—for
primary names are not thus composed. All names are composed out of syllables and let‑
ters: while derivative names are also composed out of further names, primary names are
only composed out of syllables and letters. Accordingly, primary names are analyzed by
applying mimetic techniques whereby the single letters in them are phonetic imitations
of basic features of reality: for instance, the sound “r” is “a tool for all motion” (426c1–2)
because the tongue “is most agitated and least at rest in pronouncing this letter”
490   Paolo Crivelli

(426e4–5); since “the tongue glides in the highest degree in pronouncing ‘l’ ” (427b2–3),
this sound reveals gliding and smoothness. The whole primary name describes its
nominatum as exhibiting the basic features of which its letters are phonetic imitations.
Not every imitation, indeed not every vocal imitation, is a name: otherwise “we would
have to agree that those who imitate sheep, cocks, or other animals are naming the
things they imitate” (423c4–6). Socrates (423d4–424a6) distinguishes the vocal imita‑
tions that are names from those that are not by assuming that the former imitate the
being (ousia) of their nominata, whereas the latter imitate their models’ sounds, shapes,
or colors—their qualities, as we would call them. Socrates does not explain what the
being of the things imitated by names is. The likeliest hypothesis is that the being of a
thing is whatever can be truly and appropriately mentioned in answering the “What is
it?” question asked about it (even if my car is white, it is not appropriate to say “It is white”
to answer the question “What is it?” asked about my car). Given that this is correct, the
being of the things imitated by names is their essence, in a somewhat weak sense of
“essence” (“weak” in that it ignores matters of identity over time). So, Socrates’ attempt
to distinguish the vocal imitations that count as names from those that do not comes to
the requirement that names should imitate the essence of their nominata.
The natural correctness of a name therefore consists in its describing the essence of its
nominatum. The descriptive content of a name, however, cannot be identified with its
ordinary meaning, if by meaning we understand what competent speakers would men‑
tion in answering the question “What does it mean?”: for the essence described by the
name can only be discovered by means of an art, not by simply examining the intuitions
of competent speakers.
Many of the Cratylus’s analyses of names are awkward, and modern linguists would
regard them as ridiculous. This induces many commentators21 to think that Plato is not
serious when he produces these analyses. However, throughout antiquity the analyses of
the Cratylus were regarded as serious. Moreover, passages from other dialogues where
Plato seems earnest present analyses that sound to us no less implausible than those of
the Cratylus.22
Socrates’ criticism of naturalism. In the last part of the dialogue (427d4–440e7),
Cratylus is persuaded to join the discussion. He agrees with Socrates’ account of the
natural correctness of naming (428c1–8). In particular, he agrees that the correctness of
a name consists in “displaying what the object is like” (428e2), that “names are therefore
spoken for the sake of instruction” (428e5), and that there is an art of giving names,
namely that of the legislator (428e7–429a1).
The natural correctness of naming introduced by Socrates might seem to amount to
a descriptive theory of naming, according to which a name n names whatever has the
characteristics revealed by n. But, in fact, Socrates and Cratylus disagree on whether
the natural correctness of naming is to be understood in this way (429a2–433b7).

21 E.g., Schofield, “Dénouement,” 63; Williams, “Theory,” 92; Reeve, Cratylus, xxx‒xxxiii.
22 Cf. Sedley, Cratylus, 25–50.
 Plato’s Philosophy of Language   491

Cratylus thinks that it should, and is therefore committed to the claim that whatever is
named by a name n, is named correctly by n. (Since the characteristics revealed fix the
nominata, the nominata must have the characteristics revealed.) Socrates instead is
committed to denying that his account of the natural correctness of names amounts to a
descriptive theory of naming: he is committed to denying that a name n names whatever
has the characteristics revealed by n. Instead, Socrates seems to believe that what n
names does not depend on the characteristics revealed by n. For this reason, there are
better and worse names: they are better to the extent that their nominata enjoy the char‑
acteristics they reveal, worse to the extent that their nominata fail to do this. And there
are two ways in which a name can be poor: the characteristics it reveals either constitute
a partially unfaithful portrait of its nominata, or fail completely to belong to its nominata.
In this last case, the name is downright false.
Although some commentators23 criticize Plato’s view that names can be false, the view
is actually reasonable. An analogy (admittedly far from what we find in the Cratylus) can
help to illustrate the competing views of naming defended by Cratylus and Socrates.
Cratylus’s position is analogous to that of someone recognizing only the “attributive” use
of definite descriptions: if a definite description denotes anything, what it denotes is the
only thing satisfying the condition expressed by its descriptive component. Socrates’
position is instead analogous to that of someone allowing the “referential” use of definite
descriptions. This may be explained by means of an example. Suppose that you and I are
at a party. We are looking at a man, Smith, who is holding a glass of yellowish liquid.
Since we had earlier heard Smith uttering the sentence “Could I have some beer?,” we
assent to the sentence “That man over there is drinking beer” (uttered while pointing at
Smith). Later I make a statement to you by using the sentence “The man who is drinking
beer is German.” My utterance successfully refers to Smith, and this successful reference
is achieved by using the definite description “the man who is drinking beer.” But, as a
matter of fact, Smith is not drinking beer (he is drinking champagne like everyone else,
as his earlier request to have some beer had been unsuccessful).24 My use of the definite
description “the man who is drinking beer” may be fairly described as a case of false
naming, and seems analogous to the sort of case Socrates has in mind.
Cratylus also agrees with another thesis of the theory set out earlier by Socrates and
Hermogenes: that names reveal what their nominata are like by imitating them by
means of their component letters (433b8–c2). But he dissents from Socrates on one point:
while Socrates holds that names can be more or less accurate in their imitation of their
nominata, Cratylus states that every name imitates perfectly its nominatum, and that
there is no place for a name being an inaccurate imitation of its nominatum (433c3–10).
In order to refute this position, Socrates first (433d1–434b9) rehearses together with
Cratylus the main claims of the mimetic account of the natural correctness of names.
He then (434b10–e1) focuses on the name “sklērotēs” (“hardness”). He points out that

23 E.g., R. Robinson, “A Criticism of Plato’s Cratylus,” Philosophical Review, 65 (1956), 324–41 at 328.
24 For the distinction between the “attributive” and the “referential” use of definite descriptions, see
K. Donnellan, “Reference and Definite Descriptions,” Philosophical Review, 75 (1966), 281–304 at 285–89.
492   Paolo Crivelli

the Eretrians pronounce it “sklērotēr.” In order to avoid admitting that the Attic version
of the name is less accurate than the Eretrian, Cratylus claims that the sounds “r” and
“s” imitate the same characteristic. By making this move, Cratylus implicitly commits
himself to regarding “sklērotēs” as a correct name of hardness. But now, when Socrates
points out that the sounds “l” and “r” give opposite indications (“l” imitates smoothness,
“r” hardness),25 Cratylus finds himself forced to concede that “sklērotēs” is not after all a
perfectly accurate name of hardness: “skrērotēs” would have been more accurate. This
concession already suffices to refute Cratylus’s position that every name imitates perfectly
its nominatum. Socrates however seems to think that he has found a loophole in the
naturalist position, and pushes the argument further (434e1–435d1). He remarks that we
do understand one another when we use “sklēron” (“hard”): how does this come about?
The background to this question is probably the admission that while “r” and “s” indicate
hardness, “l” indicates smoothness: “sklēron” thus contains two indicators of hardness
and one of smoothness. Cratylus replies that the reason we understand one another when
we use this expression is “because of habit” (434e4). This answer damages naturalism
because it points toward acknowledging a role for habit, and perhaps also for conven‑
tion, at the very heart of the naturalistic theory, that is, in the link of primary names to
their nominata. Cratylus could retreat by claiming that “sklēron” is not a name of what is
hard (just as, in his view, “Hermogenes” is not the name of Hermogenes, cf. 383b4–7).
Such a line would however clash with his earlier commitment to “sklērotēs” being a cor‑
rect name of hardness, and if “sklēron” is after all a name of what is hard, it is difficult to
see how this could be the case otherwise than through habit.
In the last section of the dialogue (435d1–440e2), Socrates subjects Cratylus’s views to
further criticism. Cratylus claims (435d1–436a8) that names are not merely “the best
and only way” of giving instruction (didaskalia), but also the only means of making
discoveries (heuresis) about things. This adds an epistemological dimension to Cratylus’s
naturalism. It is on this epistemological dimension that the last part of the dialogue
focuses, with a battery of three objections (436a9–437d8, 437d8–438c3, and 438c4–440e2).26

3. Semantics and Ontology

in the Sophist

Why an account of statement in the Sophist? The Sophist’s explicit purpose is to define the
sophist (218b7–c1). An Eleatic Visitor and Theaetetus pursue this project by applying
the method of division, but they encounter difficulties connected with the concept of
falsehood. For they attempt to define the sophist as someone who produces false statements

25 Earlier (426d3–e6) “r” was said to imitate motion, whereas hardness was not mentioned (cf. Schofield,
“Dénouement,” 74).
26 A discussion of these objections may be found in F. Ademollo, The Cratylus of Plato. A Commentary
(Cambridge, 2011), 427–86.
 Plato’s Philosophy of Language   493

that seem to be true, and therefore lead those who hear them to make false judgments
(234b5–d1).27 Thus, the sophist speaks falsely and induces false judgments. This description
of the sophist clashes with the falsehood paradox, which is summoned by way of objection
(236d5–241b4). The falsehood paradox is a family of arguments whose conclusion is
that it is impossible to speak falsely and to make false judgments. I say “a family of
arguments” because there are many subtly different arguments with this same counter-
intuitive conclusion. Accordingly, I shall often speak of a “version of ” the falsehood
paradox. Versions of the falsehood paradox appear in other dialogues (Euthd. 284b1–c6;
Cra. 429c6–430a5; Tht. 167a6–8; 187c7–200c7, esp. 188c9–189b9, cf. R. v 478b5–c2).
Only in the Sophist does Plato solve one of them, but his earlier presentations already
suggest some awareness of the disarming strategy.28
As for statements, the main version of the falsehood paradox in the Sophist goes as
follows:

[1] To speak falsely is to state what is not.

[2] It is impossible to state what is not.
[3] Therefore it is impossible to speak falsely.

A subordinate argument supports premise [2]:

[2.1] Stating what is not implies not stating anything.

[2.2] Not stating anything implies not accomplishing an act of stating.
[2.3] Therefore it is impossible to state what is not.

Parallel steps lead to the result that it is impossible to make false judgments.
The Sophist’s core section (236d5–264b10) is devoted to showing that it is possible to
speak falsely and to make false judgements. Since the main version of the falsehood par‑
adox deals with not being, a crucial move is the development of an account of not being.
This is done by bringing in the concept of difference (257b10–c3): roughly, for σ not to be
π is for σ to be different from everything that is π (where “σ” and “π” are schematic
letters to be replaced by syntactically appropriate expressions).29

27 In Plato’s late dialogues, “doxa” is used in two ways: for acts of a certain type and for a state by
being in which a person in given circumstances consistently performs acts of that type. For this reason
I have opted for “judgment” as a translation of “doxa”: “judgment” in English can be used both for acts
(“Judgement is that act of mind whereby the relation of one concept to another is determined”) and for
a state (“In his judgement conflict could be avoided”). Some commentators render “doxa” by “belief,” a
rendering that is suitable in contexts where a state is referred to.
28 Cf. M.F. Burnyeat, “Plato on How Not to Speak of What Is Not: Euthydemus 283a–288a,” in
M. Canto-Sperber and P. Pellegrin (eds.), Le style de la pensée: Recueil de textes en hommage à Jacques
Brunschwig (Paris, 2002), 40–66.
29 Plato’s account of not being in terms of difference is variously interpreted. The exegeses are recorded
by F.J. Pelletier, Parmenides, Plato, and the Semantics of Not‑Being (Chicago and London, 1990), 45–93,
and P. Crivelli, Plato’s Account of Falsehood. A Study of the Sophist [Account] (Cambridge, 2012), 177–204
(at pp. 192–96 I defend an interpretation close to that in the main text above).
494   Paolo Crivelli

After the account of not being has been developed, the Visitor declares that he and
Theaetetus must “agree what statement is” (260a7–8), that is, define statement. Theaetetus
wonders why this is needed (260b3–4). The Visitor explains (260b5–261c10) that since to
state, or judge, a falsehood is to state, or judge, what is not, the sophist could still adopt a
last defense based on denying that not being combines with statement and judgment:
only by defining statement and judgment will it be possible to show that not being
combines with them. At first glance, this last defense of the sophist seems a silly and des‑
perate move. An account of not being in terms of difference has been offered; we have
been hearing all along that to state, or judge, a falsehood is to state, or judge, what is not;
why on earth should we doubt that not being combines with statement and judgement?
The subsequent discussion divides into three parts: a definition of statement
(261d1–262e10), a proof that statements can be false (262e11–263d5), and a definition of
judgment on the basis of which it can be easily established that judgments can be false
(263d6–264b5).
Words, names, and verbs. When it comes to defining statement, the Visitor and
Theaetetus do not apply the method of division. Instead, they describe statements of
the simplest kind. Does this suffice to yield a definition? The Socrates of Plato’s early
dialogues would probably have denied it.
The Visitor distinguishes (261e4–262a8) two kinds of words (onomata, 261d2), or
vocal indicators (tē(i) phōnē(i) dēlōmata, 261e5): verbs (rhēmata) and names (onomata).
Verbs signify30 actions (praxeis).31 Names signify performers of actions, or agents (as I
shall normally call them).32 If only names or only verbs are uttered in succession, in neither
way does the whole utterance signify anything: only if verbs are uttered in combination
with names is the whole utterance endowed with signification (261d9–e2). More specifi‑
cally, if only names or only verbs are uttered in succession, in neither way does the whole
utterance constitute a statement (logos) (262a9–11): if only verbs are uttered in succession,
the whole utterance fails to be a statement (e.g., an utterance of “walks runs sleeps” is not
a statement) (262b2–8); if only names are uttered in succession, the whole utterance again
fails to be a statement (e.g., an utterance of “lion stag horse” is not a statement) (262b9–c2).
Only if verbs are uttered in combination with names does the whole utterance constitute
a statement (262c4–6). For instance, an utterance of “Man understands” is a statement
(262c9–d1), and it is obtained by uttering the verb “understands” in combination with
the name “man.”
Let me pause for a few remarks. (1) “Onoma” is used in two ways, a narrow and a broad
one. On its narrow usage, on which it is best rendered by “name,” “onoma” denotes the

30 The expressions used to describe the relation of words to what they stand for are the nouns
“indicator” (“dēlōma,” 261e5, 262a3), and “sign” (“sēmeion,” 262a6, 262d9) and the verb “to signify”
(“sēmainein,” 262b6).
31 262a3–5, 262b5–6, 262e13–14.
32 262a6–8, 262b10–c1. At 262e13–14, while describing the entity contributed by the verb as an action,
the Visitor calls “object” (“pragma”) the entity contributed by a name.
 Plato’s Philosophy of Language   495

vocal indicators that signify agents. On its broad usage, on which it might be translated
by “word,” it denotes all vocal indicators (including those that signify actions as well
as those that signify agents). (2) The distinction between actions and agents is unclear.
Is it an exhaustive ontological classification, so that every entity is an action just if it
is not an agent? (3) The distinction between verbs and names is also unclear. Is it con‑
trasting the grammatical categories of verbs and names, or the syntactic categories of
predicate expressions and subject expressions? Both alternatives face difficulties.
For, not every member of the grammatical category of verbs signifies an action (consider
“is carried,” to which in Greek there corresponds a single word, “pheretai”). Analogously,
not every member of the syntactic category of predicate expressions signifies an action
(the point is proved by the same example). (4) Although verbs signify actions, the
contribution made by a verb to a statement of which it is a component cannot be
exhausted by its signifying an action: otherwise the statement produced by uttering
“Man understands” would be perfectly equivalent to an utterance of the string of names
“man, understanding.”33
Naming and stating. The Visitor and Theaetetus agree that one name and one verb
make up a statement that is shortest (smikrotatos, elachistos, brachutatos) and primary
(prōtos) (262c5–d1, 263c1–4). This presupposes that there are statements of other kinds,
in particular, longer and non-primary statements that do not consist of merely one name
and one verb. These other kinds of statement are not described. The use of “primary”
suggests that statements of other kinds are composed out of primary ones (whose
components are not statements, but names and verbs), much in the same way as, according
to the Cratylus (422a1–e1), derivative names are composed out of primary ones (whose
components are not names, but syllables and, ultimately, letters). Statements can concern
not only the present, but also the past and the future (262d2–3).
The Visitor remarks that when a string composed only of names, or one composed
only of verbs, is pronounced, “either way the utterance reveals no action nor inaction
nor being of what is or of what is not” (262c2–4). The part of this remark about the failure
to reveal the “being of what is or of what is not” is obscure. It probably involves the pred‑
icative elliptical use of “to be” (whereby “to be” is employed as a copula to be completed
with a predicative expression, which, however, is suppressed and remains understood):
the point made is probably that a string of words of the sort described fails to signify the
being so‑and-so of what either is so‑and-so or is not so‑and-so. This sounds like an
anticipation of the account of truth and falsehood given later.
The Visitor says that a speaker34 producing a primary statement, that is, a statement
composed of one name and one verb, “does not only name something [onomazei],
but accomplishes something [ti perainei]” (262d3–4),35 that is, brings a speech‑act to

33 Cf. Denyer, Falsehood, 164–67.

34 Taking “tis” at 262c9 as the grammatical subject of “ti perainei” at 262d4.
35 Cf. Cra. 425a2–3. For the phrase “ti perainein,” cf. Grg. 472b8; Smp. 217c1–2; R. iv 426a2; Tht. 180a6–7.
496   Paolo Crivelli

completion.36 He adds that in producing a primary statement, a speaker “does not only
name something [onomazei], but also states something [legei]” (262d5). This remark
presupposes that naming and stating are different. No explanation of what they are is
offered. The Visitor and Theaetetus then agree (262e4–8) that every statement must be
“of,” or “about,” something. Their later observations (262e13–263a11, 263c1–12) on the
primary statements introduced as examples, “Theaetetus is sitting” and “Theaetetus is
flying,” show that the item a primary statement is about is the agent signified by its name.
In the face of these data, let me indulge in some speculation.

(1) What is it that a speaker producing a primary statement names? Does such a
speaker name both the agent signified by the primary statement’s name and the
action signified by its verb? Or only the agent? The Visitor and Theaetetus do not
address this problem, but the etymological link between “onomazein” (“to name”)
and “onoma” (“name”) suggests that the last alternative is the right one: a speaker
producing a primary statement names only the agent signified by its name.37
(2) If this result is correct, a further point can be plausibly inferred: in a primary
statement, the name is what mainly contributes to the speaker’s performing an
act of naming, whereas the verb is what mainly contributes to the speaker’s per‑
forming an act of stating.38
(3) Given that a speaker producing a primary statement names only the agent signi‑
fied by its name, and given that the item a primary statement is about is the agent
signified by its name, it follows that a speaker producing a primary statement
names only the item the primary statement is about. On the plausible assump‑
tion that a speaker producing a primary statement refers only to the item the
primary statement is about, a further inference can be plausibly drawn: for a
speaker who produces a primary statement to name an item is to refer to it.
(4) One reason for insisting that every statement must be about something is proba‑
bly the need to do away with an assumption made by one version of the falsehood
paradox, that is, the assumption that a false statement is about nothing (because
it is about what is not). Note that there are two different uses of “about” here: that
whereby “about” expresses the relation of a statement to its referent (in this use,
“Theaetetus is flying” is about Theaetetus), and that whereby “about” expresses the

36 Cf. G. Nuchelmans, Theories of the Proposition: Ancient and Medieval Conceptions of the Bearers of
Truth and Falsity (Amsterdam and London, 1973), 15–17. Other commentators (e.g., G. Rudebusch,
“Does Plato Think False Speech Is Speech?,” Noûs, 24 (1990), 599–609 at 601–02) take the Visitor to be
claiming that a speaker producing a primary statement ti perainei in the sense of limiting something:
such a speaker limits both the agent signified by the primary statement’s name (by specifying what
action it is performing) and the action signified by the primary statement’s verb (by specifying which
agent is performing it).
37 It is less likely that “onomazein” should be connected to “onoma” in its broad usage (in which case
a speaker producing a primary statement would probably name both the agent signified by the primary
statement’s name and the action signified by its verb): for the wide usage of “onoma” appears only at
the beginning of the linguistic section (261d2, 261d4), and is then superseded by the narrow usage.
38 Cf. M. Frede, “Plato’s Sophist on False Statements” [“Statements”], in R. Kraut (ed.), The Cambridge
Companion to Plato (Cambridge, 1992), 397–424 at 413–14.
 Plato’s Philosophy of Language   497

relation of a statement to what is said (in this use, “Theaetetus is flying” is about
Theaetetus’ flying).
(5) Plato is probably committed to the claim that utterances that one might be inclined
to describe as singular predicative statements with empty subject expressions
(e.g., an utterance of “Pegasus is flying”) are not genuine statements. Some
modern philosophers of language (e.g., Frege) explicitly endorse this claim.
(6) According to the Visitor, a speaker makes a primary statement “by putting an
object together with an action [suntheis pragma praxei] by means of a name and
a verb” (262e13–14). So, when a speaker carries out an act of stating by producing
a primary statement, what he or she does is to put the object, that is, the agent,
signified by the primary statement’s name together with the action signified by
the primary statement’s verb. The act of stating is two-pronged: the speaker
operates with two entities, namely an agent and an action, and does something
with them, namely puts the first together with the second.
(7) What is it to put two entities together in the way required by an act of stating?
An answer may perhaps be gleaned from some points made in the Cratylus, in
connection with a version of the falsehood paradox (Cra. 429c6–430a5). Socrates
and Cratylus agree (430a6–431c3) both that one can assign39 names to objects,
and that such an assignment40 can be carried out correctly, and therefore truly, as
well as incorrectly, and therefore falsely. One is tempted to modify this Cratylus
account to fit the situation of the Sophist by substituting actions for names. The
result is: when a speaker carries out an act of stating by producing a primary
statement, what he or she does is to assign the action signified by the primary
statement’s verb to the agent signified by the primary statement’s name.41
(8) Many modern philosophers of language think that the act of stating is single-
pronged in that it is directed to a single entity that corresponds to the whole
statement produced. Plato’s account is different: in Plato’s view, an act of stating
carried out by producing a primary statement is two-pronged in that it is directed
to two entities (an action and an agent) that correspond to two parts of the primary
statement produced (utterances of a verb and a name). For instance, according to
the modern view, the act of stating carried out by uttering the words “Theaetetus is
flying” is directed to a single entity, namely flying-Theaetetus or that-Theaetetus-
is-flying; according to Plato, it is directed to two entities, namely flying and
Theaetetus. The reason Plato does not offer an account of stating akin to the
modern one could be, of course, that he never thought of it; the reason could,
however, instead be that he did think of it but feared that it would give rise to a
version of the falsehood paradox. For, if one explains the act of stating carried out
by uttering the words “Theaetetus is flying” as directed to a single entity, namely

39 The verbs used are “dianemein” (430a7–8, 430e1), “prospherein” (430a8), and “apodidonai” (431b4).
40 The nouns used are “dianomē” (430d3, 431b1), “dosis” (430d6), and “epiphora” (430d6).
41 In the waxen block account of false judgement of the Theaetetus (190e5–196d2), Socrates and
Theaetetus again speak of an assignment, in this case of a perception to a memory imprint: they use the
verbs “apodidonai” (193c3), “prosarmozein” (193c4, 194a8), “prosballein” (193c6), “sunagein” (194b5),
“dianemein” (194d5), and “aponemein” (195a6), and the noun “sunapsis” (195d1).
498   Paolo Crivelli

flying-Theaetetus or that-Theaetetus-is-flying, it is hard to avoid admitting that

this single entity is not, and such an admission seems to plunge one into the dif‑
ficulties of the falsehood paradox. There is a way out for the modern account:
distinguish a veridical from an existential use of “to be.” On the basis of such a
distinction, an advocate of the modern account can say that the single entity to
which an act of stating falsely is directed “is” in that it exists but “is not” in that it
is not true or does not obtain. Whatever its merits, this is not Plato’s strategy.42
(9) As I pointed out earlier (cf. (2)), the verb is what mainly contributes to the
speaker’s performing an act of stating carried out by producing a primary state‑
ment. It is therefore natural to assume that an act of stating (legein) carried out
by producing a primary statement should be described by means of an expres‑
sion where “to state” (“legein”) takes a direct object that denotes the action sig‑
nified by the verb uttered within the primary statement (logos) produced: “σ
states α” (“σ legei α”), where “σ” is replaced by a noun-phrase denoting a speaker
or a statement and “α” by one denoting an action. A bit of reflection, however,
shows that a proper description of such an act of stating should be more com‑
plex. For, given that an act of stating carried out by producing a primary state‑
ment is two-pronged because it is directed both to the action and to the agent
signified, respectively, by the verb and the name uttered within the primary
statement produced, it is more appropriate to describe such an act of stating by
embedding “to state” within a three-place construction whose direct object
denotes the action signified by the verb whereas the agent signified by the name
is denoted by an additional factor: “σ states α about π” (“σ legei α peri π”), where
“σ” is replaced by a noun-phrase denoting a speaker or a statement, “α” by one
denoting an action, and “π” by one denoting an agent. In fact, Plato uses con‑
structions of this sort in his discussion of true and false statements.

True and false statements. The Visitor and Theaetetus agree that a statement must be
“of a certain quality,” that is, either true or false (Sph. 262e9–10, 263a12–b3, cf. Phlb.
37b10–c2). Plato occasionally contrasts the qualities of a thing with what it is, that is, its
essence (cf. Men. 71a1–b8; 86d8–e1; 87b3; Grg. 448e6–7; Tht. 152d3–4; 152d6). Therefore,
by saying that truth and falsehood are qualities of statements, the Visitor and Theaetetus
are probably hinting that neither truth nor falsehood is essential to statements as such
(some statements are true and not false, others false and not true).43
Two examples are brought in: the true statement “Theaetetus is sitting” and the false
statement “Theaetetus is flying.”44 Here is the relevant passage:

42 Plato’s account of statement as a two-pronged act recalls Russell’s account of judgment as a “multiple
relation”: cf. B. Russell, “On the Nature of Truth and Falsehood,” in id., Philosophical Essays, 2nd edn
(London, 1966), 147–59 at 150–58. Russell’s account was also motivated by the need to explain falsehood.
43 Cf. Frede, “Statements,” 417.
44 The English phrases “is sitting” and “is flying” render the Greek words “kathētai” and “petetai”
(the English words “sits” and “flies” convey the wrong sense).
 Plato’s Philosophy of Language   499

262e Vis. Now let us fix our attention on ourselves.

Tht. We will.
Vis. I shall make a statement to you by putting an object together with an
action by means of a name and a verb. You tell me what the statement
belongs to.
263a Tht. I shall do my best.
Vis. “Theaetetus is sitting”—not a lengthy statement, is it?
Tht. No, of a just length.
Vis. Now your job is to say what it is about and what of.
Tht. Clearly it is about me and mine.
Vis. And this one?
Tht. Which one?
Vis. “Theaetetus—with whom I am now speaking—is flying.”
Tht. This one also can only be described as mine and about me.
Vis. Besides we say that it is necessary for each of the statements to be of a
certain quality.
263b Tht. Yes.
Vis. Of what quality must one then say each of these is?
Tht. One is, somehow, false, the other true.
Vis. And the true one states [legei] the things which are [ta onta] to be
[hōs estin] about you [peri sou].45
Tht. Certainly.
Vis. Whereas the false one <states> things different from the things which
are [hetera tōn ontōn].
Tht. Yes.
Vis. It therefore states [legei] the things which are not [ta mē onta] as things
which are [hōs onta].
Tht. I suppose so.
Vis. But things which are different from things which are about you
[ontōn46 . . . onta hetera peri sou]. For, in a way, we said that about each
there are many things which are and many which are not.47
Tht. By all means. (Pl. Sph. 262e11–263b13)

45 Alternative translation: “And the true one states the things which are as they are about you.” For a
defense of the rendering in the main text above, see D. Keyt, “Plato on Falsity: Sophist 263B” [“Falsity”],
in E. N. Lee, A. P. D. Mourelatos, and R. M. Rorty (eds.), Exegesis and Argument: Studies in Greek
Philosophy Presented to Gregory Vlastos (Assen, 1973), 285–305 at 288–91, and Frede, “Statements,” 418.
“About you” is to be construed in common with “states,” “things which are,” and “are” (cf. D. Robinson,
“Textual Notes on Plato’s Sophist” [“Notes”], Classical Quarterly, n.s. 49 (1999), 139–59 at 159). “About
you” must be mentally supplied in the next two remarks by the Visitor.
46 I adopt Cornarius’s emendation ontōn (263b11), printed also by all recent eds. The main MSS read
ontōs (cf. Robinson, “Notes,” 159).
47 The Visitor is making a claim “about each” (263b12) of the “things [sc. kinds] which are different
from things which are about you” (263b11). The claim is that about each of these kinds “there are many
things which are and many which are not” (263b11–12). The Visitor regards this claim as a consequence
500   Paolo Crivelli

The Visitor offers three examples of primary statement: they are utterances of
“Man understands,” “Theaetetus is sitting,” and “Theaetetus is flying.” The name “man”
probably signifies the Form man, the name “Theaetetus” probably signifies the boy
Theaetetus (a perceptible particular), and the verbs “understands,” “is sitting,” and “is
flying” probably signify the Forms understanding, sitting, and flying. Near the beginning
of the passage translated above, the Visitor says that he will make a statement “by putting
an object together with an action by means of a name and a verb” (262e13–14). In view of
these facts, it can be plausibly inferred that if a speaker produces a primary statement by
putting an agent, signified by a name, together with an action, signified by a verb, the
action is always a Form (albeit one of a special type, like the Forms understanding, sitting,
and flying),48 whereas the agent can be either a form or a particular. It can also be plausibly
assumed that “to be about” (in some of its occurrences in the passage translated above)
expresses the relation that in modern philosophical jargon is expressed by “to hold of.”49
As for true statement, the passage suggests the following account: a primary state‑
ment produced by uttering a name n and a verb v is true just if the action signified by v is
about the agent signified by n. For instance, an utterance of “Theaetetus is sitting” is true
because sitting, the action signified by the verb “is sitting,” is about Theaetetus, the agent
signified by the name “Theaetetus.”
The account of false statement is controversial. At least four different exegeses have
been suggested.50

of another claim that he and Theaetetus had agreed upon in an earlier passage (256e6–8), namely of the
claim that “about each of the kinds what is is a lot whereas what is not is of indefinite multitude”
(256e6–7) (the relationship between the two claims is not that of strict identity, as is shown by the
expression “in a way,” “pou” at 263b12, which I take to qualify “we said,” “ephamen,” in the same line).
Consider the negative components of these two claims. When in the earlier passage the Visitor and
Theaetetus agreed that “about each of the kinds [. . . ] what is not is of indefinite multitude” (256e6–7),
the ground for this was that each kind is different from indefinitely many kinds (for it is different from
all other kinds). It may then be plausibly inferred that when in the later passage the Visitor claims that
“about each” (263b12) of the “things [sc. kinds] which are different from things which are about you”
(263b11) “there are many things [. . .] which are not” (263b11–12), the ground for this claim is that each
of the “things [sc. kinds] which are different from things which are about you” (263b11) is different from
(indefinitely) many kinds. I suggest that the Visitor is bringing up this ground: he is bringing up the
fact that each of the “things [sc. kinds] which are different from things which are about you” (263b11) is
different from (indefinitely) many kinds. This provides some justification (note the “for,” “gar,” at
263b11) for speaking of each of the “things [sc. kinds] which are different from things which are about
you” (263b11) in the way in which the Visitor is actually doing, namely for describing it as different
from certain kinds—specifically, as different from all the kinds that are about Theaetetus, the “things
which are about you” (263b11) (these kinds are many).
48 Although some commentators (e.g. O. Apelt, “Platons Sophistes in geschichtlicher Beleuchtung,”
in id., Platonische Aufsätze (Leipzig and Berlin, 1912), 238–90 at 258–59) doubt that the actions signified
by verbs are Forms, the cross-reference at 263b11–12 seems to require it. The author of the Seventh
Epistle mentions “all actions [poiēmata] and affections [pathēmata]” (342d8) in his list of entities of
which there are Forms.
49 Cf. M. Frede, Prädikation und Existenzaussage: Platons Gebrauch von “ . . . ist . . . ” und “ . . . .
ist nicht . . . ” im Sophistes [Prädikation] (Göttingen, 1967), 52–5, 94–5; “Statements,” 418; B. E. Hestir,
Plato on the Metaphysical Foundation of Meaning and Truth (Cambridge, 2016), 202–08.
50 Classifications of the interpretations of Plato’s account of false statement are also in Keyt, “Falsity,”
293–95 and Crivelli, Account, 233–41.
 Plato’s Philosophy of Language   501

(1) According to the “Oxford interpretation,” a primary statement produced by

uttering a name n and a verb v is false just if the action signified by v is different
from everything that is about the agent signified by n. For instance, an utterance
of “Theaetetus is flying” is false because flying, the action signified by the verb “is
flying,” is different from everything that is about Theaetetus, the agent signified
by the name “Theaetetus.”51
(2) According to the “incompatibility interpretation,” a primary statement produced
by uttering a name n and a verb v is false just if the action signified by v is incom‑
patible with something that is about the agent signified by n (two forms are
incompatible just if their very nature makes it impossible for them to be about
the same particular at the same time). For instance, an utterance of “Theaetetus
is flying” is false because flying is incompatible with something that is about
Theaetetus (in particular, it is incompatible with sitting).52
(3) According to the “incompatibility range interpretation,” a primary statement
produced by uttering a name n and a verb v is false just if the action signified by
v is different from, but in the same incompatibility range as, something that is
about the agent signified by n (an incompatibility range is an exhaustive set of
incompatible forms of the same level). For instance, an utterance of “Theaetetus
is flying” is false because flying is different from, but in the same incompatibility
range as, something that is about Theaetetus (in particular, it is different from,
but in the same incompatibility range as, sitting).53
(4) According to the “extensional interpretation,” a primary statement produced by
uttering a name n and a verb v is false just if the agent signified by n is different
from everything that the action signified by v is about. For instance, an utterance
of “Theaetetus is flying” is false because Theaetetus is different from everything
that flying is about.54

The incompatibility interpretation may be ruled out because it implausibly presupposes

that at some points in the Sophist the Greek word “heteron” expresses (not difference, as
it does elsewhere in the dialogue, but) incompatibility. The extensional interpretation
cannot be easily reconciled with the wording of the passage (although, as I have argued

51 Cf. D. Peipers, Ontologia Platonica: ad Notionum Terminorumque Historiam Symbola

(Leipzig, 1883), 173–77; Ross, Theory, 116; Frede, Prädikation, 95; G. E. L. Owen, “Plato on Not-Being”
[“Not-Being”], in G. Vlastos (ed.), Plato: A Collection of Critical Essays, vol. 1 (Garden City, N.J., 1971),
223–67 at 237–38; Frede, “Statements,” 419; J. van Eck, “Plato’s Theory of Negation and Falsity in Sophist
257 and 263: A New Defense of the Oxford Interpretation,” Ancient Philosophy 34 (2014), 275–88 at
284–86.
52 Cf. K.M. Sayre, Plato’s Late Ontology: A Riddle Resolved (Princeton, N.J., 1983), 236–83
(cf. 229–34).
53 Cf. M. T. Ferejohn, “Plato and Aristotle on Negative Predication and Semantic Fragmentation,”
Archiv für Geschichte der Philosophie, 71 (1989), 257–82 at 258–62, and Lesley Brown’s contribution of
chapter 13 to this volume.
54 Cf. P. Crivelli, “Plato’s Sophist and Semantic Fragmentation,” Archiv für Geschichte der Philosophie,
75 (1993), 71–74 at 73–74.
502   Paolo Crivelli

elsewhere,55 a reconciliation is not impossible).56 It is difficult to decide between the two

remaining contenders. The issues here are too many and too complicated to be
addressed in this essay.57 I restrict myself to expressing my preference for the Oxford
interpretation, a preference due to the fact that this interpretation fits well with the
earlier account of not being in terms of difference: given that for σ not to be π is for σ to
be different from everything that is π,58 it follows that for a certain action not to be about
a certain agent is for that action to be different from everything that is about that agent.59
The idea that to state a falsehood is to state what is not resurfaces—but, thanks to the
analysis of statement and the account of not being in terms of difference, it is now innoc‑
uous. The venom is taken out in two steps. In the first, we realize that a false statement
states what is not in that it states about an agent what is not about it. This first step is based
on the analysis of stating as a two-pronged act (an utterance of “Theaetetus is flying”
states flying about Theaetetus) and on the replacement of the intractable concept of
“complete” not being with the manageable one of “specific” not being (e.g., not being
about Theaetetus). At last we understand why there was a need to define “statement” in
order to realize that not being combines with it. The sophist’s last defense was not so silly
as it appeared at first—what the sophist wanted was an explanation of whether and
how the manageable concept of “specific” not being is relevant to false statement. In the
second step, we capitalize on the manageability of the concept of “specific” not being: we
realize that to state about Theaetetus what is not about him is to state about Theaetetus
what is different from everything that is about him. This second step relies on the
account of not being in terms of difference.
Paradoxes that arise from confusing naming with stating. As I noted earlier, a remark
made by the Visitor presupposes that stating and naming are different. Moreover, the
Visitor and Theaetetus stress that the two statements offered as examples, “Theaetetus
is sitting” and “Theaetetus is flying,” are about the same agent, that is, Theaetetus.
These facts are probably due to one version of the falsehood paradox being based on
confusing stating with naming. Were one to confuse stating with naming, one would be
likely to identify the statement “Theaetetus is flying” with the complex name “the flying
Theaetetus,” and to conclude that “Theaetetus is flying” is not false, but meaningless
(because the absence of a flying Theaetetus allegedly makes the complex name “the
flying Theaetetus” meaningless).60 Hence the point of indicating that stating and naming

55 Cf. note 54 of this chapter.

56 We obtain two (equivalent) analyses of what it is for π not to be about σ that fit the account of false
statement: first, for π not to be about σ is for π to be different from everything that is about σ; second,
for π not to be about σ is for σ to be different from everything that π is about.
57 An extensive discussion may be found in Crivelli, Account, 233–49.
58 Cf. text to note 29 of this chapter.
59 In Crivelli, Account, 261–74, Plato’s remarks on true and false primary statements are developed
into a theory of truth and falsehood covering a considerably large group of statements (a group that
includes the “categorical propositions” of Aristotle’s logic).
60 According to some commentators (e.g., Owen, “Not-Being,” 245), Plato himself earlier succumbed
to such a version of the falsehood paradox.
 Plato’s Philosophy of Language   503

are different. Moreover, were statements identical with complex names, then not only
would “Theaetetus is flying” be identical with “the flying Theaetetus,” but also “Theaetetus
is sitting” with “the sitting Theaetetus.” The two statements would then be about differ‑
ent items (the flying Theaetetus and the sitting Theaetetus), and would therefore not
contradict one another. Hence the point of emphasizing that the two statements are
about the same agent.
From false statement to false judgment. The Visitor and Theaetetus define thought
(dianoia) as “the inner conversation of the soul with itself that occurs without voice”
(263e4–5), and judgement (doxa) as the soul’s inner silent (affirmative or negative) state‑
ment that concludes an inner silent conversation (263e10–264a3, cf. Tht. 189e6–190a6;
196a4–7; Ti. 37a2–c5; Phlb. 38c5–e8). Given that an account of false statement has been
attained, an account of false judgment comes as a bonus.
Several important questions remain unasked: Is the identification of judgment with
inner silent statement to be taken literally or metaphorically?61 What is the language of
the soul’s inner silent conversation? Is it a divine language? A language of images? Or the
language the speaker feels most comfortable with at the moment?62 Can thinkers always
tell, by introspecting their consciousness, what the contents of their judgments are?63

4. Conclusion

Some of Plato’s views in philosophy of language had a remarkable impact on later

thinkers. For, Aristotle picked up and developed some of Plato’s views in this area, for
example, the idea that truth and falsehood are qualities of statements (SE 22, 178b27–28),
the distinction between primary and non-primary statements (Int. 5, 17a8–9, 17a20–24),
and the analysis of primary statements into names and verbs (chapters 2–5 of
de Interpretatione); others he criticized, for example, the claim that linguistic expres‑
sions are tools (Int. 4, 17a1–2). Through Aristotle, some of Plato’s views became integral
components of the philosophy of language of Western thought. Moreover, the false‑
hood paradox never surfaced again as a serious threat: the Sophist laid it to rest.
Alongside these elements of success in the area of philosophy of language, a conspicuous
absence in a neighboring field should be mentioned: the study of argument and of its
validity does not appear in Plato. Its development is perhaps the most notable of the many
achievements of his most distinguished pupil.

61 Cf. M. Duncombe, “Thought as Internal Speech in Plato and Aristotle,” Logical Analysis and
History of Philosophy, 19 (2016), 105–25 at 106–07.
62 Cf. P. Crivelli, “ΑΛΛΟΔΟΞΙΑ,” Archiv für Geschichte der Philosophie, 80 (1998), 1–29 at 21–23.
63 Cf. N. P. White, “Plato (427–347),” in M. Dascal, D. Gerhardus, K. Lorenz, and G. Meggle (eds.),
Sprachphilosophie–Philosophy of Language–La philosophie du langage, vol. 1 (Berlin and New York,
1992), 234–44 at 241.
504   Paolo Crivelli

Bibliography
Ackrill, J. L., “Language and Reality in Plato’s Cratylus” [“Language”], in Ackrill, J. L. , Essays
on Plato and Aristotle (Oxford, 1994), 33–52.
Ademollo, F., The Cratylus of Plato. A Commentary (Cambridge, 2011).
Apelt, O., “Platons Sophistes in geschichtlicher Beleuchtung,” in Apelt, O., Platonische Aufsätze
(Leipzig and Berlin, 1912), 238–90.
Borsche, T., “Platon,” in P. Schmitter (ed.), Sprachtheorien der abenländischen Antike, 2nd edn
(Tübingen, 1996), 140–69.
Bostock, D., Plato’s Phaedo [Phaedo] (Oxford, 1986).
Burnyeat, M. F., “Plato on How Not to Speak of What Is Not: Euthydemus 283a–288a,” in
M. Canto‑Sperber and P. Pellegrin (eds.), Le style de la pensée: Recueil de textes en hommage
à Jacques Brunschwig (Paris, 2002), 40–66.
Crivelli, P., “Plato’s Sophist and Semantic Fragmentation,” Archiv für Geschichte der Philosophie,
75 (1993), 71–4.
Crivelli, P. “ΑΛΛΟΔΟΞΙΑ,” Archiv für Geschichte der Philosophie, 80 (1998), 1–29.
Crivelli, P. Plato’s Account of Falsehood. A Study of the Sophist [Account] (Cambridge, 2012).
Denyer, N., Language, Thought and Falsehood in Ancient Greek Philosophy [Falsehood]
(London and New York, 1991).
Donnellan, K., “Reference and Definite Descriptions,” Philosophical Review, 75 (1966), 281–304.
Duncombe, M., “Thought as Internal Speech in Plato and Aristotle,” Logical Analysis and
History of Philosophy, 19 (2016), 105–25.
Ferejohn, M.T., “Plato and Aristotle on Negative Predication and Semantic Fragmentation,”
Archiv für Geschichte der Philosophie, 71 (1989), 257–82.
Fine, G., On Ideas: Aristotle’s Criticism of Plato’s Theory of Forms [Ideas] (Oxford, 1993).
Frede, M., Prädikation und Existenzaussage: Platons Gebrauch von “ . . . ist . . . ” und “ . . . ist nicht . . . ”
im Sophistes [Prädikation] (Göttingen, 1967).
Frede, M. “Plato’s Sophist on False Statements” [“Statements”], in R. Kraut (ed.), The Cambridge
Companion to Plato (Cambridge, 1992), 397–424.
Hestir, B. E., Plato on the Metaphysical Foundation of Meaning and Truth (Cambridge, 2016).
Irwin, T. H., “The Theory of Forms” [“Forms”], in G. Fine (ed.), Plato 1: Metaphysics and
Epistemology (Oxford, 1999), 143–70.
Keyt, D., “Plato on Falsity: Sophist 263B” [“Falsity”], in E. N. Lee, A. P. D. Mourelatos, and
R. M. Rorty (eds.), Exegesis and Argument: Studies in Greek Philosophy Presented to Gregory
Vlastos (Assen, 1973), 285–305.
Kretzmann, N., “Plato on the Correctness of Names” [“Correctness”], American Philosophical
Quarterly, 8 (1971), 126–38.
Nuchelmans, G., Theories of the Proposition: Ancient and Medieval Conceptions of the Bearers
of Truth and Falsity (Amsterdam and London, 1973).
Owen, G. E. L., “Plato on Not-Being” [“Not-Being”], in G. Vlastos (ed.), Plato: A Collection of
Critical Essays, vol. 1 (Garden City, N.J., 1971), 223–67.
Peipers, D., Ontologia Platonica: ad Notionum Terminorumque Historiam Symbola (Leipzig,
1883).
Pelletier, F. J., Parmenides, Plato, and the Semantics of Not-Being (Chicago and London, 1990).
Reeve, C. D. C. (trans.), Plato: Cratylus [Cratylus] (Indianapolis, Ind., and Cambridge, MA, 1999).
Robinson, D., “Textual Notes on Plato’s Sophist” [“Notes”], Classical Quarterly, n.s. 49 (1999),
139–59.
 Plato’s Philosophy of Language   505

Robinson, R., “A Criticism of Plato’s Cratylus,” Philosophical Review, 65 (1956), 324–41.

Ross, D., Plato’s Theory of Ideas [Theory] (Oxford, 1951).
Rudebusch, G., “Does Plato Think False Speech Is Speech?,” Noûs, 24 (1990), 599–609.
Russell, B., “On the Nature of Truth and Falsehood,” in Russell, B., Philosophical Essays, 2nd
edn (London, 1966).
Sayre, K. M., Plato’s Late Ontology: A Riddle Resolved (Princeton, N.J., 1983).
Schofield, M., “The Dénouement of the Cratylus” [“Dénouement”], in M. Schofield and
M. Craven Nussbaum (eds.), Language and Logos: Studies in Ancient Greek Philosophy
Presented to G.E.L. Owen (Cambridge, 1982), 61–81.
Sedley, D., Plato’s Cratylus [Cratylus] (Cambridge, 2003).
Sedley, D., “Plato and the One-over-Many Principle,” in R. Chiaradonna and G. Galluzzo
(eds.), Universals in Ancient Philosophy (Pisa, 2013), 113–37.
Sharma, R., “On Republic 596a,” Apeiron, 39 (2006), 27–32.
Smith, J. A., “General Relative Clauses in Greek,” Classical Review, 31 (1917), 69–71.
van Eck, J., “Plato’s Theory of Negation and Falsity in Sophist 257 and 263: A New Defense of
the Oxford Interpretation,” Ancient Philosophy 34 (2014), 275–88.
White, N. P., Plato on Knowledge and Reality (Indianapolis, Ind., 1976).
White, N. P., “Plato (427–347),” in M. Dascal, D. Gerhardus, K. Lorenz, and G. Meggle (eds.),
Sprachphilosophie–Philosophy of Language–La philosophie du langage, vol. 1 (Berlin and
New York, 1992), 234–44.
Williams, B., “Cratylus’ Theory of Names and Its Refutation” [“Theory”], in M. Schofield and
M. Craven Nussbaum (eds.), Language and Logos: Studies in Ancient Greek Philosophy
Presented to G.E.L. Owen (Cambridge, 1982), 83–93.
 chapter 21

Pl ato on th e Sou l

Hendrik Lorenz

Plato’s central contribution to psychology is his theory of the tripartite soul. This is at
once a theory about the nature of the embodied human soul and a theory of human
motivation. Its implied theory of motivation was accepted with little or no modification
by Aristotle. It remained influential into the later ancient period and beyond, not only
among Platonists but also in the Aristotelian tradition. The theory is introduced and
put to extensive use in the Republic. As a theory of motivation, it has noteworthy ante-
cedents in the Phaedo and supersedes an incompatible theory that is in play in earlier
dialogues such as the Protagoras, the Meno, and, arguably with notable signs of strain,
the Gorgias. As a theory about the nature of the human soul, it significantly departs from
ideas presented in the Phaedo, raising questions about the immortality of the soul of
which Plato is keenly aware. The Phaedrus and the Timaeus, both of them written after
the Republic, revisit the theory and bear witness to Plato’s ongoing reflection about the
nature of the soul.
In this chapter, I focus on the Republic, whose psychological theory is discussed in
considerable detail in Section 4. I begin by discussing the Protagoras (Section 1), the
Gorgias (Section 2), and the Phaedo (Section 3), insofar as speakers in those dialogues
express views about human motivation or about the nature of the soul.

1. Protagoras

In the course of an elaborately constructed discussion with Protagoras, Socrates presents

an outline of a remarkable theory of human motivation. The central and most famous
tenet of that theory is that “if someone were to know what is good and bad, then he
would not be forced by anything to act otherwise than knowledge dictates”1 (352c4–6).

1 Translations of Plato’s works are as in J. M. Cooper (ed.), Plato: Complete Works (Indianapolis,
Ind., 1997), with some modifications.
508   Hendrik Lorenz

Let us call this the claim that Knowledge Reigns Supreme (KRS). Socrates takes that
claim to be one that is not shared by the majority of people. The majority view, he thinks,
is that people can, and frequently do, act contrary to knowledge of good and bad, being
mastered by emotions such as anger, pleasure, pain, lust, or fear (352b3–8). KRS under-
lies Socrates’ strikingly optimistic assessment that knowledge of good and bad, if only
we had it, “would save our lives” (356d7–e2), at least ensuring that our lives would go as
well as, given the circumstances, they possibly could.
KRS, in turn, rests on two key assumptions: first, that knowledge of good and bad
would be consistently effective in affording its bearer an accurate view of how it would
be best to act in the circumstances; second, that a given person’s knowledge of how to act
would be fully in control of his or her actions. Socrates says disappointingly little about
the nature and structure of the knowledge of good and bad (cf. 357b5–6) or about how it
might be attained. It is clear, however, that he thinks of it in quantitative terms: as an
expertise of measurement concerning matters of value. He says that such expertise
would give us “peace of mind firmly rooted in the truth” (356d7–e2) and would render
powerless any mistaken appearances about good and bad that may arise in the varied
circumstances of life. One thing this makes clear is that he thinks the person of practical
knowledge may, on occasion, find herself with inaccurate, preliminary impressions
about how to act in a given situation, in a way that he likens to perceptual illusions.
Having a dull, run-of-the-mill doughnut right now might strike even a sage as a greater
pleasure than having her favorite dessert tomorrow after dinner, much as the truck in
the distance does look smaller than the car near at hand. However, Socrates thinks that,
whereas those who lack practical knowledge tend to be fooled by illusions about matters
of value, the person who is knowledgeable about such things is in no danger of acting on
illusory appearances. “The art of measurement,” he says, “would make the appearances
lose their power by showing us the truth” (356d7–e1).
For the knowledge of good and bad to save our lives, or at least to ensure that they go
as well as they can in the circumstances, this knowledge must not only afford a steady,
accurate view of how to act; it must also be the case that the bearer of such knowledge
acts entirely in accord with that steady, accurate view. Socrates clearly thinks that
people can be fully relied on to act in accord with such practical knowledge as they may
have. He goes further than that, holding that “no one who knows or believes there is
something else better than what he is doing, something possible, will go on doing what
he had been doing when he could be doing what is better” (358b6–c1; emphasis added).
Thus he takes the view that no one will pursue a given course of action if at the time she
knows, or merely believes, that the circumstances allow a better course of action.
On some conceptions of akrasia, that amounts to a denial of akrasia. In fact, Socrates
seems to go further even than that, holding that practical knowledge or belief governs
not only what one ends up trying to do but even what one wants to do. “No one,” he
claims, “goes willingly towards the bad or what he believes to be bad,” adding the note-
worthy further claim that “it seems not to be in human nature, either, to want to go
towards what one believes to be bad instead of to the good” (358c6–d2). On that view,
believing that φ-ing is bad, and that a good, or better, course of action is available,
 Plato on the Soul   509

ensures that one is not even going to want to φ, at any rate as long as one retains the
relevant beliefs.
It is worth noting that one can accept KRS without accepting what is known as
“Socratic intellectualism.” For the sake of clarity, let that be the view that every desire to
do something or other aims at the pursuit of a course of action that the person in question
knows or believes is at least no worse than any other course of action she takes to be
available. One could reject Socratic intellectualism and nonetheless uphold KRS. One
might hold that people can, and frequently do, want to do things that they realize it
would be better not to do, but that genuine knowledge of good and bad is a psychological
condition in which, among other things, one’s motivational and affective states have
settled into unison with one’s thoughts about good and bad. Something like this may be
Plato’s view in the Republic. However, the basis on which the Socrates of the Protagoras
endorses KRS would seem to be a simpler one. As we have seen, he presents it as a fact of
human nature that people will not even want to do something they believe to be bad
when they believe a good, or better, course of action is available. Presumably, his basis
for holding that is the thought that desires to act in some way or other always aim at the
pursuit of what the person in question knows or believes to be the best course of action
available, or at least a course of action no worse than any other option he or she takes to
be available. It would seem, then, that the Socrates of the Protagoras endorses KRS
because he subscribes to Socratic intellectualism.
As has been noted already, the Protagoras theory of human motivation rules out
akrasia, if akrasia is conceived of as a matter of acting contrary to one’s currently held
judgment about how it would be best, or anyhow better, to act. But that is by no means
the only way one can conceive of akrasia. In the classical tradition, it is not before
Aristotle that we get a conception of akrasia according to which akratic action is specifi-
cally a matter of action that runs counter to a practical judgment, currently held by the
agent. In earlier writers such as Xenophon, the language of akrasia is tied to a rather
loose notion of being mastered by emotions such as pleasure, pain, anger, or fear, as
well as by sources of pleasure such as food, drink, or sex (e.g., Memorabilia 1.5). It is not
required by that notion that the person who is mastered or overcome in this way, in
acting as he should not, all the while retains a judgment that he should not act as, in fact,
he does. It may well be that he experiences a lapse in judgment, brought on by intense
emotion or the prospect of indulgence. At the time of action, he may be deluded enough
to believe that acting as he does is best. Or, at the time of action, he may have no belief at
all about the rather complicated matter of how it would be best for him to act in the
circumstances, busy as he is enjoying himself or being in the grip of fear that focuses all
his attention on whatever it may be that terrifies him.
Against that background, Socrates in the Protagoras should not be interpreted as
arguing against the possibility of akrasia, understood as being overcome by emotions or
sources of pleasure. Rather, his main target is the view that people can know that some-
thing is less good than something else they might pursue, and they pursue it all the same.
This, of course, is the view he is ascribing to “the many.” His view is not, then, that there
is no such thing as being overcome by emotions or by things that are pleasant or painful.
510   Hendrik Lorenz

Rather, his view is that being overcome in this way is always a matter of adopting, or
finding oneself in the grip of, false practical beliefs and hence is always a manifestation
of ignorance (cf. 357e2, 358c1–3). Thus knowledge of good and bad, if only we had and
maintained it, would reliably protect us from acting as we should not. It would do this,
moreover, by reliably affording us accurate views of how we should act in the circum-
stances in which we find ourselves. With those views steadily in place, we would never
even want to act as we should not.

2. Gorgias

The Socrates of the Gorgias seems to operate within a robustly intellectualist framework,
at least for much of the dialogue. He assumes that anyone who knows what is just and
unjust cannot possibly act unjustly or even want to act unjustly. He also seems to assume
that every desire to do something or other aims at the good in the sense that every such
desire rests on a judgment that it is (in the circumstances) better to do the thing in
question than not to. Presumably, the latter assumption underlies the former: knowing
what is just brings with it knowledge that acting unjustly is always worse than not to, and
so no one who knows what is just can ever even want to act unjustly.
Gorgias tells Socrates that if a student of rhetoric lacks knowledge about matters of
value, such as what is good, fine, and just (459c8–e1), “he will learn those things from me
as well” (460a3–4). But, says Socrates, just as a person who has learned carpentry, music,
or medicine is a carpenter, musician, or doctor, so a person who has learned what is just
is a just person (460b7). Moreover, a just person wants to do just things, and, in fact, he
will never want to act unjustly (460c1–3). Since a trained orator knows just things, then,
he is incapable of using oratory unjustly or even of wanting to act unjustly in any way at
all (461a4–7). Thus Gorgias is contradicting himself when he holds that orators must
know what is just and also that some orators use their art unjustly. Socrates evidently
sees no gap at all between having learned what is just and unjust, in the way one might
learn such things by a teacher’s instruction, and being wholeheartedly motivated to
pursue justice and avoid injustice. This is best explained by the conjecture that, much
like the Socrates of the Protagoras, the Socrates of the Gorgias assumes that since people
already are wholeheartedly motivated to pursue the good, knowledge of what precisely
effective pursuit of the good involves and requires is fully sufficient to ensure proper
motivation.
That the Socrates of the Gorgias operates with some such picture of human nature is
confirmed by what he says in his discussion with Polus. During that discussion, he
claims that whenever we do something that we do not take to be good in itself, we are
acting “for the sake of the good,” by which he seems to mean that we do the thing in
question because we suppose that doing it is better for us than not doing it. Thus he
holds that “it’s because we pursue the good that we walk whenever we walk; we suppose
that it’s better to walk” (468b1–2); and again: “we put a person to death, if we do, or banish
 Plato on the Soul   511

him and confiscate his property because we suppose that doing these things is better for
us than not doing them” (468b4–6). He also operates with a conception of desire that
serves to explain and underwrite his remarkable claims about human behavior and
motivation. Moreover, that conception of desire explains why people such as tyrants
“do just about nothing they want to do” (466d6–e2). According to that conception, all
desires to do something or other aim at doing something that it is good for the person to
do, presumably in that it is better for the person to do the thing in question than not to.
Since every human desire aims at doing something that it is good for one to do, Socrates
seems to think, one can only do what one wants to do by doing something that it is, in
fact, good for one to do. In doing something that it would be better for one not to do, one
is inevitably frustrating the very desire on which one is acting. People who rarely, if ever,
do something that it is good for them to do, such as tyrants, will thus rarely, if ever, do
what they want to do, even if they are in a position to do everything they see fit.
This is a reasonable view for Socrates to take, if he thinks that every desire to do some-
thing or other aims at the good in the strong sense that one thing that any such desire is
so constituted as to aim at is the doing of something that it would be genuinely good for
the person to do in the circumstances. On that view, forming a desire to do this or that
necessarily involves envisaging doing this or that as being something that it is, or would
be, good for one to do in the circumstances. Moreover, it is the relevant course of action
envisaged in this particular way that constitutes the object of the desire in question.
On that view, if it seems best to a tyrant, say, to kill a dissident, a full specification of the
tyrant’s object of desire will be something like “killing that dissident, as something that
it is good for me to do in the circumstances.” Of course, there is no way for the tyrant
to perform an action that matches that specification. But that is as it should be. It is pre-
cisely why the tyrant quite definitely fails to do what he wants to do, even though he may
well succeed in killing the dissident.
Interestingly enough, we have independent evidence that it is part of the Socratic
conception of desire, at any rate as Plato thinks of it, that desire aims at the good in
something like the strong sense just explicated. In arguing for the theory of the tripartite
soul in book IV of the Republic, Plato relies on what is known as the Principle of Relatives
(Republic 4, 437d7–439b1). The principle says, in effect, that if A and B are a pair of
relatives, A and B bear the same degree of complexity. For instance, knowledge with-
out qualification and what can be learned without qualification are relatives, and so
are medical knowledge and what can be learned about health and disease. Likewise,
thirst and drink are relatives, and so are, say, small thirst and little drink. Thirst, then,
is a kind of desire that by itself is simply for drink but not, says Socrates, “for much or
little, good or bad drink, or, in a word, for drink qualified in some way or other”
(Republic 439a4–7).
The Socrates of the Republic rejects an imaginary interlocutor’s claim that thirst is
desire for drink qualified as good, “on the grounds that everyone desires good things”
(Republic 438a3–4; cf. Meno 77c1–2). When Glaucon remarks that “the person who says
that has a point” (Republic 438a6), Socrates enunciates and explains the Principle of
Relatives. He then refutes, as being a violation of that principle, the idea that thirst by
512   Hendrik Lorenz

itself is desire for drink qualified as good. On what is philosophically the most natural
and attractive construal of the view that is being rejected, any desire to obtain this or
that (say, food) necessarily involves envisaging the thing in question as being a good
thing to obtain. On this view, it is the relevant thing envisaged in this particular way that
constitutes the object of the desire in question. But that view of how desire aims at
the good looks to be a notational variant of the view employed by the Socrates of the
Gorgias, put in terms of desires to obtain this or that rather than in terms of desires to
do this or that.
It may seem that the Socrates of the Gorgias, in his discussion with Callicles, abandons
Socratic intellectualism by allowing that at least some desires do not aim at the good but,
for instance, at pleasure. In his discussion with Callicles, Socrates speaks of the part of
the soul to which the person’s desires belong (493a3–4), and stresses the need to main-
tain control over one’s pleasures and desires (491d4–e1). He recommends self-control
(491d10–e1) and warns of its lack (525a3–6). Moreover, Callicles advocates enlarging
one’s desires as much as possible and satisfying them with bravery and intelligence
(491e6–492a3; cf. 491b2–4), apparently thinking that without that bravery one might
find oneself with a suitably enlarged desire for some gratification, with access to the
object of one’s desire but without the determination or ruthlessness needed to achieve
(what one would take to be) satisfaction.
However, nothing that Socrates says in the Gorgias entails the rejection of Socratic
intellectualism. The jar that stands for the part of our soul that houses our desires or
appetites is presented as part of a story that Socrates says he heard some clever person
tell (493a5–c3). In turn, it seems to be an elaboration of something Socrates claims to
have heard from some wise person (493a1–5). Socrates notes that “the story is on the
whole a bit strange” (493c3–4). The story does not contain a determinate psychological
theory, and in any case Socrates keeps plenty of distance from it both by reporting it as a
story told by someone else and by noting its strangeness.
As we have seen already, Socratic intellectualism is compatible with accepting the
possibility of lack of self-control, understood as weakness in relation to affects such as
lust or anger or in relation to pleasant things such as food, drink, or sex. The Socrates of
the Protagoras does not deny the possibility of akrasia, understood in this way. Rather,
he is rejecting the view of “the many” that people can be overcome by emotions or
objects of desire all the while knowing, or correctly believing, that what they are doing is
bad for them to do. Being overcome in this way, Socrates holds, is always a matter of
finding oneself in the grip of a false practical belief, and hence is always a manifestation
of ignorance.
In his discussion in the Gorgias with Callicles about the superior individuals who
Callicles thinks should rule their cities, Socrates asks him whether they should rule
themselves, as well as their fellow citizens (491d4–8). Callicles claims not to understand
what Socrates means by self-rule, and Socrates clarifies: “Nothing very subtle. Just what
the many mean: being temperate and in control of oneself, to rule over the pleasures
and desires within oneself ” (491d10–e1). But in the Protagoras, Socrates has offered
“the many,” and us, an analysis of what control over oneself really is: nothing other than
 Plato on the Soul   513

wisdom, the knowledge of good and bad (Protagoras 358c1–3).2 Wisdom ensures that
not only in one’s pursuit and enjoyment of pleasures but quite generally in all one’s
actions, one is invariably guided by one’s own accurate, stable, and unified view of what
is good and bad. The unwise person, by contrast, tends to find himself riddled with false
beliefs about what is good and bad, arising from misleading appearances about plea-
sures and pains that he is unable to correct in anything like a reliable way. He is always
vulnerable to having pleasures and desires forced on him by false appearances about
good and bad that he fails to render powerless.
This intellectualist construal of self-control and its lack is on display in the Protagoras.
If Socrates’ remarks about self-rule in the Gorgias are interpreted along such intellec-
tualist lines, as they certainly can be, what he says in his discussion with Callicles chimes
in well with what he has said in the earlier discussions with Gorgias and Polus. On the
alternative, nonintellectualist interpretation of those remarks, Socrates acknowledges,
in his discussion with Callicles, the existence of human desires that do not aim at the
good and that stand in need of being controlled or repressed. In that case, however, it
becomes mysterious why it should be the case, as Socrates claims it is in his discussion
with Gorgias, that it is impossible for anyone who has learned by instruction what is
just and unjust even to want to perform an act he or she recognizes is unjust (461a4–7;
cf. 460c3). If it is a fact of human nature that we find ourselves with desires that aim not
at the good but at pleasure, what could possibly guarantee that Jones, who has learned
what is just and unjust, cannot form a pleasure-directed desire, say, to eat the last piece
of chocolate cake when he knows perfectly well that in the circumstances it would be
unjust for him to do so? Worse still, Socrates falls into outright incoherence if he
allows it to be a psychological possibility that Jones can, say, steal his neighbor’s money
while judging that doing so is worse, and worse for him, than not doing so, having
been overcome by a pleasure-directed desire for greater wealth. This is because in his
discussion with Polus, Socrates plainly holds that even when we do things that we do
not take to be good in themselves, we still always act for the sake of the good, doing
what we do “because we suppose that doing these things is better for us than not doing
them” (468b1–8).
Better, then, to interpret Socrates’ remarks about self-rule in the Gorgias along the
intellectualist lines familiar from the Protagoras. That is not to say, however, that
Callicles is likely to interpret Socrates’ remarks about self-rule along those lines.
Naturally enough, Callicles’s idea of self-control is one of abstaining from pleasures
and repressing desires because of lack of resources, a sense of shame or scruple, or con-
ventional and misguided thoughts about matters of value. But that is Callicles’s outlook,
not Socrates’. To say that the Socrates of the Gorgias operates within a robustly intellec-
tualist framework is not to deny that the dialogue has a notable undercurrent of psy-
chological complexity. Callicles’s interest in bodily pleasures and desires—in particular,
his idea that one should enlarge one’s desires by indulging them—raises awkward

2 That the Protagoras articulates a “Socratic” conception of akrasia is duly noted in J. Cooper, Reason
and Emotion (Princeton, N.J., 1999), 63 n.54.
514   Hendrik Lorenz

questions for Socratic intellectualism. Since desires such as hunger and thirst plainly
have both physiological and habitual aspects, it seems clear that one can find oneself
with such desires, even intense, pressing ones, regardless of whether or not one thinks
that it would be good for one to eat or drink in the circumstances. Socrates considers the
bodily desires of sick people, noting that doctors often do not allow them to fill them-
selves with what they want (505a6–10). What to say about a sick patient who fully under-
stands and appreciates that he should not now drink, intensely thirsty though he is?
Furthermore, both Callicles and Socrates show interest in people’s sense of shame
and, more broadly speaking, in their sensitivity to the values of their community.3
Callicles envisages a psychologically interesting case of mental conflict, in which some-
one has an enlarged desire for gratification but cannot bring himself to act on it “because
of softness of spirit,” by which he presumably means some kind of scruple or sense of
shame (491b2–4). How do such psychological factors relate to one’s thoughts about good
and bad? Is it a psychological possibility for Jones to think, firmly and without any
wavering, that his eating the last piece of cake would be best and yet to be unable to get
himself to take it because he is worried that he might be perceived as greedy? Those and
other such questions seem close to the surface of the Gorgias. It is tempting to think that
Plato, by foreshadowing the psychological complexities of the middle dialogues, means
to prepare the reader for the developments that culminate in the Republic’s theory of the
tripartite soul.

3. Phaedo and the Immortality

of the Soul

The Socrates of the Phaedo accepts the possibility of psychological conflict without
accepting that the soul, even the embodied soul, is a thing of parts. One of his arguments
against the harmonia theory of the soul, put forward by Simmias (85e3–86d3), relies on
the occurrence of conflicts between desires and also of conflicts between how one
decides to act and how anger or fear incline one to act. Among other things, he points
out that people who are thirsty may nonetheless be averse to drinking (94b8–10).
According to his characterization of such a conflict, the soul opposes the affection in
question and impels the person “towards the opposite, not drinking.” It is presumably
part of the picture that being thirsty involves experiencing an impulse toward drinking,
which the soul may thwart by opposing it.
Socrates adds that the soul may oppose the body and its parts harshly and painfully, as
in athletic exercise or medical treatment, or, more gently, when it converses with one’s
emotions by issuing threats or exhortations (94c9–d6). To illustrate that more gentle
form of opposition, Socrates refers to a passage from Homer’s Odyssey (20, 17–18), which

3 Socrates’ appeals in the Gorgias to his interlocutors sense of shame are discussed in J. Moss,
“Shame, Pleasure, and the Divided Soul,” Oxford Studies in Ancient Philosophy 29 (2005), 137–70.
 Plato on the Soul   515

also features in the argument for the tripartition of the soul in Republic 4. The Socrates of
the Republic takes the passage to provide an example of a conflict between reason and
the spirited part of the soul (Republic 441b5). Confronted with the disloyal behavior of
Penelope’s maids, Odysseus is furious and is sorely tempted to punish them there and
then. However, he also realizes that it would be unwise to reveal himself at this time as
the legitimate king of Ithaca. In an effort to control his anger, he addresses his heart, say-
ing “Endure, my heart, you have endured worse than this” (94d7–e1). The more gentle
form of opposition is characterized as a matter of the soul’s conversing with desires,
anger, or fear, “as one thing that talks to another” (94d5–6).
It is important to note that in the picture of conflict that Socrates presents, the con-
flicting parties are the soul on the one hand and the parts of the body on the other.
Socrates construes the harmonia theory as claiming that the soul is a certain kind of
arrangement or composite of bodily parts (92a7–b2, 92e5–93a9). He holds that no
arrangement could oppose the parts that compose it (93a8–9, 94c3–7). Since the soul
frequently opposes the various bodily parts (94c10–d2) and the affections that reside in
them (94b7–c1), it follows, Socrates argues, that the soul could not be any kind of
arrangement of the parts that compose the body. The argument presupposes that the
bodily parts that compose our organisms are able to form impulses to eat or drink, as
well as emotions such as anger or fear. This idea is nicely illustrated by the Odyssey pas-
sage that Socrates is quoting: there it is Odysseus’s heart that is said to growl in anger and
that Odysseus addresses with soothing words.
The Socrates of the Phaedo, then, takes the body, at any rate while it is ensouled, to
be the subject of what we would call “mental states” of various kinds. In doing so,
he may seem to foreshadow Aristotle’s psychological theory, according to which it is
the ensouled organism, not the soul, that is the subject of mental states (De Anima
1.4, 408b12–15). However, it should be noted that Socrates takes the soul, too, to be a sub-
ject of mental states and acts. After all, he takes the soul to oppose the bodily affections
in various ways—for instance, by pulling one away from drinking when one is thirsty or
by confronting one’s anger or fear with threats and exhortations.
In arguing against the harmonia theory, the Socrates of the Phaedo is recognizing
forms of psychological conflict that Socratic intellectualism cannot accommodate.
In the Odyssey passage to which Socrates refers, Odysseus is determined not to pounce
on the maids there and then, plainly because he is aware that the time has not yet
come to disclose his presence in Ithaca to Penelope’s suitors. Nevertheless, he still feels
driven to kill the maids right away, furious as he is at their flagrantly disloyal and shame-
less behavior. That is why he has to oppose his anger, exhorting his heart to endure. To
put things schematically, Odysseus finds himself with a desire to kill the maids at once,
without in the least believing that doing so would be as good as, or better than, any other
course of action that is available in the circumstances. He has a firm and unwavering
grasp of the fact that killing the maids at once would be immeasurably worse than not
doing so. Nonetheless, he feels a pressing urge to do so. To accept this as a genuine psy-
chological possibility, as the Socrates of the Phaedo shows every sign of doing, is to aban-
don Socratic intellectualism. It seems out of the question that Plato is not aware of this.
516   Hendrik Lorenz

It is unclear whether the Socrates of the Phaedo is prepared to accept that impulses of
the body may prevail over contrary impulses of the soul. Might Odysseus’s soul have
failed to master his heart’s anger? Might he have proceeded to slaughter the maids, all
the while knowing or believing that it would have been better to keep quiet? We do not
know, but Socrates does say that the soul leads all the bodily parts, “opposing nearly all
of them throughout life and mastering them in all sorts of ways” (94c10–d2). Thus he
might think that although the body can form various kinds of emotion and desire, it is
always the soul that determines what is done, accommodating or thwarting the body’s
inclinations as it sees fit.
In that case, he might think that people can, and frequently do, want to act as they
recognize they should not but that they never act on such urges, at any rate not without
revising their views about how it would be best, or better, to act in the circumstances.
As we shall see, this view would amount to a halfway house between Socratic intellectu-
alism and the psychological theory of the Republic. One thing to note about a view along
such lines is that it seems unstable. If people more or less frequently find themselves
with irrational urges, what guarantees that they never act on them, contrary to their
better judgment? What ensures the control of people’s better, more thoughtful, desires
over how they end up behaving? If Plato favored such an intermediate view while writ-
ing the Phaedo, it seems easy to see how that view gave way to the psychological theory
of the Republic, according to which people are capable not only of irrational desire but
also of acting contrary to their own better, more thoughtful, desires.
One striking difference between the Phaedo and the Republic is that while the latter
famously introduces the claim that the embodied soul is a composite of three parts, the
Socrates of the Phaedo seems committed to thinking that even the embodied soul is
incomposite. In the context of the so-called affinity argument (78b–80b), Socrates gets
his interlocutor Cebes to agree that

it is naturally appropriate for what has been combined and what is a composite to
undergo this, to be divided up in the way in which it has been combined. But if
something turns out to be incomposite, for that alone, if for anything, it is appropri-
ate not to undergo these things. (Phaedo 78c1–5)

Socrates is aware that the affinity argument is not sufficient to prove that the soul is
indissoluble. But he is firmly committed to the immortality of the soul, of course. He
goes along with Simmias and Cebes (77b–e) in thinking that for a given soul to be
divided up, or dispersed, is to be destroyed (80d5–e1). Thus he must think that souls will,
for whatever reason, not be divided up. If he thinks that souls are composites, he owes an
explanation of why they will not be divided up, even though, being composites, they are
naturally such as to be vulnerable to division. He shows no concern at all to offer such an
explanation. This seems best explained by supposing that he takes each soul to be
incomposite. In any case, there is no sign in the text that the Socrates of the Phaedo takes
the soul to be a combination or composite of constituent parts.
At this point, it is important to recall that the Phaedo treats the body as a subject of
impulses and emotions. Conflicts that the Republic conceives as taking place within the
 Plato on the Soul   517

soul and hence as showing the existence in the soul of distinct and conflicting parts are
treated in the Phaedo as conflicts in which the soul opposes suitable parts of the body, such
as the heart. On this picture, such conflicts may leave the soul itself entirely undivided.
One might think that the Phaedo can preserve the unity of the soul only by sacrificing
the unity of the psychological subject, ascribing some psychological states to the soul
and others to a distinct subject, the body or some suitable part of it. One might also
think that forming even the simplest desire is a task that could never be accomplished by
a mere body or bodily part, but only by a soul (cf. Philebus 34d10–35d6). We do not know
whether these considerations were among Plato’s reasons for abandoning the Phaedo’s
picture of motivational conflict; however, we do know that the Socrates of the Republic
ascribes all psychological states to the soul and none to the body, conceiving of the soul,
at least in its embodied state, as a composite of three parts.
When he turns to the soul’s immortality in Republic 10, Socrates confronts the ques-
tion how a thing of parts could turn out to be everlasting. “We must not think,” he says,
“that the soul in its truest nature is full of multicoloured variety and unlikeness or that it
differs with itself. . . . It isn’t easy for a composite of many parts to be everlasting if it isn’t
composed in the finest way, yet this is how the soul now appeared to us” (Republic
611a10–b7). What he goes on to say (at 611b–612a) is indeterminate between two signifi-
cant alternatives. Once separated from the body, the soul might be a composite that is
composed finely enough to be everlasting; or it might then be incomposite and wholly
rational, because reason is the only part of it that is immortal. The Timaeus operates with
the second of these alternatives, conceiving of reason alone as immortal and of appetite
and spirit as together constituting the soul’s mortal part (Timaeus 69c5–d4). The
Phaedrus, by contrast, seems to explore the first alternative: it likens the disembodied
human soul to a charioteer in charge of two horses—one good, the other bad—which
plainly stand for spirit and appetite (Phaedrus 246a–250c). In the disembodied condi-
tion, there may be no room for conflict between the driver’s directives and any desires
the horses might have. In the case of human souls, as opposed to divine ones, the driver
has a hard time maintaining control over his chariot (247b3–6, 248a1–6). But the driver’s
troubles may arise simply from the disparity of his horses and the heaviness of the bad
horse (247b3–5) rather than from any irrational desires that the horses might have.

4. Republic

It is time to turn to Republic 4, in which Socrates attempts to establish by argument

that the embodied human soul is composed of three parts. The argument is better
than its reputation.4 Its conclusion is momentous. It follows validly from premises

4 “No tutor,” wrote Gilbert Ryle in 1947, “would accept from a pupil the reasons given by Plato
for . . . the doctrine that the Soul is tripartite.” That statement is from his review of Popper’s The Open
Society and Its Enemies. Ryle’s dim view of the argument is revived by Myles Burnyeat, in “The Truth of
Tripartition,” Proceedings of the Aristotelian Society 106 (2006), 1–4.
518   Hendrik Lorenz

that Socrates provides. And each of those premises would have seemed to an intelligent,
well-informed ancient reader at least a plausible candidate for truth.

4.1 The Principle of Opposites

The argument relies crucially on a principle whose truth Socrates claims is clear (436b9),
though he is careful enough to note that, for the purposes of the discussion with Glaucon
and the others, it has and retains the status of a hypothesis (437a3–8). This is the so-
called Principle of Opposites (PO; 436b9–10):

PO: The same thing cannot do or undergo opposites in the same respect, in relation
to the same thing, and at the same time.

The principle stated, Socrates proceeds to defend it against two apparent counterex-
amples, thereby clarifying significantly what it is meant to come to. The first apparent
counterexample is a person who stands still but moves her arms and head. Socrates
rejects this as a counterexample to PO, analyzing it not as a case of the same thing being
simultaneously at rest and in motion but, rather, as a case of one part of the person being
at rest while another is in motion (436d1–2). Just as in the parallel case of the archer,
whose one arm pushes the bow in one direction while the other arm pulls it the other
way (439b8–c1), Socrates accepts that the predicates in question are opposites and that
they apply at the same time. In both cases, he holds that the opposite predicates belong
to distinct bearers. Strictly speaking, it is the person’s arms and head (or whatever) that
are in motion, while other parts of her body remain at rest. This need not mean that it is
false that the person is in motion and at rest at the same time, only that by saying this,
one does not succeed in accurately pinpointing what it is that is, respectively, in motion
or at rest. As we might say, by saying this, one fails to pick out the proper subjects of the
predicates in question.
The second apparent counterexample to PO that Socrates considers is a spinning top,
which, according to an imaginary interlocutor, is as a whole in motion and at rest at the
same time (436d4–6). The point of this more sophisticated example (note 436d4) seems
precisely to be that it cannot be resolved by identifying distinct subjects for the opposites
motion and rest.5 Socrates rejects this, too, as a counterexample to PO. He points out
that a top has a certain complexity and can undergo motion in more ways than one.
Given that it has an axis, it can incline or wobble in a certain way. Given that it is suitably
curved, it can rotate without coming to occupy a different place. On Socrates’ analysis of
the spinning top example, the top is “at rest with respect to its axis” (436e1–2), which is to
say that it does not incline or wobble. At the same time, he notes, it is “in circular motion
with respect to its curved surface” (436e2–3).

5 This is pointed out by C. Bobonich, Plato’s Utopia Recast: His Later Ethics and Politics (Oxford,
2002), 226–35.
 Plato on the Soul   519

This analysis distinguishes between two different ways in which a top may be in
motion or at rest: it may undergo motion that affects the axis of its rotation, or it may
undergo motion that affects its curved surface, sending the surface around in circles
(cf. Aristotle, Physics 6.9, 240a28–b7). This is not to say that what it is about the top
that is in circular motion is specifically and exclusively its curved surface. It is plainly
not the case that the top’s surface is going around in circles, whereas the other parts of
the thing remain at rest. Rather, Socrates quite reasonably identifies rotation, as opposed
to inclination, in terms of its being a form of motion that affects the top’s curved sur-
face—namely, by sending it around in circles. This allows him to accept the imaginary
interlocutor’s nice point that, at the same time, a spinning top is as a whole at rest and in
motion. As far as inclination or motion with respect to the top’s axis is concerned, it is at
rest and is at rest as a whole. As far as rotation or motion with respect to the top’s curved
surface is concerned, it is in motion and is in motion as a whole.
By disarming these two apparent counterexamples to PO, Socrates has made available
two importantly different methods for dealing with cases in which a given subject does
opposites at the same time. One method treats the subject as a composite and ascribes
the one opposite to one part of it and the other to another. The other method distin-
guishes between different respects in which something or other can do the opposites in
question, so that one can say that the subject as a whole does the one opposite in one
respect and the other in another. We may add that the first method treats the opposites
in question, such as unqualified motion and unqualified rest, as full-on opposites, tak-
ing it that nothing can simultaneously serve as their proper subject. The second method,
by contrast, amplifies the predicates in question by attaching appropriate qualifications
and thereby shows the situation not to involve a clash of full-on opposites. Nothing
prevents a suitable object from simultaneously serving as the proper subject of both
members of such a pair of qualified opposites.
The argument for tripartition presupposes that desire and aversion are opposites in
the sense of the word that is in play in PO. Desire and aversion are not logically incom-
patible, and so it will not do to interpret PO as a purely formal truth to the effect that
logically incompatible predicates cannot apply to the same subject at the same time, in
relation to the same thing, and in the same respect. Socrates is quite explicit about the
fact that he does take desire and aversion, as well as assenting and rejecting, to be oppo-
sites in the relevant sense (437b1–c5). It is somewhat unclear on what basis he takes that
view. It may well be that he takes them to be opposites in the relevant sense simply
because they are extremes that delimit a range of states or attitudes, much as Aristotle
takes black and white, for instance, to be opposites because he thinks that they delimit
the range of colors. In that case, Socrates’ acceptance of PO commits him to the view
that, for any such range, nothing can at the same time serve as the proper subject of both
extremes in the same respect and in relation to the same thing.
Moreover, Socrates evidently takes desiring something and being averse to it to be
full-on opposites, treating a case of desire for something and simultaneous aversion
to the same thing like the case of the archer’s arms pulling the bow in one direction and
at the same time pushing it the other way (439b3–c8). One thing this makes clear is that
520   Hendrik Lorenz

he thinks of the parts of the soul that the argument brings to light as the bearers of
­psychological states such as desires and aversions.6 That this is the way he conceives
of the parts of the soul is corroborated when later on in the Republic he also ascribes to
them emotions, pleasures, and beliefs.7
Why does he treat desiring something and being averse to it as full-on opposites?
There is some indication that he takes desires and aversions to involve motions or
impulses of the soul toward, or away from, the object of desire or aversion. He repeatedly
describes parts of the soul as pulling or dragging the rest of the soul toward the object of
desire (439b4, d1, 604a10, b1–2, etc.), sometimes with another part pulling the other way
(e.g., 439b3). When Aristotle uses the same kind of language (e.g., Nicomachean Ethics
1.13, 1102b16–18), we know not to take it literally because he takes care to make clear in
book I of the De Anima that he takes the soul not to effect motion by itself engaging in
some kind of motion (1.3, 406b24–25, 1.4, 408b30–31). But in the same context, he criti-
cizes Plato, among others, for having held precisely the view that the soul imparts motion
to the animal by itself engaging in motion (De Anima 1.3, 406b26–28). Thus it is probable
that Plato conceived of desires and aversions in straightforwardly directional terms, as a
matter of the soul impelling you toward something or away from it. This seems a rather
natural thought, and it would make it reasonable for Socrates to analyze his cases of
psychological conflict like the case of the archer’s arms simultaneously pulling the bow
one way and pushing it the other way, rather than like the case of the spinning top, which
is as a whole in motion in one respect and at rest in another.
Now, one might think that if desire and aversion are opposites in the sense relevant
to PO, and if desiring something and being averse to it are full-on opposites, then the
psychological conflicts to which Socrates appeals in the course of the argument for
tripartition prove PO to be false. For if desiring to φ and being averse to φ-ing are full-on
opposites, then the fact that people can at the same time desire to φ and be averse to
φ-ing just shows, one might think, that there are things that can at the same time
do opposites in the same respect and in relation to the same thing. After all, a person or
soul plainly is one thing.8 However, having attended carefully to the way Socrates
resolves his first apparent counterexample to PO, we can see that this objection depends
on a questionable understanding of what PO is meant to come to. Considering the case
of a person standing still and, at the same time, moving her arms and head, Socrates
takes care to pinpoint what precisely it is that, respectively, is in motion or at rest. Strictly
speaking, it is only some parts of her body that are in motion, while other parts remain
at rest. Thus the case is no genuine counterexample to PO. Note that Socrates’ analysis
does not require denying that a person is one thing. It only requires treating the person
as a composite, so that one can identify distinct parts that will serve as the subjects of the
relevant predicates.

6 Contra T. Irwin, Plato’s Ethics (Oxford, 1995), 204–05.

7 Emotions are ascribed as follows: excitement at 439d7; anger at 441c2; indifference at 581b8;
pleasures at 580d6–7; beliefs at 442c9–d2, 571d1–4, 603a1–2.
8 Thus Burnyeat, “Truth of Tripartition,” 4.
 Plato on the Soul   521

Likewise, Socrates can accept that a given soul is one thing, without having to accept
that his examples of psychological conflict, as he analyzes them, show PO to be false. He
will only have to say that if people do sometimes desire this or that and are at the same
time averse to it, this shows that the soul is a composite, with distinct parts of it available
to serve as the bearers of conflicting desires and aversions. This he is ready to say. What
PO, strictly speaking, rules out, then, is not that opposites can simultaneously apply to
the same thing in the same respect and in relation to the same thing. Rather, what it rules
out, strictly speaking, is that one and the same thing can simultaneously serve as the
proper subject of two opposite predicates, which apply to it in the same respect and in
relation to the same thing. There is no reason to think that any of the examples of psy-
chological conflict considered in the argument for tripartition, as Socrates analyzes
them, shows PO to be false, once it is understood along those lines. It is to Socrates’
examples of psychological conflict that we now turn.

4.2 The Examples of Conflict

The first example concerns someone who simply is thirsty, so that Socrates can apply the
Principle of Relatives to obtain the result that what the person wants is simply to drink.
It happens frequently, Glaucon agrees, that someone who is thirsty in this way is averse
to drinking (439c5). Suppose that the person understands that, given his medical
condition, he should abstain from drinking anything at all. Given that the person is
averse to precisely the thing he wants, Socrates has a case of something simultaneously
doing opposites in relation to precisely the same thing. If desiring to φ and being averse
to φ-ing are full-on opposites, then Socrates can at this stage infer, and infer validly, that
the soul must be a composite of at least two parts. Moreover, he holds that the two parts
that have come to light are reason and appetite, on the grounds that the aversion to
drinking, in the case considered and in others like it, stems from reasoning or calcula-
tion, while the desire to drink arises from “affections and diseases,” by which he means
physiological imbalances such as dehydration (439c10–d2).
Later remarks, in book IV and beyond, yield a rather detailed picture of reason and
appetite, as well as of spirit, to which we turn presently. According to that picture,
reason, at any rate if it is tolerably well developed, is “always wholly straining to know
where the truth lies” (581b6–8), and so Socrates also calls it the learning-loving and
philosophical part of the soul (581b10–11). It is also the natural ruler of the soul: a prop-
erly developed reason is knowledgeable about what is advantageous for each part and
for the soul as a whole (441c4–7). It issues directives about how to act and how not to,
and it gives rise to corresponding desires and aversions, which are desires and aversions
of reason itself. The desires of appetite are related both to physiological imbalances
(439d1–2) and to indulgences and pleasures (439d6–8). The objects of appetitive
desire prominently include the pleasures of food, drink, and sex (436a10–b2, 442a7–b1),
but if and when the person in question comes to take pleasure in a broader range of
activities, appetite’s interests expand as well. Thus people may form appetites to have a
522   Hendrik Lorenz

bit of fun by participating in philosophical discussion or by sounding off in political

debates (561c6– d4). What unifies appetitive attitudes is that they are impulses toward
perceived or expected pleasures or away from perceived or expected pains.
In arguing for the existence of the spirited part of the soul, Socrates relies on two more
examples of psychological conflict to show the distinctness of spirit first from appetite
and then from reason. Although he does not specify or analyze those examples as explic-
itly as one might wish, it seems clear that they are meant to follow the pattern set by his
first example of psychological conflict: in all three cases, a person experiences a conflict
between wanting to do something or other and at the same time being averse to doing it.
Leontius, probably a notorious necrophile, wants to take a close look at some corpses
lying by the side of the road and is, at the same time, averse to doing so (439e5–440a4).
Odysseus can hardly restrain his impulse to punish the maids there and then, but he
does manage to control himself, knowing that it would be unwise now to reveal himself
as the legitimate king of Ithaca (441b3–c2).
It would have been easy enough for Plato’s contemporaries to see that Leontius’s
aversion to gazing at the corpses is a spirited attitude, one that springs from his sense of
what is and what is not praiseworthy and respectable. Before running up to the corpses,
Socrates says, Leontius feels disgust, turns himself away, and covers his face. For a while
he does battle with himself, then appetite defeats him (440a1–2). As he runs toward the
corpses, he angrily addresses his eyes, saying, “Look for yourselves, you evil wretches,
take your fill of the beautiful sight!” (440a3–4). Socrates takes the story to indicate that
anger sometimes does battle with the appetites—and hence that anger and appetite are
distinct from one another, since it takes two parties to do battle (440a6–7). This, in turn,
he takes to show the distinctness of spirit from appetite. To understand what he has in
mind, we need to recall that he has already noted the psychological link between anger
and the character trait of spiritedness, by associating being impelled to act from anger
(436a10) with the spiritedness that he takes to be particularly notable in the ideal city’s
military class (435d9–436a3).
How Plato and his contemporaries conceive of spiritedness is a complicated matter.9
For present purposes, a sketch may suffice. It is a character trait or psychological ten-
dency that includes as a central aspect eagerness to distinguish oneself and to gain and
maintain the esteem and respect of others in one’s community. It crucially involves an
awareness of one’s social position and of one’s merits. It also involves solicitude that one’s
status be duly noted and respected by others and that such merits as one might have be
recognized and honored. Hence the spirited person’s sensitivity to slights, insults, and
the like, and the connection between spiritedness and anger. One thing that being
notably spirited must involve is a vivid sense of what does and does not count as praise-
worthy and respectable by the lights of one’s community. Socrates and Glaucon are
agreed at the outset of the argument for tripartition of the soul that each of us has this
characteristic, at least to some extent (435d9–436a3). The idea is presumably that every

9 A good place to start is A. Hobbs, Plato and the Hero: Courage, Manliness and the Impersonal Good
(Cambridge, 2000), 1–31.
 Plato on the Soul   523

ordinarily constituted and developed human being is spirited, at least to some extent.
This entails, among other things, that we all have a more or less vivid sense of what does
and does not count as praiseworthy and respectable by the lights of our community, and
also that we are all motivated to behave as that sense demands, at least to some degree,
and to avoid acting in ways that would violate it.
Against this background and given Socrates’ purpose at this stage in the argument,
the Leontius anecdote is exquisitely chosen. Leontius’s angry exclamation, addressed to
his own eyes, dramatizes the frustration of his passionate aversion to the very course
of action that at the same time attracts him irresistibly. Like the anger that wells up in
him as he fails to maintain control over himself, the aversion that in the end is overcome
by appetite is naturally and plausibly thought of as being intimately connected with
Leontius’s spiritedness. It is an aversion that springs from his awareness that gratifying
oneself by gazing at the corpses would be an ugly and disreputable thing to do. That is
why he feels disgust as he struggles with himself (439e8), and that is why he refers to the
sight of the corpses, with grim irony, as a beautiful one.

4.3 Comments on the Argument

It is worth noting that reason is not even mentioned in Socrates’ analysis of the Leontius
episode. Leontius is brought in to show that spirit is not, contrary to what Glaucon is
initially inclined to think (439e1–4), an aspect of the appetitive part, but is a part of the
soul distinct from appetite. For all that Socrates says, the rational part of Leontius’s soul
might have been entirely inactive, both at the time of conflict between appetite and spirit
and when Leontius ran toward the corpses, impelled by appetite. Perhaps the idea is that
Leontius’s reason was incapacitated by intense emotion. Perhaps it did take the view that
it is better to abstain and lent support to spirit’s aversion.
In any case, there is no indication at all in the text that Socrates takes reason somehow
to assist appetite in forming the desire that Leontius proves unable to control. Socrates’
analysis treats appetite as being capable both of forming determinate, situation-specific
desires and of getting a person to perform a certain action. The appetitive desire that
overcomes Leontius is not a mere craving for pleasure or sexual gratification. It is a
desire to take a close look at the corpses lying by the side of the road (439d7–8). All it
takes for Leontius to start running toward the corpses is for this appetitive desire to
overcome the aversion that resists it. Socrates goes on:

Don’t we often see in other cases, too, that when appetite compels someone contrary
to reasoning, he reproaches himself and gets angry with that in him that’s doing the
compelling, so that of the two factions that are fighting a civil war, so to speak, spirit
allies itself with reason? (440a9–b4)

It is part of the picture that appetite may find itself at war with the allied forces of reason
and spirit, and that it may compel a person to act contrary to reason, much as Leontius’s
524   Hendrik Lorenz

appetite compels him to run toward the corpses for sexual gratification. Again, this
presupposes that appetite can by itself form determinate impulses to act in certain ways.
This, in turn, requires that appetite is equipped with cognitive resources that enable it to
apprehend suitably determinate and specific objects of desire.
That the psychological theory of the Republic treats even appetite as being equipped
with considerable cognitive resources, such as are required for the apprehension of quite
specific objects of desire, becomes perfectly clear at the beginning of book IX, when
Socrates describes some of the lawless desires that arise in people while they are asleep:

Then the brutish and savage part, full of food and drink, casts off sleep and seeks to
find a way to gratify itself. You know that there is nothing it won’t dare to do at such
a time, free of all control by shame or reason. It doesn’t shrink from trying to have
sex with the person’s mother, as it supposes, or with anyone else at all, whether man,
god, or beast. (571c3–d3)

The Socrates of the Republic, then, takes even the appetitive part of the soul to be able
to form impulses to do specific things, as well as to get a person to act in certain ways
without having to be assisted in some way or other by the other parts of the soul. It is
important, and has been duly stressed, that the psychological theory of Plato’s Republic
takes reason not to be inert but to have its own attachments and desires.10 It is equally
important that the theory takes appetite, and no doubt spirit as well, not to be blind but
to have their own forms of sensitivity and cognition.
This is not to say, however, that appetite and spirit are equipped with the same cogni-
tive resources as reason. For instance, there is good reason to think that Plato’s theory
takes neither appetite nor spirit to be capable of means-end reasoning.11 To begin with,
Socrates refers to the appetitive part as alogiston—that is, as being such as not to engage
in logismos (439d7; cf. 441c1–2). In Plato’s Timaeus, a later dialogue that restates the the-
ory of the tripartite soul, the main speaker Timaeus says that appetite has no share in
logismos at all (Timaeus 77b3–6). Since Plato treats straightforward cases of means-end
reasoning as cases of logismos,12 not engaging in logismos excludes engaging in means-
end reasoning.
Moreover, if the non-rational parts of the soul can reason about how best to satisfy
their desires, and can form desires and aversions on the basis of such reasoning, there
seems to be no satisfactory way for Plato to rule out the simultaneous occurrence within
appetite or spirit of both a desire and an aversion in relation to the same thing. For
instance, having a burger right now may seem a very pleasant thing to do, but it may also
seem an obstacle to one’s full enjoyment of the exquisite dinner party one expects to

10 Cooper, Reason and Emotion, 121–26; M. Frede, Introduction to M. Frede and G. Striker (eds.),
Rationality in Greek Thought (Oxford, 1996), 5–7.
11 Contra (among others) J. Annas, An Introduction to Plato’s Republic (Oxford, 1981), 129–30;
Bobonich, Plato’s Utopia Recast, 244.
12 For instance, Republic 553d2–4, 604d4–5; cf. Timaeus 30a6–b6, 33a6–b1.
 Plato on the Soul   525

attend in an hour’s time. Suppose that appetite, on that basis, forms a reasoned aversion
to having the burger now. If the person in question nonetheless continues to have a plea-
sure-directed desire to have the burger right away, that desire must belong to appetite.
Notice that we now have a situation that is relevantly like the example to which
Socrates appeals in order to show the distinctness of reason and appetite (439b3– d2): a
physiologically based desire for something or other and a simultaneous reasoned aver-
sion to the same thing. Socrates’ argument presupposes that conflicts of this kind show
that what undergoes the conflict, the soul or whatever it may be, must have at least two
distinct parts, which constitute the conflicting parties. Furthermore, he takes the con-
flicting parties in question to be appetite and reason. This crucial step in his argument
would be undermined if he thought that appetite by itself can form desires and aversions
based on its own reasoning about how best to satisfy its desires. In that case, he would
have no basis for holding that his first example of psychological conflict brings to light
appetite on the one hand and reason on the other. It might equally well be taken to show
the complexity of appetite. On the other hand, Socrates can reasonably hold that the
example brings to light appetite and reason, if he takes reason, but not appetite, to be
responsible for reasoned desires for, and aversions to, this or that as a means, or obstacle,
to the achievement of some given goal.
The strongest prima facie support for the view that the Republic takes even appetite to
be capable of means-end reasoning is that Socrates holds appetite to be a lover of money
(or property) and profit, explaining that it is most of all through money that appetite’s
desires for food, drink, sex, and the like are satisfied (580e2–581a1). In saying this, he
may have in mind that appetite loves money specifically as a means to the satisfaction of
its primary desires. But the idea may also be that appetite tends to form and maintain an
intense, pleasure-directed attachment to money itself, because it, more than anything
else, provides access to the pleasures of eating, drinking, and the like; as a result, suitable
psychological mechanisms bring it about that thoughts of money come to be intimately
associated with thoughts of such pleasures.
Now, if appetite cared about money specifically as a means to the satisfactions of its
other desires, that would make it hard to resist the conclusion that it is capable of at least
basic means-end reasoning. But Socrates never says that appetite cares about money
specifically as a means. In fact, he makes clear that his paradigmatic man of appetite, the
moneymaker, loves money not as a mere means but in the pleasure-directed way that is
characteristic of appetite. Comparing the philosopher, the honor-lover, and the money-
lover in terms of how truly pleasant their lives are, he notes that the moneymaker will
say that “the pleasure of being honoured and that of learning are worthless compared to
that of making profit, if he gets no money out of them” (581c11–d3).
The moneymaker is endowed with reason and capable of means-end reasoning, of
course. He knows that money is an effective means to the satisfaction of all sorts of
pleasure-directed desires, and his interest in money no doubt rests in important part on
his awareness of this fact. However, what Socrates says makes clear that the moneymaker’s
concern for money rests not only on his awareness that it is an effective means, the
526   Hendrik Lorenz

way one might care for some pills only because one knows that they are effective against
headache. Socrates says that the moneymaker takes pleasure in making profit, the way
an honor-lover takes pleasure in being honored and a lover of wisdom takes pleasure in
learning. The thought is that the moneymaker is passionate about money and its acqui-
sition in part because it is something that is, for him, in itself a potent source of pleasure.
Moreover, Socrates holds that while the love of wealth is particularly prominent in the
moneymaker, it is a psychological tendency that is to some degree or other present in
everyone (435d9–436a3). This chimes in well with the fact that he takes money and profit
to be among the canonical objects of appetitive desire and pleasure (581a3–7). In every
ordinarily constituted and developed human being, Plato seems to think, appetite is at
least somewhat attached to money as something that is, for it, in itself a source of pleas-
ure. One thing this presupposes is that there are psychological mechanisms by which
reason-endowed creatures, if they live in a suitable cultural environment and participate
in suitable practices, will typically form attachments to money and its acquisition as
things that in themselves give them pleasure.
The Republic does not offer anything like a systematic account of the psychological
mechanisms that are at work when people form such attachments. Plato may well think
that such attachments tend to be prefigured in early childhood, as people internalize the
beliefs and values of the culture that surrounds them—for instance, by way of hearing
the stories and myths in which those beliefs and values are given concrete expression. In
his remarks about poetry in books II and III of the Republic, Socrates emphasizes that in
the corrupt culture of his contemporary Athens, even young children are already busy
absorbing false and damaging beliefs, including ones that are apt to turn them into
people in whom the love of wealth is unduly prominent (390d7–391a2; cf. 377a11–b8).
In this way, even young children may already find themselves with, say, the unreflective
belief that wealth is a delightful and wondrous thing. Later on in their lives, as they begin
to reason about how to achieve their various goals (note 441a7–b2), their own recogni-
tion that money is an effective means to the satisfaction of this or that appetite may not
just provide their attachment to money with a new basis. It may also serve to reinforce
their appetitive, pleasure-directed attachment to money, as it establishes fresh connec-
tions in their minds, and perhaps strengthens old ones, between money and various
pleasant things that money can buy: delicious meals, fine wines, and so forth.
In any case, once one accepts that reason-endowed creatures, given a suitable cultural
environment, will tend to form a pleasure-directed attachment to money, it is easy to see
how the appetitive part of the soul can come to be attached to money and its acquisition,
as Socrates evidently holds that it can do and, in fact, very much tends to do. Note that
the formation and maintenance of such appetitive attachments does not require that
appetite itself be capable of means-end reasoning or of recognizing that money is an
effective means to the satisfaction of certain other desires it may have. It may well be the
case that from appetite’s own point of view, eating, drinking, having sex, making a profit,
and the occasional bit of philosophizing or speechifying are all on a par: at the appetitive
level, all of these activities seem attractive, if and when they do, simply because, at the
time, they seem pleasant things to do.
 Plato on the Soul   527

5. Desire and the Good

The psychological theory of the Republic is a dramatic departure from the intellectual-
ism of earlier dialogues. From the point of view of the later theory, it may seem as if the
Socrates of the Protagoras and the Gorgias is blind to the fact that humans are not only
driven to pursue whatever they take to be the good but also find themselves with desires
and aversions that flow directly from their natural aspiration to distinguish themselves
and from their equally natural anxiousness to maintain and enhance their social status,
as well as from their natural and immediate attraction to what they perceive or expect to
be pleasant and their equally natural and immediate aversion to what they perceive or
expect to be painful.
But it is not only that the Socrates of the Republic sharply disagrees with the views
expressed by the Socrates of the Protagoras and the Gorgias in holding that adult, ordi-
narily developed humans frequently find themselves with desires, even intense, pressing
ones, for various sorts of things that they know, or anyhow believe at the precise time, it
would be better for them to keep away from. This in itself would be a significant theoret-
ical development and surely one to be welcomed as representing a cluster of important
insights about the human condition. However, the Socrates of the Republic goes further
beyond the earlier dialogues than this, by taking it that appetitive and spirited desires
can by themselves get even adult, ordinarily developed humans to act in certain ways,
either because the desire in question is not resisted by the person’s rational part or
because it overcomes any resistance that reason might put up. On this view, then, it is
not just that not every human desire aims at the good or even at what seems to the
person to be the good. It is also that not every human action aims at the good in either
of these ways. It is a fact of human psychology, the Socrates of the Republic is committed
to thinking, that adult, ordinarily developed humans are quite capable of doing things
that they think are thoroughly bad, even at the precise time of action, all the while
believing that a better course of action is available to them.
One thing this makes clear is that a much-quoted passage in book VI of the Republic
needs to be handled carefully. Many readers have thought that the passage restates at
least a somewhat qualified version of Socratic intellectualism.13 In that passage, Socrates
says about the good that

every soul pursues it and does everything for its sake. It divines that the good is
something but is perplexed and cannot adequately grasp what it is or acquire the
sort of stable beliefs it has about other things, and so it misses the benefit, if any, that
even those other things may give. (Republic 6, 505e1–5)

13 For example, G. Carone, “Akrasia in the Republic: Does Plato Change His Mind?,” Oxford Studies
in Ancient Philosophy 20 (2001), 132–35; M. Anagnostopoulos, “The Divided Soul and the Desire for
Good in Plato’s Republic,” in G. Santas (ed.), The Blackwell Guide to Plato’s Republic (Oxford, 2006),
181–83.
528   Hendrik Lorenz

It is worth noting that Plato’s Greek in the first sentence quoted can quite naturally be
interpreted as meaning that every soul exerts itself immensely for the sake of the good,14
rather than that every soul does whatever it does for the sake of the good.15 On the first
interpretation, Socrates is making the substantial claim that one thing that everyone is
immensely concerned to obtain or promote is the good. This leaves it open that there
might be things other than the good that people pursue and also that they sometimes
do things without, in doing the thing in question, acting for the sake of the good. It thus
leaves open the possibility that people sometimes do this or that without supposing
that, in doing what they are doing, they are making progress in obtaining or promoting
the good.
On the alternative interpretation, the claim is that everyone acts for the sake of the
good in everything he or she does. In other words, people always do whatever they do
because they suppose that doing the thing in question will conduce to obtaining or pro-
moting the good, however they may conceive of it. But no one who has read book IV
with the care it deserves will be inclined to prefer this alternative interpretation to the
other. For it is plainly part of the psychological theory that is presented and argued for in
book IV that people are capable of doing something or other not because they think that
acting in this way conduces to obtaining or promoting the good, however they may con-
ceive of it, but, for instance, because they are overcome either by a powerful pleasure-
directed impulse or by an angry desire to retaliate or inflict punishment. In acting in
such ways, people may do things that they themselves realize are seriously detrimental
to their pursuit of the good, as they conceive of it.
Given the context of the Republic, then, Socrates’ remark in book VI about every
soul’s pursuit of the good is best understood as the claim that the good is something that
every soul pursues and for the sake of which it exerts itself immensely. In saying this, the
Socrates of the Republic is not reverting to the Socrates of the earlier dialogues. He is
only highlighting the fact that even though the psychological theory of the Republic
does depart dramatically from the intellectualism of earlier dialogues, it nonetheless
preserves and accommodates what certainly is one of its central and most important
commitments: that all of us, in virtue of being endowed with reason, are naturally ori-
ented toward the good, no matter how misguided or confused our views about it may be.

Bibliography
Anagnostopoulos, M. “The Divided Soul and the Desire for Good in Plato’s Republic,” in
G. Santas (ed.), The Blackwell Guide to Plato’s Republic (Oxford, 2006), 166–89.
Annas, J. An Introduction to Plato’s Republic (Oxford, 1981).
Bobonich, C. Plato’s Utopia Recast: His Later Ethics and Politics (Oxford, 2002).

14 On panta prattein at 505e1–2, see T. Irwin, Plato’s Moral Theory (Oxford, 1977), 336 n.45.
15 The notion of acting for the sake of the good is not explicated in the context of the remark in
Republic 6. In the Gorgias, it is introduced as apparently equivalent to the notion of doing something
supposing it to be better—better, that is, than not doing it (Gorgias 468b1–8).
 Plato on the Soul   529

Burnyeat, M. “The Truth of Tripartition,” Proceedings of the Aristotelian Society 106 (2006),
1–22.
Carone, G. “Akrasia in the Republic: Does Plato Change His Mind?,” Oxford Studies in Ancient
Philosophy 20 (2001), 107–48.
Cooper, J. (ed.) Plato: Complete Works (Indianapolis, Ind., 1997).
Cooper, J. Reason and Emotion (Princeton, N.J., 1999).
Cooper, J. “Plato’s Theory of Human Motivation,” in his Reason and Emotion, 118–37.
Cooper, J. “Socrates and Plato in Plato’s Gorgias,” in his Reason and Emotion, 29–75.
Frede, M. “Introduction,” in M. Frede and G. Striker (eds.), Rationality in Greek Thought
(Oxford, 1996), 1–28.
Hobbs, A. Plato and the Hero: Courage, Manliness and the Impersonal Good (Cambridge,
2000).
Irwin, T. Plato’s Ethics (Oxford, 1995).
Irwin, T. Plato’s Moral Theory (Oxford, 1977).
Moss, J. “Shame, Pleasure, and the Divided Soul,” Oxford Studies in Ancient Philosophy 29
(2005), 137–70.
 chapter 22

Pl ato’s Ethics

Julia Annas

Ethics, in the sense of a concern to act rightly and to live a good life, is pervasive in
Plato’s work, and so we find Plato’s ethical thinking throughout the dialogues. Even the
Sophist, whose major theme is the problem of being and not-being, examines this in
the context of discovering what is distinctive about sophistry, which can corrupt our
attempt to live well. For Plato philosophical inquiry, however far it may get from the
immediately practical, is never detached from the framework of living a good, worth-
while life.
Ethical concerns are found in dialogues of the most varied types, from those in which
Socrates shows other people that they fail to understand the claims they make about
courage, friendship, or virtue to those in which he, or another person, gives long,
sometimes uninterrupted speeches on a variety of topics. And we find ethics treated
sometimes on its own, sometimes in political contexts, and sometimes in a framework
of metaphysical theorizing.
Given this, it is clearly always important, when examining Plato’s ethical arguments,
to pay attention to their role in the dialogue in which they occur. Nonetheless, it is legiti-
mate to extract ethics, as a subject, from the dialogues and to outline a Platonic theory of
it. Differing as they do between dialogues, his discussions of ethical concerns do fall
into patterns which can be brought together and seen to have a distinctive structure.
In antiquity, Plato is (apart from fragmentary ideas in Democritus) the first philosopher
to form a tradition of specifically ethical theory. We find recurring themes, which Plato
can reasonably be seen as the first to unify into a recognizably ethical theory. This, of
course, does not exclude noticing differences between dialogues, and some of these are
best explained as suggesting changes of mind on Plato’s part.
With some issues, it makes a large difference whether an interpreter takes there to be
an overall development in Plato’s thought or, rather, a continuing overall concern allow-
ing for particular changes. In this chapter, I do not take a stand on this very contentious
issue. The so-called Socratic dialogues in which Socrates displays what is wrong with the
thoughts of others on virtue are often taken to be an early stage of Plato’s thought, one in
532   Julia Annas

which he had not yet developed positive ethical views of his own.1 They can just as well,
however, be read as complementary, rather than prior to, the passages where Socrates
puts forward positions in his own right, and this is how I read them, especially since they
contain passages important for understanding the positive exposition. I will continue to
refer to as “late” the dialogues generally so labeled, but nothing depends on chronologi-
cal claims about these or other dialogues.
Plato’s ethical thought begins, rather surprisingly, from something everyone accepts.
We all seek to be happy; we seek everything for the sake of this, while we do not seek this
for the sake of anything further.

‘‘Suppose someone . . . asked,’’ says Diotima to Socrates in the Symposium. ‘‘Now,

Socrates, the lover of good things has a desire; what does he desire?’’ ‘‘That they
become his,’’ I said. ‘‘And what will he get, when good things become his?’’ ‘‘That’s
easier to answer,’’ I said. ‘‘He’ll be happy.’’ ‘‘It’s by the possession of good things,’’ she
said, ‘‘that happy people are happy, and there’s no need to ask the further question,
‘‘Why does someone want to be happy?’’ The answer has the final end.’’
(Symposium 204e–205a).2

In the Euthydemus, in a passage famous in the ancient world as a “protreptic” to philoso-

phy, Socrates and a young boy agree that it makes no sense even to ask whether every-
one wants to do well and be happy; nobody would deny it. It is equally obvious that
everyone thinks that they will be happy by having (and making use of) good things. The
discussion proceeds from there, showing a deep difference between Socrates and con-
ventional opinion on what things are good but leaving unchallenged the point that we
all seek happiness (Euthydemus 278e–282d).3
Elsewhere, Plato is completely unimpressed even by unanimous common belief.
Here, however, he does not, as elsewhere, find fault with it or challenge it philosophi-
cally. Rather, he holds that everyone is, indeed, on to something important; our problem
is, rather, that our grasp of happiness is extremely dim. For most, if not all, people go
radically wrong in their views about what things are good and thus about how we shall
actually succeed in becoming happy. Much of Plato’s effort in discussions of value is
devoted to improving our views about goodness and thus about what things are good.

1 Some scholars take these dialogues to present positive “Socratic” views, which differ from
“Platonic” views in dialogues such as the Republic, taken to be later.
2 All translations are my own. I am very grateful to the large number of excellent contemporary
translations of Plato that have become available in the last two decades.
3 At Philebus 20b–23a, 60a–61a, Socrates puts forward as obvious the ideas that a good life must
meet the conditions of being complete (we seek it for its own sake and not for the sake of anything
further) and self-sufficient (it lacks nothing we have reason to seek). As with the other two passages,
Plato takes it that nobody would deny this, so that no argument is needed. See Daniel Devereux’s essay
on Socratic ethics, chapter 17 of this volume, for a more detailed treatment of the Euthydemus’s
eudaimonistic framework. For the importance of this framework, see G. Vlastos, Socrates, Ironist and
Moral Philosopher (Cambridge, 1991); T. Irwin, Plato’s Ethics [Ethics] (Oxford, 1995); and J. Annas,
Platonic Ethics, Old and New [PEON] (Ithaca, N.Y., 1999). For a dissenting view, see N. White,
Individual and Conflict in Greek Ethics (Oxford, 2002).
 Plato’s Ethics   533

Why does Plato take most people to be drastically wrong about goodness but not
about happiness? Might we not expect him to take the sorry state of most people’s views
about goodness to show that they are no authority on happiness, either, or its place in the
way we think about our lives? Given Plato’s radically anticommonsensical approach to
metaphysics and epistemology, why is the framework of his own ethical thinking found
in the few passages where we can be confident that he is telling us what most people
actually think?
The answer here lies in the notion of happiness, which is how we have hitherto ren-
dered eudaimonia. Since “being happy” in these passages is treated as interchangeable
with “doing (or acting) well,” it obviously does not fit closely onto the predominant
modern notion of happiness. What the ancients took us uncontroversially to be seeking
in everything we do is something achieved over life as a whole, not a feeling or episode,
as happiness is often now construed. Importantly, for the ancients, eudaimonia is itself a
highly indeterminate notion. That everyone aims at eudaimonia leaves open that there
are many divergent and, indeed, mutually incompatible ways of so doing. In noting the
fact that everyone agrees that eudaimonia is our universal aim, Plato does not recognize
any great achievement on our part. We cannot go wrong because what we agree on is so
indeterminate that there is extreme disagreement as to how to achieve it.
Plato’s ethical thought is, then, structured by a broad eudaimonist assumption. His
main concern is to challenge the views most people have about goodness, for it is here
that they go disastrously wrong in trying to live happy lives. Most people think that
virtue is a minor good, or even an impediment to living a happy life. Plato thinks this
utterly wrong; it is only by being virtuous that we can hope to be happy. This is a radical
claim, one that demands that we reject a lot of things intuitively considered needful for
happiness, and can on its own seem to reconceive happiness on theoretical grounds. But
Plato is in fact never prepared to pull happiness away from the intuitive idea that the
happy life is a pleasant one. His preoccupation with virtue is matched by a preoccupa-
tion with pleasure and its relation to virtue. The main lines of Plato’s ethics are thus best
followed by doing the following: looking at his theoretical answer to the question about
virtue and happiness, then examining the way he discusses virtue, and then exploring
his positions on pleasure. I shall conclude by looking at the relation of Plato’s ethics to
his political and his metaphysical thought.
In the Euthydemus passage already mentioned, Socrates gets the young boy Cleinias
to list things most people would take to be good and thus components of a happy life:
health, beauty, power, status, and the like; he includes the virtues but gives them no
particular prominence. Socrates argues him into accepting that there is a crucial differ-
ence between virtue and all these conventional goods. The value of conventional goods
in a life is conditional on their being used wisely, as a craftsperson makes wise use of
her material in producing the results of her skill. Thus there is a radical distinction
among things that we call good: no conventional good will do what we expect of a
good—namely, benefit us and so make us happy—unless it is put to proper use by the
kind of wisdom that will make us good. A similar distinction is drawn in two passages
of the Laws (631b–d, 661a–e) between human goods—health, beauty, strength, and
534   Julia Annas

wealth—and the divine ones, the virtues, on which they depend. For both a city and an
individual, the goodness of, and so the benefit from, conventional goods depends on the
possession of the virtues, which depend, in turn, on the wise use of reason.
Clearly, Plato is making the point that most people, who think that they will achieve
happiness by becoming rich, or famous, or powerful, are fundamentally mistaken and
need to revise their priorities radically. As Socrates tells his fellow Athenians in the
Apology, they are wrong to care for wealth, for wealth does not bring about virtue,
whereas “from virtue wealth and all other things become good for people, both privately
and publically” (Apology 30a–b). But what exactly is the relation between living virtu-
ously and being made happy by having health and wealth?
Once raised, this question dominated ethical philosophy; ethical theories were dif-
ferentiated by their answers to it. Plato does not distinguish between two distinct posi-
tions clearly separated by later debate. Plainly, he always thinks that being virtuous is
necessary for eudaimonia. Happiness cannot be a matter of how wealthy, good-looking,
or powerful you are; it must lie in what we do with these things in our lives. If we take
them as our ultimate goals, we have no hope of being happy; without the divine goods,
every person and city loses the human ones, too (Laws 631b–c). Plato’s position, however,
allows of two interpretations, both with textual support.
One is that, while conventional goods do not make us happy independently of virtue,
they can make a virtuous life better. Thus in the Euthydemus, we find that the correct
account of conventional goods is that “if ignorance guides them, they are greater evils
than their opposites . . .; but if practical intelligence and wisdom are the guides, they are
greater goods” (281d). And in the Laws, the Athenian says that “although what are called
evils are in fact evil for the just, they are good for the unjust; and what are called goods,
while actually good for the good, are evils for the evil” (661d). This may seem common-
sensically appealing: if a virtuous person with a chronic disease is cured, it seems obvi-
ous than her life has become better just by acquiring health. The consequence is very
uncommonsensical, however: that if you are vicious, you are actually better off being
poor and ill than rich and healthy—indeed, it is best for you also to be stupid and deaf
(Euthydemus 281b–d). The claim that the value of conventional goods in a life depends
on the use to which they are put commits Plato to thinking not only of health, for
example, as making a good life better but also as making a bad life worse, by providing
more opportunities for the wicked person to live wickedly.
Sometimes, however, Plato claims that virtue is the only good thing. “[Of the other
things,] none of them is either good or evil, but of these two, wisdom is good and igno-
rance evil” (Euthydemus 281e). Hence virtue, being the only thing which is good, and
thus of benefit to us, will be sufficient for happiness. The only correct account, says the
Athenian, is the one that insists that “the good man, being temperate and just, is happy
and blessed, whether he is big and strong or small and weak, rich or poor; and thus that
even if he is richer than Midas . . . but is unjust, he is miserable and lives wretchedly”
(Laws 660e–661a). On this view, nothing besides virtue should properly be called good
at all (and the same for evil and vice).
 Plato’s Ethics   535

These two positions have different implications for the role in the happy life of health,
wealth, and power, later called external goods or goods of fortune, implications which
Plato does not distinguish. Plato notably does not face a problem which would have
forced him to make a distinction: How, if nothing but virtue is good, can the virtuous
person have a reasonable basis for selecting among other things? If health does not make
a virtuous life better, why should the virtuous person bother trying to be healthy? In
practice, much of what Plato says about virtue and happiness suggests that health and
wealth, while they cannot in themselves make us happy, do make good lives better,
suggesting the weaker interpretation. Yet the language of some of the more uncom-
promising passages about virtue shows that later Stoics had a point in claiming affinity
with Platonic ethics:4 often, Plato suggests that there is such a gulf between the kind of
value that virtue has and the kind found in conventional goods that we would be wrong
to think that we should just care more about virtue than we do; we need a complete
change of perspective on value. To do justice to Plato’s ethical thought, we need to find
both positions in the dialogues and to recognize that they are not distinguished and
often juxtaposed.5
Uncompromising insistence that we need to change utterly to be happy chimes with
many passages where Socrates asserts radical ethical positions most people reject
(though he claims that if they submitted to Socratic questioning, he could show them
that they do, deep down, believe them—Gorgias 471e–472c). A good person cannot be
harmed (Apology 41c–e); other people can unjustly accuse and execute him, as hap-
pened to Socrates, but if he is virtuous, this does not harm him. If you cannot live virtu-
ously, there is nothing to be gained by staying alive; life in itself has no value for you
(Crito 47d–48a). Doing wrong is worse for you than suffering wrong (Crito 49a–d;
Gorgias 472d–476e); it is so much worse that if you do wrong, the best thing for you is to
seek out punishment for it (Gorgias 476a– 479e). It is irrelevant to a person’s happiness
whether he is rich and powerful, only whether he is wise and good (Gorgias 470c–471a).
There is a reason these passages became famous in antiquity: they give us a stark
choice between conventional views of happiness and an utterly different perspective.
The starkness of the choice shocks us into asking whether we are, as Socrates thinks we
are, committed to this extreme divergence from convention.
The stronger claim, that virtue is sufficient for happiness, is marked in the Apology,
Crito, and Gorgias. In the Republic we find it too, but not distinguished from the weaker
claim. The Republic is structured as an answer to the question, how one ought to live
(Republic 344e). Glaucon and Adeimantus make a challenge to Socrates, which he
answers, referring back to it at the end of book 9 (366d–367e, 588b–592b). He is challenged

4 Antipater, one of the heads of the Stoa, wrote a three-book work claiming that Plato converged
with the Stoics on many points, particularly in taking virtue to be the only good, and sufficient for
happiness. Annas, PEON, argues that this is a reasonable interpretation of Plato’s ethical thought
overall. See Irwin, Ethics, for interpretations, particularly of the Republic, defending the weaker view.
5 Hence there is debate whether the evidence supports scholars who take the stronger view to
characterize dialogues taken to be early and the weaker view to characterize later work.
536   Julia Annas

to show that virtue6 is worth having “in itself ” and for what it does for us “in itself,” as
opposed to external rewards which could equally well be obtained from a successful
cynical pretense. The challenge is put in two forms, juxtaposed but not equivalent.
Glaucon describes two extreme cases: the virtuous person who suffers every misfor-
tune, including that of being traduced as wicked, and the wicked person who success-
fully puts up a front of being virtuous (360e–362c). In the case of the virtuous person,
virtue has been stripped of everything that it could possibly owe to appearance and
pretense; and so when we ask whether it is in our interests to live justly, we can appeal
only to what virtue can do for us by its own nature. Socrates is challenged to show that
even in the worst circumstances the virtuous person is happier than the wicked person
(361d). This is compatible with the virtuous person not being happy in terrible circum-
stances, only happier than he would be were he not virtuous. The argument proceeds by
way of Plato’s sketching the structure of an ideal society, whose structure exhibits the
same form as the virtuous person’s soul; those brought up in such a society, Plato claims,
would be completely virtuous and, so, happy. In such a society virtuous people would, it
appears, be happy because they are virtuous in circumstances where virtuous people
have circumstances organized for them to flourish. It looks, therefore, as though Plato
is claiming that, while virtue is necessary for a happy life, it is not sufficient, since the
virtuous are happy only in a society where they have favorable external circumstances.
Still, even in the worst circumstances of the actual world, the virtuous person will always
be happier than the wicked, however favorable the circumstances of the wicked.
Other passages in the Republic, however, appear to defend the claim that virtue is
sufficient for a happy life. In the very passage where the challenge is posed, Socrates
states it in terms of virtue simply benefiting you and being in your interests (367c). And
when Socrates refers back to this challenge at the end of the main argument, this is the
claim defended. It is in your interests to be virtuous, it has been argued, because this is
the best way for your soul to be organized, without reference to your being in ideal or in
actual circumstances (588b–591b); indeed, the virtuous person will be concerned with
conventional goods and evils only insofar as they conduce or the reverse to their being
in this internal state (591b–d). The happiness of the virtuous in the ideal state is thus
not a part of Socrates’ answer to the challenge to show that being virtuous is in your
interests even in the worst conditions of the actual world.7
At the theoretical level of ethical thinking, then, we find that Plato does not have a
single determinate answer to the question whether virtue will make us happy; for he
does not distinguish between the claim that being virtuous is sufficient for living a happy
life and the claim that it also needs favorable circumstances, although even in the worst
circumstances, it will render you happier than the most prospering wicked person.

6 The Republic is about dikaiosunê, often translated “justice,” since part of the argument concerns the
ideal society, but the main argument concerns the individual, where “justice” may be misleading and
“virtue” is a safer rendering.
7 In the bulk of the Republic, there are many passages suggesting the stronger, sufficiency thesis: for
example, 387d–e, 472b–e, 427d, 444e–445a, 580c, 613a–b. Irwin, Ethics, and Annas, PEON, give different
interpretations of these passages.
 Plato’s Ethics   537

The sufficiency claim can reasonably be seen as lying behind the most uncompromising
claims by Socrates in some dialogues, but it is blurred with the weaker claim in Plato’s
best-known work, as well as in the passages where he comes closest to the kind of theo-
retical discussion of virtue and happiness that we find from Aristotle onward.
Even the weaker claim conflicts dramatically with the conventional assumption that
health and wealth contribute to happiness in their own right. We find Socrates continu-
ally calling into question ordinary assumptions about virtue. This is a two-pronged
attack. Positively, Socrates develops a conception of virtue in which it is taken as having
the structure of a practical skill, accepting the prominence this gives to the cognitive and
intellectual elements in virtue. And negatively he shows a variety of people with such
strong views about virtue that they do not comprehend what they are talking about,
since they cannot “give an account” (logon didonai) of virtue or a given virtue. What they
fail to provide is the kind of ability to explain and justify what they are doing, which is
typically found among those with mastery of a practical skill. The positive and negative
approaches work together.
In the Euthydemus, virtue was readily identified with mastery of a productive skill,
and, although it sometimes surprises modern readers, the idea that being virtuous is like
having mastery of a practical skill is quite intuitive. Becoming good is learning to act
well, as a skilled person does things well; in both cases, what is exercised is practical
knowledge, which has the feature that it is normally learned from someone who can
convey expertise to the learner, while also requiring of the learner that she come to mas-
ter the relevant field of activity for herself. Where practical expertise can be conveyed,
we have more than a natural talent or happy “knack”; we have a cognitively structured
way of thinking. The impetus for Plato’s “intellectualist” account of virtue thus comes
from virtue’s sharing with practical skill familiar features of practical reasoning.
Plato engages with the skill analogy for virtue in a variety of ways. Virtue is practical,
thus a matter of getting down to the task of living; Socrates’ anxiety about his fellow
citizens’ drift and lack of concern for living well is an urgent one. What they need to do,
he insists, is to start paying attention to themselves; they need to exercise epimeleia,
proper attention and diligence (Apology 29d–30b). One entire dialogue, the Alcibiades,
is devoted to this point. The young Alcibiades is confident of happiness, backed as he is
by looks, wealth, and influence. Socrates convinces him that all these apparent advan-
tages are utterly worthless until he begins to pay attention to himself. Alcibiades comes
to see that he needs to get working on his own lazy and self-satisfied state before he can
understand virtues such as justice and their importance for happiness. The historical
Alcibiades did not do this; he went spectacularly to the bad. The reader is encouraged to
do a better job of paying attention to his own deficiencies.8
Living virtuously is, in the light of the skill analogy, thought of as actively living in a
way which is the product of exercising your own intelligence, taking charge and working

8 The notion of “care” here is dominant in the interpretation of Plato, particularly his use of the
figure of Socrates, in the work of Hadot (for example, P. Hadot, What Is Ancient Philosophy?
(Cambridge, Mass., 2002).
538   Julia Annas

at living with a clear overall goal in view, rather than drifting along with uncritical views
of what matters in life. The skill analogy emphasizes the cognitive side of virtue: the
ability to learn how to perform a task, to figure out how to respond to diverse situa-
tions, and to explain what you are doing—all analogous to what mastery of a practical
skill requires.
The kind of understanding required to be virtuous emerges from the Laches, where
Socrates turns a discussion of the educative value of fighting in armor into an examina-
tion of two generals’ understanding of courage. (One of them, Nicias, remarks that a
conversation with Socrates inevitably turns into an examination of your own life
(187e–188b).) Laches begins by characterizing courage as standing firm, and then, more
widely, as endurance, but fails to grasp the kind of understanding we require in endur-
ance that we admire. Nicias then suggests that courage is a kind of understanding. This
is found intuitively bizarre, but the objection which sinks it is that if courage is a kind of
understanding, it cannot be limited to its intuitive area, that of what is and is not to be
feared. A right understanding of what is to be feared requires understanding what is to
be feared or not in general, at any time; and to know in general what is worth fearing
amounts to knowing what in general is valuable; and this kind of understanding would
amount to virtue, not just one aspect of it, as courage is usually taken to be.
Whatever puzzles the reader is supposed to take away from this dialogue, we can see
here the force of the thought that what underlies a virtue is practical understanding—
the ability to discern and to respond to what is truly valuable—and that this will be
something all the virtues will share. An account of virtue which starts by taking it to be
cognitively rich will tend toward some form of the unity of the virtues.
It is then no surprise that in other dialogues various interlocutors are shown by
Socrates that they fail to understand what they think they know about virtue. In the dia-
logues named after them, Euthyphro fails to understand piety, Charmides and Critias
fail to understand temperance, Meno fails to understand virtue, and Hippias fails to
understand the fine or kalon, what virtuous people aim at.
In all these cases, the interlocutors fail because they cannot “give an account” of what
they think they know; they can neither explain what it is nor justify their own judgments
about it. Their failure is persistently likened to failure to possess mastery of a practical
expertise. In the case of virtue it involves a failure to appreciate that having a virtue
involves more than just learning from society how to behave, even reliably, in certain
contexts. This kind of piecemeal socialization may leave the person with no clue as to its
point and hence no way of linking in his understanding the different circumstances in
which the virtue is exercised. Coming to understand what virtue, or a specific virtue, is,
thus requires, if it is to pass the test of “giving an account,” an ability to give a unifying
and explanatory account of actions and reasons over your life in general, not merely in
one area conventionally associated with, say, courage or justice. In the virtuous person’s
reasonings, virtue is dominant over her life as a whole, in a way that unifies her priorities
over a variety of different circumstances. The virtuous life has the kind of unity charac-
teristic of the product of expertise (Gorgias 500e–501b), suggesting the metaphor used
by Stoics and later Platonists of virtue as the skill of molding the material provided by
 Plato’s Ethics   539

your circumstances into a life unified by its overall achievement of eudaimonia by way of
virtuous activity.9
In the Protagoras, we find that this kind of distinctive way of thinking about your life
might be amenable to some degree of formalization. The sophist Protagoras has
expounded his own view: that people are socialized into virtue the way they pick up
their native language. Socrates pushes him into discussing the idea that it might, rather,
be something more like a precise expertise, “the skill of measurement.” He also intro-
duces the idea that what we are trying to measure is pleasure and pain (a controversial
move we shall return to). Socrates develops the idea that what will “save our lives,” make
us able to live securely whatever life faces us with, is a skill which will measure the plea-
sures we want and the pains we want to avoid, objectively, in a way that avoids our bias
toward what is here and now, as well as our bias toward favoring what presents itself or
“appears” to us as appealing or unappealing at a given moment:

The power of appearance confuses us and often makes us take and then regret the
same things back and forth in our actions and choices of both great and small
things, while the art of measurement would have taken authority from this appear-
ance and, by showing us the truth, would have given us peace of mind resting on
truth, and would have saved our lives. (356d–e)

The Protagoras is the only dialogue in which Plato even entertains the idea that pleasure
might be an adequate final goal for our lives, and he is careful to have Socrates introduce
it merely as a thesis to be discussed. But the wider thought has continuing appeal for
him: that our lives should be lived in a way unified by pursuit of an overall final goal
that has objective value, a pursuit to which we have to be summoned away from our
tendency to be misled by the power of “appearances,” the way things attract or repel us.
Living well requires us to reject being at the beck and call of our likes and dislikes and
the ways things affect us, and to organize our life overall in a way that has objective value.
For Plato, it is always our reasoning powers which enable us to do this, and to counteract
our tendency to follow our desires, which are fixated on their own fulfillment in a way
unresponsive to wider concerns. To varying degrees in different dialogues, he is open to
the idea that this achievement of our reasoning powers might ultimately be, or rest
upon, thinking that is rigorous and precise beyond the achievement of everyday arguing
of the Socratic kind—indeed, as rigorous and precise as that of mathematicians.
In the Gorgias, living a good life requires imposing on our life the kind of structure
that an expert imposes on her materials to come up with a unified product; there are
hints, however, that it requires a deeper kind of understanding of a formal, even mathe-
matical kind (507c–508a). In the central books of the Republic, Plato explores the idea
that a proper grasp of ethics requires profound mathematical and metaphysical study,
which takes years to acquire. But, despite the different level of background demanded,

9 The development of the idea of skill (technê) in Plato is well set out in P. Woodruff, “Plato’s Early
Theory of Knowledge,” in S. Everson (ed.), Epistemology: Companions to Ancient Thought, vol. 1
(Cambridge, 1990), 60–84.
540   Julia Annas

the basic thought remains constant: we come to live virtuously by coming to have overall
understanding, expressed in practice, of what is worth pursuing as our aim in life;
achieving this is difficult and requires detaching ourselves from caring about the kinds
of desire that most people invest most of their time and energy in gratifying. It is no sur-
prise, then, that Socrates shows his interlocutors that where virtue is concerned, they
have little or no idea what they are talking about; the kind of understanding that virtue
really requires demands attention and effort that most people are not prepared to put in,
and puts you at odds with most people’s views of what life is all about and what is worth
doing and having (Gorgias 485a–e).
Grasping what is truly valuable in life will, then, pull us away from uncritical identifi-
cation with the satisfaction of our desires for particular things and our pursuit of what
we find appealing. How radical is the required detachment? Plato does not have a single
answer to this. Throughout a good part of his work, he is tempted in two different direc-
tions on this, and this leads him to conceive of the dominance of reason in the life of
virtue in two discrepant ways, both powerfully presented.
The first account is most memorably put forth in the Republic, where the virtuous
person is the person whose soul is rightly ordered. In this and some other dialogues,10
Plato distinguishes three parts of the soul—reason, spirit, and desire—rather than
simply contrasting reason with desire, but this does not alter the general point made
here. Each part has its proper function; that of reason is to rule, since only it can discern
what is good for the whole soul, whereas the other parts can register only what is attrac-
tive to them (441d–443e). Reason’s rule does not imply that the other two parts should
be repressed but, rather, that they should play the parts assigned to them by reason,
given its superior understanding of what is good for the whole which the parts make up.
Thus reason aims not to extirpate desires but to bring it about that they are fulfilled only
in ways which encourage, rather than conflict with, the overall pursuit of goals set by
reason (588b–591b).
Famously, there are two ways in which the obedience of the other two parts to reason
is construed, each reflected in metaphorical descriptions of the person. One is that of
education.11 The parts of the soul can agree, in a harmonious way, and thus are taken to
be capable of the kind of communication that is capable of agreement and disagreement,
as opposed to blank conflict of mutually uncommunicative forces. Plato develops the
idea that desire and spirit, in particular, which in actual societies are recalcitrant in
accepting the conclusions of reason, can, in principle, and in ideal conditions, be formed
and, if necessary, reformed, by education of varying kinds, ranging from the attractive to
the coercive. This depends on the thought that reason and the non-rational parts of the
soul share enough by way of cognitive structure that communication is possible (442c–d,
443c–444a, 586c–587a). It also depends on the thought that non-rational aspects of

10 Timaeus and Phaedrus. There is considerable scholarly discussion of the differences, if any, made
to other central ethical claims by the tripartition of the soul. C. Bobonich discusses the issue in depth
in Plato’s Utopia Recast (Oxford, 2002). See also Hendrik Lorenz’s essay, chapter 21 of this volume.
11 On Plato’s complex views about education, see Rachana Kamtekar’s essay, chapter 25 in this
volume.
 Plato’s Ethics   541

ourselves are plastic and can be successfully molded by social forces. This assumption
underlies the highly transformative program of education in the early books of the
Republic and becomes explicit in the similar program of the Laws, where even sexual
desires are taken to be so plastic that a regulated society could eliminate same-sex sexual
attraction (838a–841e).
The other way of construing the dominance of reason over other parts of the soul is in
terms of force. Reason is a stronger kind of item than the non-rational parts of the soul;
they cannot grasp what reason can, but the kind of force that reason can produce is more
effective than what they can bring to bear. This idea is often expressed by representing
reason as a little person within the whole person whose good it alone discerns, the other
parts being represented as nonhuman animals, which reason is capable of controlling.
This produces Plato’s most famous metaphors: the soul as inner person controlling
beasts and as a chariot whose rider controls two powerful horses (Republic 588b–589b;
Phaedrus 246a–e, 253d– 256e). Clearly, reason and the non-rational parts do not on this
construal share enough cognitive structure for communication and agreement to be
possible.
These are not only different but mutually incompatible ways of thinking about reason’s
dominance in the soul. That Plato does not clearly make up his mind between them
shows that he continues to be drawn both by the thought that good psychological func-
tioning is a matter of putting into practice, in differing ways, shared overall principles of
organization, and by the thought that, even in ideal psychological functioning, there are
elements which are potentially resistant to overall organization and direction and thus
have to be coerced. In the Phaedrus passage, the bad horse pleads and argues with the
other horse and the charioteer but is deaf (253e, 254c–e)! As has often been pointed out,
Plato’s ambivalence here reflects his ambivalence as to whether the producer class in the
Republic fit their assigned roles unconstrainedly or whether force is always needed to
keep them from usurping the rule of the Guardians.12
There is a second picture of virtue as reason’s dominance in the soul: reason’s exercise
reveals to us the utter insignificance of human life. If we take this seriously, we become
detached from all worldly things. In the Phaedo, Plato vigorously stresses that to achieve
grasp of the Forms, the philosopher must “practice dying” (64a); he must detach himself
from the everyday way in which we identify with our beliefs and, especially, our desires.
Relatedly, true virtue does not deal with the matters of everyday life but consists in an
escape or “purification” from them (69a–d).
On this austere construal, virtue consists not in dealing with the material circumstances
of life but in rising above them, aspiring to living a life in which they have no part. A life
free from the encumbrances and drawbacks of the human condition is the Greek idea of
the life of the gods, and Plato does not hesitate to say that a life of virtue, construed as
fleeing or being purified from the everyday world, is a life which is godlike or aspires to

12 The relation between the partition of the soul in the Republic and the structure of the ideal state is
much disputed. G. R. F. Ferrari, City and Soul in Plato’s Republic (Sankt Augustin, 2003), is a recent
contribution to this debate.
542   Julia Annas

become like god. The most famous statement of this is in the “digression” in the
Theaetetus, where Socrates says that human life must unavoidably contain evil, “so we
should try to flee from here to there as quickly as we can; and flight is becoming like
God to the extent that we can. And becoming like God is becoming just and pious, with
wisdom” (176a–b; cf. Republic 612e–613b).
In these passages, flight to escape our world rests on the sharp metaphysical and
epistemological divide between the world of our experience and the true world of the
Forms revealed to the person willing to use his mind in a rigorous way and pursue
lengthy courses of mathematical and dialectical thinking.13 We cannot, however, miss
the point that grasping what is transcendent to our world renders us aware of the insig-
nificance of human concerns. In the Laws, this awareness of insignificance comes with
awareness of God rather than Forms (644d, 803c–804c).
In the later dialogues, the idea of the individual’s becoming like god is found in
another context: the idea that reason in the person is a small-scale version of what
reason is in the cosmos as a whole (Philebus 28c–30e). Reason’s function continues to
be that of unifying and ordering, and thus living our own lives in a rational way comes to be
seen as participating in the large-scale workings of reason in the cosmos, which is seen
in different ways in the later dialogues (as a causal element in the world in the Philebus,
as the rational plan of a divine Craftsman in the Timaeus, as the rationality of a cosmic
soul in the Laws), but always as divine. The work of the divine reason is to organize
things for good, since the divine is good and, lacking envy, seeks to spread goodness in
the way it orders things. Becoming like God thus comes to be construed as aspiring to
identify with the goodness-producing works of the divine reason, which makes it less
surprising that this is a characterization of virtue. Virtue continues to be seen as impos-
ing rational order on potentially refractory materials, but this is linked to, rather than
contrasted with, transcendent divinity (Timaeus 90a–d).14
These two construals of “becoming like God” are not unified in Plato’s thought. We
can see how they might be put together. The flight idea stresses our need to detach from
everyday thoughtless engagement with our everyday world and to recognize a new kind
of value, by comparison with which our everyday values are dross. This is then followed
by the realization that we respect this new kind of value by identifying with its workings
in the cosmos and cooperating, as far as we can, in its functioning; in our own lives, this
consists in living in a way in which this value is dominant in our lives, and reflection
shows us that this is what the life of virtue is. Stoic ethics brings these two ideas together,
stressing both the distinctive nature of virtue’s goodness and the way in which we honor
its value in the way we live our lives, and setting this in a cosmic rather than specifically
political framework. We can thus agree with the Stoics that their ethical ideas converge

13 It is controversial how determinate is the idea of the objects of such thinking in the Theaetetus
passage, and hence how appropriate it is to call them Forms.
14 For a recent development of this aspect of the idea, see J. Armstrong, “After the Ascent: Plato on
Becoming Like God,” Oxford Studies in Ancient Philosophy 26 (Summer 2004), 171–83, criticizing
Annas, PEON, and D. Sedley, “The Idea of Godlikeness,” in G. Fine (ed.), Plato, vol. 2 (Oxford, 1999).
 Plato’s Ethics   543

with Plato’s, even though they reject his transcendental understanding of the divine and
of reason in the world.
Plato’s various claims about virtue all have in common that virtue is either the only or
the dominant element in a life which can achieve what we all dimly grope for—happiness.
We might expect, given these strong claims, that he would defend a much reconfigured
view of what happiness is, one in which the role of pleasure is reduced or even elimi-
nated. But we find that, for Plato, pleasure presents a continued invigorating challenge
to his ethical thinking, producing a variety of spirited responses. In the ancient world,
this was taken to show that Plato was the source of a whole range of later theories of
pleasure; modern scholarship has tended to use the differences between the accounts to
support a variety of accounts of Plato’s alleged overall development.15 Plato’s engagements
with pleasure show us different responses in his continuing concern with the central
issue of how we are, through living a virtuous life, to achieve eudaimonia, which is hap-
piness, and thus in some way satisfying and enjoyable.
In some passages, pleasure appears as what we seek by fulfilling our desires, where
these are taken to be paradigmatically short-term urges for fulfillment. In the Gorgias,
Philebus, and the treatment of pleasure in Republic 9, this goes with the idea that pleas-
ure is the replenishment of a lack (Gorgias 492e–495a; Philebus 42c–47d; Republic
585b–586b). Standard examples are hunger, thirst, and sexual need: these are natural to
the kind of beings that humans are, and will in the normal course of things standardly
recur. If we accept this picture of pleasure, it is clear why it would be a mistake to seek
pleasure as our overall end. For pursuing pleasure will simply be a matter of filling the
lacks coming from our recurring needs to eat, drink, and so on. This is not an organizing
principle for a life, but simply a way someone might drift through life, reacting to the
pressures of felt need but never giving her life overall direction from within. Only reason
can provide the organizing power to get the person to focus on an end which will give
the life a shape. Thus it is rational organization, not the fulfillment of desires, which
should shape our lives.
Plato makes the significant claim that having reason give overall direction does not
lead to a loss of pleasure, as the pleasure seeker might fear. The overall dominance of
reason will stop or inhibit the fulfillment of many desires which without it would just
have gone ahead, so the person will lack many pleasures which she would otherwise
have had. But this should not be construed as a lessening of pleasure, since it is crucial
for the happiness of a life what kind of pleasures are had.

So when the entire soul follows the philosophical part, and there is no civil war in it,
it comes about that each part does its own work and is just, and in particular each
has enjoyment of its own pleasures, the best and truest pleasures it can have. . . . But
whenever one of the other parts gets control, it comes about that it cannot even dis-
cover its own pleasure, and it compels the others to pursue a pleasure that is alien to
them, and not true. (Republic 586e–587a)

15 We find the ancient claim in Aulus Gellius Attic Nights 9.5.

544   Julia Annas

Behind this picture lies Plato’s conviction that our capacity to have pleasure is plastic
and can be formed and, if necessary, re-formed, by education and training of various
kinds, ranging from teaching and practice to punishment. This conviction underpins,
as we have seen, the educational programs of the Republic and Laws; the latter work
especially focuses on the differences between kinds of pleasure and our capacity to be
educated to enjoy activities and entertainments quite different from those we start with.
Legislation and social opinion have brought it about that people not only stifle incestu-
ous desires but do not feel frustrated by this and, indeed, are appalled by the thought of
finding incest enjoyable; Plato claims that the same could be true of homosexual sexual
desires (Laws 838a–841c). Similarly, he thinks that a rightly ordered and focused educa-
tional program could produce citizens who not only perform the approved songs and
dances that they are taught but unforcedly enjoy them and feel no need for innovation,
the young enjoying the same songs and dances as the old (652a–660c). Much of Plato’s
confident authoritarianism in political matters rests on his conviction that our capacities
for enjoyment are plastic, as well as on his conviction that firm understanding is possible
on the issue of what humans ought to enjoy if they are to lead happy lives.
Hence alongside passages which stress the dominance of virtue we find passages,
sometimes surprising to the modern reader, such as: “We choose less pain with more
pleasure, do not choose less pleasure with more pain and when they are equal find it
hard to be clear about what it is we want.” Pleasure and pain influence our wishes, and so
our decisions, because of their “number, size, intensities, equalities and the opposites.”
“Since things are thus ordered,” we desire a life in which pleasure predominates over
pain, whether the feelings are frequent and intense or few and weak: “We should regard
our lives as all being naturally bound up in these; and therefore if we say that we wish
for anything beyond these, we are speaking as a result of some ignorance and lack of
experience of lives as they are” (Laws 733b–d). Plato sees a problem here, which is solu-
ble once we realize that most people have little clue about which activities are really
pleasant. It is the task of education to change people’s priorities so that they find living
virtuously to be their overriding aim in life. In giving up attachment to their pursuit of
pleasures they may think that they are losing out, but they find themselves mistaken; the
virtuous life turns out to be the most pleasant they could lead.
In the light of this thought, we can see why some of Plato’s discussions of pleasure
treat it as a competitor with our aspiration to be virtuous, while others claim that the
greatest pleasure comes from being virtuous. Plato is not confused here; he is develop-
ing the view that pleasure is not a feature of your life independent of your activities and
the value these have. People whose characters differ will take pleasure in different things;
and Plato holds that some of these, the virtuous, are right and the others wrong. Hence
his continuing detailed concern with the education of enjoyment and the censorship
and filtering of culture this requires. Aiming at happiness in the right way, by living vir-
tuously, will bring it about that you find a different way of life, and different activities,
enjoyable, and reveal to you that your previous view of what activities were pleasant was
profoundly mistaken.
 Plato’s Ethics   545

Plato also thinks, interestingly, that conventional understandings of pleasure actually

lead to underachieving it. Most people take pleasure to be nothing more than release
from pain (an idea obviously linked to the thought that it is a replenishment of a lack).
But, then, most pleasures will be “mixed” with pain: experiencing the pleasure of drink-
ing when thirsty will involve being aware of the disappearance of the thirst, for example.
Plato vehemently denies that this exhausts the nature of pleasure. It is not true of pleas-
ure, he holds, that it is simply release from pain; the claim that it is comes from limited
experience and reflection. In dense and fascinating passages of the Republic and
Philebus, he argues that pleasures which depend for their pleasantness on release from
pain are “false” and “impure.” Pure pleasures are those which do not rely for their pleas-
antness on any release from pain and thus will not come from filling lacks or needs:

There are many others, but a particularly good example to notice are the pleasures
of smell. They suddenly become very intense without your having had any preced-
ing pain, and when they cease they leave no pain behind. (Republic 584b)

The Philebus adds the pleasures of experiencing pure colors, shapes, and sounds, not to
be confused with the pleasures of enjoying these as representations of something. Plato
also adds the pleasures of learning, which do not rely on release from any pain (Philebus
51e–52a). Pure pleasures enhance and improve the state of freedom from pain, without
being more of whatever produces that (Republic 584c–585a).
Epicurus later defended the position that the extreme of pleasure just is freedom from
pain (bodily and mental), a state he identified with “tranquillity.” Plato already takes this
to be inadequate as an account of all that pleasure contributes to the happy life. One of
the most fascinating aspects of Plato’s ethical thought is his combination of a rigorous
insistence on the dominance of virtue in a happy life, and the dominance of reason in
virtuous living, with equally tenacious insistence not only that the happy life is pleasant
but also that pleasure forms a positive contribution to it.
Clearly, the account of pleasure in the Protagoras is not part of Plato’s general project
of locating the place of pleasure in the happiness achieved by a virtuous life. In that
dialogue, the assumption is that pleasure can be a final goal in a way that pays no regard
to the values of the people seeking it, something quantifiable without regard to the activ-
ities in which pleasure is taken. I have suggested that in the Protagoras Socrates puts the
idea forward just in the service of a larger argument. If, however, Plato does want
Socrates to be committed to this view of pleasure, and does so because he is himself
committed to it, the Protagoras illustrates a change of mind from his predominant view
of pleasure.16
Happiness, virtue, and pleasure are in most of the dialogues in which they figure dis-
cussed in the context of an individual life, but in Plato’s two longest and most magisterial

16 For a recent treatment of pleasure in Plato’s work overall, see D. Russell, Plato on Pleasure and the
Good Life (Oxford, 2005).
546   Julia Annas

works, Republic and Laws, the happiness of the individual’s life, and the role in it of virtue
and pleasure, are set in an ideal society, one that in the Republic is lightly sketched and in
the Laws thoroughly, indeed ponderously, worked out.
In one way, Plato’s thoughts about the individual’s happiness are unaffected by this. As
we have seen, there are continuities in the argument about happiness and virtue in the
Republic from the Apology, Crito, and Gorgias. Plato clearly, when writing the Republic,
thought that the argument that an individual lives happily only by living virtuously is
given additional support by his analogy of individual and state, which shows that virtue
has the same form in both. For, given this, virtue in soul and state cannot be adequately
understood merely by studying only one of them; it requires a more abstract level at
which the virtue which is studied will be applicable to both. Hence the ethical argument
remains but is supported by discussion of the ideal state and of the Forms which both
state and soul exemplify.
The context of the ideal state does, however, serve to enable Plato to develop his views
on education of character. In both Republic and Laws, he sets out a program where indi-
viduals are, from infancy onward, to be socialized and educated in ways that encourage
attraction to virtue and repulsion from vice. In both dialogues, Plato goes into some
detail as to how the culture of his time will have to be modified to do this, in ways that
lead to large-scale censorship and rethinking of the contemporary arts. Thus Plato needs
to show us individuals as citizens of a society in order to show us the kind of education
and formation that would make us virtuous, and so livers of pleasant and happy lives.
Do we, however, need the ideal state to make this point? Could Plato not have noted
the importance of education for character and left the reader to think of the appropriate
development? He seems to be thinking that we will not fully understand either the
power or the importance of education in the formation of character unless we are given
a good example, and for this we need the ideal state. This also serves to weed out the irrel-
evancies that any example of non-ideal education would bring along.
Some features of Plato’s political thought have dominated consideration of his
­ethics—in particular, the elitism defiantly present in both his longest works. In the
Republic, Plato insists that the individual’s life must be ruled by reason, his or her own if
it is adequate, but that otherwise

in order that someone like that should be ruled by what is similar to what rules the
best person, we say that he should be the slave of that best person who contains
within himself a divine ruler. It is not for the harm of the slave that we think he should
be ruled . . . but because it is better for everyone to be ruled by divine intelligence,
preferably his own that he has within himself, but otherwise imposed from outside,
so that as far as possible we may all be alike and friends, steered by the same thing.
(Republic 590c–d)

Since Plato thinks that only a very few people will have adequate reason, properly
trained, this means that for most people the living of their lives will be structured largely
by the deliberations of others. Only a few, then, will be living lives which are virtuous in
a way that depends on their own thinking. If we take it that the living of an ethically good
 Plato’s Ethics   547

life excludes such radical dependence on another’s thinking, then Plato is excluding
all but a few from living a genuinely ethical life. The Republic, however, is a sketch of an
ideal state based on the principle of rule by those with ideal understanding, and Plato is
clear that in the actual world nobody can come up to this standard. This aspect of the
ideal society is thus not part of the main argument that any individual in the actual
world will be happy by being virtuous even in the worst possible conditions. The elitism
of the ideal state is not part of a practical political proposal about the actual world
(a point that has been frequently misunderstood).
This does not get rid of all legitimate worries about Plato’s elitism, however. The Laws
depicts a society which is ideal but in which there are no longer the Republic’s huge divi-
sions of understanding among the citizens. All of them will be educated in the same way,
will live in fairly egalitarian circumstances in which extremes of wealth and poverty will
be avoided, and will live under institutions which encourage them to see themselves as
members of a community of equals taking turns at ruling and being ruled. But this
community whose efforts are so strenuously devoted to making themselves virtuous
itself depends for its conditions of existence on the labor of others who are sharply kept
outside that community, the resident aliens and slaves who do the work that enables the
citizens to enjoy the leisure to educate themselves. Plato never envisages a community in
which all are equal members of the entire functioning community, and so it can reason-
ably be argued that his conception of the individual’s virtue is always based on the
assumption of this being supported by someone else’s labor and efforts. There is thus a
serious and ineliminable elitism even in Plato’s more egalitarian ideal state. His concep-
tion of virtue always requires leisure and education, which has to be supported by the
work of others.
This is important, for it is an ironical result for a philosopher who insists so strongly17
on the irrelevance for happiness of material goods, bodily well-being, and social goods
such as status. It was left to the Stoics to see that insisting on the sufficiency of virtue
for happiness leaves us with a conception of virtue that does not actually require for its
exercise circumstances of leisure or the specific context of the Greek city state, and thus
does not require the exploitation of others who are thereby excluded from the pursuit of
virtue and happiness.
It is also in the Republic that we find Plato’s ethical position in the context of perhaps
his most striking and elaborate presentation of Forms. As with the political background,
there is a sense in which this does not make a profound difference to the ethical ideas
that have already been developed in other contexts. Plato always takes virtue to require
an overall grasp of what is good to take as a guide for one’s life as a whole, and takes this
to be an objective matter which requires intellectual effort. As between various dia-
logues, his view differs as to how much such effort is needed and what form it takes. This
varies in stringency from mastery of a practical skill to mastery of mathematical and
dialectical thinking that requires years of effort. The increased intellectual demand

17 Though not always consistently, since Plato does not clearly distinguish between the weaker and
stronger positions about the relation of virtue to happiness (see above, pp. 270–73).
548   Julia Annas

comes from putting ethics in the larger context of metaphysics, where a different range
of intellectual skills will be required.
The metaphysical background is itself not something which itself remains static as a
background to Plato’s ethical thinking. In the later dialogues, as already touched on, the
individual’s life is related to the workings in the universe of divine reasoning. In our
small way, each of us tries to do in our own life what cosmic reason does in the universe
as a whole: organize things for the better to the extent that we can. In some ways, this
metaphysical picture makes the larger whole of which we are parts less remote than
Forms; ethics is seen in the dynamic larger whole of the cosmos rather than being
located on the far side of a lot of mathematics and abstract reasoning.18
In other ways, however, the picture makes us, as individuals, smaller, since, though
the important educative and sustaining role of society continues, the individual’s basic
ethical relation is with the cosmos and with others in the cosmos. This is another way in
which Plato, at least in the later dialogues, can be seen as a precursor of the ethics of the
Stoics. It is a notable difference, however, that Plato never envisages the kind of cosmo-
politanism which we find in Stoic ethics, in which we are related ethically to all rational
humans and the context of the city-state recedes in importance. Plato’s stress on our
relation to the cosmos coexists with a continued and, indeed, in the Laws, strengthened
insistence on the crucial importance of the culture of the city-state in developing and
sustaining the individual’s ethical development.19

Bibliography
Annas, J. An Introduction to Plato’s Republic (Oxford, 1981).
Annas, J. Platonic Ethics, Old and New (Ithaca, NY, 1999).
Annas, J. “Wickedness as Psychological Breakdown,” Southern Journal of Philosophy 43
(Supplement, Spindel Conference 2004) (2005), 1–19.
Armstrong. J. “After the Ascent: Plato on Becoming Like God,” Oxford Studies in Ancient
Philosophy 26 (Summer 2004), 171–83.
Bobonich, C. Plato’s Utopia Recast (Oxford, 2002).
Bobonich, C. Plato’s Laws: A Critical Guide (Cambridge, 2010).
Ferrari, G. R. F. City and Soul in Plato’s Republic (Sankt Augustin, 2003).
Gill, C. “Plato and the Education of Character,” Archiv fuer Geschichte der Philosophie 67
(1985), 1–26.
Gill, C. Greek Thought, Greece and Rome. New Surveys in the Classics 25 (Oxford, 1995).
Hadot, P. What Is Ancient Philosophy? (Cambridge, MA, 2002).
Irwin, T. Plato’s Ethics (Oxford, 1995).
Kahn, C. Plato and the Socratic Dialogue (Cambridge, 1996).
Kraut, R. (ed.) Plato’s Republic, Critical Essays (Lanham, MD, 1997).
McPherran, M. Plato’s Republic: A Critical Guide (Cambridge, 2013).

18 Mathematics is still needed in order to discern the regular and mathematically comprehensible
structure of the universe, as the Laws insists, but the dialogue also suggests a certain division of
intellectual labor between the experts and the rest of the citizens.
19 I am grateful to Gail Fine for helpful comments.
 Plato’s Ethics   549

Russell, D. Plato on Pleasure and the Good Life (Oxford, 2005).

Santas, G. (ed.) The Blackwell Guide to Plato’s Republic (Oxford, 2006).
Sedley, D. “The Idea of Godlikeness,” in G. Fine (ed.), Plato, vol. 2, Oxford Readings in
Philosophy (Oxford, 1999), 309–28.
Taylor, C. C. W. “Platonic Ethics,” in S. Everson (ed.), Ethics: Companions to Ancient Thought
vol. 4 (Cambridge, 1998), 49–76.
Vlastos, G. Socrates, Ironist and Moral Philosopher (Cambridge, 1991).
White, N. Individual and Conflict in Greek Ethics (Oxford, 2002).
Woodruff, P. “Plato’s Early Theory of Knowledge,” in S. Everson (ed.), Epistemology:
Companions to Ancient Thought, vol. 1 (Cambridge, 1990), 60–84.
 chapter 23

Pl ato On L ov e

Richard Kraut

1. Eros and Philia

Two families of Greek words, each with its distinctive semantic content, can properly be
translated as “love.” On the one hand, there is the verb philein and its cognates (philia is
the noun, philos the adjective)—a word we use all the time when we talk about philan-
thropy, philosophy, philharmonic, and the like. On the other hand, “to love” is also the
proper translation of the verb eran (eros is the name of this psychological force, erastês
designates a lover, and erômenos is the one who is loved). Those Greek terms are of
course no less familiar to us than the “phil . . . ” family: from them we have “erotic,”
“erogenous zones,” and so on. Although our terms “erotic” and “sexual” are by no means
interchangeable (a depiction of sexual organs, for example, need not be erotic), no dis-
cussion of the place of eros in human life could possibly ignore sexual desire and sexual
activity. Similarly, Greek texts that concern eros—the subject of Plato’s Symposium, and
one of the central topics of Phaedrus—must address themselves to sexuality, though
they can encompass far more than that. Eros, unlike philia, picks out a type of desire that
drives people, under certain conditions, to physical contact—to touch, to kiss, to
embrace, to “make love”—and also to think obsessively of the person who is loved and to
be filled with longing when he or she is absent.
But philia is not necessarily low in affect. Although it can be applied to nearly any
group of cooperative associates, it is the word that would most naturally be used to name
the strong feeling and close relationship that exists among family members and also
among close friends, whether or not they are sexually attracted to each other. To call two
people philoi is to suggest neither that there is nor that there is not an erotic component
to their relationship.
So there is no semantic oddity in the thesis, which Plato endorses (Section 4 of this
chapter), that there is philia in the best kind of erotic relationship. He does not single out
sexless relationships for special commendation—although we will see how such a mis-
reading of his thought might arise. To call a friendly relationship that is devoid of sexual
552   Richard Kraut

attraction or interaction “Platonic,” as we often do, is therefore to misrepresent him.

On the contrary, he is especially interested in sexually charged relationships and is
aware of their potential to do great good—though also great harm. One of his principal
motives, in writing about eros, is to teach us how to make that distinction.
For the best and fullest discussion of philia in antiquity, we should turn not to Plato
but to books VIII and IX of Aristotle’s Nicomachean Ethics. But there is a remarkable gap
in those books: Aristotle sees no need to say much about eros, for he assumes that his
theory of philia can be applied to those bound together by sexual love no less than to
those who are “just friends.” He remarks at one point that when philia is strong, it can be
felt toward only a few people, and then he adds that eros is an extreme form of philia,
because it is felt toward one person (IX.10 1171a10–12). The sexuality of eros (as opposed
to its intensity and narrow focus) has little interest for him, as a philosophical writer.
We would have to say the same about Plato, had he written only Lysis, but not Symposium
and Phaedrus, for there is hardly a word about eros in this short work (widely assumed to
have been composed before the other two). Instead, it searches but does not find a
proper account of philia. The problems it raises about that notion have nothing to do
with eros but apply to all philoi whatsoever—whether or not their relationship is erotic.
But Plato, unlike Aristotle, is fascinated by eros. He considers it a subject from which
philosophers have a great deal to learn and upon which philosophy can cast much light.
His greatest contributions to the study of intimate human relationships, which are con-
tained in Symposium and Phaedrus, have to do with eros and with philia only insofar as
erotic relationships should involve philia as well. He does not have, as Aristotle does, a
robust and systematic theory about philia in its own right. Instead, he offers us, in his
Lysis, a series of puzzles about philia; and then, in Symposium and Phaedrus, a series of
contrasting speeches about eros. The heart of his theory of intimacy, affection, and sexu-
ality lies in those speeches.1
Every Platonic dialogue presents its own riches and difficulties of interpretation.
Readers of his Symposium cannot be criticized for paying special attention to the speech
delivered by Socrates (that is the approach I adopt here), for it is reasonable to suppose
that this is the most important segment of the dialogue—the one that contains the cor-
rect theory of love, which Plato himself accepts and recommends to his readers. That
supposition can be supported by the fact that the speech within that speech—particu-
larly the ascent to the form of beauty described by Diotima—fits hand in glove with the
metaphysical scheme propounded in such other dialogues as Phaedo, Phaedrus, and

1 For further study of the speeches of these two dialogues, I recommend G. R. F. Ferrari, Listening to
the Cicadas: A Study in Plato’s Phaedrus (Cambridge, 1987); C. Griswold, Self-Knowledge in Plato’s
Phaedrus (New Haven, Conn., 1986); R. Hunter, Plato’s Symposium (Oxford, 2004); Frisbee Sheffield,
Plato’s Symposium: The Ethics of Desire (Oxford: Oxford University Press, 2006); James Lesher, Debra
Nails, and Frisbee Sheffield (eds.), Plato’s Symposium: Issues in Interpretation and Reception (Cambridge
Mass., 2006); and P. Destrée and Z. Giannopoulou (eds.) Cambridge Critical Guide to Plato’s Symposium
(Cambridge: Cambridge University Press), 2017.Valuable discussions of both works are presented by
T. Irwin, Plato’s Ethics (New York, 1995), 298–317; and M. Nussbaum The Fragility of Goodness: Luck and
Ethics in Greek Tragedy and Philosophy (Cambridge, 1986), 165–233. A comprehensive treatment of Lysis
is offered by T. Penner and C. Rowe, Plato’s Lysis (Cambridge, 2005).
 Plato On Love   553

Republic. The middle books of the Republic are an especially close match to the ascent
Diotima prescribes in Symposium, for there too we are told about the process by which a
philosopher-in-training will be led to a great discovery: the form of the Good. But how-
ever important the speech of Socrates is for an understanding of Symposium, Plato must
want his readers to pay careful attention to the other speeches as well—to see how they
bear on each other, complementing each other in some respects but conflicting in others.
Similarly, although the party-crashing of Alcibiades adds immeasurably to the dramatic
impact and comedy of the dialogue, we must ask ourselves what that denouement has to
do with the content of the other speeches. Can it be that Plato gives the final word to
Alcibiades because his encomium contains some corrective to what has gone before?
The problem of understanding Symposium as a philosophical work is in part a problem
of seeing how all of its material hangs together.
Phaedrus creates rather different obstacles. It seems to be a work about two subjects—
love and rhetoric—and yet it contains a theory of composition that insists that every
discourse must contain an organic unity, each part contributing to a larger organiza-
tional plan (264c). Plato is apparently prodding his readers to ask themselves what the
principle of organization of this dialogue is. In this essay, I must bypass that question
and so give short shrift to Phaedrus (though see Sections 9 and 10 of this chapter) in
order to concentrate on Diotima’s speech.

2. Eros and Desire

The centrality of Diotima’s ascent (210a–212b) in the scheme of Symposium is undeniable,

but some of the ideas it contains cannot be understood on their own, since they build on
earlier material. For our purposes, the best place to begin an examination of this dialogue
(if that is the right word for it) is the passage in which Socrates, by cross-examining
Agathon (199e–201c), shows that there is a close connection among eros, desire, need,
and futurity. More precisely, when someone loves (erai), there is something or someone
he wants and, therefore, something or someone he needs and lacks. Plato realizes that
this thesis requires defense, and so he has Socrates show Agathon that apparent counter-
examples to it can be redescribed in ways that make them conform to it. We say: “I want
to be healthy, and I have what I want.” But can we really mean what this seems to mean?
If I already am healthy at the present time, what would be the point of wanting to be
healthy at the present time? A desire is a motivator; its role is to move us to do some-
thing. Properly understood, then, when someone who already is healthy says, “I want to
be healthy,” he should be taken to mean that he wants to continue to be healthy. He wants
to have his health in the future, but that is something that he does not yet have. He lacks
and needs future health.
Eros is characterized here as a desire. That does not mean that whenever someone
wants something, he loves it. The relationship goes in the other direction: whenever
someone loves, he wants. This thesis says nothing about what kind of desire one has,
554   Richard Kraut

when one loves. It may be a desire that Plato would locate in the appetitive part of the
soul, but it need not be. The word Plato most often uses for desire in this passage, as so
often, is epithumia. But an epithumia can be any sort of desire—it is not necessarily an
“appetitive” desire for food, drink, or sex.2
Plato might be accused of making a mistake here: people can want things that have
nothing to do with themselves, and so desiring is not the same thing as needing and
lacking. Suppose I want someone else’s needs to be fulfilled. That does not show that I
have a need or a lack. If I act on my desire, that is not because I need or lack something
but because someone else does.
Perhaps this criticism of Plato can be sustained. But even if it can, it does not reveal a
defect in his conception of eros. Whether or not there are desires that reflect no need or
deficiency in the subject, what matters to Plato, in Symposium and Phaedrus, is the dis-
tinctive psychological phenomenon that goes by the name of eros. He can insist that this
type of desire always arises out of the subject’s needs, even if he were to concede that
other desires might not.3

3. Birth in Beauty

Let us pick up the thread of the conversation several pages later, where Diotima, in her
cross-examination of Socrates, insists on a connection between eros and good. Up to
that point, it is assumed that the object at which eros is directed is beautiful—or, at any
rate, is taken by the lover to be beautiful. Agathon claimed in his speech that love is
never of ugliness but always of beauty (197b), and that assumption is allowed to stand in
Socrates’ conversation with him (201a–b). But then Diotima asks Socrates a series of
questions about the relationship between eros and good, and this interchange eventually
leads to the thesis that “eros is wanting to possess the good forever” (206a11–12).4 Is Plato

2 See, e.g., Republic 431d1, 554e1, 580d7, 587b1–4. Plato’s division of the soul into three parts (reason,
spirit, appetite) is most fully presented and defended in Republic 435e–441c, but, to understand what spirit
and appetite are, it is important to read his critique of defective political regimes and their corresponding
character types in 545a–580a. His conception of the rational part of the soul and the values with which it
is associated are contained in his depiction of philosophical training in books VI and VII. A helpful
introduction to this subject is provided by H. Lorenz, “The Analysis of the Soul in Plato’s Republic,” in
G. Santas (ed.), The Blackwell Guide to Plato’s Republic (Malden, Maine, 2006), 146–65.
3 I address this issue more fully in “Eudaimonism and Plato’s Symposium.” in Destrée and
Giannopoulou (eds.), Cambridge Critical Guide to Plato’s Symposium,” pp. 235–52. For a different
approach, see S. Obdrzalek, “Moral Transformation and the Love of Beauty in Plato’s Symposium,”
Journal of the History of Philosophy 48 (2010), pp. 415–44.
4 There is no word corresponding to “wanting” in the Greek text, and so a translation that mirrors
only what is on the page would be “erôs is of the good to be one’s own forever.” But the insertion of
“wanting” is justified by the context. See 204d5, where one of Plato’s several terms for “want” (erai) is
used to express a similar idea, and the conversation then moves back and forth between this and
several other such terms (bouletai at 205a3, epithumia at 205d2). I am grateful to Tushar Irani for
discussion of these passages.
 Plato On Love   555

here inviting us to infer that because goodness is the object of love, beauty is not—that
beauty has no bearing on eros? That would be too great an about-face.
It is more plausible to take him to mean that the relation between love and beauty is
not quite as simple as Agathon and Socrates had been assuming (206e2–3). The object of
one’s desire, when one loves, is always something that is good—but it is never claimed,
during the remainder of the dialogue, that it must be something beautiful or taken to be
such. Diotima insists (204e–205d) that each person’s desire for the goods that he equates
with happiness (the ultimate object of desire: 205a1–3) should be classified as eros, and
that since everyone wants happiness, everyone should be classified as a lover, despite the
fact that this term (erastês) and its cognates are normally reserved for only one kind of
love—the kind that consists in “making love” (as we would put it). Diotima is complaining
that ordinary ways of using eros and related terms are arbitrary and should therefore be
reformed. Someone who loves money, or wisdom, or sports is no less a lover (erastês)
than someone who seeks sexual intercourse, even though that term is normally reserved
for lovers in the conventional sense (205d). The Greeks use philein rather than eran to
talk about the love of philosophy (philosophia) or sport (philogumnastia), but Diotima sees
no reason to avoid the semantic field of eros to describe our passion for these activities.
Presumably, then, she would also say that it is arbitrary to deny that a parent is an
erastês of his children or to insist that his feeling for them is philia but not eros. That would
not commit her to saying that parents should have sex with their children but only that
eros and kindred terms should be applied more broadly than common usage allows.
Note that since Diotima drops Agathon’s idea that the object of eros must be taken to be
beautiful (206e2–3), she would not be forced to concede that in having eros for their
children, parents find them beautiful, alluring, or attractive.
The thesis that “eros is wanting to possess the good forever” is obviously an extension
of the point Socrates makes in his discussion with Agathon. What someone who loves
health desires is not that he be healthy now but that he be healthy in the future. If health
is part of his conception of happiness, surely he will want to be healthy not just for some
short period of time but for a very long one—in fact, indefinitely into the future. Being
mortal, he cannot have his health for as long as he would like, and so his love of health
leads to the pursuit of some approximation to his possessing health eternally. This is the
thought process that lies behind Diotima’s thesis that every lover seeks “birth in beauty”5
(206b7–8)—the striking formula by means of which she allows beauty to be readmitted
to the theory of love. She draws on the assumption that as sexual beings we cannot help
being responsive to beauty (206c–d) and that the outcome of this responsiveness, in
sexual intercourse and pregnancy, is the production of a new generation that extends
the lives and often the projects of its forebears.

5 The Greek term translated “birth,” tokos, applies both to a process and its product: that is, both to
giving birth and to the child. Diotima immediately explains herself by ascribing to all human beings a
desire to go through the process (tiktein, 206c4), but she also uses her theory to explain our love of
children, both those made of flesh and blood and those composed of ideas (see, e.g., 208b and 209c–d).
556   Richard Kraut

She then (208e–209e) broadens the basic ideas of her theory by elaborating on the
notion that someone can be pregnant in soul and not merely in body. Someone who
loves wisdom and justice, for example, cannot possess these qualities forever, but even
so he can get closer to this goal by inculcating them, through reasoning and education
(209b–c), in the next generation, which will, in turn, reproduce its virtues in others. In
this way, one can come as close as mortality allows to the eternal possession of what one
takes to be good. But, Diotima insists, one cannot do this in the absence of a sense of
beauty (209b). Just as the desire for sexual intercourse is aroused by the sight of physical
beauty, so the desire to give birth to discourse in the education of a younger person is
aroused by some perceived beauty in that person’s soul.
She even applies her theory to animal reproduction: eros exists among these winged
and many-footed creatures as well (207a–d). She is not merely making the obvious point
that they copulate. Rather, her idea is that for them sexual activity serves a purpose that
they know nothing about: the production of a new generation and, therefore, member-
ship in a chain that extends without end into the future. She might even be assuming
that they too are responsive to beauty—if her statement that “it is impossible to beget in
ugliness, but only in beauty” (206c4–5) is meant to be exceptionless. Animals do not
have a conception of happiness or good, but if the thesis that “eros is wanting to possess
the good forever” (206a11–12) is meant to apply to them as well, then the good that they
want forever must be life itself.
What of human beings who are not pregnant in soul but only in body? Is Diotima
saying that the function—whether acknowledged or not—of their sexual intercourse,
like that of all animals, is the eternal possession of life itself? Perhaps, but she must also
make room in her theory for the fact that nearly all parents want to transmit their values
to their children. If what they love (and what they therefore wish they could have for-
ever) is not merely life but also money or health or sports, then they will do what they
can to reproduce the love of these goals in their children. To have a simulacrum of life
forever merely requires being part of an unending generative chain, but human beings
can, in a way, also have health or wealth forever by giving to their children a passion for
these goals and the resources needed for sustaining them. There is no reason, in fact,
why Diotima should deny that ordinary parents (pregnant in body but not soul) might
also love justice and virtue in general and try to possess these goods eternally by engen-
dering this love in their children.
She also mentions a way of eternally possessing the good that bypasses sexual repro-
duction but falls short of the pregnancy of soul that issues in reasoned education. The
great heroes of the past, she claims, sought honor and glory for their acts of courage and
wanted to possess these goods forever (208c–e). They do not give birth to flesh-and-blood
children or to reasoned discourse but, rather, to beautiful deeds, and if these are celebra-
ted by poets whose words live on, then the fame they seek will live forever, as will the
reputation of the poets who memorialize those deeds (209d). We seem to have here a
threefold division that corresponds to the tripartition of the soul (into appetite, spirit,
and reason) in Plato’s Republic: appetitive people seek such low-level goods as eternal
life, or eternal wealth, or eternal physical well-being; those who love honor in battle or a
 Plato On Love   557

reputation for excellence—goods that Plato associates with the spirited part of the
soul—can possess these goods eternally by being the subjects or creators of song; but the
best sorts of people employ philosophical methods of reasoning to pass along a full
understanding of what is good.

4. Self and Others

Eros, so conceived, is necessarily self-involved. Even when it is directed at another person,

it nonetheless expresses a desire for the lover’s own good. The self-involvement of eros
is implicit in the thesis that to love is to want to possess the good forever; for that must
be taken to mean that to love is to want oneself to possess the good forever. But it must be
emphasized that, according to Diotima, the self that is involved in all love is not genu-
inely single because it is ever-changing: the stuff of the body is constantly being replen-
ished; new desires, beliefs, and habits replace old ones; and even knowledge depends
on fresh memories (207d–208a). When the present self plans for the satisfaction of its
desires, it is already reaching out to something that is merely similar to rather than iden-
tical to what it is now.
That, Diotima, claims, is why all living things care for their offspring (208b). An animal,
she notes (207b), will die for her young out of eros, and in doing so it obeys the law of
nature that whatever ages seeks to replace itself with younger versions of itself (208b).
Love of life (one’s own life, that is) lies behind efforts to preserve not only one’s own body
and soul but also the lives of one’s offspring. Similarly, if one loves justice—that is, if one
wants it to be the case that one possesses this good forever—then one will not only see to
it that one’s future soul possesses this good but also look for ways in which other people
who bear some similarity to oneself and will take one’s place also possess this good.
If one propagates what one loves by having and raising a family (that is, if one is preg-
nant in body but not soul), one will try to inculcate a love of justice in one’s children. But
Diotima claims that there is a superior way of wanting to possess the virtues eternally.
Some are pregnant in soul rather than body; in other words, they have within them
notions about justice that they want to expose to the light of public discourse, in the
hope that they can create just conditions for others who will outlive them. There is some-
thing self-involved even in that moral drive, because its beneficiaries are thought of as
extensions of the self and not utterly alien to it.
Diotima’s conception of eros is a far cry from the self-forgetting kind of love that cares
only for others and is devoid of all thought of oneself—the kind that does not care
whether it is I who helps others but only that they be helped by someone. That selfless-
ness, whether praiseworthy or not, is rare among human beings and therefore accounts
for very little of the good that they have achieved. By contrast, eros as Diotima conceives
it—a complex combination of self-involving and other-regarding motives—has great
motive power and has achieved more impressive results. Although Diotima does not
propose that we regard all humans as members of a single family, her theory of eros does
558   Richard Kraut

rest on the idea that some sense of likeness to others (and not only to blood relations)
elicits a willingness to forgo comforts, resources, and even life itself.
The idea that all human beings, when they reach a certain age, are pregnant in one way
or another, either in body or soul, contains the suggestion that we all are overfull with
self-love—in other words, that our love for something within us eventually leads to our
dedication to something outside us, as pregnancy leads to the birth of a new individual
who receives his or her mother’s loving care. Here too, as in her reflections on the conti-
nuity of body and soul, Diotima finds a form of self-involvement in the love of other
people. Her tacit assumption is that a mother loves her child because that child was once
inside her. When Diotima likens a poet’s songs or a statesman’s laws to their children
(209d–e), she is drawing on the idea that those products were once inside their minds
and is suggesting that they are loved at least partly for that reason.
But the main idea that she is driving at when she uses the metaphor of the pregnant
soul is that the best form of love among two human beings is one in which there is rea-
soned discussion about sophrosunê (moderation) and justice—and presumably all of the
other topics that Socrates loved to discuss (209a). In such a relationship, the erastês gives
birth to ideas that have been within him for a long time by finding someone capable
of philosophical discussion—someone who, in this sense, has a beautiful soul. Their
relationship will be all the more intense if the erômenos is also physically attractive,
although it is not necessary that he be so (209b). What they produce in their philosophi-
cal discussions, if it is nurtured well, will itself become a thing of beauty, for if a poem
can be beautiful so too can other forms of discourse when the ideas in them fit together
harmoniously. Diotima claims that the relationship between these two—the philosoph-
ical erastês and his erômenos—is a firmer love (here her word is philia) than parents
have, because they have far more in common with each other than those whose only
bond is their joint production of children (209c). It is important not to lose sight of the
fact that these two philosophers love each other—far more so than many other couples
do—and not only their discursive offspring, however beautiful their jointly produced
philosophical theory may be. In this respect, at least, Diotima’s theory of love has
become familiar and widely accepted: couples who talk to each other about serious mat-
ters and arrive at a meeting of minds enjoy a better form of love than do those whose
relationship rests on nothing but the physical attraction that initially brought them
together and their responsibilities as parents.
Diotima briefly indicates, at one point, that the well-nurtured offspring of this philo-
sophical couple has more to be said in its favor than its beauty and the way in which it
binds them together. When she starts to explain the notion of the soul’s pregnancy, she
says that the offspring that it is most fitting for a soul to produce is wisdom and that by
far the greatest and most beautiful form of wisdom is the one that organizes the affairs of
cities and households—namely, justice and moderation (209a). This suggests that the
beautiful product that arises from the discussions of a philosophical couple is not merely
of interest and value to them—rather, it can also benefit the entire political community.
The way in which two people love each other matters to everyone else, not only when the
offspring of such love is a child who will enter the political community and affect it for
 Plato On Love   559

better or worse but also when the offspring is a theory about how the community should
arrange its affairs. Presumably, one of Diotima’s reasons for claiming that this product of
a pregnant soul is most beautiful by far is precisely the potential it has for improving the
life of the whole city. Of course, the philosophical couple is unlikely to have had this
motive for establishing their relationship. But Diotima’s reason for thinking so highly of
their bond has to do, in part, with the great good it can do for others, not only for them.
The communal benefits of the love felt by a male couple is a theme that enters Plato’s
Symposium almost from the start. It plays an important role in Phaedrus’s encomium to
Erôs (178d–179a); although it drops out of sight for a while, it returns in Diotima’s speech,
and she continues to dwell on it when she includes such civic founders as Lycourgos and
Solon among her examples of individuals who love the products of their own minds
(209d–e). We are briefly reminded once again of the connection between eros and politics
when Diotima describes the ascent to the form of beauty: the beauty of laws is one of the
steps of her ladder, and that connection cannot be far from the reader’s mind when
Alcibiades crashes the party. It cannot be an accident that Plato chooses to bring the
Symposium to a close by bringing on stage a political figure and that the theme of that last
speech is the failure of his erotic pursuit of Socrates. Plato is perhaps suggesting that
Alcibiades’s failure to understand what eros is and how to be an erastês is connected in
some way to his larger failure in the political arena. I return to Alcibiades in Section 9 of
this chapter.

5. Aristophanes and Diotima

The conception of eros contained in the speech of Aristophanes (189d–193d) is, in some
respects, the converse of the one that Diotima proposes. She conceives of the erastês as
overfull—as containing within himself so much that he must, with the help of another,
get it outside of himself. The additional life within him is reasoned discourse that can
benefit another person of the right sort—the erômenos whose soul is beautiful—but, as
we just noticed, it can be of value to the entire community as well. Throughout the dura-
tion of the relationship of these two friends, each continues to have a mind of his own;
that, surely, is what enables their discussions to be worthwhile and capable of leading to
a beautiful product. By contrast, Aristophanes takes love to be nothing but an effort to
overcome the burdens of distinctness and separation. It is a sense of isolation that drives
distinct individuals to want to meld into one and to do nothing for each other but
embrace. For them, fusion, not sexual satisfaction, is the goal of physical contact; the
satiation of their sexual desires happens to serve a useful purpose, in that it temporarily
induces them to separate and attend to their quotidian tasks (191c). Their desire to elimi-
nate all physical distance between themselves is an expression of a deeper longing to be
melded into one body with one soul. This is the opposite of the Socratic inquiry that
Diotima assumes will take place when the erastês educates his erômenos; it is the unex-
amined life par excellence—the life of someone who wants nothing more than to lose
560   Richard Kraut

his mind and to become one with the person he loves. Remarkably, the simple point that
offspring of some sort—flesh-and-blood children, or psychological transformation—is
the product of love is a matter of no significance for Aristophanes. He entirely ignores
the political implications of his conception of eros.
The Aristophanic lover is seeking one person in particular: the one to whom he was
once joined. But, of course, neither Aristophanes nor anyone else believes that each of us
was once literally joined to another half. When his allegory is interpreted, it must be
taken to mean that although we long to be joined to some one person, there is no way to
articulate why we long for fusion with precisely this person and no other. We just have a
strong sense that this is the right person for us, and our longing to become one with him
or her is a brute force that can have no justification.
But the unique appropriateness of the object of love is not an idea for which Diotima
has any use. What a lover teeming with ideas is looking for is someone who can help
nurture those ideas and turn them into something substantial, and beauty of soul con-
sists in those qualities of mind that make someone a good conversational partner in this
endeavor. A lover should have no trouble articulating the features of the person to whom
he is drawn and whom he finds attractive. (Admittedly, he has nothing to say about why
he finds certain bodies beautiful: he simply does.) Nothing Diotima says suggests that
for each of us there is one uniquely appropriate partner in love. In fact, the multiplicity
of suitable objects of affection is already implicit in Socrates’ idea that to want something
is to want all of the many future replacements for one’s present self to have it. The “wide
sea” of objects of love then becomes one of the themes of Diotima’s description of the
ascent to the form of beauty (210d4). If there is anything in her speech that provides an
analogue to the one object of love that an Aristophanic lover seeks, it is the form of
beauty. The Phaedrus has a great deal to say about the way in which eros reunifies us with
the forms that we observed when, in a previous life, we were able to see them far more
clearly than we can now. So, some truth can be salvaged from Aristophanes’s speech, but
it must be transformed almost beyond recognition before Diotima can accept it.

6. The Ascent to Beauty

We are now ready to examine the final stage of Diotima’s presentation, in which she
describes a series of steps by which a lover-in-training is educated and brought to the
recognition of an “amazing beauty” and thereby to the achievement of the “goal” of his
erotic education (210e4–5). At that highest point of the ascent, the lover arrives at an
understanding of what beauty is (211c8–d1). That should be taken to mean that he can
articulate and defend a theory that explains what makes all beautiful things beautiful.
But Plato’s way of talking about this unchanging object and its location at the pinnacle of
a series of objects, each of which is beautiful, implies that the unchanging entity about
which the fully educated lover achieves an understanding is itself a supremely beautiful
object. It is pure, unmixed, divine, uniform, and devoid of the great silliness that mars
 Plato On Love   561

the beauty of human things; their beauty is a mere image of its true beauty (211e–212a).
All of this suggests that the form of beauty, untainted by any imperfection, has a beauty
that surpasses the sullied or short-lived appeal of all else. That is why the life of the lover
who reaches this stage is greatly enhanced, so much so that it becomes godlike (212a).
Not only can the lover explain why imperfectly beautiful things are beautiful; he has
gazed on the greatest beauty of all.
The first step of this staircase (epanabasmos: 211e3) is to love one body and, in doing
so, to beget beautiful ideas (logoi). The next step is to generalize—to recognize that there
are many other beautiful bodies as well and that there is something identical in the
beauty of them all (210a–b). The outcome of this second stage will be that the lover-in-
training’s extreme fascination with the single body that occupied his attention during
the first stage will diminish, and he will realize that this was a small thing (210b). Diotima
does not say, in her all too brief description of the first stage, what the learner’s words
will be about or what their purpose is. Are they simply the lover’s verbal depiction of
the beauty of the person whose form he finds so alluring? Another question we wish
Diotima had answered is how the transition from loving just one body to loving many is
to be brought about by the teacher who is guiding this process (mentioned at 211c1).
How does that guide induce the lover-in-training to broaden the field of things he loves?
And what transpires between the lover and the one he loves (and later with the many)?
We are perhaps given some help by an idea Socrates expresses in the Republic: an eroti-
cally inclined man will be attracted to many different physical types and not only to a
few (474d–e). Presumably, then, Plato thinks it will not be difficult, at least for someone
who has a receptivity to physical beauty, to see what is alluring in bodies of different
types and to put his appreciation of each type into words. The more difficult task that
must be accomplished, to arrive at the second stage, is to describe what all of these
kindred kinds of physical beauty have in common. That, of course, is a project akin to
the one pursued in several of Plato’s shorter ethical dialogues. Even if it is easy for some-
one to appreciate the allure of many different beautiful bodies, it is a task for philosophy
to put into words the common element in them all.
Diotima never claims that the lover who has moved from the first to the second stage
is no longer a lover of bodies. On the contrary, she says that he becomes “a lover of all
beautiful bodies” (210a4–5). Presumably, that means that he recognizes and enjoys the
perception of the beauty of any body that is beautiful. It is not the thing that all beautiful
bodies have in common—the property they share—that is loved, but those concrete
bodies themselves. Physical beauty is not to be treated as something that is entirely with-
out value.6 Rather, what occurs when the lover moves beyond the first stage of the ascent
is a diminution in his extreme concentration on one body. He has become a lover of

6 Though some translations unfortunately imply as much. Thus the generally excellent translation by
A. Nehamas and P. Woodruff, Plato, Symposium (Indianapolis, Ind., 1989), has Diotima say at 210c5–6
that the ascending lover thinks that “the beauty of bodies is of no importance.” So too R. Waterfield’s
translation, Plato, Symposium (Oxford, 1994): “he comes to regard physical beauty as unimportant.” The
Greek, smikron ti—“something small”—does not imply that the lover’s estimation of physical beauty
sinks to zero. He never becomes completely indifferent to physical beauty.
562   Richard Kraut

many things and is no longer transfixed by any one of them. He now sees how defective
his initial response to beauty was because it excluded so much. These points are important
to recognize because doing so keeps us from mistakenly supposing that in the ascent to
the form of beauty the lover, as he moves from each stage to the next, stops loving and
finding beautiful what he appreciated at some earlier stage. What the lover constantly
learns as he ascends is that his outlook at each earlier stage was defective because it
included too little. So, when he reaches the final stage and recognizes the greatest beauty
of all, he does not stop thinking that other things—including bodies—are beautiful as
well and responding affectively to their attractiveness. Their beauty may be small by
comparison with that of the form of beauty, but they nonetheless participate in its beauty
and because of that participation they too have some degree of beauty, however small.
Having arrived at stage two, the lover-in-training now moves up to a new ontological
level by turning his attention to the beauty that a soul can have—a beauty that he should
recognize to be superior to that of the body (210b). Here again Diotima tells us less
than we would like to know. How precisely is this recognition achieved? Perhaps she
is assuming that since the lover-in-training has been admitted to the mysteries of love
because of his suitability he has the philosophical talent and character that will make it
possible for an experienced guide to show him, through philosophical discussion, that
having a well-ordered soul is far more important than having a good-looking body. It is
better to have the virtues of the soul than those of the body—but best to have both kinds.
Someone who can be persuaded of that can also be brought to place the beauty of a
virtuous person in a higher order than the beauty of an alluring body.
Diotima does not explicitly say that the student of eros will go through soul-loving
stages that recapitulate the numerical difference between body-loving stages, but that is
clearly what she has in mind. Having passed through stages one and two, the next step is
to fix one’s attention on one soul, to care for that one individual, and to give birth to the
kind of discourse that can educate young minds (210c). Diotima then describes a
number of other items—practices, laws, types of knowledge—whose beauty the lover-in-
training will come to appreciate. Just as he was able to say what made all beautiful bodies
alike in their beauty, so too with these noncorporeal objects of appreciation. During
these stages, the student of love is engaged in the study of politics; that is what is implied
by Diotima’s reference to practices and laws. He is not becoming a lover of many actual
souls, for very few of the people he encounters will have beautiful souls, but he is inquiring
into the ways in which the lives of all citizens might be improved, and in this respect he is
recapitulating the transition he has already made—a transition from one to many.
In effect, the lover-in-training is becoming a student of political philosophy. Diotima
calls him a philosopher when she refers to the “many beautiful words and thoughts” that
he will beget in his unstinting love of wisdom (philosophia: 210d5–6). Significantly, she
never speaks of the lover-in-training who has reached the stage of political theorizing as
someone who undergoes a loss of enthusiasm for the one beautiful soul with whom he
discusses practices, customs, and branches of knowledge. She never hints that there is
any defect in remaining, at every subsequent stage of the ascent, an erastês of one and
only one soul—namely, the erômenos whom the lover fills with beautiful discourses
 Plato On Love   563

about political matters. She criticizes Socrates (presumably she does not know him very
well) and others for their obsessive desire to look at and be with young boys, and she
associates this with a fascination with gold and clothing (211d), but in doing so she is
merely elaborating on her earlier critique of overvaluing the beauty of bodies. A beautiful
soul, she implies, cannot be loved to excess—so long as one goes about loving that person
in the right way.
Since Diotima claims that the form of beauty is the most beautiful object there can be,
it might be inferred that the lover-in-training, having beheld that magnificent sight, has
nothing further to do as a lover beyond continuing to savor his understanding of that
supreme object. But that cannot be what she means, because it is a consequence of her
theory of love that the earlier discourses constructed by the lover-in-training and his
efforts to educate his erômenos were defective and therefore need to be improved. After
all, this erastês did not know at those less than ultimate stages of the ascent what beauty
is; his search for it was as yet incomplete. With his new understanding of beauty, he can
now construct more beautiful discourses, and these will do a better job of explaining
what makes laws well designed, which practices should be adopted, and which branches
of knowledge should be studied.
That is what Diotima implies when she says, at the end of her speech, that someone
who has seen the form of beauty will “beget not images of virtue . . . but true things”
(212a3–4). There are more children, composed of words, to be nurtured. Surely the
erastês, having seen and understood the form of beauty, will want to engage in conversa-
tions with his erômenos that will bring him to the same vision of beauty itself that he, the
erastês, has had. The notional children they nurture together, as they re-examine laws
and other social institutions in the light shed by the form of beauty, will be even finer
than their earlier offspring, and because they now have all the more to share with each
other their friendship will be even more firmly established than it was before (209c). It is
implicit in Diotima’s allusion to the “true things” begotten through the vision of
beauty itself that the ascent to that form is at the same time an intensification of the love
that exists between two individuals. And yet it remains a love that can benefit others as
well, because what these two have collaboratively achieved is an understanding of laws
and institutions that will, under favorable circumstances, lead to the improvement of
civic life.
The ideal relationship, then, is one in which two people care for and are friends to
each other; one in which they are receptive to much of the beauty of the world, ranging
from the beauty of human bodies to the beauty of beauty itself; and one in which they
work out, with well-crafted words, ways in which the world can be made more just. This
is not a relationship that must be devoid of sexual allure; on the contrary, Diotima makes
it clear that responsiveness to physical attraction is always a welcome (though not a
necessary) component of such relationships (209b). The Phaedrus tells us a great deal
about the struggles a lover must endure to prevent sexual allure from playing too large a
role in a good erotic relationship. The Symposium’s ladder of love affirms that human
beings cannot learn how to handle their sexuality except by going through a period in
which they are overly responsive to the erotic enticements of a beautiful body. It is left to
564   Richard Kraut

the Phaedrus, however, to depict the psychological conflict that must take place in all of
us as we learn through our mistakes to control our response to sexually attractive people.

7. Equal Relationships:
Diotima Transformed

Diotima’s conception of ideal eros is, no doubt, far too narrow. The model she proposes
is the relationship of a homosexual male couple in which one (the erastês) is the more
active and older partner, and the other (the erômenos) is reactive, younger, and in some
way beautiful. Only one of them—the erastês—is pregnant with ideas; the physical
appearance of only one of them—the erômenos—is a matter of significance (the erastês
can be ugly, as Socrates is ugly). It is assumed that one of them (the erastês) is far older
than the other and that he plays the role of educator, whereas the other plays the role of a
student. What of women? What of heterosexual couples? What of equal relationships
among people of roughly the same experience and education? The speech of Aristophanes,
to its credit, is more inclusive and egalitarian: it applies to every sexual proclivity, and
the lovers who seek reunification, each being the other’s halved counterpart, are equals—
equal in their need to lose their parthood and their ability to repair that loss.
Diotima’s assumption that erastês and erômenos are male is probably a mere conven-
ience of exposition. By choosing a woman to be the expositor of this theory of love
(a woman who may have been his own invention7), Plato in effect acknowledges that
women can be experts in this area. And since he sees they can be experts, he must have
realized that two women, no less than two men, can enter into ideal relationships. But
the male-female sexual relationships with which Plato was familiar were, for the most
part, marriages. A man typically sought a marriage partner not to have conversations in
which he could unburden his mind and pour out his ideas but to have children. It is only
to be expected, then, that Plato’s template for ideal erotic relationships should be the
erastês-erômenos institution with which all of the dialogue’s symposiasts were at home
and with which all of Plato’s contemporary readers were familiar.
Nonetheless, it may seem that Diotima’s theory requires there to be a significant dif-
ference in age, and therefore in experience and education, between the two partners.
That is because she assumes that eros leads us to have an effect on the world that will still
be in place after we die—it inevitably leads, in other words, to an attempt to influence
the younger generation in some way, either by bringing children into existence or by

7 As D. Nails, The People of Plato (Indianapolis, Ind., 2002), 138, notes: “All extant later references
to Diotima are derived from Plato.” But Nails is noncommittal about whether Diotima is a Platonic
invention. If she is not modeled on a real person, she would be a rarity in Plato, because nearly every
named character in his dialogues is a representation of someone he knew. (Callicles of the Gorgias is,
like Diotima, a difficult case; here too scholars disagree about whether he is a Platonic invention.) For
reflections on Plato’s choice of a female expert on erotics, see D. Halperin, “Why Is Diotima a Woman?”
One Hundred Years of Homosexuality (New York, 1990), 113–51.
 Plato On Love   565

educating a young mind. A lover wants to possess justice (for example) as far into the
future as possible. He cannot possess it forever, but he can create discourses that lead a
young person to become just; and as each new generation adds another link to the chain,
the lover does, by way of approximation, possess justice eternally. So it might seem
that it is essential to Diotima’s way of thinking about ideal eros that it be a relation-
ship between people who are a generation apart, and that will inevitably be a relationship
between people who are unequal in education and experience. Those may indeed be
very good relationships, perhaps even very good erotic relationships. But surely they are
not the only good kind.
In fact, however, all the ingredients that Diotima claims are present in ideal erotic
relationships can exist in a heterosexual couple of the same age. Consider this contem-
porary scenario: a man and a woman of roughly equal age (and therefore equal educa-
tion and experience) fall in love, their relationship is sustained by their conversations
about the most serious matters, and they have children because they are eager to nurture
and educate people who will make their world more just. They are, to use Diotima’s met-
aphor, pregnant in both body and soul; that is, they have ideas about how the world
should be improved, and they want to put those ideas into effect by bringing up their
children in a certain way. This relationship differs from the ideal erotic relationship
Diotima explicitly describes in only one important respect: for her, there is a division of
labor between heterosexual and homosexual love; the former ensures that there will
always be a fresh generation, and the latter ensures that some of the members of each
new generation will be well educated. But there is no reason propagation and education
cannot be tasks performed by the same couple. That, in fact, is a model of a good erotic
relationship that has become prevalent in our society. By rearranging the several ingre-
dients of Diotima’s theory, but without altering what is most fundamental in it, we
emerge with an ideal of eros that is now widely taken for granted.
If we go a step further, and think of God as occupying a similar role to the one played
by the form of beauty (which is, after all, a divine thing), the resemblance between her
ideal and one that has become familiar to us is even greater. The ideal erotic couple, in
that case, would be two equal human beings searching together for a way of extending
their love of justice into the future, educating their children to be just and treasuring that
additional bond between them, arriving at a fuller appreciation of the beauty of divine
existence, and using their enhanced understanding of God to improve their efforts to
comprehend and improve the world.

8. The Love of Individuals

There is another way in which the Aristophanic ideal may seem, at first sight, more
appealing than Diotima’s. Each half of an Aristophanic couple loves and is loved as an
individual. For each of them, no one else but his (or her) other half will do. A loves B not
because B plays some role in his life that another person might play equally well or
566   Richard Kraut

better. It was from B and B alone that A was severed, and so A’s relationship with no
other person can fill A’s need for completion. It is a need that can be completed only by
reunification with a particular individual, not by a person of a certain type.
That may seem appealing, but, at the same time, the Aristophanic ideal is burdened by
its commitment to the thesis that for each of us there is one and only one other person
who can give us what we seek when we look for someone to love and by whom we will be
loved. It is far more plausible to suppose (as many people do) that each of us needs a
good match and that, although some matches are better than others, no one person is a
uniquely best match. It is hard to believe that for each person there is only one other
person in the world who is right as a lover.
For Aristophanes, the mythical history of our relationships explains why we each seek
one and only one person who will meet our erotic needs. A was once united with B; it was
that past state of affairs that accounts for what each is doing, as he or she searches for love.
By contrast, Diotima’s model can acknowledge the importance of facts about what has
actually transpired between a couple. Once a good relationship has been established, a
couple nurtures notional children together, and that shared experience is what ties them
together so firmly in their friendship (209c). Although Diotima does not point this out, it
is obviously true of each lover that there is only one person with whom he has had these
fine discussions and produced these fine children. At this point in their relationship, then,
A will care for B and B for A in a different way from that in which each cares about anyone
else. If B dies, A will feel the loss of this particular sexual and conversational partner; he
will not treat this death in the same way that he reacts to the death of just any human
being. If A meets someone even more beautiful in soul and body than B, he will not think
about ending his relationship with B, because he is not in the business of loving the most
beautiful person he can find. It is B whom he loves, and although he can say what it is
about B that he finds so appealing, he is not trying to become the lover of every person
who has those qualities or with the person who most fully exemplifies them.
We can find confirmation in his Phaedrus that Plato thinks of the best erotic relationships
as ones that continue throughout a person’s life and even beyond. The couple that is most
fully in control of their sexual appetites remains intact even after each has died. Even the
second-best sort of relationship, in which the partners are ruled by the spirited part of the
soul rather than reason, is one of lifelong fidelity (256a–d). It would be appropriate, then, to
say that such lovers as these love each other “as individuals.” What each loves is that other
individual human being and no other; they do not treat each other as dispensable instru-
ments by which they achieve their goals. And yet there is no illusion, in these relationships,
that before each met the other, there was one and only one person who would have been
right. Diotima’s model of love is able to avoid falling into the trap that undermines
Aristophanes’s theory: its commitment to a single right lover, one’s other half. At the same
time, because it recognizes the importance of the history of a relationship, the kind of love it
admires can be described as the love of individuals “as individuals.”8

8 In this section, my thoughts have been shaped by reflection on G. Vlastos, “The Individual as an
Object of Love in Plato,” Platonic Studies (Princeton, N.J., 1973), 3–42, a great essay with which I profoundly
disagree. I have also learned much from Frisbee Sheffield, “The Symposium and Platonic Ethics: Plato,
Vlastos, and a Misguided Debate,” Phronesis 57 (2012), pp. 117–141.
 Plato On Love   567

9. Bad Eros

Diotima says nothing that even suggests that eros can be a destructive force in human
relationships—a desire that leads both lovers and those they love to great harm—and yet
we know that Plato was fully aware of this possibility. The tyrannical soul Socrates
describes in book IX of Republic is ruled by the eros that resides in the lowest part of the
soul, and he is a model of everything we should not be. In Phaedrus, Socrates says that
lovers recapitulate in their earthly existence the character of the god with whom they
traveled when they were disembodied. Someone who served Ares (the god of war), for
example, will become murderous when he thinks he is being wronged by his erômenos
and will be ready to kill himself or his boyfriend (256c). Phaedrus offers a sketch of a the-
ory of love that recognizes the duality of eros. It holds that a proper dialectical treatment
of this subject must use a system of divisions that contrasts sinister (a word whose Latin
counterpart means “left-handed”) forms of love with those that are “right-handed” and
divine (266a). Here Socrates characterizes his first speech (about why one should succumb
to a nonlover) as an exercise in the service of left-handed love, and his second speech
(depicting the charioteer and his two horses) as the praise of correct love. Returning to
the Symposium, we note that several of its speakers make a distinction between a good
and a bad form of eros (180d–181d, 186a–c), although others (Phaedrus, Aristophanes,
Agathon) do not. So Plato was reflecting on this distinction when he composed this
dialogue. But, curiously, Diotima does not mention it. What are we to make of that?
The distinction she draws between those who are pregnant in soul and those who are
pregnant in body does not seem to be a division between good and bad forms of love. It
is better to be pregnant in soul, she holds; but that does not show that being pregnant in
body—conceiving and giving birth to children—is bad. The eros in animals drives them
to procreate, and no one could plausibly believe that they go astray in doing so. It would
be equally absurd to hold that no human being should engage in sexual intercourse or
have children, and there is no evidence that Plato disagrees.
Another possibility that should be considered is this: since “erôs is wanting to possess
the good forever” (206a11–12), perhaps good forms of human love are those that are
based on an understanding—or, at any rate, a true belief—about what is good; and bad
forms are those that are false. This proposal is acceptable, I believe, but it does not go far
enough. What needs to be added is that Symposium contains a paradigm of bad eros:
that paradigm, I suggest, is Alcibiades. He is Plato’s way of portraying, in this dialogue,
how badly one can go astray when one’s sexual desires are allied with deeply mistaken
assumptions about how one should one lead one’s life. Not all mistakes about what is
good are equally harmful. For example, someone who thinks that well-being consists in
physical health and tries to possess this good far into the future by having healthy
children and teaching them to love health is not someone whom Plato wishes to assign
to the lowest depths of human misery; nor would anyone say that it is eros that has led
such a person astray. There are far worse lives, and Socrates claims in Republic that eros
can lead us to those depths. This is what has happened to the tyrannical man. He allows
the sexual waywardness that is a potential source of misery within all human beings
568   Richard Kraut

(Rep. 571b–572b) to become the leading element in his psychology. His appetite for sex is
what shapes his conception of what is worth pursuing, and the results are disastrous.
Alcibiades, as portrayed in the Symposium, is not at this pitch of depravity, but he is the
vehicle that Plato uses, in this dialogue, for showing the reader how badly eros can dis-
figure us. Diotima’s theory of love is not in error, but it is radically incomplete; it must be
supplemented by a portrayal of what eros looks like when it goes badly astray.
Plato sometimes uses eros to name the desire that impels us toward sexual activity.
When he divides the soul into three parts in the Republic, for example, he describes the
lowest (appetitive) part as the one by which one loves (erai), is hungry, and is thirsty
(439d6). Here eros is simply the brute desire to have sex, an instinctive conative force
that inheres in our nature no less than thirst and hunger. One would be missing a funda-
mental component of Plato’s psychological outlook if one failed to notice the persistence
with which he emphasizes that this desire is an extremely dangerous feature of the appe-
titive part of the soul—far more so than our desire to eat and drink. That is the lesson he
clearly means to convey in his portrait of the tyrannical man. But we should not over-
look the warnings that his Symposium and Phaedrus also issue regarding the destructive
potential of eros. What he is suggesting, in these two dialogues, is that human relations
inevitably become poisoned when they are dominated by the desire for sexual satisfac-
tion. His Symposium is by no means a bouquet laid at the feet of Love: the duality of eros
is never far from Plato’s mind, as the entrance of Alcibiades makes clear.
Before we consider this episode in greater detail, and before we examine as well Plato’s
warning, in Phaedrus, about the duality of eros, we should linger a bit longer on the
tyrannical man portrayed in Republic. Can we apply to him the two formulae Diotima
proposes: that love is a desire for the eternal possession of good, and that its goal is to
give birth in beauty? What is the good that the tyrant wants forever, and in what way
does he seek birth in beauty? How these questions are to be answered is by no means
certain, but perhaps Diotima’s theory and Socrates’ portrait of the tyrant can be fit
together in this way: As noted in Section 2 of this chapter, Diotima applies her theory
to animals no less than to human beings. Something in them makes them have sex,
although they, of course, have no idea what goal that instinct serves. Something in them
wants to replicate life forever by producing copies of itself, and the copulation of animals
is its modus operandi. We can reasonably take Plato to be suggesting, in his portrait of
the tyrannical man, that his sexual drive, like that of any brute, aims at the replication of
his life. The tyrannical man almost certainly does not think of himself as having sex
because he wants to live forever, but the impulse to which he gives free rein is a genera-
tive force that has eternal life as its goal. He seeks to give birth in beauty: that formula
applies to him because his sexual appetite, like everyone else’s, is attentive to visual cues
and responds to alluring bodies. That does not mean that he wants to have children and
raise a family; rather, he sets aside whatever sexual inhibitions most people have and
promiscuously chases after young girls and boys, taking every possible opportunity to
bed attractive young things (574b–c). It might be asked: Why is it that the sexual instinct
of animals leads not only to copulation but also to self-sacrifice, whereas the tyrannical
man presumably would not lift a finger for any babies he happens to produce? Plato can
 Plato On Love   569

reply that human beings are influenced by all sorts of notions about what is worth
pursuing, whereas the child-oriented instincts of animals do not have to compete with
ideas in their heads, since they have none. Whatever feelings a tyrannical man might
have for his offspring can be overpowered by his belief that to be a real man is to move on
to the next erotic episode.
Alcibiades does not have every feature that Socrates ascribes to the tyrannical man in
Republic, but there are some similarities. He is presented as a party-crasher, a drunk, a
man in love with flute girls, men, and boys; and all of Plato’s contemporary readers would
have present to their minds, when they read the Symposium, the notoriety Alcibiades
achieved as a man in love with power, an untrustworthy traitor to Athens, and a suspected
mutilator of religious statues. “Does a drunken man have, in a way, a tyrannical mind?”
Socrates asks in Republic (573b9–c1), and Adeimantus answers affirmatively. A few lines
later (d2–5), Socrates notes the tyrannical mind’s love of feasts, revelries, girlfriends, and
things of that sort. When Alcibiades first notices Socrates, he immediately becomes jeal-
ous and angry, because he sees Socrates reclining next to Agathon (213c). Socrates reacts
to Alcibiades’s outburst by expressing his fear of the violence that may be done to him
(213d). He asks Agathon to protect him, because Alcibiades is full of verbal abuse and
can scarcely keep his hands off him: “If he tries to use force, keep him away, because I am
very much afraid of his madness and his philerastia” (213d5–6). Philerastia combines
both Greek words for love: it is love of love. Alcibiades, in other words, makes sexual
love his favorite pastime. The madness Socrates refers to here is by no means the divine
sort of madness praised in the Phaedrus—the beauty-inspired ecstasy that is under the
control of philosophical reason and is expressed by indifference to wealth and other
ordinary human concerns (249d, 251e). Alcibiades’s madness is that of the violent tyrant
who tries to rule over not only human beings but gods as well (Republic 573c)—an apt
description for a powerful politician who dares to mutilate religious statues.
The encomium of Socrates given by Alcibiades allows Plato’s readers to see that someone
can be both an erastês and an erômenos. No one has more beauty of soul than Socrates, and
that is why Alcibiades loves him; and yet he is no merely passive object of sexual interest but
an active lover—that is why Alcibiades expects Socrates to make sexual advances toward
him. What he discovers is that as an active lover Socrates is highly selective and controlled,
allowing himself not the slightest physical expression of affection when such expression
would be inappropriate. Plato’s dialogues by no means p ­ resent Socrates as a man who is
indifferent to the physical allure of beautiful young men (see, for example, Charmides
153e–154d). His refusal even to touch or embrace Alcibiades (217a–219e), despite
Alcibiades’s best efforts to seduce him, does not arise out of insensitivity to physical
allure or a commitment to the avoidance of every physical expression of sexual desire.
His second speech in Phaedrus acknowledges with approval a lover’s longing for physi-
cal contact and never says a word in criticism of the lover’s kisses and embraces (255e).
The bad horse’s desire for homosexual copulation must be restrained (253d–254e), but
physical expression that falls short of that is accepted as normal and harmless.
So, we must regard Socrates’ coldness toward Alcibiades as a refusal to become
entangled in the sexual contract that Alcibiades implicitly proposes. Alcibiades assumes
570   Richard Kraut

that Socrates, lover of attractive young men that he is, wants to penetrate him and that,
in exchange for such intercourse, Socrates will pour his wisdom into him. Those assump-
tions are a colossal misunderstanding of what Socratic wisdom is and how those who
love and are loved by Socrates can benefit from it. (The whimsical suggestion that
wisdom might flow through physical contiguity, made earlier by Agathon at 175c–d, is a
close cousin of Alcibiades’s idea that Socratic wisdom might come easily to him, in
exchange for sex. A more distant cousin is the hero worship of those who tell the tale of
that marvelous night in 416 B.C.—Apollodorus and Aristodemus: they hang on every
word Socrates uttered, as though that will bring them understanding.)
The speech of Alcibiades is Plato’s device for holding up to our eyes an especially per-
nicious form of bad eros. When sexual attraction and activity is treated as a mere means
to a further end, a chip that can be traded in exchange for something else one wants, the
soul becomes corrupted. Symposium implies that the public offenses of which Alcibiades
was guilty are akin to his private waywardness. A man who would trade sexual penetra-
tion for wisdom thinks of human relationships in purely instrumental terms, and little
can be expected of a political leader who uses others as stepping stones to success. At the
same time, we cannot help finding something good in Alcibiades. After all, it is wisdom
that he seeks from sex with Socrates, and he has enough sense to recognize how remark-
able a human being Socrates is and to pay him fitting tribute. He is not the complete
tyrant. But Plato’s goal, in writing this portion of Symposium, is not to show Alcibiades
in a good light but to pursue the apologetic agenda of Plato’s Apology and Crito. In effect,
he tells the reader: yes, it is true that Socrates loved Alcibiades, but this was a love that
came close to making Alcibiades a better person; and it was a love that refused to treat
sex as an item to be traded in exchange for a successful career.

10. Seduction in the Phaedrus,

Homosexual Intercourse in the Laws

Although we have to do some thinking to see that the gap created by Diotima’s omission
of a discussion of bad eros is filled by Alcibiades, the Phaedrus directly confronts us with
bad eros. It creates the category of “left-handed love” and makes this one of its major
themes. Plato’s way of handling the dangers of sexuality in this dialogue is in line with the
approach he takes in his Symposium. “Left-handed” eros is at play in the speech of Lysias
and the first speech of Socrates, for each of these discourses is a device by which a man
tries to persuade a boy to prefer being sexually penetrated by him to having intercourse
with someone who, because of his passionate longing, is a genuine erastês. These speeches
propose a cold-hearted exchange: here are the many benefits you will receive from me,
because I am a calculating person who would not profit from harming you; in exchange
for what I can give to you, all that I ask is that you give me your sexual favors. The trade
 Plato On Love   571

being proposed is, in this way, rather like the transaction that Alcibiades thinks Socrates
is willing to enter with him.
Although Phaedrus treats the theme of a sexual contract at far greater length than
does Symposium, it does not say, in so many words, what is objectionable about such an
exchange of benefits. But it can safely be assumed that Plato expects his readers to find
the speech of Lysias a mere piece of cleverness (227c) rather than a truly convincing
demonstration that boys should have sex with men who do not love them. Phaedrus is
not presented as someone whose sexual attitudes have changed as a result of his admira-
tion for Lysias; rather, Phaedrus admires the speech of Lysias because of its audacious
advocacy of a paradoxical thesis that most people would consider shameful. The non-
lover who claims that he will confer great benefits merely in exchange for sexual satisfac-
tion is simply not to be believed. He offers no reason to suppose that he has any genuine
concern for the well-being of the boy he is trying to seduce. Nothing about the boy
attracts him but his physical beauty, which excites the nonlover’s desire to have an
orgasm with the boy’s aid (although he would never express himself so indelicately).
Would any father want his son to have sex with someone like that? The nonlover of
Phaedrus’s speech is transparently a sex-starved and clever manipulator.
This feature of the dialogue goes a long way toward explaining why Socrates assumes,
in his second speech, that the charioteer must restrain the bad horse’s eagerness to jump
all over the erômenos. How are an erômenos and his father to be assured that an erastês
really does take the boy’s well-being to heart and is not merely making fine speeches in
order to relieve his sexual urges? If the lover claims that he has something to teach the
boy, that he is offering a boy a philosophical exploration of the most serious matters, that
the boy’s beauty reminds him of the beauty of the Forms he once saw in a previous life,
he will not be taken seriously if at the same time he is trying to devise ways to get the boy
into bed. It will be difficult to tell him apart from someone who is merely saying these
things for the purpose of sexual satisfaction.
Elsewhere, Plato offers other reasons for refraining from homosexual copulation. The
speakers of Laws agree that in the well-governed city for which they are drafting legisla-
tion, such intercourse, though allowed elsewhere, will be banned. Why so? The dialogue’s
principal speaker, an unnamed visitor from Athens, says that the pleasures of male-male
and female-female intercourse are “contrary to nature” (636c6), and this is supported, at
a later point, with the claim (one that we now know to be false) that homosexual inter-
course is not found among animals (836c). It is difficult to believe, however, that this
point by itself carried a great deal of weight with Plato, for nowhere else does he claim
that human beings should take animal behavior as a model for their own way of life. It is
more likely that what carries most weight for Plato is the assumption that organs
should not be used to defeat the purpose for which they are suited by their nature. The
production of sperm, we are told in Laws (838e–839b) has a generative purpose, and
male homosexual intercourse wastes the reproductive potential and weakens the affec-
tive ties among husbands and wives. Plato elsewhere rejects the idea that the only
sexual intercourse that should be permitted is procreative, for he has Socrates say in the
572   Richard Kraut

Republic that couples who are past their fertile years may have sexual relations more or
less as they please (461b–c).
It is certainly possible that these ideas are at work in Phaedrus, for in Socrates’ sec-
ond speech the erastês whose disembodied vision of the Forms is long past or
obscured is likely to “pursue pleasure that is contrary to nature” (251a1), and the con-
text indicates that this is the pleasure of homosexual copulation. But Plato cannot
believe that those words by themselves convey a convincing argument for restraint
from homosexual intercourse. Rather, the dramatic structure of Phaedrus indicates
that such restraint is to be practiced because of the way it secures the friendship
between erastês and erômenos. The first two speeches of the dialogue (that of Phaedrus
and that of Socrates) are intended to show us how suspicious anyone will be if he
claims to be a friend and an educator but, at the same time, pursues sex as a quid pro
quo. The argumentative strategy of the dialogue is to show the value of homosexual
restraint by presenting it as a safeguard against misunderstanding the motives that
lie behind a sexual but not merely sexual relationship. Just as a professional teacher
might accept a rule that forbids sexual relations with students, Socrates insists that
the highest erotic ideal is one in which the complete physical expression of eros is
foresworn. But it should not be forgotten how lenient he is when he discusses those
devoted homosexual couples who are occasionally mastered by their desire for genital
intercourse (Phaedrus 256b–e).

11. Final Thoughts

Cephalus, the first interlocutor Socrates examines in Plato’s Republic, says that he agrees
with Sophocles about one of the great benefits of old age: released from the tyranny of
sexual desire, one can at last find peace and freedom (329b–c). Plato, I believe, has some
sympathy for this attitude. At any rate, he has Socrates endorse the idea that there is
something inherently transgressive in human sexuality. In the dreams of even the best of
us, lawless sexual urges—to have intercourse with one’s mother, or with gods or beasts
or any human being—make themselves manifest (571a–572b). But Plato’s recognition of
the dark side of human sexuality, which emerges most fully in his portrait of the tyrannical
soul, does not blind him to its great value. Without it, there would no future generations,
and our deep longing to perpetuate ourselves by making a long-lasting difference in the
world—by having children, or transmitting our conception of the good, or both—would
be fruitless. Furthermore, receptivity to the sexual allure of the human body is one of
the modes by which we take pleasure in the beauty of the physical world. That beauty
is not as great as the beauty of souls (which, in turn, pales in comparison with the
beauty of the form of beauty), but that does not mean that it would be best for us to be
indifferent to it. If old man Cephalus is no longer an erastês of all beautiful bodies, that
has to be counted as one of the ways in which old age is a period of decline. For the
beauty of the human form is one of the ways in which the sensible world offers us
reminders of beauty itself.
 Plato On Love   573

Plato is perfectly aware that genital intercourse can be intensely pleasant (Philebus
45d–e), but he does not take the pleasantness of an experience, in isolation from its cause
or object, to be a point in its favor. It is good pleasure—pleasure that it is good for some-
one to feel—that we should seek (Gorgias 495d– 499d), and so the intensity of pleasure
that can be achieved in genital intercourse is not by itself a reason for engaging in this
practice. Plato’s denial that every pleasure is good simply because it is a pleasure may be
defensible, but even if it is, we should recognize a blind spot in his thinking about sexu-
ality. He recognizes that such gestures as embracing and kissing are appropriate expres-
sions of one person’s sexual interest in and love of another (Phaedrus 255e–256a); these
are things we naturally do, when we are responsive as lovers to another person’s physical
attraction, and it would be insane to suppose that these expressions of eros are always
to be suppressed. Sexual intercourse is a more intense way in which eros is naturally
expressed, and Plato knows how strongly we desire it. But he cannot bring himself to
believe that because of the greater pleasure it gives sexual intercourse in a homosexual
couple it can be a fuller expression of affection than kissing and embracing, and that
when it carries with it this meaning it is welcome and healthy. What lies behind his dis-
approval, I have suggested, is his fear that nonprocreative intercourse compromises the
trust and affection that people should have for each other: it leads each partner in an
erotic relationship to suspect that everything he receives from the other is a mere means
to the relief of imperious sexual urges.
We must not take Plato to suppose that we should treat other human beings, beautiful
in body or soul or both, as mere stepping stones on the way to the vision of beauty itself.
The best sort of lover is someone who is bursting with ideas about how to improve human
life. Because he cannot fully understand or develop those ideas on his own, he needs a
conversational partner who will help him nurture those inchoate theories. That he needs
a partner in order to fulfill his need to give birth to a better world does not show that it is
only his needs that matter to him. In the best kinds of eros, self-regard and dedication to
others mutually reinforce each other; in the worst kinds, a lover destroys himself as he
goes about destroying others.
Above all, Plato insists that the erotic tendencies of human beings—their sexual appe-
tites, their yearning for immortality through propagation, their receptivity to beauty—
need to be educated, because they will never lead to anything of great value if they are
put in the service of mistaken conceptions of what is truly good and truly beautiful. The
desire to change the world so as to invest the future with something of ourselves will
merely replicate and rearrange its defective furniture if it is allied to common misunder-
standings of what is genuinely good for human beings. Love needs to be turned into
something more than an inarticulate yearning for a sexual life partner or a procreative
force. It needs to become something better than the intense alliance of two people who
care not at all for the larger world—or even for their own families (Phaedrus 252a)—but
only for their own togetherness and satisfaction. For that to happen on a grand scale,
Plato believes, we will need a new kind of political community.9

9 For their comments on an earlier draft of this essay, I am grateful to Elizabeth Asmis, Gail Fine,
Tushar Iranai, and Gabriel Richardson Lear.
574   Richard Kraut

Bibliography
Destrée, P. and Z. Giannopoulou (eds.), Cambridge Critical Guide to Plato’s Symposium
(Cambridge, 2017).
Ferrari, G. R. F. Listening to the Cicadas: A Study in Plato’s Phaedrus (Cambridge, 1987).
Griswold, C. Self-Knowledge in Plato’s Phaedrus (New Haven, Conn., 1986).
Halperin, D. ‘‘Why Is Diotima a Woman?” One Hundred Years of Homosexuality (New York,
1990), 113–51.
Hunter, R. Plato’s Symposium (Oxford, 2004).
Irwin, T. Plato’s Ethics (New York, 1995).
Kraut, R., “Eudaimonism and Plato’s Symposium,” in Destrée and Giannopoulou (eds.),
Cambridge Critical Guide to Plato’s Symposium (Cambridge, 2017), pp. 235–52.
Lesher, J., D. Nails, and F. Sheffield (eds.), Plato’s Symposium: Issues in Interpretation and
Reception (Cambridge, Mass. 2006).
Lorenz, H. ‘‘The Analysis of the Soul in Plato’s Republic,” in G. Santas (ed.), The Blackwell Guide
to Plato’s Republic (Malden, Maine, 2006), pp. 146–65.
Nails, D. The People of Plato (Indianapolis, Ind., 2002).
Nehamas, A., and P. Woodruff (trans.) Plato, Symposium (Indianapolis, Ind., 1989).
Nussbaum, M. The Fragility of Goodness: Luck and Ethics in Greek Tragedy and Philosophy
(Cambridge, 1986).
Obdrzalek, S., “Moral Transformation and the Love of Beauty in Plato’s Symposium,” Journal
of the History of Philosophy 48 (2010), 415–44.
Penner, T., and C. Rowe. Plato’s Lysis (Cambridge, 2005).
Sheffield, Frisbee, Plato’s Symposium: The Ethics of Desire (Oxford: Oxford University
Press, 2006).
Sheffield, F. “The Symposium and Platonic Ethics: Plato, Vlastos, and a Misguided Debate,”
Phronesis 57 (2012), 117–141.
Vlastos, G. ‘‘The Individual as an Object of Love in Plato,” Platonic Studies (Princeton, N.J.,
1973), 3–42.
Waterfield, R. (trans.) Plato, Symposium (Oxford, 1994).
 chapter 24

Pl ato’s Politics

Christopher Bobonich

1. Introduction

The dialogues that are most obviously important for Plato’s political philosophy
include: the Apology, the Crito, the Gorgias, the Laws, the Republic, and the Statesman.
Further, there are many questions of political philosophy that Plato discusses in his
dialogues. These topics include, among others: (1) the ultimate ends of the city’s laws and
institutions; (2) who should rule, the forms of constitution, and their ranking; (3) what
institutions and offices there should be; (4) the nature and extent of citizens’ obligation
to obey the laws; (5) the proper criterion of citizenship; (6) the political and social status
of women; (7) the purposes of punishment; (8) private property; and (9) slavery.
In such a short essay, I cannot explore all or even the most important works in detail
and cannot sketch Plato’s views on all the relevant topics without resorting to brief,
potted summaries. Although I hope to provide an overall picture of Plato’s political
philosophy, my aim is not to give a précis of the dialogues to be read in their stead, but
rather to concentrate on some of what seem to me to be the most fundamental and
persisting issues. In doing so, I shall focus on three moments: the “Socratic” dialogues,
including the Apology and the Crito; the great middle-period work, the Republic,
along with the Phaedo; and finally, two works from Plato’s last period, the Statesman
and the Laws.1

1 I take Apology, Crito, and Gorgias to antedate Phaedo, which is followed by Republic, Statesman,
and Laws (in that order). None of Apology, Crito, and Gorgias appeal to philosophers as knowers of
Platonic Forms, while Phaedo and Republic do. I count as “Socratic dialogues”: Apology, Charmides,
Crito, Euthydemus, Euthyphro, Gorgias, Hippias Major, Hippias Minor, Ion, Laches, Lysis, Menexenus,
and Protagoras. But it would not greatly affect my argument if some of these, for example, Euthydemus
and Menexenus, were later. A longer discussion would say more about the Meno, which I take to be
transitional between the Socratic dialogues and those of the middle period. I use “Socrates” and “Plato
in the Socratic dialogues” interchangeably, and intend no claim about the historical Socrates’ views. For
further discussion of chronology, see Irwin, chapter 3 in this volume.
576   Christopher Bobonich

2. The “Socratic” Plato

In the Apology, Socrates compares himself to a gadfly set upon the great horse of Athens
whose job it is to rouse and persuade his fellow citizens (30e6–31a2). Socrates confers on
each citizen

what I say is the greatest benefit, by trying to persuade him not to care for any of his
belongings before caring that he himself should be as good and as wise as possible.
(Apol. 36c4–8, cf. 29d–e and 36d–e)2

What is especially worth noting here is that Socrates claims to benefit Athens by benefit-
ing its citizens, and that this benefit consists in getting them to examine themselves and
their lives with regard to virtue. Since Plato, throughout his career, believed that virtue
was by far the most important contributor to happiness (although not its only compo-
nent), such encouragement to virtue seems a reasonable way to proceed for anyone
really seeking to benefit his fellow citizens.3
Socrates concedes that it may seem odd that while he gives advice privately

I do not venture to go to the assembly and there advise the city . . . Be sure, men of
Athens, that if I had long ago attempted to take part in politics, I should have died
long ago, and benefited neither you nor myself . . . no man will survive who genuinely
opposes you or any other crowd and prevents the occurrence of many unjust and
illegal happenings in the city. A man who really fights for justice, must lead a private,
not a public, life if he is to survive for even a short time.
(Apol. 31c4–32a3, cf. 32e–3a, 36b–c)

In sum, Socrates is a divine gift to the city and benefits it. Nevertheless, the current state
of politics is bad and apparently must remain so, and Socrates contrasts his own activity
with practicing politics.
But in the Gorgias, Socrates accepts that the proper task of a statesman is to benefit the
city maximally and that the best way of doing this is by making the citizens virtuous
and happy.4

I believe that I am one of a few Athenians—so as not to say I am the only one, but
the only one among our contemporaries—to attempt to practice the true political

2 My translations of Plato draw upon Cooper (1997), except for the Laws where I draw upon Pangle
(1980). On Plato’s political philosophy, see Annas (1995), Barker (1960), Berges (2009), Brickhouse and
Smith (1999, 185–229), Irwin (1995), Johnson (2015), Kraut (1984), Laks (2005), Lane (1998), Rowe
(1995a), (1995b), Saunders (1981), Samaras (2002), Schofield (2006), and White (2007). Schofield (2017)
is a useful Laws bibliography.
3 See Bobonich (2002), (2011), and Irwin (1995). Here, and throughout, I cite work giving further
references.
4 Cf. Hipp. Maj. 284b–d, Lysis 209c–e.
 Plato’s Politics   577

art and perform the true politics. This is because the speeches I make on each occasion
do not aim at gratification, but at what is best. (Gorg. 521d6–e1, cf. 502e–503b)

What is best is making the citizens as good as possible (Gorg. 513e5–7) and unless the
city is tended in this way, no other action can benefit the citizens (513e8–514a3).5 Here,
unlike the Apology, Socrates claims to attempt to practice the true art of politics.
The Socratic dialogues do not attempt to describe the best possible city. In the
Republic, this becomes a fundamental task of political philosophy and Plato revisits the
issue in the Statesman and the Laws. But we can try to work out, drawing on claims
found in the Socratic dialogues, what the best possible city would look like. In doing so,
we shall both more deeply understand the political implications of the ethical views of
these dialogues and see how their unresolved issues help set an agenda for the following
dialogues.
But first we must consider Socrates’ famous denial of knowledge. Throughout the
Socratic dialogues, Socrates claims not to know the definitions of the virtues or ethical
truths, such as that it is better to suffer than do wrong, and, I think, such denials are not
merely “ironic.”6 In ethical matters, Socrates has nothing analogous to the knowledge
that craftsmen have of their craft; he does not possess definitional knowledge of ethical
properties and explanatory accounts based on such definitions. So how should he pro-
ceed while lacking such knowledge? Socrates provides a general answer in the Crito
when he responds to Crito’s advice to escape from jail.

We must therefore examine whether we should act in this way or not, as not only
now but at all times I am the kind of man who listens to nothing within me but the
argument that on reflection seems best to me [moi logizomenôi beltistos phainêtai].
I cannot, now that this fate has come upon me, discard the arguments I used;
they seem to me much the same. I revere and honor the same arguments as before,
and if we have nothing better to say at this moment, be sure that I shall not agree
with you. (Crito 46b3–c3)

The Crito does not make explicit exactly what an argument seeming best on reflection
consists in, but Socrates subjects at least some of the relevant claims to examination
using Crito as his answerer (47a–48b) and gives Crito the opportunity to object (48d–e)
to Socrates’ answers to the speech of the laws of Athens. In the Gorgias, Socrates claims

5 This is related to a fundamental claim of Plato, the Dependency Thesis, that is, nothing benefits a
person who lacks wisdom or knowledge of the good; see Euthydemus 278e–282a and Meno 87c–89a.
6 Socrates accepts, I think, the possibility of a human attaining such knowledge even if he himself
lacks it. Socrates does occasionally make knowledge claims about ethical matters, e.g. Ap. 29b6–7. It is
controversial whether we should explain such claims by positing that he recognizes a weaker sort of
‘knowledge’ (e.g., a form of non-explanatory justified true belief) and claims to have such knowledge of
some ethical propositions or whether his occasional knowledge claims are merely a loose way of
speaking. I am inclined with Benson (2000, 236) to think that Socrates is speaking loosely, since he
never offers even a rough and partial characterization of how this weaker sort of knowledge differs
from the stronger sort that he lacks and, in fact, never explicitly distinguishes a weaker from a stronger
conception of knowledge. For more discussion, see Benson, chapter 5 in this volume, Fine (2008),
McPartland (2015), and Matthews (chapter 16) and Taylor (chapter 18) in this volume.
578   Christopher Bobonich

that his ethical beliefs have never been refuted in his examination of them, while beliefs
inconsistent with his, when examined, are inconsistent with other beliefs held by the
one examined and that the interlocutor responds to this conflict by giving up the claim
that is inconsistent with Socrates’ beliefs (474a–b, 508e–509a).7 So Socrates will con-
tinue acting on these beliefs at least until something better comes along. Although we
receive no worked-out argument that this is the only (or the most) rational way of pro-
ceeding, it is, I think, a reasonable reaction to Socrates’ circumstances. Socrates can, it
seems, reasonably act in this way without supposing that he possesses knowledge.8
In what follows, I shall take Socrates’ epistemic limitations as a constraint, that is,
I shall ask what is the best sort of city possible given that no one in it possesses ethical
knowledge and no one is (significantly) epistemically better off than Socrates. I shall
also assume that most citizens are below Socrates’ ethical and epistemic level. As we
shall see, the epistemic and ethical capacities of the citizens affect what sort of cities are
possible and the goodness of the resulting cities. To begin, we face the issue of what
changes are actually possible. There are reasons for pessimism.9 First, Apology 31c–32a
suggests that Socrates thinks that it is impossible for him, or others like him who would
act justly, to come to rule in Athens and that even any attempts to change laws and insti-
tutions for the better will lead to the destruction of those trying to do so before they can
achieve anything.
Second, in the Crito, Socrates claims that

One should never do wrong in return, nor harm any man, whatever harm one has
suffered at his hands . . . I know that only a few people hold this view or will hold it,
and there is no common deliberation [boulê] between those who hold this view and
those who do not, but they inevitably despise [kataphronein] each other’s resolu-
tions and designs. (49c10–d5)

Socrates’ word for deliberation, boulê, in addition to meaning “advice” or “deliberation,”

is politically highly loaded: it is the name of the Council or Senate in Athens.10 It is

7 See the seminal Vlastos (1994a) and (1994b) and the helpful survey, Wolfsdorf (2015). What
counts as an “elenchus” and how common elenchus is in the Socratic dialogues are controversial; see
Scott (2002). Benson holds that (1) the elenchus is a unique form of argument such that its only
necessary condition of premise acceptability is that the interlocutor believes the premise and that this is
also a sufficient condition (2002, 105), and (2) such a method can only establish inconsistency among
the premises the interlocutor accepts and cannot provide differential support for the denial of the
refutand. For criticism of Benson’s “non-constructivism,” see Santana (2011). Although I cannot defend
the view here, I accept a constructive interpretation of some of Socrates’ ways of arguing, including
Vlastos’ “standard elenchus.” Here I shall speak of Socrates’ “examination” of others while allowing that
this may take different forms.
8 Pace Benson, not all the premises the interlocutor accepts have the same epistemic status: the
refutand is rejected by the interlocutor and Socrates and in at least some cases both of them accept the
premises entailing the refutand’s denial. Further support could come from Socrates’ examination of
himself (e.g., while reflecting about what is best) and of others who deny the refutand if in both cases
he finds no inconsistency in the relevant set of beliefs. Cf. Santana (2007), (2011), and McPartland
(2015, especially 134–35).
9 Cf. Kraut (1984, 194–309).
10 Other political language: 49d2–3; a common boulê for Socrates and Crito: 47c11.
 Plato’s Politics   579

sometimes thought that Socrates only means here that most people will reject the “no
harm” principle until they undergo Socratic examination; but that doing so will lead
them to adopt the principle and to retain this true belief. Yet given the actual effects of
examination on many of Socrates’ interlocutors (they grow angry, leave, answer insin-
cerely, or, even if they do change their minds, this may not last long), it is not clear how
much optimism is warranted. Thus Socratic examination may not only fail to bring about
knowledge or stable true belief in those on whom it is practiced, it may not prevent a
return in the long run to a false conceit of wisdom. (It is difficult, for example, to imagine
Euthyphro or Hippias maintaining an awareness of his ignorance.) But unless most peo-
ple’s beliefs about doing wrong do stably change, those disagreeing over this principle will
be unable to engage in the sort of deliberation needed for good collective decision-making.
They might not only disagree over a wide range of political issues, but insofar as they
“despise each other’s resolutions and designs,” they may not even be able to sustain
productive rational discussion. (Even if Socrates and those accepting the “no harm”
principle are willing to engage in rational discussion, their opponents may not be.)
This is a good instance of how an unresolved issue in the Socratic dialogues leads
to political indeterminacy. To go further, we need answers to questions such as: (1) if
enough experience of Socratic examination would persuade all (or most) citizens, is it
permissible to coerce the participation of those who are unwilling to undergo that
amount of examination?; (2) is there some new form of education that would make citi-
zens more likely to participate in and benefit from such examination?; (3). if not, and
some cannot improve via examination, should they be citizens, and what, if any, politi-
cal functions could they have, and what restrictions should they live under in the best
city?; (4) is there some sort of education not employing Socratic examination that bene-
fits such people sufficiently so as to allow them to exercise political functions?
If educational success in extant cities is sufficiently unlikely and the costs of trying
sufficiently high, Socrates might reasonably recommend withdrawal from public activity.
But that only answers the specific question of what Socrates and others like him should
do in circumstances such as those in which Socrates finds himself. It does not answer
the more general question of what the best humanly possible city would be on Socrates’
principles.
So what would such a city look like? We find no answer to this in the Socratic dia-
logues and, as I have suggested, this is not accidental, since we would need to come to
some conclusions about issues that they leave unaddressed. The first, minimalist, option
is that Socrates (and others like him) carry out examinations, but have no greater hopes
than persuading probably only a few.
This is not, however, the only option compatible with Plato’s commitments in these
dialogues. A number of Socrates’ ethical beliefs have passed examination. These include:

a) it is better for a person to suffer than to do injustice (Gorg. 469a–475e),

b) if one commits an injustice it is better to be punished than to escape punishment
(Gorg. 476a–479e),
c) virtue is fine and good for its possessor (fine: Charm. 159e, La. 192c; good for
possessor: Charm. 160e, Meno 87d),
580   Christopher Bobonich

d) no one does wrong willingly (Gorg. 466a–468e, 509c–e), and

e) virtue is necessary for happiness and is the most fundamental part of
happiness.11

In these dialogues, Plato also accepts some claims as obvious with little disagreement
and thus little defense, such as the Principle of Rational Eudaimonism, that is, that the
ultimate aim of all of a person’s rational actions is her own greatest happiness. It is rea-
sonable to ask, even if Plato does not do so in the Socratic dialogues, what a city might
look like if these principles were embodied in its laws. Doing so leads us to four espe-
cially important questions.

(1) If these principles are embodied in laws and institutions, they will require
sanctions and will sometimes require the people subject to them to give up
decision-making authority over various aspects of their lives. What legitimates
such coercion and removal of decision-making authority? (The Crito has
arguments for an obligation to obey the law in general. Here I want to consider
what specific grounds there may be for coercing people in accordance with
Socratic principles.12)
(2) In implementing Socratic principles, the aim is to make the citizens virtuous and
happy. Will all citizens benefit equally or will the benefit be highly uneven?
(3) Plato is pessimistic about the possibility of sustained public action on behalf of
justice. How stable could a city based on such laws be?
(4) We must also consider questions of motivation; does political activity compete
with the development of one’s own virtue so that Socrates and others like him
could not rationally pursue activity in support of such laws?

2.1 Coercion
To simplify, I shall focus on the costs and benefits to the person to whom the law applies
and shall not attempt to characterize coercion precisely. If coercion results in the person
coming to have knowledge, it does not seem that the Plato of the Socratic dialogues
would object. In the Euthydemus, Socrates says, “Let him destroy me, or if he likes, boil
me or do whatever he wants, but he must make me good” (285c4–6).13

11 Since a person lacking all goods cannot be happy, virtue is necessary for happiness. Since other
goods are dependent on wisdom for their being beneficial to their possessor, wisdom is the
fundamental component of happiness. Further, although the interpretation of the following passages is
controversial, they appear to make virtue sufficient for happiness: Gorg. 470e, 507b–c, Euthyd. 279d and
282c–d.
12 See Johnson (2015, 236–51).
13 One could argue that the boiling and so on are justified in part by Socrates’ agreement, but there
is no such agreement in the following Gorgias passage.
 Plato’s Politics   581

For the more common and politically more important cases in which the person does
not come to possess knowledge, consider the following passages.

In those areas where we really understand something everybody . . . will turn them
over to us, and there we shall act just as we choose, and nobody will want to get in
our way. (Lysis 210a9–b4)

In the Charmides, Socrates considers the implications of a proposed definition of mod-

eration as knowing what it is that one does and does not know and what others do and
do not know. If this definition were correct,

it would be of the greatest benefit to us to be moderate, because those of us who had

moderation would live lives free from error and so would all those whom we ruled.
Neither would we ourselves attempt to do things we did not understand—rather we
would find those who did understand and turn the matter over to them—nor would
we trust those whom we ruled to do anything except what they would do correctly,
and this would be that which they understood. (Charm. 171d5–e5)14

One might argue that epistemic deference to an ethical expert is only justifiable if one is
rationally justified in thinking that the option that she recommends is correct and that
one is only so justified if one possesses a good argument, independent of the expert’s
opinion, for that option. This suggestion is not, however, obviously correct. Suppose
that on many previous occasions you and I have differed over what the right action is
and on all of them I have eventually come to think that your view was correct and my
view wrong. On the next occasion on which we disagree on a close case, it is not clear
that it would be wrong for me to infer that there is more reason than not to think that
you are correct even if I cannot identify a flaw in my reasoning.15
In the Gorgias, Socrates likens the true statesman to a doctor who would be convicted
by a jury of children on charges brought by a cook (521d–522c). Such a doctor would act
appropriately although he is unable to justify to the children the cuttings and burnings
that are involved in the proper course of treatment. Similarly, Socrates’ examination of
his fellow citizens, despite being unpleasant to them, is just (Gorg. 522bc). Although
Socrates does not explicitly say so, the analogy suggests that yet harsher methods than
Socratic examination could be justifiable if they were in the citizens’ best interests even if

14 The difficulties surrounding this definition do not affect the point with which I am concerned.
15 Charm. 170d–171c does not exclude such deference, since deference does not require that I know
that you have ethical knowledge. Prot. 313e–314a is more difficult, but we have reason to be skeptical of
the case on which the analogy is based: it (1) seems to justify having an expert examine whatever one
eats, and (2) faces the problem of identifying the food expert. If observed past success is sufficient in
this case, why not in the ethical case? Moreover, it seems concerned with mistakenly adopting an
opinion rather than undermining one’s decision-making capacities. On ethical experts in Plato, see
Benson (2015, 20–47) and Hatzistavrou (2005), which underestimates the expert identification
problem; for contemporary discussion, see Enoch (2014).
582   Christopher Bobonich

the citizens do not consent to them.16 (Indeed, as we shall see, the Gorgias is optimistic
about the ethical efficacy of traditional punishments.)
What is striking in these passages is that Socrates does not seem to take into account
the costs of giving up one’s decision-making power to someone else. One important cost
that should concern Socrates is that deferring to an ethical expert or (perhaps to a lesser
extent) being subject to coercion may tend to atrophy one’s own reasoning and decision-­
making capacities. But Socrates clearly does accept that it can be permissible for the laws
to coerce the citizens, since he accepts that it is appropriate for the laws to use punishments
to deter wrongdoing and that such punishment benefits the wrongdoer (e.g. Gorg.
476a–479e).17 Neither actual coercion nor simply giving up such decision-making
power would be objected to by Plato as a violation of one’s rights or autonomy if these
are understood independently of a person’s good. But even in the Republic, and especially
in the Laws, Plato does take it to be vastly better, if one is right, to be ruled by oneself.
One might think that something like this is suggested by Crito 46b–c quoted previously,
but Socrates there does not distinguish the value of acting on one’s own reasoned judg-
ments from the idea that such a way of proceeding gives one the best chance at correct-
ness. Plato faces a further worry: if virtue is knowledge and one benefits through
approximating it by having fewer false, and especially, more true beliefs, then doing the
right thing for the wrong reason (which would seem to be the typical position of one
who is coerced) stands in particular need of justification as a benefit.18 Plato could
appeal to two lines of thought: (1) avoiding injustice through fear of the consequences is
less bad than actually doing injustice, and (2) one who refrains out of fear of the conse-
quences may still come to acquire the right reasons (this is, after all, what parents hope
for in coercing children).19 In any case, Socrates in the Gorgias despite lacking explicit

16 The harsh treatment seems to be directly justified by the benefit to the coerced. The Gorgias holds
that each person, in some way, believes the truth at least in ethical matters (474b8), regardless of that
person’s conscious beliefs, and can be brought via examination to assert it. Nevertheless, Socrates does
not appeal to the idea (e.g., in Pitkin (1965)), that imposing a policy upon a person can be justified by
the fact that she would consent to it in good discursive conditions.
17 Some deny that Plato of the Socratic dialogues thought that painful punishments, as opposed to
intellectual persuasion, could benefit a wrongdoer, e.g. Moss (2007, 232 n.8). I agree with Brickhouse
and Smith (2010) that the Gorgias’s text is so emphatic on the benefit of painful punishments that we
cannot interpret it away.
18 The Dependency Thesis renders benefitting from false belief about the good problematic, cf. note
5 of this chapter.
19 Brickhouse and Smith (2010) and (2015, 200–07) make the important argument that although a
person’s non-rational motivations cannot directly move her to act, they can (1) prevent her from
attaining knowledge of the good, and (2) lead her into diachronic belief akrasia by making objects
appear better than they are. They also argue that (3) a purely cognitive account cannot explain how
wrongdoing damages the soul. With respect to (2), it is not clear that non-rational desires are needed,
for example, , to explain the “power of appearance” in the Protagoras. There errors about objects’ sizes
owing to spatial proximity are a paradigm case of a “cold”, that is, cognitive error. Temporal proximity
can also explain misestimation cognitively: the agent’s attention focuses on the near-term good and
away from long-run consequences, cf. Nisbett and Ross (1980, 45–62). For the Protagoras, see Bobonich
(2007, 47 n.17) and Price (2011, 264–69). With respect to (3), a cognitivist can argue that the initial
belief prompting the wrongdoing will not remain isolated from the person’s other beliefs. It is plausible
that, ceteris paribus, an agent’s psychic system tends to increase its coherence. Beliefs that are active are
 Plato’s Politics   583

answers to these worries seems confident that ordinary legal punishments for crimes
such as theft typically improve those subject to them (e.g. 478d–480d).
Some of Socrates’ ethical beliefs can be more easily translated into law than others: for
example, it is easier to imagine how a law code could aim at inculcating virtue than to
see how it might reflect the belief that it is better to suffer than to do injustice. But even
with respect to the latter, there are legal implications. The view that punishment is bene-
ficial to one doing injustice will, ceteris paribus, encourage the legal regulation of more
of a citizen’s life and law in both Kallipolis and Magnesia (the ideal cities respectively of
the Republic and the Laws) will cover aspects of citizens’ lives that are not legislated
about in Athens. Although an ideal legislative system should be designed to minimize
mistaken verdicts, it would seem better—since doing injustice while avoiding punish-
ment is the worst outcome—to err on the side of mistaken punishment rather than
allowing unpunished wrongdoing.20
Insofar as our sketch of Socrates’ ideal city respects his own epistemic limitations, the
laws will be unable to promote virtue and virtuous action if these require ethical knowl-
edge. In determining what character states and types of action are to be encouraged,
however, Socrates should be able to rely on judgments about those that have survived
Socrates’ examination.21 But it should also be a pressing matter to discover what virtue
is, and we should expect the city’s institutions to provide some help. In the early dia-
logues, unlike those of the middle- and late-period, Plato does not seem to think that
progress will be facilitated by specialized knowledge, for example, intensive training in
higher mathematics, so the city will not need such institutions (for that reason anyway).
It might, however, support Socrates’ inquiries in various ways, for example, by free meals
in the Prytaneum (Apol. 36d). In addition, some sort of city-supported education would
seem a good idea if such education could help remove citizens’ false conceit of wisdom
and make them more amenable to and better at participating in Socratic examination.

2.2 Benefit
Determining who can benefit and in what ways requires answering some challenging
philosophical questions. Even if citizens cannot attain ethical knowledge, how closely

used to justify other actions and support other beliefs and tend to gather support from other beliefs the
agent has or acquires. This larger and more coherent set of beliefs will be harder to dislodge than the
initial single belief.
20 If mistakenly punishing is doing an injustice, this is a significant cost for this option. Perhaps the
psychic harm done to well-intentioned and epistemically responsible agents who wrongly punish is
much less than the psychic harm suffered by those left unpunished since their wrongdoing is typically
neither well intentioned nor well informed.
21 Our examples of examination tend not to focus on such questions, but there is no obvious reason
that this is impossible. (The Crito considers the justice of a particular action, escaping from jail, as
well as the action type, disobedience to law.) Law prescribes and proscribes certain action types, but
Socratic examination suggests that it is not possible to specify action types such that they will always
be, for example, just, without using ethical terms. Thus a good legal system needs a doctrine of equity.
584   Christopher Bobonich

can they approximate it? Would Euthyphro, or Callicles really be improved by more
examination? And why are there so many failures? It might be that the defect is simply
epistemic; the knowledge necessary for virtue may be as hard to attain as knowledge that
a proof of Poincaré’s Conjecture is correct, so even with the best efforts, few can attain it.
Or is the defect owed in part to the effect that non-rational motivations can have in
preventing and irrationally changing true beliefs?22 At least an important part of the
problem is that Socrates’ examination of others takes people as they are, but is there any
sort of education that might enable people to make greater progress? To settle these
questions, Plato needs to answer various questions in psychology, epistemology, and the
theory of education and learning.
Finally, and perhaps most important, Socrates does not offer worked-out definitions
or accounts of virtue and happiness in the Socratic dialogues. Even if we grant, for
example, that virtue is knowledge of what is good and bad for human beings and that
happiness is one’s optimal condition, these characterizations remain purely formal with-
out an account of the good. If, for example, virtue and thus happiness require possessing
holistic explanatory understanding, it is reasonable to think that few can achieve them.
Without a more substantive account of the good, it is especially difficult to determine
who can benefit, and how much they can benefit, from the various ways in which
Socratic principles might be embodied in laws and institutions.23

2.3 Stability
Here again, determining the stability of such laws and institutions requires answering
some of our previous questions. How far is there a coincidence of interests among the
citizens and how far can the citizens come to realize this? Even if they do accept such a
coincidence, will their acceptance rest on what Socrates would count as good reasons?
We shall also need to draw on answers about what benefits the citizens in order to deter-
mine what are the effective and just means of reducing conflict in the city. Political sta-
bility, for example, might be fostered by encouraging false beliefs among the citizens
(e.g., the Republic’s Noble Lie) and the costs of such a policy will only be clear if we
understand how bad having such false beliefs is.

22 If there are such motivations in the Socratic dialogues, cf. note 20 of this chapter.
23 Granting that the Protagoras advocates hedonism, it does not specify the life that maximizes
pleasure. Bobonich (2011) argues that on the account of virtue that the Socratic dialogues seem to be
moving toward, that is, knowledge of the good, it is difficult to sustain the claims we find there of the
priority of virtue, the necessity and sufficiency of virtue for happiness, and the Dependency Thesis.
Plato's middle-period conception of human beings as most fundamentally rational creatures, and his
understanding of rationality as involving love and knowledge of the truth help furnish one response to
this gap.
 Plato’s Politics   585

2.4 Perfection
Some scholars think that in the Socratic dialogues there is an especially strong tension
for someone like Socrates between his own perfection and undertaking political
action.24 There are two specific worries here: (1) since Socrates does not have knowledge,
would he not be better off attending to his soul rather than engaging in political action?;
and (2) since Socrates lacks knowledge, should he be confident enough to enforce his
principles on others?
To take up (2) first, although there are some passages that suggest that one should not
undertake politics until one possesses the relevant knowledge, given Socrates’ confi-
dence in the Athenian penal system, there is no reason to think that it would always be
wrong to enforce laws embodying principles that have repeatedly survived Socratic
examination.25 With respect to (1), some have argued that the improvement of Socrates’
own soul takes nearly absolute priority over benefiting others. Here again, to see whether
this is so, we need answers to questions that the Socratic dialogues do not fully address.

1. Without a theory of happiness, we cannot determine how far others’ happiness

might compete with mine or whether it could be consistent with, or even a part
of, mine.
2. We need to know how far others can benefit from the political activity I might
undertake. If the mere enforcement of laws against, for example, theft, constitutes
a significant benefit to the Athenians, how much more would they benefit from
laws based on Socratic principles? Other things being equal, it seems that the
greater the benefit to others, the stronger will be Socrates’ reasons for helping.
3. How much improvement is possible for Socrates? Does he think it possible that
human beings can attain ethical knowledge? If not, how much does he benefit
from engaging in further examination? How long has it been since his ethical
beliefs have changed? Would just more examination with the sorts of inter-
locutors he has already had discussions with really be of significant benefit to
him: Would he not eventually just come to see them make the same mistakes
again or others at least as easy to refute? Are there other routes to epistemic
improvement?
4. Finally, how valuable to Socrates is it simply to think about the beliefs and argu-
ments that he already possesses? In later dialogues, Plato clearly thinks that there
is enormous value in contemplating the truth. Is the same true for Socrates think-
ing about his own system of ethical beliefs?

24 E.g., Kraut (1984, 207–15).

25 E.g., the disputed Alcibiades I 118A ff., and Xenophon Memorabilia 4.2.2 and 4.6.
586   Christopher Bobonich

3. The Middle Dialogues

In the Republic, Plato invokes philosophers’ knowledge of Forms to explain why

­philosophers should rule.26 But the Phaedo, which is seen as politically unimportant,
invokes the Forms in distinguishing two basic groups of people and in giving more content
to virtue and happiness. The Phaedo’s significance is clearer if we see it as a successor of
the Socratic dialogues and a predecessor of the Republic. In particular, the Phaedo gives
first answers to some of the questions that the Socratic dialogues left unanswered.

3.1 The Phaedo

In the Phaedo, Plato thinks that knowledge of Forms is possible (and may come in
degrees); knowing, for example, the Form of Justice will involve knowing the “real” defi-
nition of justice.27 The Forms allow Plato to distinguish two basic kinds of people:
philosophers who seek, and may partially attain, knowledge of Forms and believe that
having such knowledge is a prominent part of their ultimate end, and non-philosophers,
none of whom accept the existence of Forms, and who have as their ultimate ends bodily
goods, that is, wealth and honor (e.g., Phd. 68b–c and 82d–83e).28 Such a conception of the

26 I take the Form of F to be the non-sensible property F (rather a non-sensible thing having the
property F). Other entities are F in virtue of standing in the appropriate relation (or having the
appropriate non-relational tie) to the Form of F. I take no position on whether Forms are the only
objects of knowledge or whether one who knows the Form of F (and perhaps some other Forms as
well) can also know Fa where a is a sensible particular.
But if virtue is a kind of knowledge, then it will require knowing some Forms, and if virtue is
necessary for happiness, then being happy will also require knowing some Forms. See Harte
(chapter 19) and Matthews (chapter 16) in this volume for further discussion of Forms. On the role
of Forms in ruling, see Bobonich (2016).
27 For the possibility of gradations in knowledge, see Phd. 65c–e, 66b, 66d–e, 67b, and 68b. Socrates
and others know the Form of the Equal (Phd. 74b2–4). Vasiliou holds that Socrates does not know the
Form of the Equal (2012, 23–24) because he “never endorses” Simmias’s claim that “we” know the Form
(Phd. 74b2–3). But if Socrates did not endorse it, he would not then immediately ask “Whence have we
acquired knowledge of it?” (74b4). For a possible attribution of knowledge of value of Forms, see 76b.
As Butler (2012, 122 n.27) rightly notes, 79d1–7 along with 84a2–b3 entail that wisdom is in principle
available while embodied, although it may not be “pure” until the soul separates after death (66e2–67a2,
68b3–4). For a very helpful extended discussion of relevant issues, see Fine (2016).
28 Weiss (1987, 62–63) suggests that loving, rather than having, wisdom suffices for virtue. This is
hard to reconcile with 69b3–4’s requirement that virtue be “with wisdom” (meta phronêseôs). Weiss
suggests that meta here means “with the aim of ” (65, n.18). For meta, LSJ gives “in pursuit of ” only
with the accusative, not the genitive. Weiss’s Apol. 32b8–c1 is not an example; there, Socrates is not
acting “aiming at” the law, but rather in cooperation with it. Even if wisdom were not possible while
embodied, philosophers would differ radically from non-philosophers in their ultimate ends.
Philosophers take wisdom as a prominent part of their ultimate end and this end, unlike that of
non-philosophers, has non-sensible content. Also, pace Weiss, a love of wisdom is not sufficient for
 Plato’s Politics   587

good is radically mistaken because it attributes purely sensible content to the good.29
True virtue requires one to value wisdom, that is, knowledge of the Forms for its own
sake, and to use wisdom to guide one’s choices. Thus only philosophers possess real
virtue, while non-philosophers have only a “shadow-painting of virtue that is really
slavish and contains nothing healthy or true” (Phd. 69b7–8).30 Since happiness for Plato
crucially depends on real virtue, only philosophers can be happy. The lives of non-
philosophers are necessarily unhappy, and in the afterlife “they lie in mud” (Phd. 69c3–6),
which is, regardless of Plato’s views about the afterlife, another way of characterizing the
value of their earthly lives.
Nothing except being a philosopher can significantly ameliorate this bad condition,
and it does not seem that Socratic examination or any other education can succeed in
turning most people into philosophers.
We can now make some progress on the questions left by the Socratic dialogues.

3.2 Coercion and Benefit

There seems to be a significant limitation to what a city can do to improve the lives of
the vast majority of its citizens; no non-philosopher can have a life that is really worth
living for a human being. When they undergo reincarnation, the very best of the non-
philosophers return (Phd. 82a–d) as wasps, bees, ants, members of some other tame and
political race, or again as respectable people.31 If coercion makes some of their charac-
ters less bad than they might otherwise be, it can still be justified by minimizing the
harm to the person coerced, but the laws and other institutions can bring about no
significant positive benefit for them.

virtue, since one might value non-sensible properties that are bad-making. Some success in valuing is
necessary, although Plato does not give details.
29 In the Phaedo (and the Republic), it is sufficient for pessimism about non-philosophers’ virtue and
happiness that the content of their ultimate ends does not have non-sensible content and it is sufficient,
but not necessary, for this they do not recollect Forms. For such failure in their ultimate ends, see Phd.
68b–c, 82d–83e and Bobonich (2002, 14–31). Scott (1995, 53–73) persuasively argues that non-philosophers
do not recollect in the Phaedo. For later discussion, see Harte (2006), Kelsey (2000), and Woolf (2000,
especially 128–31).
30 Vasiliou thinks that those with the “popular and political virtue” (82a12–13) act virtuously for its
own sake (2012, 13). Such optimism faces two problems Vasiliou does not discuss. First, political
virtue’s possessors merit insect reincarnations (cf. note 29 of this chapter). Second, the souls of the
reincarnated (including those with political virtue) are “polluted and impure” at death because they
have “served” the body and been “bewitched” by it (Phd. 81b–c). Thus they are lovers of the body
and so lovers of money, honor, or both (68b–c) and so do not value virtue in itself. This is why the
eincarnated souls are “not at all those of the good, but of the bad [phaulôn]” (81d6-7). Kraut has a more
optimistic view of slavish virtue: “their souls contain a shadowy form of wisdom—a partial grasp of the
good . . . They have some true opinions about what is valuable” (2010, 56). This is hard to reconcile with
Plato’s characterization of slavish virtue as containing “nothing [ouden] healthy or true.”
31 On the textual difficulty at 82b7, see Bobonich (2002, 484–5, n. 15). There is no suggestion that
returning as a respectable person is vastly better than the other reincarnations.
588   Christopher Bobonich

3.3 Stability
Since philosophers and non-philosophers differ radically in their understanding of
virtue and happiness, such a city cannot be a common association aiming at furthering a
shared conception of happiness among the citizens. It could not realize what is com-
monly thought of as the goal of the city in classical political philosophy: it could not be a
shared association in which all the citizens aim at the genuine common good, that is,
developing and fostering virtue in each other. Indeed, it is a challenge to see how such a
city could be stable for long.

3.4 Perfection
On the Phaedo’s understanding of knowledge and the place of reason in human nature,
there is great value to the individual philosopher not only in improving her knowledge,
but also in continued contemplation of her existing knowledge. Even if she advances
no further, contemplation’s value seems to compete strongly with political action even
if Socrates does not explicitly draw attention to this fact.32 Also, since so little improve-
ment is possible for non-philosophers, the tension between seeking one’s own good and
seeking the city’s good is all the greater.
Since the Phaedo’s views about non-philosophers rest on its epistemology and
­psychology, we cannot dismiss them as a mere aberration.33 If the Republic’s position

32 Philosophers value not only the pursuit of wisdom, but also its possession (Phd. 66d3–e4).
Attaining it has been Socrates’ “chief preoccupation” (67b9–10) in life and we can infer that its
comparative value to him is at least very high from the fact that his belief that pure knowledge
(68b9–10) is available in Hades, if anywhere, is sufficient to make him glad to go there (68a7–b3).
Socrates does explicitly say that “war, factional struggle and battles” (66c6–7) interfere with the leisure
necessary for philosophy. It seems reasonable to think that other political activities besides factional
struggle also compete with philosophy for leisure time.
33 The Phaedo does not draw all the distinctions that we would like. The cases it is clearest about are
philosophers who possess wisdom (phronêsis) as described in the “right exchange” passage (Phd.
69a–d) and non-philosophers with “slavish” virtue (andrapodôdês, 69b) or “popular and political”
virtue (dêmotikên and politikên, 82a–b). These philosophers are happy; these non-philosophers have
lives that are not worth or are barely worth living. (Crito 47d–e and Republic 445a–b are sometimes
thought to suggest that a sufficient degree of ill health can make one’s life not worth living no matter
what else is true of one, for example, even if one possesses virtue. To avoid complications, I shall
assume that each type of person I discuss has an adequate degree of non-virtue goods.) Possessing the
best humanly possible knowledge (epistêmê) of the good and related value Forms, it is reasonable to
suppose, satisfies the knowledge requirement for wisdom (phronêsis). Cf. note 28 of this chapter. Phd.
86b may attribute some sort of knowledge of value Forms to Socrates; if it does, this may suffice for
wisdom. Phd. 74b attributes some knowledge of non-value Forms to Socrates and this along with, for
example. well-grounded true beliefs about the value of knowing Forms and a grasp of what this
involves based on his own knowledge of Forms, may suffice for wisdom. It is less attractive to hold that
wisdom involves no sort of knowledge. If Socrates lacks wisdom (because he lacks knowledge or the
right kind of knowledge), then he might still benefit from some lesser psychic condition that is both
good in itself and such that it allows him to benefit from other goods. (Meno 96d–98d suggests that
 Plato’s Politics   589

fundamentally differs, this would be a significant change for which we would need
an explanation.

3.5 The Republic

Kallipolis is divided into three classes: (1) philosopher-rulers, (2) auxiliaries, and
(3) producers. All political and legal authority rests with the philosophers. The auxilia-
ries protect the city from enemies within and without, and the producers perform the
economic activities needed for the city’s material life. In the two upper classes (and
only there) private property is abolished, along with the private family. Similarly,
education differs for the three classes: philosophers receive a long education in math-
ematics, culminating in knowledge of the Forms. The auxiliaries’ musical education has
­educated them

through habits [ethesi]. Its harmonies gave them a certain harmoniousness, not
knowledge [epistêmên]; its rhythms gave them a certain rhythmical quality; and its
stories whether fictional or nearer the truth, cultivated other habits akin to these
(Rep. 522a4–9).

But nothing in such an education leads people to have any grasp of the Forms that make
things fine or good.34 Such a grasp of non-sensible entities is initiated by the study of

true belief may allow one to benefit from one’s other goods, although 100a may call into question not
merely true belief ’s value in itself, but also its value in guiding other goods.) A plausible candidate for
such a condition is one resulting from recollection, since without recollecting some non-sensible
concepts, it is not clear that one has a sufficient grasp of the non-sensible to love something other than
sensible goods, such as wealth.
Finally, one could bite the bullet and hold that Socrates’ epistemic state is such that his life is not
worth, or is barely worth, living (although it might become well worth living either in the afterlife or
in another reincarnation in which he makes epistemic progress). Such an interpretation might be
supported by (1) the idea that few days of any human’s life, including Socrates’, are better than
dreamless sleep (Ap. 40c–e), and (2) the clear implication of Hipp. Maj. 304d–e that without
knowledge of the fine, it is no better to live than to die. Even if such pessimism is warranted, one
should hold that Socrates is much less badly off than non-philosophers who are unaware of their own
ignorance and value only bodily goods and that Socrates is much more likely than they are to make
progress in the afterlife.
34 Harte (2007) attributes to non-philosophers an “implicit conception” of F to explain: (1) the
sameness of reference of philosophers’ and non- philosophers’ terms, and (2) non-philosophers’
success in classifying F things. But the reference of non-philosophers’ terms may not be determined by
“what is in their heads.” We might use “F” to refer to whatever is ultimately responsible for certain
experiences. Or, drawing on the Cratylus, we might hold that names (including terms for general
properties) have informational content depending on the name’s etymology, which can be true of, and
pick out, the referent although it is not grasped by some speakers. Pace Harte, 515b7–10 makes it at least
as likely that the prisoners speak Greek as that they speak their own language. Harte objects that such
accounts make it a remarkable brute fact that non-philosophers divide up the world corresponding to
its real divisions and classify objects correctly. But non-philosophers go wrong with respect to natural
kinds (e.g., “barbarian”) and misclassify many just actions. Further, F particulars are non-accidentally
590   Christopher Bobonich

mathematics, and culminates in the study of the Forms. Only philosophers receive such
an education, and thus the auxiliaries lack any grasp of non-sensible value properties.
The producers do not seem to receive even the musical education that the auxiliaries
do.35 The philosopher-rulers’ knowledge of the Forms is the main justification for
their ruling.36
In book IV, Plato gives accounts of the four virtues—courage, justice, moderation, and
wisdom—and bases them on the parts of the soul. The soul, like the city, has three parts:
the Rational part, the Spirited part, and the Appetitive part. These parts are the ultimate,
non-derivative bearers of things such as beliefs, desires, and emotions. Their conven-
tional names may be misleading: not all beliefs are located in the Rational part, and all
parts have desires.37 Plato defines the virtues in terms of the features of these parts.
Courage, for example, consists in the power of the Spirited part to preserve, through
pleasure and pain, the Rational part’s correct orders; justice requires that all three parts
do their own job and thus requires the presence of the other three virtues (Rep.
429c–435c). Since there is no wisdom (sophia) without knowledge (epistêmê), only phi-
losophers can possess the virtues.38 (Whether or not the Republic restricts knowledge to
Forms, it is widely accepted that Plato thinks that at least knowledge of some Forms is
required to know anything at all.39) Plato does not offer a complete account of happiness.
Nevertheless, the Republic’s understanding of human nature places at its center the ability
to know the truth and the love of the truth, and both genuine virtue and genuine happiness
require the realization of these most fundamental aspects of human nature. Thus genu-
ine virtue requires the possession of knowledge, and philosophic contemplation is a major
component of human happiness.

3.6 Coercion and Benefit

The Republic holds that the city’s ultimate aim is the greatest happiness of all the citizens
(421b–c) and these include members of all three classes. So we should expect that the

like one another because they all participate in the Form of F. With respect to terms such as “ox,” such
similarity is apparent at the perceptual level. With respect to, for example, “just,” there will be overlap
between actions associated with the right psychic state and those promoting social cohesion and so on.
35 Hourani (1949) and Rep. 456d. 36 See Bobonich (2016).
37 See Bobonich (2002, 216–57) and chapter 21 by Lorenz in this volume.
38 Kahn (2004, 349–50) holds that the only individual virtue of book IV that requires knowledge
(epistêmê) is wisdom. For example, he takes 430c3 to define the courage of a citizen based on true
opinion. But it is better to see this as characterizing the character state in the citizens that makes a city
courageous. Plato is defining the city’s virtues at 427e–434c and defining an individual virtue here would
violate his methodological principle of first characterizing the city’s virtue (368c–369b). tauta (442c5)
shows that the commands courage preserves (442b10–c2) are those of a wise Rational part. Justice
consists in each part of the soul doing its own job (441d11–e1, 442d5–6, 443b1–2, 443c9–444a2) and the
Rational part only does its job when it is wise and thus possesses knowledge (441e3–5); justice thus
requires knowledge. Moderation is friendship and concord among the three parts (442c9–d2). All three
parts must be doing their own jobs in this case (441e3–442b3) and so moderation requires knowledge.
39 See Taylor’s chapter 18 in this volume.
 Plato’s Politics   591

members of all three classes are better off in the ideal city than in others. It is clear that
the philosopher-rulers benefit by possessing the virtues and engaging in contemplation.
What of the other two classes? They benefit, rather, by approximating in some way the
condition of the philosophers (and Plato gives an ordinal ranking of lives that increase
in badness with distance from the philosophic life, Rep. 580a–c).
The lower two classes’ citizens are not ruled by their own reason. Instead their Spirited
and Appetitive parts are trained by the philosopher-rulers so that such people are better
off, or at least less badly off, than non-philosophers in ordinary cities.40 The auxiliaries’
education leads them to love some subset of fine things, although they do not love them
for what actually makes them fine.41 The producers are educated and regulated so
that they reliably pursue the orderly satisfaction of their decent appetitive desires.42
Whatever coercion is involved in this education and regulation as adults, along with the
complete denial to both groups of political decision-making authority as well as the
elimination or at least vast reduction in decision-making authority over a huge range of

40 Vasiliou (2008, 212–46, 259–67) takes Rep. 590a–591c to show that non-philosophers can
possess Socrates’ human wisdom and, thanks to their upbringing, pursue virtue for its own sake. But
a producer’s Rational part is “by nature weak” (590c2) so that it cannot rule the other parts “but serves
them and is only able to learn what flatters them” (590c4–5). Because of this condition, which is a
“reproach” (590c1) to its possessor, he must for his own benefit be a “slave” (doulon, 590c8) to the
philosopher-rulers.
Obedience to their orders does not allow the producer’s Rational part—which remains weak—to
rule him. A slave is one who properly receives orders, not explanations that he cannot understand
(cf. Laws 720c–e, 777e). Since the philosopher’s Rational part cannot bring the slave’s own Rational
part to appreciate the correct ultimate end, it will rule by giving commands justifiable in terms of the
lower parts’ ends. Producers, for example, will believe (rightly) that the satisfaction of necessary
appetites is best for them. Reason rules producers and auxiliaries because it is philosophers’ Rational
parts that give commands in both cases, not because non-philosophers’ Rational parts grasp the
correct ultimate ends.
41 Bobonich (2002, 58–72) and Irwin (1977, 202–04). More optimistic are Kamtekar (1998), Vasiliou
(2008), and Wilberding (2009).
42 Vasiliou thinks (2008, 234–46) that auxiliaries and producers have an intelligent appreciation of
the book V argument (476e–480a) and of Platonic Forms, their non-identity with sensibles and their
basic metaphysical and epistemological features.
(1) Socrates never says that the lower classes receive the book V argument. They are “naturally
suited to leave philosophy alone and follow their leader” (474c1–3). The auxiliaries’ education
had “nothing in it” drawing them toward being (522a9) and both classes lack the rigorous
mathematical training that the Republic thinks necessary for learning about Forms.
(2) Socrates contemptuous dismissal of the producers’ educational attainments (405a, 456d) is hard
to understand if Vasiliou is right.
(3) Plato is not sanguine about persuading even the lover of sights and sounds (who is much more
sophisticated than the average producer) via the book V argument. His inability to see the Form
or follow arguments leading to it (476b–c) is not easily overcome, if at all, and Plato’s last word is
not that such a person has been intellectually convinced, but that he will be “shamed into
agreeing with us if nothing else” (501e6–502a2).
(4) Intelligently appreciating the Book 5 argument with an explicit de dicto recognition of Forms
should put one outside the Cave, and even Vasiliou accepts that non-philosophers remain within
the Cave (2008, 244). Plato identifies Kallipolis’s non-philosophers with the bound prisoners at
the Divided Line’s lowest level (517d–e, 520c–d).
592   Christopher Bobonich

other activities, should be justifiable as conducing to their own benefit primarily because
of this sort of improvement of their characters.
But how good or bad are non-philosophers’ lives really? They (1) fail to satisfy book
IV’s definitions of the virtues, (2) they do not grasp the Forms at all, and (3) their ends
are set, directly or indirectly, by the lower parts and do not include that which is genuinely
good in itself. The deeper explanation of these limitations rests on Plato’s epistemology
and metaphysics: since non-philosophers do not grasp the Forms at all, they do not grasp
at all the properties that really make anything fine or good. Both philosophers and non-
philosophers might, for example, value a beautiful painting as kalon. Even if we grant
that the non-philosopher values the painting in itself as kalon and does not, for example,
value it because it produces memories or anticipations of bodily pleasures, she values it
for having sensible properties, for example, bright colors (cf. Rep. 476b), and this is not
the actual finemaking property. If we accept this analysis of the case, it does not matter
whether we describe it as philosophers and non-philosophers valuing the same thing
for different reasons or their valuing different things. What is important is that non-
philosophers do not value what is genuinely good as such and have the further deep
misfortune of valuing what lacks genuine value.
These claims about the Republic’s epistemology, metaphysics, and psychology are
controversial. But even leaving aside such controversial claims that explain why Plato
is so pessimistic about the two lower classes’ virtue and happiness, one well-known
passage makes it clear that he is pessimistic:

when he [one who has left the Cave] reminds himself of his first dwelling place
and what passed for wisdom there, and of his fellow prisoners, do you not think that
he would count himself happy because of the change and pity the others? . . . would
he not feel with Homer that he would greatly prefer to “work the earth as a serf to
another, one without possessions” and go through any sufferings, rather than share
their opinions and live their life? (Rep. 516c4–d7)

This is an echo of the famous passage in Homer’s Odyssey, in which Achilles’s shade tells
Odysseus of how deeply undesirable life in Hades is (Od. 11.488–91). In the Republic, the
philosopher and the philosopher alone is ever outside the Cave. The Cave analogy is
not only an epistemological analogy, but is intended to give us a picture of human life,
and the picture for non-philosophers remains bleak.43 That the philosopher has to go
back into the Cave to rule and that her ruling activity consists in large part of making

43 Even if the Cave were merely an epistemological image, given the centrality of knowledge and
reason in Plato’s conception of happiness, it would have significant eudaimonic consequences. The
Republic does not clearly endorse the Dependency Thesis (Bobonich (2002, 185–94))—and, without it,
there is less pressure to deny that at least some auxiliaries or producers lead lives worth living. But
arguments from silence are of limited strength and it would be odd if Plato endorsed the Dependency
Thesis before (Euthyd. 278e–282a and Meno 87c–89a) and after (Laws 631b–d, 660e–661e) the Republic,
but rejected it there. In any case, the Cave passage suggests a very pessimistic evaluation of the
auxiliaries’ and producers’ lives even if they are much less badly off in Kallipolis than anywhere else.
The arguments in the text, however, do not require that their lives are not worth living.
 Plato’s Politics   593

judgments about shadows there (Rep. 520c1–5) suggest that this is where political activity
takes place and so is where the bulk of the city is.
There is no suggestion that ruling involves getting most of the others out of the Cave
(cf. Rep. 516e–517a), although, of course, those who will become the next generation of
philosophers will be helped by the rulers to ascend out of the Cave. That the bulk of the
city remains in the Cave is confirmed by Socrates’ remarks to the philosophers justifying
the requirement that they go back into the Cave:

you have been better and more completely educated than the others [ekeinôn], and
you are more capable of sharing in both ways of life. So you must go down, each in
his turn, into the common dwelling of the others [tên tôn allôn sunoikêsin] . . . you will
see ten thousand times better than the those there [tôn ekei]. (Rep. 520b7–c4)

The referent of “the others” and “those there” is the same in each instance and, as the first
occurrence shows, it must be the auxiliaries and producers: there is no hint in the pas-
sage or elsewhere in the Cave image that they ever leave their “common dwelling place.”

3.7 Stability
The Phaedo’s basic problem remains: the citizens have very different and incompatible
views about happiness, and the large majority of them lack genuine virtue. But in the
Republic, Plato makes a remarkable effort to show that despite this, a just city can exist
and remain in existence for some time. There are two especially important ways that
Plato tries to ensure the city’s stability. First, he thinks that the abolition of private prop-
erty and families will, with the proper habituation, lead people to extend outward to
others the caring attitudes they typically have for family members. Unsure of who one’s
biological brother is, one will treat all those in the appropriate age group as brothers
(Rep. 461c–464d). Whether this could be successful is controversial, and it only applies
to the first two classes, since producers have private families. Second, Plato thinks that
each class benefits from the political association and that this will tend to unite the city
(Rep. 463a–b). A significant worry here is whether the two lower classes, given their dif-
ferent conceptions of happiness, can recognize this coincidence of interests.
In both these lines of defense, Plato is helped by the Republic’s acceptance of non-
rational motivations. Love of one’s family and affection for benefactors are both emo-
tions that one can feel independently of reason’s distinct activity of determining what is
best using its own resources. Such love does not require loving them for possession of
the relevant non-sensible value property, for example, being fine (kalon) or just, and
notoriously in the case of attachment to one’s kin, city, or country it is independent of
these qualities.44 These emotions, in those habituated from birth onward, Plato might

44 Plato’s conception of perception is broad in the Republic and seems to include judgments such as
“This is a finger.” Perception should have the resources necessary for conceptualizing something as “my
city” or “my sister” without employing concepts based on one’s grasp of Forms. Cf. Bobonich (2017).
594   Christopher Bobonich

plausibly think, can lead many in the just city to have concern for their fellow citizens,
even if they have the wrong reasons for doing so. The Republic tries to show that a good
city is possible, even while holding epistemological and metaphysical views that are,
broadly speaking, similar to those of the Phaedo. The worry remains, however, that the
lower classes’ faulty conceptions of happiness and inability to truly understand why they
are benefited will ultimately prove destabilizing.

3.8 Perfection
This is one of the most controversial issues in the Republic.45 At 519c–521c, Socrates
announces that for the just city to come into being, philosophers, once they have grasped
the Form of the Good, must be “compelled” to return to the Cave and rule. There has
been much debate over whether Plato here requires philosophers to make a sacrifice of
their happiness to benefit others, or whether he thinks that ruling in these circumstances
is what actually most conduces to their happiness.
Whichever interpretation we adopt, we should note that the problem is especially
acute because the philosopher must return to the Cave. He will only be able to bring a
few up to the light (i.e., to grasp the Forms), but for the rest all he can do is make their
existence in the Cave less bad. It remains, however, difficult to see how a life spent
entirely in the Cave could be in itself a significant good for a human being, and thus one’s
own perfection, for philosophers, seems still to compete strongly with any political
activity open to them.

4. The Late Dialogues

Throughout most of the twentieth century, the ethical and political aspects of the late
dialogues have received comparatively little attention. But from around the turn of the
twenty-first century, interest and scholarship have grown and show few signs of abating.

4.1 The Statesman

The Statesman’s declared goal is to provide an account of the ideal statesman (politikos),
and it proceeds by trying to find an account of the science that he possesses. Nevertheless,
especially toward its end, the dialogue takes up basic issues such as the nature of a good
political community, the appropriate relations among citizens, and the revisability of
laws, as well as important topics in Plato’s psychology and pedagogy.
There are, the Statesman tells us, two basic kinds of people, the “courageous” and the
“moderate”: the former are quick and spirited, while the latter are slower and more
intent on leading a private life. Both of these character types, before receiving a highly

45 E.g., Brown (2000) and Sedley (2007).

 Plato’s Politics   595

specific sort of education, are prone to ethical and political errors: the courageous tend
to violence and the moderate are unwilling to assert themselves even when they should
(Stsmn. 307d–308b). The fact that such characters go wrong in these ways shows that
these are not genuine virtues. Indeed, the seriousness of these mistakes shows that the
original forms of courage and moderation do not even closely approximate genuine
virtues. Plato’s views about how such characters must be improved are bound up with
significant changes in his conception of citizenship from the Republic.
One of the most fundamental differences between the Republic and the Statesman is
that in the latter, citizenship has much greater ethical significance, and the qualifications
for it are much higher. Kallipolis, as we saw, counted as citizens the members of the two
lower classes, and none of them had just characters. The Statesman, however, gives a
criterion for citizenship in terms of virtue: only just people can be citizens in a good city
(Stsmn. 309e–310a). Those who cannot become genuinely virtuous are entirely excluded
from citizenship (Stsmn. 308e–309a). Thus those having the original character types of
“courage” and “moderation” cannot be citizens; only those whose characters have been
moderated and improved so that they come to possess genuine virtue are citizens. The
most important way of effecting such improvement is by education, specifically one that
results in their having “really true and secure opinion about what is fine, just, and good”
(Stsmn. 309c5–7). This education, unlike that in the Republic, is common to all the citi-
zens of the good city. Because of this sameness of education and similarity of finished
characters, the just city is not stratified by class as it was in Kallipolis.
Indeed, except for the single ruler who possesses knowledge, there are no classes
of citizens differentiated by their conceptions of happiness or the kind of virtue they
can attain.
In this city, all citizens have private families and property, possessions are not held
in common, at least all able-bodied male citizens perform military service, and both
the courageous and the moderate serve in political offices on roughly equal terms
(Statesman 311a). All citizens are expected to possess a high level of virtue. Given the
important place of virtue in happiness, the same conclusion holds, mutatis mutandis,
for happiness.
Thus regarding citizenship, the Statesman takes a crucial step. It redraws the city’s
boundaries so that the just political association becomes a community of the virtuous.
Political science takes as its task drawing the citizens of a good city together “by concord
and friendship into a common life” (Stsmn. 311b9–c1). It is only in this way that political
science can bring about a “happy city” (Stsmn. 311c5–6). In such a city, the citizens share
the same ultimate goal of fostering virtue in all the citizens. This is only possible because
all citizens receive the same education that aims at giving them “really true and secure
opinion about what is fine, just, and good.” We cannot determine precisely what such
opinions are or how they are inculcated, but they appear to rest much more on reasoned
explanations than the auxiliaries’ musical education in the Republic.46
We thus see in the Statesman significant changes in Plato's political theory regarding
the citizens’ education, the political activity allowed to them, and the city’s goal. If it is
right, we need some explanation of them. Are they simply freestanding changes in

46 Bobonich (2002, 412–16) and Cooper (1999).

596   Christopher Bobonich

Plato’s political philosophy, or do they rest on developments in his ethics, psychology,

and epistemology? I return to this issue in the Laws section. But we saw that Plato’s
middle-period political views depended on deeper aspects of his psychology and episte-
mology. It would be surprising if his late-period political views did not also rest on other
aspects of his philosophy.

4.2 The Laws

In addition to stylometric evidence, Aristotle tells us (Politics 2.6) that the Laws post-
dates the Republic. The Laws’ text shows signs of being unfinished; there is some ancient
evidence that Plato died while composing it (Diogenes Laertius 3.37) and, since it is the
longest of the dialogues, its composition may have overlapped with that of some of his
other late works. Let us begin with the Laws’ most quoted passage in which Plato charac-
terizes the city that he will proceed to describe.

Anyone who uses reason and experience will recognize that a second-best city is to
be constructed . . . That city and that constitution are first, and the laws are best,
where the old proverb holds as much as possible throughout the whole city: it is said
that the things of friends really are in common . . . If the constitution we have been
dealing with now came into being, it would be, in a way, the nearest to immortality
and second in point of unity . . .
First, let them divide up the land and the households, and not farm in common,
since such a thing would be too demanding for the birth, nurture, and education
that we have now specified. (Laws 739a3–740a2)

Scholars have sometimes assumed that this passage settles the question of the relation
between the Republic’s political theory and that of the Laws. Plato here endorses
Kallipolis as the best possible city, but now thinks that its demands on its inhabitants are
too high: Magnesia is the second-best, although it is the best that is compatible with
human nature.
But such an interpretation misreads the passage. Plato here does not in fact endorse
the Republic’s method for making the city one by introducing a certain kind of com-
munity of property and families. In Kallipolis, these institutions are restricted to the first
two classes, but are rejected for the third class, the producers. The Laws passage presents
as the “first-best” city, not Kallipolis, but one in which there is, throughout the entire
city, a community of property and women and children. So the claim that Magnesia is
second-best does not suggest that Kallipolis still represents Plato’s ideal political
arrangement. What the Laws represents as the ideal—which is to be approximated as
closely as possible—is a city in which all citizens are subject to the same high ethical
demands. The Laws thus early on rejects the notion that Kallipolis is the ideal.47

47 Mitchell Miller (2013, 15–16), argues that Kallipolis and the Laws’ “first- best” city are the same
because in both “the things of friends are in common” (koina ta philôn) “as much as possible”
(hoti malista). But in the Laws, the proviso that the things of friends are to be in common as much as
 Plato’s Politics   597

Something new is going on here, although it is easy to overlook or dismiss as a sign of the
carelessness of old age, especially if one approaches the Laws assuming that the more
canonical Republic represents the essence of Plato’s political thought.
The Laws also announces, with great fanfare, an innovation in the relation between
the laws and the citizens:

none of the lawgivers has ever reflected on the fact that it is possible to use two
means of giving laws, persuasion and force . . . They have used only the latter; failing
to mix compulsion with persuasion in their lawgiving, they have employed unmiti-
gated force alone. (Laws 722b5–c2)

In several earlier dialogues, Plato appealed to the analogy between the statesman and the
doctor to justify harsh treatment and coercion of citizens for their own ultimate benefit.48
The doctor should employ cuttings and burnings and so on even if the patient does not
see that the treatment is beneficial and so is unwilling to undergo it. This same analogy
of the statesman as doctor is used to make a very different point in the Laws.

athenian: What pertains to the laying down of laws has never been worked out
correctly in any way . . . What do we mean by this? We did not make a bad image,
when we compared all those living under legislation that now exists to slaves being
doctored by slaves. For one must understand this well: if one of those doctors who
practices medicine on the basis of experience without the aid of theory should ever
encounter a free doctor conversing with a free man who was sick—using arguments
that come close to philosophizing, grasping the disease from its source, and going
back up to the whole nature of bodies—he would swiftly burst out laughing and
would say nothing other than what is always said about such things by most of the
so-called doctors. For he would declare, “Idiot! You are not doctoring the sick man,

possible holds of the city’s constitution and laws in general and applies “throughout the city” (pasan tên
polin, 739b8–c3). In the Republic, the claim is that among the guardians and the auxiliaries, the things of
friends—including property and women—are to be in common as much as possible (423e7–424a3). The
degree to which the things of friends are in common in the Republic’s first two classes is the same as that
recommended to hold throughout the Laws’ first-best city, so we cannot identify them. Charles Kahn
agrees that Laws 739c1 applies communism throughout the city unlike the Republic, but identifies the
Laws’ first-best city with Kallipolis because both are based on the idea that unity is the supreme criterion
for a city’s excellence (2013, 237). But even if Plato has the same criterion for excellence in the Republic and
the Laws, this does not entail that in both works he thought the same city optimally satisfied it, and even
on Kahn’s interpretation the social structure of Kallipolis and the Laws’ first-best city are very different.
Plato does not explicitly compare Kallipolis’s and Magnesia’s happiness. Although Kallipolis’s first
two classes share property and women more fully than do Magnesia’s citizens, the large majority of
Kallipolis’s citizens, the producers, practice an art for a living and are excluded from virtue, Laws
(846d–847a). (Both the Republic and the Laws make the city’s (greatest) happiness the ultimate end of
the laws, (Rep. 419a–421c, Laws 631b, 743c)). The Laws holds that unhappy citizens diminish the city’s
happiness. This is why female citizens must also be educated and be regulated by laws aiming at virtue.
Failing to do so “is to leave the city with only about half of a happy life, instead of double that” (Laws
806c5–7). So Kallipolis will, in this respect, be very much inferior to Magnesia.
48 E.g. Gorg. 515b–522b, Rep. 389b–c, and Stsmn. 296a–297b; but the Statesman does not clearly
conflict with the Laws (Cooper (1999, 188–89)).
598   Christopher Bobonich

you are practically educating him, as if what he needed were to become a doctor,
rather than healthy!”
kleinias: Would he not be speaking correctly when he said such things?
athenian: Maybe—if at any rate, he went on to reflect that this man who goes
through the laws in the way we are doing now, is educating the citizens, but not
legislating (Laws 857c2–e5).

In the Laws, Plato proposes attaching persuasive preludes to the individual laws and
the law code as a whole. One of their main purposes is to give to as many citizens as
possible a rational understanding of the laws and the political and ethical principles
underlying them.49 The rest of the citizens’ education has the same aim, and it includes
the study of calculation and arithmetic, geometry and stereometry, with special attention
to incommensurable magnitudes including the nature of their relation to commensura-
ble magnitudes (Laws 820c4–5), and astronomy (Laws 817e–818a). Such an education
fosters in all the citizen the awareness that there are non-sensible properties, in the first
place mathematical ones.
Book X’s sophisticated cosmological and theological arguments (which are a prelude
to the impiety laws) draw on this education and are meant to be studied, repeatedly, by
all the citizens (Laws 890e–891a). They are designed to bring the entire body of citizens
to recognize that souls exist and are nonmaterial and non-sensible first causes of change
in the universe, and that the universe itself has been structured by god in a fine and
orderly way in accordance with mathematical principles. The citizens should thus come
to recognize that these non-sensible principles are themselves principles of order and
value. This mathematical education goes far beyond anything that the Republic’s auxil-
iaries received, and in the Republic, it was a mathematical education that marked the
transition to grasping non-sensible value properties. (In Magnesia, those whose educa-
tion and work resemble that of the Republic’s producers are slaves or visiting workers
without political rights.)50
The differences from Plato’s middle-period positions are clear. The Laws’ view that
non-philosophical citizens can be educated to have, to some significant extent, a reasoned
grasp of basic ethical and political truths extends the Statesman’s line of thought, and it is
a crucial difference from the Phaedo and the Republic. This difference has important
implications for Plato’s political philosophy. First, in the Statesman Plato suggested a
new conception of a good city as an association in which all citizens aim at leading virtu-
ous lives and at fostering virtue in their fellow citizens. The Laws is clearer and more
emphatic in building on this conception of a good city and, as we have seen, restructures
the citizens’ education accordingly.
Second, since the citizens are more capable of exercising good ethical and political
judgment and engaging in rational discussion, they will be able to hold office. The Laws’

49 Bobonich (2002, 97–119). For other interpretations, see Annas (2010), Laks (1990) and (2005),
Nightingale (1993), Samaras (2002, 305–30) and Schofield (2006, 83–6, 319–25). Interpretations that see
the preludes as non-rational have the high cost of making Plato’s methodological justification of them
simply seem insincere. Laks (1990) and (2005) have a sophisticated response to this worry.
50 Bobonich (2002, 378). Music and dancing play a significant role in the citizens’ education and
remain important later in life, see Folch (2015), Kamtekar (2010), Peponi (2013), and Prauscello (2014).
 Plato’s Politics   599

just city is not ruled by a single philosopher or a group of philosophers. There is an

Assembly open to all the citizens and a Council elected from the citizen body. There is
also a wide variety of other political and judicial offices open to the citizens. Such politi-
cal participation is now not only possible for the citizens, but it is also good in itself for
them as a central expression of their virtue.51
There is a second important set of differences from Plato’s middle-period views. In
the Phaedo and the Republic, Plato’s characterization of non-philosophers rests on his
epistemology and psychology. The developments and changes I have suggested that we
see in the Statesman and the Laws thus should also rest on developments in Plato’s
epistemology and psychology. Although I cannot discuss these issues here, I think that
in the later dialogues we see two important shifts in Plato’s views. First, he comes to think
that a grasp of non-sensible properties is much more deeply embedded in human thought
and is not restricted to mathematical or philosophical knowledge. Thus there is no longer
such a sharp discontinuity between the ethical cognitive capacities of philosophers and
non-philosophers. Second, Plato develops a more unified conception of the soul that
emphasizes the role of reason in shaping all of the soul’s capacities.52
So let us return for a final time to our questions.

51 The Laws’ views on changing the laws, property classes, and women complicate this picture, but
do not change its essentials. I do not think that Nocturnal Council exercises sole or dominant political
authority, nor is it composed even predominantly of fully trained philosophers. See Bobonich (2002,
374–408); for a more ambitious view of the Nocturnal Council, see Klosko (2008) to which Marquez
(2011) replies. On the goodness of political activity, see Bobonich (2002, 450–73). On the citizens’
leisure, see Bobonich (2002, 389–91) and Samaras (2012). Although some important offices have a
minimum age requirement of 50 (e.g., the Superintendent of Education), it is wrong to think that
“political activity, which takes place in town, is reserved for elderly citizens,” Brisson (2005, 101).
Participation in the Assembly is open to all who have served or are serving in the military (Laws 753b);
military service starts at 20 (785b).
The age for access to offices in general is 30 for men and 40 for women (785b). “Country-Wardens”
(agronomoi) start office between 25 and 30 (760b–c) and exercise some judicial powers in the
countryside (761e).
52 For different interpretations, see Lorenz in this volume and Scott (1995). In Bobonich (2002),
I argue that a major reason for the change in Plato’s later psychological views is a change in his
conception of perception. Lorenz, in contrast, suggests that Plato denies belief (doxa) to the Appetitive
part of the soul in the Timaeus because the Theaetetus and the Sophist develop a new and more
demanding conception of belief on which all beliefs are the upshot of rational activity (2006, 91–99).
But I do not think that this explanation succeeds. In the Timaeus (70e–71e, 77b–c), Plato does not
complain that the alleged “beliefs” of the Appetitive part are not the product of rational inquiry. Instead
he says that the Appetitive part “shares not at all in belief [doxes . . . metestin to mêden],” but rather in
“perception [aisthêseôs]” (77b5–6). The Appetitive part is restricted to having images that lack
propositional content. This conception of the Appetitive part can be fully explained by two facts. First,
the Appetitive part’s epistemic resource remains in the Timaeus, as in the Republic, perception. Second,
in the Timaeus, unlike the Republic, perception—as Lorenz agrees (2006, 96)—no longer has
propositional content. These two facts by themselves explain the changed conception of the Appetitive
part without invoking any changes in Plato’s conception of belief. If changes in Plato’s conception of
belief motivated his denial of belief to the Appetitive part, he should then still allow to it propositional
thoughts that are not the upshot of rational activity. Since the Appetitive part is a natural home for such
thoughts, his not doing so suggests that it is not change in his conception of belief that accounts for his
changed understanding of the Appetitive part. I would like to thank Alex Coley, Aditi Iyer, and
Christine Kim for their comments, and I am especially indebted to Gail Fine for her many helpful
suggestions that improved this chapter.
600   Christopher Bobonich

4.3 Coercion and Benefit

If this interpretation is correct, Plato will have made considerable progress on the issues
of coercion and benefit. There will, of course, be criminals in Magnesia who must be
coerced and citizens too stupid or absorbed in their non-rational desires and emotions
to benefit from the preludes and the system of education more generally. But Plato
accepts that many citizens will benefit from the preludes and the education and acquire
some grasp, even if not a full understanding, of the ethical principles underlying the
laws (of the sort that the Republic’s auxiliaries never attained).
Although the laws still have sanctions, Plato sharply contrasts coercion with the
sort of persuasion embodied in the preludes and the citizens’ education. Citizens
benefit precisely because they are rationally persuaded and educated, and not coerced.
Moreover, the possible benefit to citizens is much greater than before. Their grasp of
ethical and value principles enables them to lead lives that are genuinely and significantly
good for them.

4.4 Stability
In the Laws Plato is perhaps more acutely aware of human frailty than ever before. He
is especially sensitive to the influence that pleasure and pain have on all human
beings’ character and choices (662d–664c, 732d–734e) and is highly doubtful that
anyone, even one possessing full philosophical knowledge, can withstand the temp-
tations of autocratic rule (689b, 691c–d, 713c–d, 875a–d, 902a–b). Nevertheless,
Magnesia should be considerably more stable than Kallipolis, since the citizens’ edu-
cation fosters a common conception of the good. In particular, it inculcates the under-
standing of a city as a shared association in which all aim at fostering virtue in their
fellow citizens.

4.5 Perfection
For similar reasons, the conflict between one’s own perfection and political activity will
be minimized even for the highest officials. Most of Magnesia’s citizens will engage in
political activities, and sometimes exercise legal and judicial authority. Since such activ-
ities are a primary way in which their virtue is expressed, they will benefit the citizens.
And insofar as political activity can help bring about genuinely virtuous states in others,
the value to the agent of so acting seems to increase. This lack of competition between
political activity and one’s own perfection is emphasized by book X’s theology. There
human ethical and political activity is seen as a form of cooperation with god in bring-
ing good order to the universe as a whole, and god acts to guarantee that it is always the
 Plato’s Politics   601

case that what is best for the individual is best, not merely for the whole city, but for the
whole cosmos (e.g. Laws 906a–b and 903b–d).

5. Conclusion

I have argued that we can gain a deeper understanding of Plato’s political philosophy by
seeing it as, at least in part, a response to certain unresolved issues and problems that
arise in the Socratic dialogues. In particular, I have argued that Plato tries to address
these issues in the political implications of the Socratic dialogues by using the resources
developed in the epistemological, metaphysical, and psychological theories found in
dialogues such as the Phaedo and the Republic. I have also argued that some of Plato’s
views on these fundamental issues—especially those concerning coercing and bene-
fiting citizens, the nature and stability of the political association, and the tension
between one’s own perfection and political activity—change in later dialogues, such as
the Statesman and the Laws. Understanding these developments more fully requires
that we no longer see the Statesman and the Laws simply as political treatises, but that we
read them in the context provided by psychology, epistemology, ethical theory, and
metaphysics (including the metaphysics of value) of the other late dialogues.

Bibliography
Annas, J. 2010. “Virtue and Law in Plato” in C. Bobonich (2010), 71–91.
Annas, J. and Waterfield, R. (eds.). 1995. Plato Statesman. Cambridge University Press.
Balot, R. 2006. Greek Political Thought. Blackwell.
Barker, E. 1960. Greek Political Theory: Plato and His Predecessors. Methuen.
Benson, H. 2000. Socratic Wisdom. Oxford University Press.
Benson, H. 2002. “Problems with the Socratic Method” in G. A. Scott (2002), 101–13.
Benson, H. 2015. Clitophon’s Challenge. Oxford University Press.
Berges, S. 2009. Plato on Virtue and the Law. Continuum.
Bobonich, C. 2002. Plato’s Utopia Recast: His Later Ethics and Politics. Oxford University Press.
Bobonich, C. 2007. “Plato on Akrasia and Knowing Your Own Mind” in C. Bobonich and
P. Destrée (eds.), Akrasia in Greek Philosophy, 41–60, Brill.
Bobonich, C. 2011. “Socrates and Eudaimonia” in D. Morrison (ed.), The Cambridge Companion
to Socrates, 293–332, Cambridge University Press.
Bobonich, C. 2016. “What’s the Good of Knowing the Forms?” in C. Arruzza and D. Nikulin
(eds.), Philosophy and Political Power in Antiquity, 67–93, Brill.
Bobonich, C. 2017. “Agency in Plato’s Republic.” Oxford Handbooks Online. Retrieved 7 Oct.
2017, from http://www.oxfordhandbooks.com/view/10.1093/oxfordhb/9780199935314.
001.0001/oxfordhb-9780199935314-e-7.
Bobonich, C. (ed.) 2010. Plato’s Laws: A Critical Guide. Cambridge University Press.
602   Christopher Bobonich

Brickhouse, T. C. and Smith, N. D. 1999. The Philosophy of Socrates. Westview Press.

Brickhouse, T. C. and Smith, N. D. 2010. Socratic Moral Psychology. Oxford University Press.
Brickhouse, T. C. and Smith, N. D. 2015. “Socratic Moral Psychology” in J. Bussanich and
N. Smith (2015), 185–209.
Brisson, L. 2005. “Ethics and Politics in Plato’s Laws.” Oxford Studies in Ancient Philosophy 28,
93–122.
Brown, E. 2000. “Justice and Compulsion for Plato’s Philosopher-Rulers.” Ancient Philosophy
20, 1–18.
Bussanich, J. and Smith, N. (eds.) 2015. The Bloomsbury Companion to Socrates. Bloomsbury
Academic.
Burnyeat, M. 1987. “Platonism and Mathematics: A Prelude to Discussion” in A. Graeser (ed.),
Mathematics and Metaphysics in Aristotle, 213–40, Haupt.
Butler, T. 2012. “A Riveting Argument in Favor of Asceticism in the Phaedo.” History of
Philosophy Quarterly 29, 103–23.
Cooper, J. (ed.) 1997. Plato: Complete Works. Hackett.
Cooper, J. 1999. “Plato’s Statesman and Politics” in J. Cooper Reason and Emotion, 165–91,
Princeton University Press.
Enoch, D. 2014. “A Defense of Moral Deference.” Journal of Philosophy 111, 1–30.
Fine, G. 2008. “Does Socrates Claim to Know That He Knows Nothing?” Oxford Studies in
Ancient Philosophy 35, 49–88.
Fine, G. 2016. “The ‘Two-Worlds’ Theory in the Phaedo” British Journal for the History of
Philosophy 24, 557–72.
Folch, M. 2015. The City and the Stage: Performance, Genre, and Gender in Plato’s Laws. Oxford
University Press.
Harte, V. 2006. “Beware of Imitations: Image Recognition in Plato” in F-G. Herrmann (ed.),
New Essays on Plato: Language and Thought in Fourth-Century Greek Philosophy, 21–42,
The Classical Press of Wales.
Harte, V. 2007. “Language in the Cave” in D. Scott (ed.), Maieusis, 195–215, Oxford University
Press.
Hatzistavrou, A. 2005. “Socrates’ Deliberative Authoritarianism.” Oxford Studies in Ancient
Philosophy 29, 75–113.
Hourani, G. 1949. “The Education of the Third Class in Plato’s Republic.” Classical Quarterly
43, 58–60.
Irwin, T. 1977. Plato’s Moral Theory. Clarendon Press.
Irwin, T. 1995. Plato’s Ethics. Oxford University Press.
Johnson, C. 2015. “Socrates’ Political Philosophy” in J. Bussanich and N. Smith (2015), 233–56.
Kahn, C. 2004. “From Republic to Laws: A Discussion of Christopher Bobonich, Plato’s Utopia
Recast.” Oxford Studies in Ancient Philosophy 26, 337–62.
Kamtekar, R. 1998. “Imperfect Virtue.” Ancient Philosophy 18, 315–39.
Kamtekar, R. 2010. “Psychology and the Inculcation of Virtue in the Laws” in Bobonich
(2010), 127–48.
Kelsey, S. 2000. “Recollection in the Phaedo.” Proceedings of the Boston Area Colloquium in
Ancient Philosophy 16, 91–121.
Klosko, G. 2008. “Knowledge and Law in Plato’s Laws.” Political Studies 56, 456–74.
Kraut, R. 1984. Socrates and the State. Princeton University Press.
Kraut, R. 2010. “Ordinary Virtue from the Phaedo to the Laws” in C. Bobonich (2010), 51–70.
 Plato’s Politics   603

Laks, A. 1990. “Legislation and Demiurgy: On the Relationship between Plato’s Republic and
Laws.” Classical Antiquity 9, 209–29.
Laks, A. 2001. “In What Sense Is the City of the Laws a Second Best One?” in F. Lisi (ed.),
Plato’s Laws and its Historical Significance, 107–14, Academia Verlag.
Laks, A. 2005. Médiation et Coercition. Pour une Lecture des Lois de Platon. Presses
Universitaires du Septentrion.
Lane, M. 1998. Method and Politics in Plato’s Statesman. Cambridge University Press.
Marquez, X. 2011. “Knowledge and Law in Plato’s Statesman and Laws: A Response to Klosko.”
Political Studies 59, 188–203.
McPartland, K. 2015. “The Priority of Definition” in J. Bussanich and N. Smith (2015), 94–135.
Miller, M. 2013. “Reading the Laws as a Whole: Horizon, Vision, and Structure” in G. Recco and
E. Sanday, (eds.), Plato’s Laws: Force and Truth in Politics, 11–30, Indiana University Press.
Moss, J. 2007. “The Doctor and the Pastry Chef: Pleasure and Persuasion in Plato’s Gorgias.”
Ancient Philosophy 27, 229–49.
Nightingale, A. 1993. “Writing/Reading a Sacred Text: A Literary Interpretation of Plato’s
Laws.” Classical Philology 88, 279–300.
Nisbett, R. and Ross, L. 1980. Human Inference. Prentice-Hall.
Pangle, T. 1980. The Laws of Plato. Basic Books.
Pelosi, F. 2010. Plato on Music, Soul and Body. Cambridge University Press.
Peponi, A.-E. (ed.) 2013. Performance and Culture in Plato’s Laws. Cambridge University Press.
Pitkin, H. 1965. “Obligation and Consent—I.” American Political Science Review 59, 990–99.
Prauscello, L. 2014. Performing Citizenship in Plato’s Laws. Cambridge University Press.
Price, A. 2011. Virtue and Reason in Plato and Aristotle. Clarendon Press.
Rowe, C. 1995a. Plato: Statesman. Aris & Phillips.
Rowe, C. (ed.) 1995b. Reading the Statesman. Academia Verlag.
Samaras, T. 2002. Plato on Democracy. P. Lang.
Samaras, T. 2012. “Leisured Aristocrats or Warrior Farmers? Leisure in the Laws.” Classical
Philology 107, 1–20.
Santana, A. 2011. “Reasons and the Problem of the Socratic Elenchus” in G. Anagnostopoulos
(ed.), Socratic, Platonic and Aristotelian Studies, 35–52, Springer.
Saunders, T. 1991. Plato’s Penal Code. Oxford University Press.
Schofield, M. 2006. Plato’s Political Philosophy. Oxford University Press.
Schofield, M. 2017. Plato’s Laws. Oxford Bibliographies, [website], http://www.oxfordbibliogra-
phies.com/view/document/obo-9780195389661/obo-9780195389661-0214.xml#firstMatch,
(accessed 20 June 2017).
Scott, D. 1995. Recollection and Experience. Cambridge University Press.
Scott, G. A. 2002. Does Socrates Have a Method?: Rethinking the Elenchus in Plato’s Dialogues
and Beyond. The Pennsylvania University Press.
Sedley, D. 2007. “Philosophy, Forms, and the Art of Ruling” in G. Ferrari (ed.), The Cambridge
Companion to Plato’s Republic, 256–83, Cambridge University Press.
Vasiliou, I. 2008. Aiming at Virtue. Cambridge University Press.
Vasiliou, I. 2012. “From the Phaedo to the Republic: Plato’s Tripartite Soul and the Possibility
of Non-philosophical Virtue” in R. Barney, T. Brennan, and C. Brittain (eds.), Plato and the
Divided Self, 9–32, Cambridge University Press.
Vlastos, G. 1994a. “Postscript to ‘Socratic Elenchus’ ” in G. Vlastos and M. Burnyeat (eds.),
Socratic Studies, 33–37, Cambridge University Press.
604   Christopher Bobonich

Vlastos, G. 1994b. “The Socratic Elenchus: Method Is All” in G. Vlastos and M. Burnyeat
(eds.), Socratic Studies, 1–29, Cambridge University Press.
Weiss, R. 1987. “The Right Exchange: Phaedo 69a6–c3.” Ancient Philosophy 7, 57–66.
White, D. 2007. Myth, Metaphysics and Dialectic in Plato’s Statesman. Ashgate.
Wilberding, J. 2009. “Plato’s Two Forms of Second-Best Morality.” The Philosophical Review
118, 351–74.
Wolfsdorf, D. 2015. “Socratic Philosophizing” in Bussanich, Smith (2015), 34–67.
Woolf, R. 2000. “Commentary on Kelsey.” Proceedings of the Boston Area Colloquium in
Ancient Philosophy 16, 122–31.
 chapter 25

Pl ato on Education
a n d A rt

Rachana Kamtekar

Concern with education animates Plato’s works: in the Apology, Socrates describes his
life’s mission of practicing philosophy as aimed at getting the Athenians to care for virtue
(29d–e, 31b); in the Gorgias, he claims that happiness depends entirely on education and
justice (470e); in the Protagoras and Meno, he puzzles about whether virtue is teachable
or how else it might be acquired; in the Phaedrus, he explains that teaching and persuad-
ing require knowledge of the soul and its powers, which requires knowledge of what
objects the soul may act on and be acted on by, knowing which in turn requires knowl-
edge of the whole of nature (277b–c, 270d); in the Laws, the Athenian Visitor says that
education is the most important activity (803d) and that the office of Director of State
Education is the most important office of the state (765d–e). Plato’s two longest works,
the Republic and Laws, tirelessly detail a utopian educational program. And Plato’s
outlook on the arts (poetry, theater, music, painting) is dominated by considerations of
whether they help or hinder correct education.1
To bring Plato’s vast and multifaceted concern with education into focus, it will be
helpful to begin by looking through the lens of his differences with those he styles
Socrates’ educational rivals: sophists such as Protagoras, teachers of rhetoric such as
Gorgias, and ultimately poets such as Homer. Plato sees the differences between these
educators and Socrates not only as a difference over what subject matter is worth learning
but also as a difference over the nature of would-be learners’ powers to learn. By under-
standing these differences, we will gain insight into the motivation for Plato’s positive
educational proposals in the Republic and Laws, for Plato’s educational proposals go
hand in hand with his psychology: his distinctive account of human capacities to learn

1 On the educational as opposed to aesthetic value of art, see C. Janaway, Images of Excellence: Plato’s
Critique of the Arts [Images] (Oxford, 1995).
606   Rachana Kamtekar

specifies both the good human condition at which his educational proposals aim, and
the methods by which this good human condition is to be achieved.2

1. Socrates and the Rival Educators

The fifth and fourth centuries b.c.e. were a period of great intellectual and cultural
­productivity in Athens, but at the same time, elite Athenians were coming to see a need for
an education beyond traditional immersion in culture and military training. We find ample
evidence of this in the writings of Isocrates in the fourth century, but also in the
­phenomenon, well documented by Plato, of itinerant teachers in a variety of subjects,
most importantly in persuasive speaking, flocking to Athens during Socrates’ lifetime.
A number of factors explain this new interest in education beyond the traditional.
Athenian political life changed radically through the fifth century, with reforms in
democratic institutions making possible greater popular participation (for example, jury
duty and assembly attendance were now compensated for by a day’s wages), at the same
time as Athens’s imperial pursuits greatly complicated its political affairs. Would-be
political leaders now had to communicate effectively with a wider cast of people than
previously and on a wider range of affairs. As Aristotle observed, successful political
leadership now called for expertise in public speaking; expertise in military strategy, once
a prerequisite for leadership, became dispensable (Politics 1305a11–15).

1.1 The Teachers

The teachers who came to Athens to meet this new demand promised tomorrow’s
politicians the means to personal and political success. According to Plato, Protagoras
claimed to teach “sound deliberation, both in domestic matters—how best to manage
one’s household, and in public affairs—how to realize one’s maximum potential for suc-
cess in political debate and action” (Protagoras 318e–319a), and Gorgias claimed to teach
“oratory” (rhêtorikê, Gorgias 449a)—that is, “the ability to persuade by speeches judges
in a law court, councilors in a council meeting, and assemblymen in an assembly or
in any other political gathering that might take place” (452e).3 An expertise in public

2 While this chapter compares and contrasts Plato’s views on education with those of the sophists,
Plato’s views, especially the educational proposals of the Republic, might also be fruitfully compared
with those of Xenophon’s Constitution of the Spartans. In “On Plato’s Politeia,” S. Menn argues that the
Republic corrects a Laconizing emphasis on physical education in the ideal constitution by insisting on
a musical education that begins with the love of wisdom and ends with the establishment of wisdom to
guide the soul’s spirited part (in J. Cleary [ed.] Proceedings of the Boston Area Colloquium in Ancient
Philosophy 21 [Leiden, 2006], 1–55.
3 In quoting from Plato I have mostly used the translations in J. Cooper and D. S. Hutchinson (eds.),
The Complete Works of Plato (Indianapolis, Ind., 1997). Where I depart from them in translating Plato I use
the Oxford Classical Texts edited by J. Burnet (1900–1907).
 Plato on Education and Art   607

speaking would involve expertise in a great many subordinate subjects. In the Phaedrus,
Socrates includes under rhêtorikê: statements of fact and evidence of witnesses, claims
to plausibility, confirmation, refutation, and implication; the power of the likely over the
true (due to Gorgias and Tisias); indirect praise and censure (due to Evenus of Paros);
preambles and recapitulations (due to Polus); and correct speech (orthoepeia); and so on
(266d–269c). This last, attributed to Protagoras, seems to have included expertise in
literary criticism and grammar (Plato, Protagoras 339a; Aristotle, Rhetoric 1407b6–9,
Sophistical Refutations 165b20–21, Poetics 1456b8–18); Protagoras is also credited with
the production of arguments for contradictory conclusions (Diogenes Laertius, IX.52, 55,
cf. Plato, Sophist 232d), and epistemology—he is still famous for the doctrine, “the human
being is the measure of all things” (Theaetetus 151e). These expertises too might reasonably
be thought to belong to the successful persuasive speaker.4
Plato’s take on the market in higher education in the Athens of Socrates’ day5 is clear
from the beginning of the Protagoras: the merchandise is potentially dangerous, and the
eager buyers are poor judges of the value of what they are getting. When Protagoras’s
prospective student Hippocrates tells Socrates of his desire to study with Protagoras (to
receive “a gentleman’s education” rather than to become a professional sophist himself),
Socrates warns, when you go to a teacher, you hand your soul over to him. But while
when you buy food in the marketplace you can take it away and test it before eating it,
“you cannot carry teachings away in a separate container. You put down your money
and take the teaching away in your soul by having learned it, and off you go, either helped
or injured” (314b). It is dangerous to study with a sophist not just because you might be
throwing away your money, but because you might end up with a damaged soul.
How could studying with a sophist damage your soul? One might think that it is
because the sophists corrupt their students by teaching them such things as that what
goes by the name “justice” is a convention established by the weak to control the strong
(Gorgias 483b–484a) or by the strong to control the weak (Republic I. 338c–339a). Yet
when Plato discusses these supposedly corrupting views, he does not put them in
the mouths of the teachers who are the targets of his criticisms, such as Gorgias and
Protagoras. Instead, he puts the charge that the sophists corrupt the young in the mouth
of Anytus (Meno 91c–92e) and shows that, like Anytus’s charge that Socrates in particular
corrupts the young (Apology 24c–25c), it is based on ignorance and unconcern for the
truth.6 In the Republic, Socrates generalizes the point and says that those who charge the

4 Even the polymath Hippias is said to be an expert in diction, ancient history and mnemonics,
along with astronomy and geometry, (Hippias Major 285c–d; cf. Protagoras 318e–319a), and there is
evidence of his expertise in the interpretation of poetry (Protagoras 347b), as well as in the crafts of
engraving, cobbling, and weaving (Hippias Minor 368b–c).
5 On Plato’s rivalry with his own contemporary Isocrates, see Schofield, chapter 2 in this volume.*
6 My observations here build on those in T. H. Irwin, “Plato’s Objections to the Sophists”
[“Sophists”] in A. Powell (ed.), The Greek World (London, 1995), 568–90. Irwin argues that Plato does
not fault the sophists for undermining the authority of Athenian moral values, but rather for being
uncritical of these values; I argue that Plato thinks the sophists dangerous because they make it
intellectually respectable to seek nothing surer than opinion, and responds to their counsel of despair
by developing a psychology to show how knowledge is possible.
608   Rachana Kamtekar

sophists with corrupting the young are themselves the real corrupters (indeed, they are
“the greatest sophists”), when they sit together in assemblies, courts, theaters, and other
public gatherings, collectively praising some and blaming others (492a–c). As a result,

Not one of those paid private teachers, whom the people call sophists and consider
to be their rivals in craft, teaches anything other than the convictions that the majority
express when they are gathered together . . . what the sophists call wisdom [is] learning
the moods and appetites of a huge, strong beast . . . how to approach and handle it,
when it is most difficult to deal with or most gentle and what makes it so, what sounds
it utters in either condition and what sounds soothe or anger it . . . [The sophist] calls
this knack wisdom, gathers his information together as if it were a craft, and starts
to teach it. In truth, he knows nothing about which of these convictions is fine or
shameful, good or bad, just or unjust, but he applies all these names in accordance
with how the beast reacts—calling what it enjoys good and what angers it bad. He
has no other account to give of these terms. (493a–c)

So the sophists are not the source of corruption but are merely reflectors of popular
opinion, and the real source of corruption is the opinion of the crowd. Our question can
be sharpened: If all the sophists teach is popular opinion, how does studying with them
make one worse off than not studying with anyone at all?
The Republic passage just quoted faults the sophists on two counts: first, they do not
know whether the popular convictions they reflect are fine or shameful, good or bad,
just or unjust (cf. Gorgias 454e, 461b); second, they call the ability they teach—to tell
which are the convictions of the majority and presumably to use these convictions to
persuade the audience of some particular course of action—“wisdom.” The sophists do
not differ from the average Athenian insofar as they too hold unexamined opinions
about the fine, good and just, but it is the sophists who make mere opinion intellectually
respectable instead of acknowledging this shortcoming. So, for example, Protagoras
argues that what appears to be true to each subject is true for that subject (Theaetetus
152a, 160c), and Gorgias brags that rhetoric enables one to persuade an audience on any
subject whatsoever even more effectively than the expert on that subject could—and
without having to bother to learn the subject oneself (Gorgias 456b–c, 459e).7 In the
Phaedo, Socrates suggests that when unskilled people experience arguments that appear
true sometimes and false other times, especially in the study of contradiction (of which
Protagoras’s work On Conflicting Arguments and Gorgias’s On Non-Being would be star
examples8), they acquire the beliefs that they are wise and that reason is not to be trusted.

7 Although Plato does not call Gorgias a sophist, and indeed in the Gorgias has Socrates
distinguish sophistic and rhetoric—sophistic making itself out to be expertise in legislation and
rhetoric expertise in justice (465c)—nevertheless, in virtue of their likeness—making themselves out
to be parts of political expertise while in fact aiming at pleasure rather than the good, and guessing
rather than knowing (464c–d)—both sophistic and rhetoric are captured by the description of the
so-called sophists at Republic 493a–c.
8 Irwin, “Sophists,” 586, suggests that Gorgias’s On Non-Being could have been written to
demonstrate that arguments as rigorous as the Eleatics’ could prove conclusions opposite to theirs, the
lesson of which would be that persuasiveness, the product of rhetoric, should be the ultimate standard
of success in speech.
 Plato on Education and Art   609

They come to believe that “there is no soundness or reliability in any object or in any
argument” and as a result become closed off to true, reliable, and understandable argu-
ments (90b–d). This is the harm of studying with the sophists.
If a sophist only reflects popular opinion, then perhaps the cross-examination by
which Socrates exposes Protagoras’s ignorance about the virtue he claims to teach is also
an examination of traditional Athenian education in values. Protagoras himself insists
that his sophist’s profession is continuous with what the familiar and celebrated poets,
prophets, artists of various kinds, and even athletes of Greece practice (316d–317a). To
the extent that Protagoras is the mouthpiece of Athenian values, contradictions within
Protagoras’s assertions about the relationship of the virtues to one another and to virtue
as a whole (brought out by Socrates at 323b–333e and 349e–360e) also reveal the inade-
quacy of Athenian ideas about the virtue that Athens claims to teach. This hypothesis
would explain at least three puzzling moments in the Protagoras. First, when Socrates
expresses his doubts as to whether the virtue Protagoras claims to teach can in fact be
taught, he gives as reasons what the Athenians must believe about virtue’s teachability,
which he infers from Athenian practices: the Athenians cannot believe that virtue is
teachable since they allow any Athenian to advise the assembly about city management
even though he cannot point to a teacher who taught him, whereas in technical matters
the assembly listens only to established experts (319d–e); good men such as Pericles
(whom Socrates calls a gratifier of the appetites at Gorgias 517b–c) provide their sons the
best possible education and so clearly value it, but yet fail, themselves and through other
teachers, to make these sons good (320b, cf. Meno 93a–94e). But the Athenians’ beliefs
only give Socrates reason to doubt that virtue is teachable if he counts them as wise
(319b).9 Why would Socrates count the Athenians wise? Second, when Protagoras says
he would be ashamed to say that the person who acts unjustly is temperate, even though
many people say just that, Socrates proceeds to examine the view of the many on the
grounds that he is primarily interested in testing the account, and regards the testing of
Protagoras and himself as a possible by-product (333c); yet just a moment earlier, while
examining the relationship between justice and piety, Socrates has said that he is not
interested in examining accounts premised on an unendorsed assumption, for “it’s you
and me I want to put on the line” (331c). Why the reversal? Third, when Socrates intro-
duces, into the examination of whether knowledge can be overpowered by anything
else, the opinion of most people that one can know what is best and yet fail to do it
because one is overcome by pleasures (352b–353a), Protagoras asks, “Socrates, why is it
necessary for us to investigate the opinion of ordinary people?” Why indeed? All three
moments fall into place if Protagoras is Socrates’ target not qua individual to be
improved by cross-examination, but qua sophist and reflector of popular opinion.

9 Plato may be having Socrates grant Protagoras’s measure doctrine, that things are (for each
community) as they appear (to that community), for the purpose of examining his/the Athenians’
views about teachability. Protagoras himself adopts as true the beliefs he infers from Athenian
practices: virtue must be teachable because we are angry at the vicious and punish them to deter them
(323d–324a); the practice of punishment for vice requires us to think that virtue is teachable (324a–c);
given the high value of virtue, it must be that everyone tries to teach it to everyone, which would
explain why the sons of the virtuous aren’t especially virtuous (324d–327c).
610   Rachana Kamtekar

Protagoras (who has more to say about traditional Athenian education than about his
own educational program) mentions three purveyors of traditional education: living
examples (children’s parents or nurses teach them, “this is just, that is unjust,” “this is
fine, that is shameful,” etc.); the aristocratic curriculum of poetry, music, and gymnas-
tics; and the laws, which constitute patterns for behavior (Protagoras 325c–326d). This
account fleshes out the idea, mooted by Protagoras and popular within the democracy,
that every Athenian teaches virtue (328a), and suggests that education is ongoing and
pervasive, not restricted to the period and methods of formal schooling.10 It will be a
short step to the outlook of the Republic and Laws, according to which every feature of
the environment—stories, works of art, fellow citizens—is a vector in the education of
citizens. Plato’s account of “musical”11 education in these two works suggests that he
agrees with Protagoras that the actual praise and blame of parents and teachers, and the
projected praise and blame of culture heroes and the law, teach people what to value and
how to behave; he differs from Protagoras because he questions whether it is virtue that
the Athenians teach.
Plato’s positive proposals for an educational curriculum in the Republic discuss the
content of poetry at length, so we may defer discussion of poetry’s content to Section 2 of
this chapter. For the moment, however, it is worth noting that Plato criticizes the poets
on grounds quite similar to those on which he bases his criticism of the sophists and ora-
tors. First, all three are indifferent to the truth, and poetry (like rhetoric and sophistry)
aims at giving pleasure to a crowd without any regard to what in fact is good (Gorgias
501d–502e, cf. 465b–c). This judgment of poetry seems to be in some tension with
Socrates saying in the Apology (22b–c), Ion (535e–535a), and Phaedrus (245a) that works
of poetry are produced by divine inspiration. This tension would be eased if Plato meant
to criticize poets but not poetry, or if he meant to exclude great works from criticism.
Second, Socrates criticizes poets on the grounds that they cannot critically assess poetry,
for although the critical assessment of poetry is part of the expertise of a rhapsode and of
a traditional aristocratic education (Ion, Protagoras 339a), it takes knowledge or at least
dialectic to do it. For example, Socrates is able to resolve an apparent contradiction
within a poem of Simonides using the distinction between becoming and being
(Protagoras 340b–d). In the Ion, Socrates argues that for Ion to be in a position to judge
Homer’s poetry good, Ion must show that the poetry gives us the truth about what it rep-
resents, and doing this would require Ion to know the truth about what it represents
(531e–532a).12 In the Republic, Socrates relegates all the critic’s concerns with rhythm,

10 Cf. Apology 24e–25a, Meno 92d–93a. Pericles’s funeral oration describes the city of Athens as a
whole as a means for the education of Greece (Thucydides 2.41). In “Culture and Society in Plato’s
Republic,” in G. B. Peterson (ed.), The Tanner Lectures on Human Values 20 (Salt Lake City, 1997),
M. F. Burnyeat discusses Plato’s idea that educational influences include one’s “total culture.”
11 A musical education, so-called after the Muses, would include learning to sing, to play an
instrument, and to recite and interpret poetry.
12 In “Socrates on the Impossibility of Belief-Relative Sciences” (in J. Cleary (ed.), Proceedings of the
Boston Area Colloquium in Ancient Philosophy 3 [Lanham, Md., 1987], 263–325), T. Penner takes
Socrates’ claim to be that Ion cannot know what Homer is saying (i.e., what Homer means) about
medicine if he does not know the relevant truths of medicine. But this is not right: Socrates is making a
 Plato on Education and Art   611

mode, diction, and so on (Protagoras’s orthoepeia) to second place by saying that these
must follow what is said, and what is said must conform to the character who suppos-
edly says it (400d)—and presumably knowledge of good and bad character (which the
poets do not have) should determine which characters appear in a poem, and how they
speak and act.
For his own part, Socrates denies that he is any kind of teacher (Apology 33a–b).
He does not charge fees as do the sophists (19d, 33b), but there is also a deeper reason:
lacking knowledge of virtue (20c), he cannot teach others,13 not even if his own beliefs
(e.g. 28b–d, 29b) are true—which they might be as a result of luck or divinity or of exten-
sive elenctic self-examination.14 Of course, he has, or tries to have, an effect on his inter-
locutors in discussion, at the very least showing them that their beliefs are inconsistent
and as a consequence that they have intellectual work to do. This is not teaching, however,
for teaching he glosses as “producing conviction with knowledge,” which he contrasts
not only with rhetoric (“producing conviction without knowledge”) but also with inquiry,
his own practice of refutation aimed at clarifying the subject at hand and rooting out
false beliefs (Gorgias 454c–455a, 458a).
Now if elenchus produces only awareness of a conflict in one’s beliefs, and hence puz-
zlement and awareness of one’s own ignorance, its difference from both teaching and
rhetoric, which produce conviction, is clear. However, Socrates does not seem to think
this is all that elenchus can do, for rooting out false beliefs requires identifying certain
beliefs as false, and so producing (negative) convictions. But this raises a famous prob-
lem about the gap between showing that someone’s beliefs are inconsistent and helping
them eliminate false beliefs: Which of their inconsistent beliefs are they to reject? On the
other hand, couldn’t a set of false beliefs be internally consistent?15
In the Gorgias, Socrates also attributes to interlocutors beliefs they expressly disavow
(466d–e, 474b–c, 495e–516d). One possible explanation for this is that Socrates attri-
butes these beliefs to interlocutors on the grounds that they are entailed by other beliefs
the interlocutors hold, implicitly or explicitly. In this case, we might expect that he
would attribute false as well as true beliefs to his interlocutors, because surely some of

claim about what Ion needs to know in order to judge whether Homer speaks well, not about what he
needs to know in order to know what Homer is saying. It’s true that Socrates begins to question Ion by
asking about his ability “to explain better and more beautifully” Homer’s than Hesiod’s verses on the
same things (531a–b), but the content of this explanation is not “What does it mean?”; rather, it is “How
is it well said?” Socrates’ point is that just as one would need to have medical knowledge to know
whether the passages on healing wounds are well composed, one would need to have ethical knowledge
to know whether the passages on the relations between humans, and between gods and humans, are
well composed.
13 Elsewhere, Socrates says that knowledge requires having an account (Meno 97e–98a, Gorgias
465a), which would seem to enable its possessor to teach others.
14 That Socrates takes some claims to be true on the grounds that they have not been refuted in his
elenctic experience is proposed by G. Vlastos, “The Socratic Elenchus” [“Elenchus”], in G. Fine (ed.),
Plato 1: Metaphysics and Epistemology (Oxford, 1999), 36–63, at 58.
15 This problem, introduced in Vlastos, “Elenchus,” is discussed in chapter 16 by G. Matthews in this
volume.
612   Rachana Kamtekar

his interlocutors’ avowed beliefs entail some falsehoods. However, it is only true beliefs
that Socrates attributes to them—which may call for the more extravagant hypothesis
that he attributes true beliefs to interlocutors on the grounds that these truths are innate,
perhaps to be recollected. This brings us to our next topic, the students.

1.2 The Students

Gorgias’s Encomium of Helen purports to demonstrate the power of persuasion by excul-
pating the widely blamed Helen of Troy; internal to the speech too are claims about the
force or magic by which speech sways its audience (8–15): words

become bearers of pleasure and banishers of pain; for, merging with opinion in the
soul, the power of incantation beguiles it and persuades it and alters it by witchcraft.
Of witchcraft and magic twin arts have been discovered, which are errors of the soul
and deceptions of opinion (10).16

According to Gorgias, our poor epistemic condition makes us dependent on opinion,

and opinion is vulnerable to persuasion, which “when added to speech can impress the
soul as it wishes”; we can see this in the way that meteorologists, skilled speechwriters,
and philosophers influence their audiences’ opinions (13). Indeed, he says, although “the
mode of persuasion is in no way like that of necessity,” “its power is the same” (12), for
“the effect of speech upon the structure of soul is as the structure of drugs over the nature
of bodies” (14).
That Plato thinks there may be something to Gorgias’s account of the soul as epistemi-
cally deprived and so as impressionable by persuasion is suggested by Socrates’ use, in
the course of arguing for self-control over the pursuit of maximum appetitive gratifica-
tion, of a wise man’s saying that “the part of the soul in which our appetites reside is
actually the sort of thing to be open to persuasion and to shift back and forth” (Gorgias
493a). In the Republic, Plato seems to accept that Gorgias’s characterization of the soul as
a whole does accurately characterize the non-rational parts of the soul, but to insist that
the soul has additional resources. Socrates’ language for describing what the rulers must
be able to resist in order to retain their educated beliefs echoes Gorgias’s: “neither com-
pulsion nor magic spells” (mête goêteuomenoi mête biazomenoi) should lead them to
give up their convictions (412e); ‘ “the compelled’ . . . [are] those whom pain or suffering
causes to change their mind, . . . [and] ‘victims of magic’ . . . are those who change their mind
because they are under the spell of pleasure or fear” (413b–c). And Socrates describes
musical and gymnastic education’s effects on the spirited and wisdom-loving parts of
the soul (411e) using images from metallurgy (410–411b) and dyeing wool (429c–430b),
saying that pre-rational souls are “most malleable and take on any pattern one wishes to
impress” on them (Republic 377a–b, cf. Laws 664a). Note that the characterization of the

16 Translation in J. Dillon and T. L. Gergel (eds. and trans.), The Greek Sophists [Greek Sophists]
(London, 2003).
 Plato on Education and Art   613

non-rational elements in the soul as easily persuaded (malleable, able to be dyed any
color) is still highly cognitive, attributing to the non-rational elements the capacity for
belief or belief-like appearance.
Plato’s real difference with Gorgias lies in his conception of reason. Whereas Gorgias
boasts that persuasion “can impress the soul as it wishes” by means of a force akin to that
of witchcraft (13), Socrates describes education as a process in which the natural capaci-
ties of the soul—and especially of reason—are awakened and developed. Thus, in the
passage immediately following the famous cave allegory of the Republic, Socrates says,

. . . the power to learn is present in everyone’s soul and . . . the instrument with which
each learns is like an eye that cannot be turned around from darkness to light with-
out turning the whole body . . . [E]ducation is the craft concerned with . . . turning
around [the whole soul until it is able to study . . . the good] . . . . It isn’t the craft of
putting sight into the soul. Education takes for granted that sight is there but that it
isn’t turned the right way or looking where it ought to look, and it tries to redirect
it appropriately. (518c–d)

Reason’s powers are not content-neutral, as Gorgias imagines. Rather, just as sight is a
power to grasp visible contents, reason is a power to grasp intelligible ones—which
is why it needs only to be directed appropriately in order to learn. By contrast with the
“so-called virtues of the soul,” which “really aren’t there beforehand but are added later
by habit and practice,” “the virtue of reason seems to belong above all to something more
divine, which never loses its power but is always useful and beneficial or useless and
harmful, depending on the way it is turned” (Republic 518d–519a).
While in the Republic our power to learn is directed by a mathematical and dialectical
education (to be discussed in Section 2 of this chapter), in dialogues earlier and later
than the Republic, Socrates proposes that the immortal soul, having acquired knowledge
in its disincarnate state, can recollect this knowledge when incarnated, for example,
when we are asked the right sorts of questions (Meno 81b–86b), or when we judge that
sensibles are deficient in the possession of some property (Phaedo 72e–76d), or when an
experience of beauty reminds us of the Form of Beauty (Phaedrus 246d–250e).17 What
all these accounts of the power to learn have in common is that they supplement the

17 The doctrine of recollection raises many questions beyond the scope of this chapter: What is the
range of things about which we have innate knowledge, allowing us to recollect (the Meno speaks of “all
truths” and does not mention Forms)? What (if any) role does this innate knowledge play in ordinary
cognition? Does anything enable us to tell which of our opinions are due to innate knowledge and if so,
what? Just what is the relationship between recollection as a result of repeated questioning and the
account of the reason that turns true opinions into knowledge (Meno 97e–98a)? For further discussion
of recollection, see Taylor, chapter 18 of this volume. Outside this volume, for an overview see C. Kahn,
“Plato on Recollection” in H. Benson (ed.), The Blackwell Companion to Plato (Oxford, 2006), and for
an account that explains recollection’s role in ordinary thought and talk as well as philosophical reflection,
see L. Franklin, “Recollection and Philosophical Reflection in Plato’s Phaedo,” Phronesis 50 (2005),
289–314. Plato’s views are considered in detail and compared to other ancient epistemological positions
in D. Scott, Recollection and Experience: Plato’s Theory of Learning and Its Successors (Cambridge, 1995).
On the Meno in particular see G. Fine, “Inquiry in the Meno,” in Plato on Knowledge and Forms
(Oxford, 2003), 44–65, and, more recently, The Possibility of Inquiry (Oxford 2014), chapters 2–5.
614   Rachana Kamtekar

soul’s impressionability before external forces, identified by Gorgias, with something

that affords the soul access to the truth. And hand in hand with this addition is Plato’s
thought that education should not be a matter of pouring doctrines into a receptive
soul, but rather a matter of providing some stimulus to a potentially active soul’s
inquiry, whether that stimulus takes the form of a questioning Socrates, inadequate or
conflicting sensory reports (Republic 523c–524d, discussed in Section 2 of this chapter),
or something else.
Plato’s difference with Protagoras is subtler. According to Protagoras, Zeus gave all
humans justice and a sense of shame (Protagoras 322c–d); however, those who live in ­cities
are much more virtuous than those who do not (327c–d), because cities educate their
citizens by the processes discussed above (Section 1.1 of this chapter). Now Protagoras
on the one hand conceives of human beings as having a capacity for virtue rather than as
being purely blank and impressionable, but on the other hand characterizes this capac-
ity as receptivity to all kinds of content: we learn what is fine and base, good and bad,
from living and literary examples and from the example of the laws. But then how do we
determine whether a new case is similar to or different from the cases we have had
identified for us? Do we generalize? Do we somehow pick up underlying principles? While
Protagoras is silent about these questions, Plato attempts to answer them by describing
the powers, objects, and activity of reason.

2. Ideal Education

While Plato’s Republic is a defense of justice against Glaucon’s challenge to show how it is
in our interest to be (and not only to appear) just, it is also a work on education, and
we see this if we follow Socrates’ interlocutor Adeimantus through the dialogue.
Adeimantus remarks about Glaucon’s challenge, “The most important thing to say hasn’t
been said yet” (362d), and goes on to explain: the things conventionally said in praise of
justice in fact undermine its claim to intrinsic value because they praise the good conse-
quences of appearing to be just, such as high reputation and all that derives from this,
and favor from the gods (362e–363e); further, poets and ordinary people alike say that
justice is hard and injustice sweet, and they willingly honor unjust people and declare
them happy; finally, they say that the gods can be bribed so as not to punish unjust deeds
(363e–364c). Adeimantus raises a general concern about the effects of culture and
education (paideia) on values, and—by contrast with Glaucon’s immoralist challenge,
which may be written off as purely theoretical18—the effects of culture and education on

18 For example, B. Williams comments on Glaucon’s claim that someone in possession of Gyges’s
ring would act unjustly, “with regard to many people, it is not very plausible . . . [for] it is likely that, if
an ethical system is to work at all, that the motivations of justice will be sufficiently internalized not to
evaporate instantaneously if the agent discovers invisibility. Moreover, it is not clear in any case how
such a thought-experiment tells one about justice in real life” (“Plato against the Immoralist,” in O. Höffe
(ed.), Platon, Politeia (Berlin, 1997), 59). One might reply that such a thought experiment tells us that
whatever our psychological propensities, it is irrational not to commit injustice when doing so would
 Plato on Education and Art   615

values are real effects. The common-sense sayings of the actual culture of Socrates and
his interlocutors are, Adeimantus suggests, a breeding ground for a casual attitude toward
the concerns of justice, an attitude that can easily slip into immoralism. While Socrates’
description of the ideal constitution aimed at the happiness not of the ruling class but
the city as a whole is likely a response to Thrasymachus’s observation that all existing
constitutions serve the interests of their rulers (338d–e), his long treatment of the educa-
tion that produces just citizens (376c–415d) and just individuals (including the perfectly
just philosopher-rulers, 514a–540c) is a response to Adeimantus’s concern; indeed, it is
at Adeimantus’s urging that Socrates begins to describe the education of the guardians
in the first place (376c–d).
Before turning to the details of Plato’s educational proposals, the reader may find it
helpful to have a synoptic view of the whole educational program of the Republic. To that
end, see table 25.1
Musical and physical education, which are designed for the whole guardian class
(future philosopher-rulers and their helpers, the military and police force), aim to
produce habituated political virtue in the soul (430a–c, 522a).19 I will say nothing about

Table 25.1 Educational curriculum in the Republic

Stage Education

First, when the soul is most malleable (377b) Poetry and music: false stories containing something
of the truth (377a and ff.); heard, perhaps enacted
(395c)
After poetry and music (403c), for 2–3 Physical training (403c)
years (537b)
Starting in childhood, in play (536d–e), but Mathematics: arithmetic, plane and solid geometry,
not during physical training (537b) astronomy, harmonics (525a–531d)
From age 20 on Synthesis of earlier studies, culminating in a unified
vision of their kinship with one another and with the
nature of what is (537c)
Ages 30–35, after the mathematical “prelude” Dialectic (532a–539e)
Ages 35–50 Practical experience (539e–540a)
50 Grasp of the good itself (540a)

go undetected and be to one’s advantage (such an argument is laid out in T. H. Irwin, “Republic 2:
Questions about Justice,” in G. Fine (ed.), Plato 2: Ethics, Politics, Religion, and the Soul (Oxford, 1999),
170–75). Still it may be that acceptance of Glaucon’s argument couldn’t actually undermine the motive
to justice because of the way in which a motive is internalized. Adeimantus, by contrast with Glaucon,
focuses on actually held beliefs that actually undermine the motive to justice.
19 In R. Kamtekar, “Imperfect Virtue” (Ancient Philosophy 18 [1998], 315–339), I argue that such
education-inculcated political virtue, although not based on knowledge as is philosophical virtue, is
nevertheless genuine in the sense that its possessors value virtue for its own sake rather than for the
sake of nonmoral consequences, for as a result of their education, they have an internalized standard of
conduct that they try to live up to even in the absence of rewards and punishments.
616   Rachana Kamtekar

physical education save that its object is also to train the soul rather than (as one might
have expected) the body (410b); in Section 3 of this chapter I say something about physical
education in Plato’s late dialogues. The higher education in mathematics and dialectic is
designed for future philosopher-rulers. Mathematical education turns the soul around
to the realm of the things that are (521d–525b). Dialectic results in knowledge of the
Forms, including the Forms of the virtues, culminating in a grasp of the good itself.
As we noted in Section 1.1 of this chapter, in the Protagoras, Socrates distinguishes
between education as mere cultural reproduction and education as cultivation of genuine
virtue by requiring knowledge for the latter. In the Republic, at the end of his account of
musical education, Socrates once again cautions,

[N]either we nor the guardians we are raising will be educated in music and
poetry until we know the different forms of moderation, courage, frankness, high-­
mindedness, and all their kindred, and their opposites too, . . . and see them in the
things in which they are, both themselves and their images. (402c)

Consequently, when Socrates acknowledges that what he has to say about the soul
and the virtues is only adequate to the standards of their present discussion (504a–b;
cf. 435d), we might reasonably conclude that his account of education, too, is subject to
revision on the basis of knowledge of the Forms of the virtues (I take it these would be,
e.g., justice itself, not merely justice in a city or soul). Since Socrates does not give (and
may well not have) accounts of these Forms, it is worth approaching his educational
proposals with the question: What, in the absence of knowledge of the Forms of the
virtues, guides these proposals?

2.1 Musical Education

Education begins, Socrates says in the Republic, with stories told to young, impressionable
children. The education is designed for guardian-children, and it is unclear whether
he imagines that the producing class will receive any part of this education, but it would
seem difficult and pointless to exclude them. To know their place in society, they would
need to hear the Noble Lie (414b–415d), and some of them might turn out to be
guardian material (415a–c). In any case, if the goal of the city is, as Socrates says repeat-
edly, to make the city as a whole as happy as possible, and if education is the route to
happiness, it would make no sense to deprive them of any education they are capable of
benefiting from.20
Socrates says that the stories told to the young are “false on the whole, though they
have some truth in them” (377a). He explains this “some truth” by adding that children

20 I argue that the Republic’s principle for the distribution of socially produced benefits such as
education is “from each according to their ability to each according to their capacity to benefit” in
R. Kamtekar, “Social Justice and Happiness in the Republic: Plato’s Two Principles,” History of Political
Thought 22 (2001), 189–220.
 Plato on Education and Art   617

should not hear stories that would cause them to take into their souls beliefs that are
“opposite to the ones they should hold when they are grown up” (377b). Because he
proceeds to censor the verses of Homer, Hesiod, and others on the basis of the vicious
behavior these verses attribute to the gods and hence license for humans, and because he
says that he lacks knowledge of the truth himself, it can seem that musical education is
only concerned with mind- and behavior-control and not at all with truth.
In fact, though, truth is the foremost concern in musical education. The first grounds
for rejecting stories is if they “give . . . a bad image of what the gods and heroes are like, the
way a painter does whose picture is not at all like the things he’s trying to paint” (377d–e).
So a bad image is not merely one that leads to undesirable consequences such as antisocial
behavior; it is also inaccurate.
But, one might wonder, if the stories as a whole are false, as Socrates says at the outset,
and if we are all ignorant of ancient events involving the gods (382c–d), what is the
objectionable inaccuracy that justifies throwing out some, but not all, the verses?
Socrates says, somewhat vaguely, that all gods and humans hate “true falsehoods”—that
is, falsehoods “about the most important things” and “the things that are” (382a–b). This
suggests that what the stories must be accurate about is how we should live, so about
happiness and what does and doesn’t conduce to it. Stories that convey moral untruths
are objectionably inaccurate and must be thrown out.
For example, Socrates’ criteria for judging stories about the gods are (1) what is pious
(which presumably requires accuracy about the gods), (2) what is advantageous to us
(which may, since the god is good, be a proxy for truth), and (3) what is consistent (380b–c).
Socrates seems to assume that these criteria, which could in principle conflict, harmo-
nize. His “patterns for stories about the gods” (379a) follow from the hypothesis of god’s
goodness: first, since nothing good is harmful, and nothing that is not harmful can do
harm, and nothing that does no harm can do or be the cause of anything bad, it follows
that a god cannot be the cause of bad things; on the other hand, since good things are
beneficial or the cause of doing well, a god is the cause of good things (379b–c). Second,
since the best things are most resistant to change, the gods would not be willing to
change, for by that they would make themselves worse (380e–381c). Finally, being per-
fectly good, the gods have no need of falsehoods, no need to change or deceive anyone;
instead, they hate falsehoods (382d–383a). Thus admissible stories about the gods are
constrained by what argument can show to be consistent with a god’s nature.
In this discussion, Plato focuses not on the truth of statements but rather on the truth
or falsehood of beliefs that people form on the basis of statements, for it is beliefs that
influence actions and form characters, virtuous or vicious. For this reason, Socrates
rejects the point that allegorical interpretations render some stories true: “the young
can’t distinguish what is allegorical from what isn’t, and the opinions they absorb at that
age are hard to erase and apt to become unalterable” (378d). The thought seems to be: the
allegorical reading of a poem requires an independent grasp of the truths that are to be
sought and found in the poem, but these poems are people’s first teachers.
It is not entirely clear whether Plato envisions one or two mechanisms by which
opinions about how to live are impressed in the soul. First, on the assumption that the
618   Rachana Kamtekar

human being to be educated desires his own happiness, and looks to gods and heroes as
models of happy living, the person who hears stories about Zeus reasons, “Zeus is good
and lives the most blessed life, so it isn’t contrary to happiness to be led by one’s lusts or
to harm one’s father, and so it cannot hinder my happiness to do these things—I am only
doing as Zeus does!” (cf. 391e). Or he might reason, “Satisfying one’s lusts and acting
on one’s anger are among the things one should do to live well—look at Zeus!” Although
the focus of the discussion is on eliminating passages of the poets that might lead listeners
astray, Socrates sometimes retains verses containing positive models for behavior. For
example, to model the virtue of self-control, he includes the passage about Odysseus’s
“exhibiting endurance in the face of everything” when he controls his angry impulse to
slaughter his maids as they carouse with Penelope’s suitors (390d).
Imitation may be a separate, second, mechanism by which people acquire opinions.
As a prelude to his prohibition on guardians taking on the roles of vicious or weak char-
acters, and ultimately on all craftsmen representing vicious characters (401b), Socrates
says that “imitations practiced from youth on become part of nature and settled into
habits of gesture, voice, and thought” (395c). He is talking about the effects of playing
the part of a character in a theatrical performance or perhaps simply in reading aloud.
Having just distinguished two kinds of narration, narration in the voice of the poet, and
imitation (mimêsis) or narration in the voice of a character (392d–394b), Socrates defines
imitation as “making oneself like someone else in voice or appearance” (394b–c). It’s not
entirely clear whether Plato thinks that we have a basic propensity to imitate (as we now
think babies do when they mimic facial expressions reflexively) which is not (at least
initially) hooked up to our happiness-seeking behavior, or whether he means that we
imitate those we regard as happy, in order to be happy ourselves—in which case imita-
tion is just an instance of the first mechanism for the adoption of opinions about how
to live. Socrates says we can’t help but imitate the things we associate with and admire
(500c), suggesting that admiration is a sufficient condition of imitation, and presumably
we admire those whom we think happy. But is admiration necessary? And is it only the
happy that we admire?
Plato deepens his claim that poetry forms our opinions because our souls are so
impressionable by his account of the divided soul. Whereas the rational part of the soul
“puts its trust in measurement and calculation” (602e–603a), obeys the law, and com-
mands us to deal with our misfortunes by fixing them as best we can rather than griev-
ing over them (604b–c), the non-rational part(s)21 persist(s) in believing appearances
despite the testimony of measurement. In other words human beings have a permanently
truth-indifferent belief-forming mechanism, and it is this that is vulnerable to the values

21 Whereas earlier in the Republic Socrates distinguishes the soul into three parts, appetitive,
spirited, and reasoning (436b–441c, 580c–581b), in this discussion he only distinguishes the rational
from the non-rational, often naming the latter by the very attitude he attributes to it in the discussion
(e.g., the lamenting thing (606b1), the pitying thing (606b7–8), the thing that judges contrary to
measurement and reasoning (603a7)). I suggest reasons for these different ways of dividing up the soul
and characterizing soul-parts in my “Speaking with the Same Voice as Reason: Personification in Plato’s
Psychology” in Oxford Studies in Ancient Philosophy, 31 (2006), 167–202.
 Plato on Education and Art   619

represented in poetry. This mechanism is responsive to the content of appearances but

not at all to whether this content accurately represents what it represents. Socrates
also says that “by nature” the non-rational part desires “the satisfaction of weeping and
wailing” (606a).22
The indifference of the non-rational part of the soul to truth explains why in Republic
II–III Socrates had so insisted on the careful screening of stories told to the young,
before the age of critical reasoning; here in Republic X he adds that even decent adults
are liable to be corrupted by contrary-to-reason values represented in poetry. When
such a person watches a tragic performance—for example, one in which a supposedly
great man grieves excessively for the loss of his son—the spectator’s reason relaxes
control over his inferior part, thinking that there is no shame in praising and pitying
the man, and that besides, watching him is pleasant. The effect of this, however, is to make
it difficult to control the desire to grieve when he himself suffers in real life (606a–b).
Critics of Plato might retort that our ability to distinguish reality from appearance,
and fact from fiction, blocks any such effect, but Plato’s claim is that the pity we feel for a
fictional character is real pity. True, we do not try to console the character, but our attitude
is not, “no shame in praising and pitying him because it’s only a play”; rather, it is, “no
shame, but rather, pleasure, in praising and pitying him.” And in doing so we segregate
our praise and pity from reason’s judgment that a good person does not lament exces-
sively—so that to praise this man is to counter that a good man may indeed lament so
much, and to pity him is to counter that his loss is indeed great enough to warrant such
lamentation. Socrates describes this as “watering and nurturing” our non-rational desires
and establishing them as “rulers in us” (606d), presumably because what we are doing is
to habituate ourselves to respond without heed to reason’s judgments.23
Thus the ban on poetry in Republic X is a ban specifically on imitative (595a) poetry,
that is, poetry that imitates appearances rather than reality and so is “third from the
truth” (597e; hymns to the gods and eulogies to good people are explicitly allowed, 607a,
cf. the poetry that is a gift of inspiration from the gods, glorifying achievements of the
past and teaching them to future generations, Phaedrus 244a–245a).24 And although the

22 It’s not entirely clear why the inferior part desires to grieve: Is it because we enjoy strong feelings?
T. Gould suggests there is a particular pleasure in the spectacle of the innocent victim and explores this
in The Ancient Quarrel between Poetry and Philosophy (Princeton, N.J., 1990). The characterization of
the inferior part by its desires also raises a more global question: What do the uncritical and the
emotional characteristics of the inferior part have to do with each other? In “Platonic Pessimism and
Moral Education,” D. Scott argues that the Republic identifies two sources of intransigent false beliefs
that account for Socrates’ failures to educate his interlocutors: the beliefs were imprinted on the very
young soul and thus became indelible, or the beliefs were caused or sustained by bad appetitive desires
(in Oxford Studies in Ancient Philosophy 7 [1999], 15–36).
23 J. Harold, “Infected by Evil” (Philosophical Explorations vol. 8 [2005], 173–87), uses a descendant
of Plato’s division of the soul, viz., of “controlled” versus “automatic” psychological processes, to suggest
that an effect of identifying with bad fictional characters is the training of automatic processes which
often determine our actions.
24 For more on this way of seeing the connection between the metaphysical and ethical criticisms of
poetry, see J. Moss, “What Is Imitative Poetry and Why Is It so Bad?” in G. R. F. Ferrari (ed.), The
Cambridge Companion to Plato’s Republic (Cambridge, 2007).
620   Rachana Kamtekar

focus in Republic II–III was on impersonation—imitating by making oneself like a

character—and in Republic X is on representation—imitating by making a product in
the likeness of a mere appearance—the former is just a special case of the latter, and the
danger in both is the same: that the uncritical belief-forming mechanism in us will form
contrary-to-reason beliefs that guide our actions and impede our moral progress. This
explains why, although Plato explains what imitation is by the example of painting,
which as a representation of how things appear to vision is an inferior thing, and although
Plato says that a painter who knows nothing of carpentry can create a visual representa-
tion of a carpenter that would take in fools and children (598b–c), he shows no interest
in banning painting.25 For it is not painters, but poets, who are taken to “know all crafts,
all human affairs concerned with virtue and vice, and all about the gods” (598d–e; but
that they do not know these things is proved by their not making, for example, real pots
or shields, or real virtue, 599a–600c, cf. Phaedrus 276c–277a).26
It may seem that Plato should acknowledge that even if the poets have no knowledge
of the truth about virtue and how to live, and can only reflect back to their audience the
uninformed opinions that circulate in society (602a–b), nevertheless, poets can also
raise critical questions about these opinions. The problem, though, is that poetry provides
no resources to answer these questions other than opinion all over again. What help can
poetry give to a reader of Sophocles’s Antigone who wonders, “What is justice, after all?
My former answers, obedience to the law or the king, and upholding the traditional
customs, conflict!” Plato offers a way ahead: dialectic.27

2.2 Mathematical and Dialectical Education

While the goal of the musical education is for citizens to acquire true beliefs about how
to live well (including such matters as which actions are courageous, and that certain
institutions are just), if our souls have the capacity to know what courage and justice are,
then it seems that we could benefit from an education beyond the musical. But the
Republic discussion of dialectic describes the corrosive effect of subjecting our values to
critical questioning: even a person who is committed to the convictions with which he
has been brought up will, when asked “what is . . . ?” questions about, for example, justice,

25 E. C. Keuls, Plato and Greek Painting (Leiden, 1978); A. Nehamas, “Plato on Imitation and Poetry
in Republic X” in J. Moravcsik and P. Temko (eds.), Plato on Beauty, Wisdom and the Arts (Totowa, N.J.,
1982), 47–78.
26 In “Plato on the Triviality of Literature,” J. Annas argues that the analogy between poetry and
painting in Republic X supports a criticism of poetry as unimportant, which is in some tension with the
criticism of poetry as dangerous (in J. Moravcsik and P. Temko (eds.), Plato on Beauty, Wisdom and the
Arts (Totowa, N.J., 1982), 1–28). The two criticisms can be reconciled as follows: like painting, poetry
(even at its best) is of little value because it can represent the truth only dimly and so it is of far less
value than legislation or teaching in making people virtuous; however, when it misrepresents the truth,
poetry is dangerous because the uncritical part of the soul accepts its misrepresentations.
27 For more on Plato and poetry, see the essays in P. Destrée and F.-G. Hermann (eds.), Plato and the
Poets (Leiden, 2011). For discussion of Platonism in the arts, see S. Halliwell, Republic Book 10
(Warminster, 1988).
 Plato on Education and Art   621

provide answers that identify the just thing in some cases but the unjust in others.28 As a
result of being repeatedly refuted, this person will start to believe that “the fine is no more
fine than the shameful, and the same with the just, the good, and the things he honored
most”; this will loosen his commitments to his former convictions and to philosophy
(538c–539c).
But avoiding “what is . . . ?” questions is not an option. If it is a natural power of the
soul to ask such questions, then an education that thwarts or even ignores it will not
be a good education. Further, Socrates says that there are certain sense experiences
that “summon the understanding”—make us ask, in other words, about the qualities we
ascribe to things, what they are. When the same object appears great or small depending
on what it is seen next to, the soul is puzzled: How can the same thing be great and
small, given that these are opposite qualities? The soul then summons the understanding
to inquire, “Are the great and the small distinct? What is the great? What is the small?”
(523b–524d).29
Socrates says that numbers are summoners: from the point of view of sense experience,
each thing is both one and many, but, since one and many are opposite qualities, the soul
is puzzled (525a–b). So arithmetic and calculation reliably summon the understanding,
turning our attention away from counting sensibles toward number itself, which can be
grasped only in thought—in other words, turning the soul soul away from becoming to
being. Next, plane and solid geometry, astronomy that seeks out the true motions
approximated by the observable heavenly bodies (529c–530c), and harmonics pursued
by way of “problems” (531b–c) all purify and rekindle the soul’s most valuable instru-
ment (527d–e), making it easier to see the form of the good (526e). Finally, the different
mathematical studies must be integrated and consolidated to “bring out their associa-
tion and relationship with one another” (531c–d). These mathematical studies are pre-
paratory for dialectic, but they are also intrinsically good for the soul.30

28 A. Nehamas argues, “Confusing Universals and Particulars in Plato’s Early Dialogues,” Virtues of
Authenticity: Essays on Plato and Socrates [Princeton, N.J., 1999] 159–75), that the problem is not that they
offer particular instances of justice when they are asked to provide universals, but that they offer the
wrong universals (presumably basing their answers on their limited experience of particular instances).
29 For an excellent discussion of how conflicting appearances can both corrupt the soul and
summon its understanding, see M. Heckel, “Plato on the Role of Contradiction in Education,” British
Journal for the History of Philosophy 25 (2017), 3–21. Notice that the puzzlement that summons the
understanding contrasts with the response of the non-rational part to conflicting appearances: it
“cannot distinguish the greater and the smaller but believes that the same things are great at one time
and small at another” (605b–c, correction of Grube-Reeve translation, which leaves out the
comparatives for meizô and ellatô; I have substituted “great” for “large” because the point applies to
moral as well as physical magnitudes).
30 M. F. Burnyeat, in “Plato on Why Mathematics Is Good for the Soul,” in T. Smiley (ed.),
Mathematics and Necessity: Essays in the History of Philosophy (Oxford, 2000), 1–81, argues that
mathematics in the Republic does not have the purely instrumental value of sharpening the mind
(as might be suggested by talk of its purifying and rekindling, and as Isocrates and others thought) but
is constitutive of ethical understanding since it provides a low-level articulation of objective value.
Concord, for example, can be understood both mathematically (in harmonics) and ethically. See also
Scott, chapter 9, and Taylor, chapter 18, in this volume.
622   Rachana Kamtekar

Properly practiced, dialectic uses reason alone to find the being of each thing and
continues until the student arrives at an understanding of the good itself (532a–b), which
enables her to give an irrefutable account of these (534b); it produces an understanding
of how all the subjects formerly studied fit together into a unified whole (537c); finally,
unlike the mathematical disciplines, it achieves unhypothetical knowledge (533c–d).
Socrates describes the good grasped at the culmination of dialectic as the cause of both
our power to know and the truth of what is knowable (508d–e). Interpreters differ
over whether the grasp of the good is a kind of direct acquaintance with a self-evident
first principle from which all the Forms may be derived, or a synoptic understanding of
a coherent teleological structure of which the Forms are parts.31 In either case, prior to
grasping the good, which is the condition of unhypothetical knowledge, students have
their dialectical studies interrupted by a 15-year practical experience requirement the
point of which is to ensure that future rulers are at least the equal of their fellow-citizens
in experience and that they remain steadfast in their values (539e–540a). Unlike the
other studies, the practical experience requirement may contribute only to competence
in political rule and not to knowledge of the form of the good.
At the level of the individual soul, knowledge of the good is a perfection of reason
and, Socrates says, in addition to providing reason its characteristic pleasure of knowing
the truth, it allows the non-rational parts of the soul “the truest pleasures possible for
them” (586d–e). These pleasures, Socrates says, are most their own [oikeias/oikeiotaton,
586e2], if indeed what is best for something is most its own. This remark flags a problem,
for if it is not the nature of the non-rational parts to seek the truth, in what sense are
the pleasures that follow the truth their own? We will see in Section 3 of this chapter how
the Laws addresses this problem. But even the truest pleasures of the non-rational parts
are inferior in truth (they fill us up with “what is never the same, and mortal”) and purity
(they do not arise without preceding and succeeding pains, and they are dependent for
their pleasantness on the preceding pains) (585b–d). Their presence in a virtuous life is
more a matter of making the best of bodily and psychological necessity than of realizing
any perfectible powers of the non-rational parts.

3. Educational Innovations
in the Laws?

In the Politics, Aristotle says that the Republic and Laws set out the same program of
education (II.6 1265a6–8)—a surprising claim, given the prominence in the Laws of
striking educational institutions absent from the Republic, such as drinking parties to
test and reinforce old men’s self-control and modesty (645d–649d, 665c–666c, 671b–e),

31 For discussion see G. Fine, “Knowledge and Belief in Republic V–VII” in Plato on Knowledge and
Forms (Oxford, 2003), 109–16, and Taylor, chapter 18 in this volume.
 Plato on Education and Art   623

the use of the Laws itself a teaching text (811c–e), and persuasive preambles to the laws
(719e–723c).32 Nevertheless the Athenian’s initial description of education echoes the
Republic conception of musical education: virtue being a harmony between reason and
passion, education channels a young child’s pleasures and pains toward the right things
before he can understand the reason, so that when he later comes to have understanding,
his reason and his passions agree (Laws 653b, cf. Republic 401c–402a). Aristotle may
think that the fact that in both texts Plato maintains that early education should prepare
our pleasures and pains to agree with our reasoning dwarfs the differences in particular
methods recommended for bringing about this condition.
Since we have found that Plato’s educational proposals in earlier dialogues are closely
informed by his conception of the soul that is to be educated, it is worth asking whether
the new educational institutions in the Laws reflect a change in Plato’s conception of the
soul’s powers to learn—or indeed whether familiar educational institutions are described
in new ways. This is of course a large topic, so what follows is a sketch.
A striking feature in this account of education is its emphasis on certain kinds of
movement, certainly in the first six years, but evidently continuing throughout life in the
form of choral dancing. The Athenian says that the souls (and bodies) of the young are
always in motion (653d, 664e, 787d), are internally agitated (791a), and are wild due to

Table 25.2 Educational curriculum in the Laws

Stage Education

Prenatal to 3 years Movement (788d–793e)

3 to 6 Play, wrestling, dancing (814e–816e), music
? to until they are old enough Choral singing and dancing (in the Chorus of the Muses)
for the chorus of Apollo?
from the time they are “tiny Arithmetic, to be followed by geometry and astronomy (819b–822c)
tots” at play to?
? to ongoing Fighting, including with weapons (813d–e)
10 to 13 Reading and writing (809e–810b), the text of the Laws (811c–e)
13 to 16 Lyre-playing (809e–810a), in unison with singing (812d–e)
18/20 to 30 Choral singing and dancing (in the Chorus of Apollo) (664c)
30 to 60 Choral singing and dancing (in the Chorus of Dionysus) (653d,
665a, 666b, 67b, 812b–c)
? “Advanced education” for guardians (965b–968c): a grasp of the
one over many and a rational account for the virtues, the good and
the beautiful; theology, the priority of soul and the reason-directed
motions of the heavenly bodies that show the existence and nature
of god—details of the curriculum to be decided by the experts

32 C. Bobonich, “Persuasion, Compulsion, and Freedom in Plato’s Laws,” in G. Fine (ed.), Plato 2:
Ethics, Politics, Religion and the Soul (Oxford, 1999), 373–403, sees the last two educational innovations
as of a piece and argues that the preambles persuade rationally, as befits a free citizen.
624   Rachana Kamtekar

the presence in them of unchanneled reason (808d–e); all this unruly internal movement
calls for external movement (e.g., dancing, and before children are mobile, being carried
about) to calm down the young soul (790c–d) and make it orderly. The soul itself, being
a self-mover, is always in motion (Laws 895a–896a; cf. Phaedrus 245c–e), and according
to the Timaeus, when it is embodied, experiences appetition and sense perception, which
distort its rational motions (Timaeus 43a–c).
Musical education in the Laws is, like musical education in the Republic, aimed at
cultivating citizens’ virtue, but the Laws’ discussion says little about the belief content
required for virtue. The Athenian makes the general point that poets and everybody
else in the city must affirm that the best life is the most pleasant (662b–664b) and that
the criterion of correctness in music—since music involves making likenesses and
imitation—is accuracy in representing its model, beauty (667c–668b), which requires
knowledge of what has been represented and of how correctly and well the copy repre-
sents it (669b). The Athenian also repeats that authors may not compose as they like
but must bear in mind the effects of their compositions on virtue and vice (656c), that
poets are bound to express the society’s notions of virtue (801d, cf. 817d–e), and that the
elder chorus, who are to guide the young citizens’ souls to virtue, must be able to accept
virtuous representations in music and reject vicious ones (812c). But compared to
Republic II–III there are few details about what conduct citizens are to believe is con-
sistent with, or required by, courage or moderation; instead, the Laws details the kinds
of dance positions that represent courage and moderation. For example, in the dance of
war, dancers must keep their bodies erect and their limbs straight as far as possible, as
they represent the defensive actions in war by dodging, retreating, jumping into the air,
and crouching, and the offensive actions by adopting the postures of shooting a javelin
and delivering blows (814e–815b). At the same time, the Laws emphasizes the pleasure
which a well-educated person takes in the right sorts of songs and dances, privileging
this pleasure over correctness (e.g. 654b–d).
These two new emphases in the Laws, on movement and on pleasure, seem to be
related.33 Although the Athenian describes the young soul as malleable and impressionable
(Laws 664a, 666c, 671b–c) as Socrates did in the Republic, he also observes that even very
young humans can perceive and delight in orderly movement (653e–654a, 664e–665a).
His point seems to be to show how even disorderly young souls, and/or non-rational
soul-parts, are responsive to and appreciative of order, so that no part of virtue is a

33 In Plato’s Utopia Recast (Oxford, 2002), C. Bobonich relates this to a rejection of the Republic view
of the soul as composed of three independently motivating parts. According to Bobonich, Plato’s late
psychology so cognitively impoverishes non-rational soul-parts that reason is needed to contribute
conceptual content to even non-rational psychic movements such as sensory pleasures; the ethical upshot
is that even sensory pleasures involve appreciation of fineness or good order in perception—rather
than being mere fulfillments of appetitive desires (350–73). By contrast, in The Brute Within (Oxford,
2006), H. Lorenz argues that in Plato’s late psychology non-rational soul-parts have their own content
and are only deprived of belief (95–110). The contrast between the Republic and Laws I have sketched is
neutral between these two views. I give my own view of the relationship between late psychology and
education in R. Kamtekar, “Psychology and the Inculcation of Virtue in Plato’s Laws,” Cambridge
Critical Guide to Plato’s Laws (ed. C. Bobonich, Oxford, 2010), 127–48.
 Plato on Education and Art   625

condition foreign to the soul and imposed on it by habituation (Republic 518d–e), but a
return of the soul to its original condition.34

Bibliography
Annas, J. “Plato on the Triviality of Literature” in in J. Moravcsik and P. Temko (eds.), Plato on
Beauty, Wisdom and the Arts (Totowa, N.J., 1982), 1–28.
Bobonich, C. “Persuasion, Compulsion, and Freedom in Plato’s Laws,” in G. Fine (ed.), Plato 2:
Ethics, Politics, Religion and the Soul (Oxford, 1999), 373–403.
Bobonich, C. Plato’s Utopia Recast (Oxford, 2002).
Burnyeat M. F. “Culture and Society in Plato’s Republic,” in G. B. Peterson (ed.), The Tanner
Lectures on Human Values 20 (Salt Lake City, Utah, 1997), 263–325.
Burnyeat, M. F. “Plato on Why Mathematics Is Good for the Soul,” in T. Smiley (ed.),
Mathematics and Necessity: Essays in the History of Philosophy (Oxford, 2000), 1–81.
Burnet, J. Platonis Opera vols. I–V (Oxford, 1900–1907).
Cooper, J. and Hutchinson, D. S. (eds.), The Complete Works of Plato (Indianapolis, Ind., 1997).
Destrée, P. and Hermann, F.-G. (eds.). Plato and the Poets (Leiden, 2011).
Dillon, J. and Gergel, T. L. (eds. and trans.), The Greek Sophists [Greek Sophists] (London, 2003).
Fine, G. (ed.), Plato 1: Metaphysics and Epistemology and Plato 2: Ethics, Politics, Religion and
the Soul (Oxford, 1999).
Fine, G. “Inquiry in the Meno,” in Plato on Knowledge and Forms (Oxford, 2003), 44–65.
Fine, G. Plato on Knowledge and Forms (Oxford, 2003).
Fine, G. “Knowledge and Belief in Republic V–VII” in Plato on Knowledge and Forms (Oxford,
2003), 109–16.
Fine, G. The Possibility of Inquiry (Oxford, 2014).
Franklin, L. “Recollection and Philosophical Reflection in Plato’s Phaedo,” Phronesis 50
(2005), 289–314.
Gould, T. The Ancient Quarrel between Poetry and Philosophy (Princeton, N.J., 1990).
Halliwell, S. Republic Book 10 (Warminster, 1988).
Harold, J. “Infected by Evil,” Philosophical Explorations 8 [2005], 173–87.
Heckel, M. “Plato on the Role of Contradiction in Education,” British Journal for the History of
Philosophy 25 (2017), 3–21.
Höffe, O. (ed.), Platon, Politeia (Berlin, 1997).
Irwin, T. H. “Plato’s Objections to the Sophists,” in A. Powell (ed.), The Greek World (London,
1995), 568–90.
Irwin, T. H. “Republic 2: Questions about Justice,” in G. Fine [ed.], Plato 2: Ethics, Politics,
Religion, and the Soul [Oxford, 1999], 170–75.
Janaway, C. Images of Excellence: Plato’s Critique of the Arts (Oxford, 1995).
Kahn, C. “Plato on Recollection” in H. Benson (ed.) The Blackwell Companion to Plato
(Oxford, 2006).
Kamtekar, R. “Imperfect Virtue,” Ancient Philosophy 18 (1998), 315–39.
Kamtekar, R. “Social Justice and Happiness in the Republic: Plato’s Two Principles,” History of
Political Thought 22 (2001), 189–220.

34 I’m grateful to Gail Fine for comments on this chapter and to Rachel Singpurwalla for commenting
on the version of this chapter that was published in the first edition of the Oxford Handbook of Plato.
626   Rachana Kamtekar

Kamtekar, R. Speaking with the Same Voice as Reason: Personification in Plato’s Psychology’
in Oxford Studies in Ancient Philosophy 31 (2006), 167–202.
Kamtekar, R. “Psychology and the Inculcation of Virtue in Plato’s Laws,” in C. Bobonich (ed.),
Cambridge Critical Guide to Plato’s Laws (Oxford, 2010), 127–48.
Keuls, E. C. Plato and Greek Painting (Leiden, 1978).
Lorenz, H. The Brute Within (Oxford, 2006).
Menn, S. “On Plato’s Politeia,” in J. Cleary (ed.), Proceedings of the Boston Area Colloquium in
Ancient Philosophy 21 (Leiden, 2006), 1–55.
Moravcsik, J. and Temko, P. (eds.), Plato on Beauty, Wisdom and the Arts (Totowa, N.J., 1982).
Moss, J. “What Is Imitative Poetry and Why Is It so Bad?” in G. R. F. Ferrari (ed.), The
Cambridge Companion to Plato’s Republic (Cambridge, 2007).
Nehamas, “Plato on Imitation and Poetry in Republic X’ in J. Moravcsik and P. Temko (eds.),
Plato on Beauty, Wisdom and the Arts (Totowa, N.J., 1982), 47–78.
Nehamas, A. “Confusing Universals and Particulars in Plato’s Early Dialogues,” Virtues of
Authenticity: Essays on Plato and Socrates [Princeton, N.J., 1999] 159–75.
Penner, T. “Socrates on the Impossibility of Belief-Relative Sciences” in J. Cleary (ed.),
Proceedings of the Boston Area Colloquium in Ancient Philosophy 3 (Lanham, Md., 1987).
Scott, D. Recollection and Experience: Plato’s Theory of Learning and Its Successors
(Cambridge, 1995).
Scott, D. “Platonic Pessimism and Moral Education,” Oxford Studies in Ancient Philosophy 7
(1999), 15–36.
Williams, B. A. O. “Plato against the Immoralist,” in O. Höffe (ed.), Platon, Politeia (Berlin,
1997), 55–67.
Vlastos, G. “The Socratic Elenchus,” in G. Fine (ed.), Plato 1: Metaphysics and Epistemology
(Oxford, 1999), 36–63.
 chapter 26

Pl ato’s Th eol ogy

David Sedley

1. Educational Theology

The term “theology”1 (theologia) makes its very first recorded appearance in book II of
Plato’s Republic (379a), although the word may at this early stage perhaps mean no more
than “telling stories (logoi) about gods.” At any rate, the context is an educational one: in
the ideal city, Socrates maintains, children must not be exposed to the traditional myths,
which misrepresent gods as capable of harm, deceit, and other bad conduct. The poets
who tell these tales must be informed as to what “outline impressions” (tupoi) about the
gods they are to convey. The starting point is that god, being essentially good, could
never be the cause of anything bad (379b–e). Since, however, bad things not only happen
in the world, Socrates observes (379e), but far outnumber the good things, some cause
other than god must be found for them. Whether Plato ever worked out what this cause
of bad things might be, and how in a god-governed world such a thing could exist, is
a question to which we will return at the end of the chapter.
The impossibility that gods should cause harm is not merely a claim about their moral
character. It is the application of a ubiquitous Platonic metaphysical thesis about causation:
just as fire, being essentially hot, can only make things hot, and never cold, quite gener-
ally too, what is essentially F can never cause anything to be or become the opposite of
F; and god is no exception. Looking ahead to Plato’s account of the world’s divine crea-
tion in the Timaeus, we may note that there the same premise, that god is essentially
good, will be the “supremely authoritative principle” from which the whole cosmogony
follows (29e–30a).

My thanks to Gail Fine for helpful comments on an earlier draft.

1 Especially significant studies of Plato’s theology include F. Karfík, Die Beseelung des Kosmos:
Untersuchungen zur Kosmologie, Seelenlehre und Theologie in Platons Phaidon und Timaios (Munich
2004); M. Bordt, Platons Theologie (Munich 2006).
628   David Sedley

Despite the fact that the like-causes-like principle is a metaphysical one, its detailed
application in the Republic context is ethical, underwriting as it does Plato’s insistence
that the gods must be paradigms of the moral standards expected of humans. Take
harming. If gods inflict punishments, for example, those must be corrective punish-
ments for the benefit of those punished, not for their harm (380a–b). This corresponds
to Plato’s non-retributive theory of punishment, as developed in the Gorgias. A theology
of vindictive gods returning wrong for wrong would undermine that Socratically
inspired insight, and myths involving divinely inflicted retribution must be rewritten to
avoid misleading the young about the morality of punishment.
The concern manifested by this educational program has little to do with establishing
truths about the gods in their own right. The morally beneficial myths prescribed by
Socrates are seen by him as deliberate falsehoods, crafted for educational purposes.
At one point (378a) he even entertains the counterfactual hypothesis that some stories
about violence among the gods might be true, and insists that, if that were so, such
stories should not be told to the young, because of the harm that would be caused.
So although we meet in this part of the Republic Plato’s first sustained engagement with
the idea of gods as paradigms for human emulation,2 the focus is as much on educational
expediency as on theological truth.

2. The Shape of God

At first surprisingly, the misrepresentations of the gods that Socrates proposes to outlaw
from the Republic’s ideal city do not include anthropomorphism. If gods are to serve as
role models for the young, no doubt the fiction of their having human form is likely to
do no harm, perhaps even to help. But we should not be misled: in more directly theo-
logical contexts Plato will make it very clear that the human form is incompatible with
divine perfection.
The text that above all others explicates this doctrine is the Timaeus,3 generally
regarded as a late dialogue, although there is little doubt that many of the ideas show-
cased in it had been maturing for decades. The eponymous speaker Timaeus sets out
what can safely be taken to be Plato’s central theological tenets, even if interpreters have
from the start been divided about how literally these should be understood.4

2 “Becoming like god” as a human aspiration in Plato is not discussed in the present chapter, but see
J. Annas, chapter 22 of this volume, 278–79; “Becoming Like God, Ethics, Human Nature, and the
Divine,” in her Platonic Ethics, Old and New (Ithaca, NY 1999), 52–71; S. Lavecchia, Una via che conduce
al divino. La “homoiosis theo” nella filosofia di Platone (Milan 2006); and D. Sedley, “Becoming
Godlike,” in C. Bobonich (ed.), The Cambridge Companion to Ancient Ethics (Cambridge 2017), 319–37.
3 Not all aspects of Timaean theology can be addressed here. See further S. Broadie, Nature and
Divinity in Plato’s Timaeus (Cambridge 2011); also T. Johansen (this volume), and id. Plato’s Natural
Philosophy (Cambridge 2004); D. Sedley, Creationism and Its Critics in Antiquity (Berkeley, CA,
2007) ch. 4.
4 See T. Johansen, chapter 12 of this volume, esp. 466–70.
 Plato’s Theology   629

To put his theology at its simplest, god is intelligence (nous).5 The ultimate reason there
is a world, universe, or cosmos (kosmos means literally “ordering”) lies in the combination
of two initial entities: intelligence and matter. Matter is inherently disorderly, whereas
intelligence—which is inherently good, and by no means in a narrowly moral sense of
the word—loves to impose order. We must therefore posit the existence of a supremely
skillful intelligence, which, confronted with the totality of disorderly matter, set out
to impose the maximum of order on it. That pure intelligence functions as Plato’s
supreme god, but Timaeus has little to tell us about him: “To discover the maker and
father of this universe is a considerable task, and to communicate that discovery to
everybody impossible” (Timaeus 28c). What Timaeus seems much more confident about
telling us is how the supreme god or divine “craftsman” (dēmiourgos, English “demiurge”)
fashioned the available matter into the entity that we now call the world. Most notably,
he made his creation intelligent (and hence also possessed of soul), unique, perfectly
symmetrical (that is, spherical), complete, self-sufficient, and everlasting, each of these
attributes being incomparably superior to its alternatives, and therefore the intelligent
choice for a world-maker starting from, as it were, a blank canvass.
As this series of engineering decisions is unfolded, it becomes increasingly clear to
the reader that the emerging cosmos is itself a god. We are thus introduced to a second
class of divinities, created gods. Of these, the world itself is the foremost representative,
but others include the world’s major components: the earth, stars, sun, moon, and
planets. Much as the world itself is bounded by a sphere, namely that of the fixed stars, so
too the earth and celestial bodies are, each of them, spherical in shape, and they also
share the world’s everlastingness and intelligence. The implied reason that these divine
attributes are bound up together is that the exercise of intelligence, focused as it is on
unchanging entities, is physically embodied in the intelligent subject’s potentially per-
petual rotation on its own axis, a kind of motion natural to the sphere.
But another significance of the spherical shape lies in self-sufficiency. Although the
essential core of a human being is the (approximately) spherical head, our seat of intelli-
gence possessed of its own naturally circular motions, we humans could not have been
designed as overall spherical beings, because survival in a material world requires
asymmetric appendages: arms, legs, mouth, sense organs etc. The world, by contrast, is a
perfect and complete being with nothing whatsoever outside it for it to ingest, perceive,
or ward off. It therefore has no need for any such asymmetries, and remains by default a
perfect sphere—the finest of all shapes according to Timaeus (33b).
Given that gods are conceived as maximally self-sufficient beings, the sphere begins
to emerge as the natural shape of a divinity. The earth and other spherical divinities
within the cosmos must be assumed to approximate the same self-sufficiency.
For Plato, then, the scientifically true shape of divinity is the sphere, the asymmetric
human frame being an expedient necessitated precisely by our lack of divine self-­
sufficiency. If he did not insist on that in the Republic, it was not because he had not

5 For discussions of what this identification amounts to see S. Menn, Plato on God as Nous
(Carbondale and Edwardsville, IL, 1995); G. Van Riel, Plato’s Gods (Farnham, UK, 2013), 68–97.
630   David Sedley

yet arrived at his rejection of anthropomorphism, a rejection that after all had been
entrenched in the philosophical tradition since Xenophanes (sixth century b.c.e.). Rather,
to repeat, it is because the error of envisaging gods as human in form is an educationally
benign error, indeed probably one that actively encourages human beings to emulate
their divine role models.

3. Other Divine Attributes

It is in any case mainly in the context of cosmological science that the shape of god
assumes such importance. From the point of view of the Platonic spectrum of values, the
world-god is a paradigm of the highest kind of happiness (eudaimonia, 34b). Unlike us,
it has no external relations with other members of its own kind, there being none
(cf. 31a–b), and therefore has neither the opportunity nor the need to exercise the
equivalent of human moral virtues. This negative attribute, far from making the world-
god deficient, is what enables it to concentrate on enjoying complete self-knowledge
and self-love, identifiable with intellectual rather than moral virtue.
That the world’s creator, a supreme intellect, should make the world as like himself as
possible (29e), reflects an approach to causality nowhere spelt out by Plato but widely
assumed and applied in his dialogues. It comes in two parts, the first of which we have
already encountered in the Republic. The dual principle is clearly in operation at Timaeus
29e–30a:

(a) Like causes like. Hence the creator god, being essentially good, naturally made
his creation good. This pattern of causal transmission is specifically reflected in
the way he passed on his own supreme kind of goodness, namely intelligence, to
the world-god he created.
(b) The cause is greater than the effect. For example, as fire regularly makes other
things hot, but not as hot as itself, so too the creator god made the world good,
but not as good as himself. Rather, in Plato’s formulation he made it as good as it
was possible for him to make it.

How and why the created world-god, though good, necessarily falls short of perfection
is a major topic in Platonic theology, to which we will return in due course. At ­present,
however, the task is a narrower one: to see how that same pair of causal principles results
in there being at least two different kinds of god. Consider the quintessentially divine
attribute, immortality. The supreme deity, being essentially immortal, must by causal
principle (a) transmit immortality to anything he creates. Therefore the world-god, being
his creation, is immortal. But by causal principle (b), the created world-god cannot have
the same degree of immortality as does its creator.
Whatever this means, it cannot be that the world-god will endure for a very long but
nevertheless finite time. Rather, it is the manner of its everlasting duration that is attenu-
ated. The creator god is represented by Plato as a supreme craftsman, and one hallmark
 Plato’s Theology   631

of craftsmanship is likely to be the durability of the product. This reaches its extreme
when the product is so durable that no one but its creator could destroy it (think of tying
a knot so tight that no one weaker than you could untie it). That is the kind or manner of
durability that the world-god enjoys. Only its creator could destroy it, and he, being
good, has no conceivable motive for doing so, now or at any future time.
Thus whereas the supreme god is essentially immortal, the world-god and its divine
components such as the earth and stars are only contingently immortal. More specifically,
the created gods possess not an essential but a derivative and conferred immortality.
Why then did there need to be gods of this secondary kind? One part of Plato’s answer
is that it is gods of this kind that must be credited with the creation of mortal animals
such as ourselves. If the supreme god had directly created us, we would be derivatively
immortal, and that would preclude the existence of an equitable life cycle in which our
souls periodically transmigrate up or down the scala naturae—a doctrine of reincarna-
tion that Plato affirms not just in the Timaeus but in many dialogues, and sees as ethically
fundamental to the existence of cosmic justice.
But that topic cannot detain us now. Instead we must focus on a second function of
the created gods, one which requires a further subdivision. At 41a Timaeus lists the
created gods as follows: “both the visibly rotating ones, and those who appear to us just
to the extent that they are willing.” The “visibly rotating gods” are the heavenly bodies,
plus almost certainly the earth too, even though the evidence for Plato’s thinking that
the earth rotates on its own axis (40b–c, quoted later on in this chapter) is controversial.
The celestial divinities are living beings made mainly of fire, visibly orbiting the earth
along trajectories that are subject to complex mathematical analysis. From the most
simple to the most complex, the observation of these rotations serves to teach the human
race mathematics, starting no doubt with the simple counting of days and nights, and
extending all the way up to the highly problematic reduction of planetary orbits to the
combined functions of regular motions. Orderliness is thus a feature of the divine that
can best be discerned and appreciated by the science of astronomy, thanks precisely to
the creator’s construction of the heavens out of these providently visible fiery gods.
Taken as a whole the celestial orbits are, in effect, the divine world-soul’s thinking made
visible, thinking that we can learn to replicate in our own heads. The orderliness of the
celestial divinities is in turn a direct reflection of divine goodness, thus making astron-
omy the privileged route toward philosophical understanding, and thereby toward the
highest form of human happiness, achieved by internalizing that same orderliness in
our own intellects.

4. Traditional Gods

Regarding Plato’s distinction between two kinds of created gods, “both the visibly rotat-
ing ones, and those who appear to us just to the extent that they are willing” (41a), we
have so far dealt with the former class of gods. They are in effect the gods of scientific
theology: paradigms of regularity, subject to empirical study and precise mathematical
632   David Sedley

analysis. The latter kind are the traditional members of the divine family, not usually
open to inspection at all, but witnessed at most only in reported epiphanies.
The passage just quoted is in effect Timaeus’s transition from scientific to nonscientific
theology. Having introduced the two major circles of celestial rotation and the visible
divinities—sun, moon, fixed stars, planets, and earth—that move in accordance with
them, he remarks that without a visible mechanical model it is impossible to go further
in charting their conjunctions, back-circlings, occlusions, and the like; and that he is
therefore bringing to an end his account of the “nature” ( phusis) of the visible gods.
To paraphrase, here ends Timaeus’s theological physics. What immediately follows
amply confirms that this is so, as Plato’s speaker turns to the gods of legend (40d–41a):

As for the other divinities, to speak about them and to know how they came to be is
beyond our capacity. Rather, we must believe those who have spoken about them in
the past, who were, on their own say-so, offspring of gods, and no doubt had clear
knowledge of their own forebears. So it is impossible to disbelieve sons of gods,
although they speak without likely and necessary proofs; but we should follow custom
(nomos) and trust them, on the ground that the things they claim to be reporting are
their own family matters.
So concerning these gods let us, basing ourselves on them, adopt and speak of the
following genealogy. The children born to Earth (Gē) and Heaven (Ouranos)
were Oceanos and Tethys, whose children were in turn Phorcys, Cronos, Rhea and
the others of their generation. Cronos and Rhea gave birth to Zeus, Hera and all
those we know of who are said to be their siblings, and yet further offspring of these.

And the creation narrative continues with the supreme creator delivering, to an audience
consisting of both sets of gods—not only the scientific divinities, but also those of
­legend—a detailed set of instructions for the creation of mortal life forms.
This is an important passage for a number of reasons. Timaeus has now moved
beyond scientific theology—the domain in which “likely and necessary proofs” are
available. “Likely” proofs are, in the Timaeus (29b–d), forms of reasoning typical of
Platonic physics as a whole, based on a well-informed reconstruction of the way a creator
god would be likely to reason when designing and constructing a world in imitation of
an eternal model. “Necessary” proofs are harder to pin down in Plato’s terminology, but
in context the expression must almost certainly refer to the kind of mathematical rea-
soning that underpins Timaeus’s complex account of celestial motion. In other words,
Timaeus is here bidding farewell to scientific theology, the mathematical study of
spheres in motion, and moving the spotlight to the gods of the traditional theogonies.
That Zeus was born the son of Cronos and Rhea, for example, may well be true, but no
amount of scientific reasoning can help to confirm it. Instead, Timaeus indicates with a
shrug, we must simply accept the divine family trees bequeathed by theogonic authorities.
What ensues is an outline synthesis of Hesiodic and Orphic theogony in five genera-
tions: Earth (Gē) and Heaven (Ouranos); Okeanos and Tethys; Phorcys, Kronos, Rhea,
etc.; Zeus, Hera, and the rest of their generation; and finally the offspring of these last.
 Plato’s Theology   633

Who are the authorities whom we are here asked to trust regarding divine genealogy?
The attribution to them of divine parentage shows that the reference is principally to
Orpheus and Musaeus (both of them offspring of Selene and the Muses, according to
Republic 2.364e), although it is hard to think that Hesiod’s divine genealogy in his
Theogony is being altogether excluded. Nearly all modern commentators remark that
the grounds offered by Timaeus for accepting the word of these supposed authorities on
divine genealogy, namely their claimed divine parentage, are “ironic,” but this should be
resisted. Timaeus’s point in the quoted words is not to mock the alleged grounds for
believing in traditional gods, but simply to make it clear that on the one hand he is piously
retaining the traditional deities in his pantheon, but that on the other he has nothing to
say about their origins beyond what can be read in the theogonies,6 since such deities are
not subject to scientific argument. If he were indeed speaking ironically, he would be
casting doubt on the reasons for believing in divinities to whom he prayed at the start of
his speech (26b–c). There, urged by Socrates to start with the conventional appeal to
gods, Timaeus agreed, saying “it is necessary that we invoke both gods and goddesses,
praying that we may say everything above all as they would wish, and secondarily as we
would.” Worse still, according to the legislation against impiety proposed in Laws book X,
mockery of traditional religious belief or practice, even by fundamentally good people, is
so damaging to social norms as to require stern punishment (908c–d).
Nor are we being offered a choice between two mutually exclusive modes of theology.
The ancestral couple who head the divine genealogy, namely Earth (Gē) and Heaven
(Ouranos), were previously implied by Timaeus to be the two primary deities from
whom the cosmos began (40b–c):

Earth ( gē), who is our nurse, but rotates around the axis that stretches through the
whole, he [the creator] contrived as guardian and fabricator of both night and day,
the first and most senior of all the gods who have come to be within the heaven
(ouranos).

If Earth is “most senior” of the gods within the Heaven, that is perhaps in the sense that
the Demiurge could not have created the celestial rotations of sun, moon, fixed stars,
and planets unless there had already been a central earth for them to orbit around. By
the same criterion, even if only implicitly, the enveloping Heaven at whose center the
earth is placed will itself be an even more senior deity, since a center presupposes a
perimeter.
This very senior divine pair become not only the primeval cosmological deities, but
also the ultimate ancestors of the divine family now headed by Zeus.
Although Plato has, as we saw, made it clear that the latter group of deities cannot be
scientifically studied, in the way that stars can, he does nevertheless find for them a

6 The same point is made at Republic 365e by Adeimantus, who a little earlier (364e) has identified
the divinely born theogonic poets as Orpheus and Musaeus.
634   David Sedley

major role in the creation process, namely the design and construction of the human
body, a task explicitly said to have been shared out among the gods of both kinds (41a–d).
Although massive stars made of fire might in principle be imagined somehow contrib-
uting to this task, Timaeus’s assignment of a role in it to quasi-personal gods of the
Olympian variety is not only more credible intuitively, but also respectful toward the
Greek tradition: in a famous passage of Hesiod’s Works and Days (53-105) Pandora,
the first woman, is constructed by the craftsman god Hephaestus, assisted by an entire
support team of other Olympians. This may well be the creative model that Plato had in
mind for his own anthropogonic myth.
Whether Plato was altogether content with the retention of these quasi-personal
divinities is a harder question to answer. The myth in the Phaedrus, in explaining people’s
differing erotic choices, goes so far as to suggest that different Olympian deities re­present
different paradigms of value, and that seeking a beloved who will follow the same god as
oneself comes down to some kind of moral choice (252c–253c): whereas followers of
Zeus are philosophical in character or aspire to leadership, those of Hera are “royal” and,
more worryingly, those of Ares vengeful.
In this passage Socrates’ myth is accounting for erotic diversity, and it may be doubted
whether Plato was, in his own theology, prepared to introduce this degree of variability
into divine goodness.

5. The Laws

Consider in this regard the opening of Plato’s last and longest dialogue, the Laws, a
conversation largely focused on the creation of a theocratic society, in which laws will
stand proxy for the commands of a divine intellect.7 In the dialogue’s first lines an
Athenian asks a Cretan and a Spartan about their own home cities: “God or some human
individual (theos ē tis anthrōpōn), visitors? Which is responsible for the assignment of
your laws?” The Greek indefinite pronoun tis here refers unambiguously to a human
individual as legislator (Lycurgus at Sparta, Solon at Athens, and so on); the same pro-
noun’s omission from the reference to god is no accident—especially in a Platonic dia-
logue’s always meticulously crafted opening words.8 Rather, it alludes to the scientific
theology that will follow, especially in book X. The wording carefully leaves open the
possibility that the reference to god, despite being grammatically singular, as often, may

7 Valuable contributions on the theology of the Laws include R. Mayhew, Plato, Laws 10 (Oxford
2008; (translation and commentary), and “The Theology of the Laws,” in C. Bobonich (ed.), Plato’s
Laws: a Critical Guide (Cambridge 2010), 197–216; J. Jirsa, 2008, “Plato on Characteristics of God: Laws
X, 887c5–899d3,” Rhizai 5 (2008), 265–85; and G. Carone, Plato’s Cosmology and its Ethical Dimensions
(Cambridge 2005), ch. 8.
8 Cf. M. F. Burnyeat, “First Words: A Valedictory Lecture,” Proceedings of the Cambridge Philological
Society 43 (1977), 1–20, reprinted in his Explorations in Ancient and Modern Philosophy 2 (Cambridge
2012), 305–25.
 Plato’s Theology   635

not be to any discrete individual. It thus subtly foreshadows the argument for theism in
book X, which will conclude that the orderly motion of the heaven is caused by divine
and ideally good “soul, or souls” (899b5–8). What Plato’s speaker thus emphasizes is not
the gods’ individuality, if indeed they are individuals, but their possession in common of
absolute goodness. Plato’s deep theological commitment is not to monotheism, a creed
with very little attestation in pagan antiquity, but to the essential unity of the divine.
At least in what I have been calling his scientific theology he, like Socrates and others
before him,9 generally avoids the common Greek practice of naming gods and thus
representing them as fully distinct individuals.
Laws book X is Plato’s final word on scientific theology, presenting a formal argument
for the existence of god, intended to be prefixed to the legislation against impiety that is
envisaged for his projected city Magnesia. In the opening pages of the book Plato shows
himself, if in passing, well aware of two existing arguments for the existence of god, both
of which remain theological classics today, and in Plato’s day were apparently already
sufficiently well known to be put into the mouth of a secondary speaker, the Cretan
Cleinias (886a), rather than that of the anonymous Athenian who leads the discussion:

1) The argument from design: the beneficial orderliness of the world, especially the
cycle of seasons and the celestial rotations on which it depends, attests the existence
of gods.
2) The argument from consensus: belief in gods is a cross-cultural universal.

The Athenian responds that these bare arguments are insufficient in the face of com-
mitted atheists, who have enough physics to think they can explain cosmic orderliness
by appeal to the regular behavior of matter, and enough anthropology to classify religion
as a human construct, its local variations running parallel to variations in another
closely related human construct, law. He therefore proceeds to construct the earliest
known extended argument for the existence of god. It at no point gives signs of aspiring
to formal validity, which it certainly lacks, but it not only exhibits a complex inferential
structure, but is densely enough theorized to serve as a major repository of Plato’s late
theological thought.
The following selective outline does not attempt to smooth over the considerable
difficulties presented by this probably unrevised text.

5.1 Kinds of Motion (893b–894c)

The Athenian proposes the following curiously heterogeneous list of kinds of motion.

9 Cf. M. F. Burnyeat, “The Impiety of Socrates,” Ancient Philosophy 17, 1–12; repr. in T. Brickhouse
and N. Smith (eds.), The Trial and Execution of Socrates (Oxford 2002), 133–45, in R. Kamtekar (ed.),
Plato’s Euthyphro, Apology, and Crito: Critical Essays (Lanham, MD 2004), and in Burnyeat,
Explorations 224–37.
636   David Sedley

Motion in one place (1) rotation

Motion in many places (2) sliding or rolling
Further kinds of motion (3) Separation
(4) Combination
(5) Growth
(6) Shrinkage
(7) Generation
(8) Destruction
A causal distinction (9) Motion that moves others but not itself
(10) Motion that moves itself and others

With this list established, he concludes (10) self-motion to be the most powerful kind,
since it is that on which the others ultimately depend. Imagine for example (895a–b)
that everything in the universe were to come to a halt: in that thought experiment, only a
motion capable of moving itself could ever get things back in motion. Thus the ranking
eventually reached is that (10) self-motion is prior to all other motion. Ranked next is
(9), motion that moves only others, since it is through this that a self-mover is able to
transmit motion beyond itself. The other eight kinds of motion, whatever order they
may be ranked in, one way or another are caused by, and depend on, the initial combina-
tion of movers.
In case all this should sound puzzlingly abstract, it is worth reflecting that Plato’s is
an immediate forerunner of Aristotle’s argument in Physics book VIII, for the existence
of the prime mover, which he elsewhere identifies as god. The difference is that Plato’s
prime mover is a self-mover, Aristotle’s an unmoved mover. For both thinkers, the
chain of motion envisaged starts from a primary deity (10), and proceeds (9) through
the circular motions of the celestial bodies (1), above all the sun. The chain then contin-
ues into the sub-celestial world in the form of the various motions (2–8) underlying
natural change.
Importantly, however, before moving to the postulation of a divine prime mover,
Plato introduces soul as the central linking concept.

5.2 The Equation of Self-Motion with Soul (895c–896c)

(i)Self-motion, whenever it is seen in matter, indicates the presence of life.
(ii)To be alive is the same thing as to have a soul (in Greek usage, a near-truism).
(iii)This enables us to equate “soul” with “self-moving motion.”
(iv) Therefore soul is first mover of all things, past, present, and future, and of these
things’ opposites. The other kinds of motion are those of body.
(v) Therefore soul is senior and prior to body, standing to it as ruler to ruled.
 Plato’s Theology   637

5.3 Soul as Cause of Both Good and Bad (896c–897b)

(i) Since soul is senior to body, things belonging to soul will be senior to bodily
things. Hence character, wish, reasoning, true opinion, concern, memory, etc.
exercise power over bodies.
(ii) Since soul is the cause of everything, it is cause of good and bad, beautiful and
ugly, just and unjust, etc. (cf. (iv) in section 5.2).
(iii) Since soul is present in and governs all things that are moved, soul must govern
the heavens.
(iv) We are speaking here of at least two kinds of soul: (a) beneficent soul and (b) the
kind capable of doing the opposite.
(v) Soul, by its wishes, right and wrong opinions, joy and grief, confidence and fear,
hate and love, etc., governs all things in the heavens, on land and in the sea; it
drives all the secondary motions, and the qualitative changes consequent upon
them. When it uses intelligence it produces correct and happy results, when
unintelligence, the opposite results.

5.4 The Goodness of Cosmic Soul (897b–898c)

(i) Which of the two kinds of soul governs the heavens, the earth, and the entire
celestial rotation?
(ii) If celestial rotation has a nature akin to intelligence, the best kind of soul
­governs it; if it is chaotic, the bad kind.
(iii) Of the 10 motions, (1) circular motion is most akin to intelligence, since it and
intelligence share complete uniformity and regularity. And irregular motion is
most akin to unintelligence.
(iv) Therefore it is the best soul—whether that be one or many—that governs the
heavens.

5.5 Cosmic Soul Is God (898c–899c)

(i) So each heavenly body is moved by soul. We cannot see anything’s soul, but we
can grasp it by intellect. Take the example of the soul that moves the sun.
(ii) The soul moving the sun is either (I) internal to it; or (II) external to it, driving it by
some material means; or (III) external to it, driving it by some immaterial means.
(iii) On any of options (I)–(III), the soul moving the sun is properly considered a god.
(iv) Likewise all celestial beings are moved by entirely good souls, and these too, on
any of options (I)–(III), should be considered gods.
(v) In short [to quote Thales, albeit anonymously], all things are full of gods.
638   David Sedley

To re-emphasize an earlier point, this Platonic defense of theism is pointedly

­ nconcerned with counting divinities, or even with conserving their individuality. In
u
that respect it contrasts strikingly with Aristotle’s argument in Metaphysics Lambda 8,
where he will set out to arrive at a precise figure for the number of divine movers needed
to account for all celestial phenomena. Nevertheless, in its own way the theology of the
Laws too is a mathematical science. Gods are primarily scientific postulates, halting the
potentially infinite regress of movers and accounting for the observed mathematical
perfection of stellar rotation.
But if that is what gods are, why does it matter to Plato that all the citizens of Magnesia
should believe in them? Why does he go so far as to propose solitary confinement and
ultimately capital punishment for unrepentant atheists? And why should the gods of sci-
entific theology even be associated with the gods worshipped in the traditional cults,
many of which will in some form be established in Magnesia? The key to Plato’s thinking
here is rational orderliness. Gods are perfect paradigms of this, and in the heaven they
manifest it in the complex combinations of perfect regularities to which astronomers in
Plato’s day, above all his close colleague Eudoxus, were with considerable success reduc-
ing the apparent irregularities of planetary motion. According to the Laws, the rational
orderliness of a good soul is essentially of this same kind. We saw at the start of the chapter
how in childhood, according to the Republic, one’s conception of the gods already serves
as a paradigm for moral emulation. In adulthood, we can now add, appreciation of god’s
perfect orderliness (albeit now intellectualized and non-anthropomorphic) seamlessly
continues that process of assimilation to the divine.

6. The Sources of Evil

In Section 1 of this chapter we encountered Plato’s concession in the Republic that there
are many bad things in the world, and that for these some cause other than god must
be found. Nowhere else does he explicitly return to this apparent gap in his theology
and supply the missing cause or causes of bad. Some candidates can nevertheless be
considered.10

6.1 Bad Cosmic Soul

Many scholars hold that in his late work Plato came to entertain a dualistic view of
cosmic soul, conceding that the world is not entirely governed by god or gods, who as we
have seen are essentially and inalienably good, but is also subject to the influence of a
second, bad kind of soul exercising power in the heaven. This, if established, would be a

10 There is still-valuable material on this controversy in H. Cherniss, “The Sources of Evil according
to Plato,” Proceedings of the American Philosophical Society 98 (1954), 23–30; repr. in G. Vlastos (ed.),
Plato, vol. 2 (1971), 244–58.
 Plato’s Theology   639

truly radical solution to the problem of evil, inviting all kinds of new questions about
whether, for example, bad soul is a necessary part of a best-possible world, or a regretta-
ble imperfection in ours.
What then is the evidence? It all lies in one short passage of Laws 10 (896d5–e7), cor-
responding to (ii)–(iv) in section 5.3, here translated verbatim:

(ii) ATHENIAN. After this, must we agree that soul is the cause of both good and
bad, both beautiful and ugly, both just and unjust, and of all pairs of opposites, given
that we are going to set it down as cause of everything?
CLEINIAS. Of course.
(iii) ATHENIAN. Since soul governs and inhabits all things that are in any way
moved, we must surely say that it governs the heaven as well?
CLEINIAS. Indeed.
(iv) ATHENIAN. One soul, or more than one? More than one, I will reply on your
behalf. Let us at the very least posit not fewer than two, namely the beneficent one
and the one capable of doing the opposite.
CLEINIAS. You are quite right to say so.

Outside these lines, the notion of a second, bad, celestial, or cosmic soul occurs nowhere
in the entire Platonic corpus. There can be no disputing that the Athenian does appear
to introduce such a cosmic dualism, and to be applauded by Cleinias for doing so.
However, it seems inescapable that something has gone wrong with the text here. The whole
argument (in section 5.3) from which the above lines are taken is devoted to showing that,
since all the motion in the world is caused by soul, soul must be of at least two kinds,
good and bad, in order to account for all explananda, bad as well as good—a tacit appli-
cation of the like-causes-like principle noted in Section 3 of this chapter. The further
question which of the two kinds governs the heaven is explicitly raised and answered
only in section 5.4, where the emphatic answer is that good soul alone does so. Hence it
makes no sense for 5.3 already to be singling out celestial soul for evaluation, let alone
announcing, without argument, that the heaven is governed partly by bad soul. The
combination of this structural incoherence and the totally un-Platonic result that would
follow makes it more prudent to assume that in 5.3 a few words are lost from the manu-
scripts, so that (iii) might for example have originally read:

Since soul governs and inhabits all things that are in any way moved, we must surely
say that it governs both the heaven <and everything within the heaven>?11

Unsupported, and even directly contradicted, by the remainder of the argument in

which it occurs, the dualistic explanation of evil should be discounted.12

11 Perhaps read καὶ τὸν οὐρανὸν <καὶ πᾶν τὸ ἐντὸς οὐρανοῦ>, 896e1.
12 I am here to a large extent agreeing with Cherniss 1954, n. 29. See also the extended discussion
of the issue in Carone ch. 8. As far as I am aware the need to emend has not previously been
acknowledged.
640   David Sedley

6.2 Bad Human Soul

What has, however, emerged from the immediately preceding discussion is that some
of the harm done in the world, perhaps most of it, is caused by the morally bad and igno-
rant souls of human beings. While human folly is unlikely to account for all the suffering
in the world, it is undoubtedly a major factor in Plato’s theodicy. The question we must
face is why, in Plato’s eyes, a good creator god should have made our souls such as to be
capable of vice.
Timaeus 39e–42e offers a bold answer. The world is an all-embracing animal, and is
therefore modeled on the Form of the genus animal, which contains all the animal spe-
cies and subspecies. In order to be as complete as possible a reflection of that model, the
world’s contents had likewise to include, not just the immortal fiery star-gods, but speci-
mens of all the mortal species that join them in constituting the genus; and souls suitable
to animate all of these specimens had also to be created. But lower-order animal species
are not suitable vehicles for the best kind of souls. These are rationally functional souls,
naturally housed at the top of the upright human frame, in its approximately spherical
head, in which the circular rotations of those souls’ reasoning activities can be developed
and in some cases perfected. Quadrupeds, reptiles, and fish, by contrast, are degenerate
descendants of the original male human archetype, reshaped so as to be occupied by
correspondingly degenerate souls that do not properly, if at all, exercise their rationality.
A system of transmigration does, in mitigation, allow souls to be promoted as well as
demoted in the scala naturae, and thus some kind of overall justice is preserved. But the
key point for present purposes is that, for the above reasons, an ideally good world is one
that includes degenerate human souls suitable for temporary reassignment as the ani-
mators of lower species. The existence of human vice is a price worth paying for the
world’s thereby maximized completeness.
But does this mean that a perfectly good god has, in creating souls, created something
at least potentially bad? That would seem to risk compromising causal principle (a), that
like causes like (see Section 3 of this chapter). Plato shows his sensitivity to such a risk
when Timaeus relates how the creator, having himself created all the individual rational
souls, delegated to the lesser gods whom he had created the task of embodying those
souls, “in order that he should not be the cause of the badness that each of them would
come to possess” (42d). This in turn implicitly relies on causal principle (b), that the
cause is greater than the effect: the created gods have a conferred rather than an intrinsic
goodness, which enables them, unlike their creator, in their own turn to create beings
with the capacity to become actually bad. Even then, it is an important point for Plato
that those who become bad are themselves, by their moral choices, the cause of their
deterioration: the lesser gods were instructed “to steer the mortal animal on the fin-
est and best path possible, except to the extent that it should itself become the cause
of its own evils” (42e). The adjective (aitios) translated “the cause of ” here can also be
translated “responsible for.” Timaeus’s care in insulating gods from causing, or being
responsible for, badness recalls an iconic maxim from the myth in Republic X, regarding
each soul’s choice of life: “Chooser’s responsibility, not god’s” (617e; aitia helomenou:
theos anaitios).
 Plato’s Theology   641

6.3 Matter
So far the emphasis has been mainly on moral badness. But how about illnesses, natural
catastrophes, and the like, and our frailty in succumbing to these and other dangers?
What causes those? Since antiquity the most favored candidate among interpreters of
the Timaeus has been: matter. When creating the world the divine intelligence (nous)
persuaded “necessity” (anankē), meaning material stuffs understood as capable of only
mechanical behavior, to cooperate in its work. For example, the creator organised huge
quantities of fire to make star-gods, light, and thereby also vision. Nevertheless, the
interpretation goes, matter has an inherent capacity to be wayward, thereby earning its
nickname “the Wandering Cause” (48a). And ultimately it is the residual intransigence
of matter that prevents divine intelligence from making the world altogether good.13
This interpretation raises difficulties of its own.14 Does Plato really think that matter,
the lowliest component of the universe, never hinted to be divine, ever successfully
resists god? His supreme creator god starts out working with an inherently featureless
and pliable matter (what Timaeus calls “the receptacle”), and shapes it just as he chooses,
structuring the particles of each of the four elements as perfect geometrical solids (tet-
rahedron for fire, cube for earth, octahedron for air, icosahedron for water). If these
chosen shapes thereafter limit what he can do with matter—for example, he presumably
cannot make a river out of earth or a mountain out of fire—that will depend on his
earlier decisions about how to shape it, not on the intransigence of matter as such.
It is therefore better to think here, not of matter that still resists divine “persuasion,”
but of Plato’s metaphysical assumption that no physical embodiment of a Form, be
it that of fire, of water, of largeness, or of beauty, can aspire to the Form’s perfection.
By replicating in matter some features of the paradigm, even the best craftsman must
sacrifice or limit other features. Timaeus’s example is the human head (75a–b), whose
divine creators had to choose whether to maximize its powers of discernment or its
durability: they rightly chose the former, as easily outweighing the latter in value. This is
not the intransigence of matter, because no kind of matter the creator had chosen or
designed could have resolved the clash of competing desiderata. What is at work here is,
rather, the inevitable need for practical engineering compromises among competing
goals. It can no more be blamed on matter than can the compromise, discussed previ-
ously, between the world’s completeness and the goodness of its inhabitants.
However, another approach is possible. In a classic passage of the Theaetetus (176a),
Socrates associates evil with the absence of god:

But it is not possible for evils to be eliminated, Theodorus, since there must always
be some opposite to good, nor for them to become established among the gods, but
of necessity they frequent human nature and this region. Which is why one should
escape from here to there as soon as possible, an escape which consists in becoming
as like god as possible.

13 Such is the prevalent reading of Timaeus 46c–e, 47e–48a, 75a–b; e.g., Mason 2010, 174–76.
14 Lennox 1985; Sedley 2007, 113–27).
642   David Sedley

The Theaetetus context is not one of physics, but it brings to mind Timaeus 53b, where
Timaeus describes the chaotic motions of matter before the world-creator imposed
order on it: “it was entirely in the condition that anything is likely to be in when god is
absent.” In the light of this we may say that when matter was earlier characterized as
the non-purposive “Wandering Cause” (48a), that portrayal of it must primarily
describe the kind of contribution to which it would revert if the ordering role of god
were absent.15
But is god in reality ever or anywhere absent? If this question were put to Aristotle, he
would say that divinely governed orderliness is more strongly present in the heaven than
in the region we ourselves inhabit. The nearest Timaeus comes to such a thought is when
he describes the cosmic intelligence as combatting the Wandering Cause by “bringing
into the best state most of the things that undergo becoming” (48a). For instance, we may
take him to mean, most but not all of the world’s fire was used up in making star-gods
and light: hence there is in the world a residue of unruly fire that may on occasion burn
out of control. The question that this raises, and that Timaeus never addresses, is the
following. Is the retention of some unruly elements in the world a sign of its divine cre-
ator having been defeated or in some other way constrained by matter, or even of his
having simply not bothered to go further in imposing order? Or was it all along part of
his preferred design for the world? If, as seems likely, the latter option is intended, then
the world was all along meant to retain a capacity for certain chaotic events, notably the
cataclysms that according to Plato (Laws 3) periodically restart and morally cleanse
degenerate civilizations. Divine intelligence perhaps is absent from the unruly air from
which hurricanes emerge, the unruly water that causes floods, the unruly earth that
permits earthquakes, and so on. But if so, that is likely to reflect a decision by divine
intelligence that, all things considered, it is better that it leave certain parts or zones of
the world’s matter in a disorderly state.

7. Conspectus

Plato inherited Socrates’ conviction that a proper understanding of the divine nature is
essential to human virtue and happiness. Hence god’s essential goodness is the motif
that runs most prominently through all Plato’s theological arguments. Since this supreme
goodness is manifested above all in the cosmic structures created by divine intelligence,
it is understandable if Plato turns out to stick resolutely to his insistence that, for all its

15 Cf. the Statesman myth. 269d–e: the world alternates between phases controlled by god and
other phases where he is absent; “it is only to the most divine things that it belongs to stay always the
same, and the nature of body does not belong to that order,” so the world cannot remain altogether
changeless. 273b–d: when god lets go, the precosmic disorder that is natural to body reasserts itself and
causes evils. The Statesman myth is a fantasy, not a literal cosmology, but may convey genuine Platonic
principles.
 Plato’s Theology   643

appearances of imperfection, from a global perspective ours is the best physical world
that could ever have been created, even by a supremely powerful being.
On the other hand, Plato shows less interest than Socrates did in the idea of divine
intervention in individual human lives. To that extent his work in theology points for-
ward to Aristotle, who would insulate god entirely from concern with the sublunary
world.
Finally, as we have seen, it is not particularly helpful to label Plato either as a polythe-
ist or as a monotheist. On the one hand, when he speaks interchangeably of “the gods”
and “(the) god,” he is following a regular Greek linguistic practice that was never felt to
imply that the gods might be in reality a singular being. And in what I have called his sci-
entific theology he does clearly enumerate distinct cosmological divinities, such as the
earth and the sun, as well as differentiating between the supreme creator god and the tier
of lesser gods that he created.
On the other hand in the Laws, which includes Plato’s final venture into theology, he
hints from the outset, and makes clear in book X, that the option of simply postulating
the existence of undifferentiated divine soul, and equating this with god, would be
explanatorily sufficient. His reason for keeping such an option open seems to lie in his
profound conviction, in the face of the mainstream religious tradition, that whatever
divine powers there may be are defined by their shared essential goodness and unity of
purpose.

Bibliography
Annas, J. 1999. “Becoming Like God, Ethics, Human Nature, and the Divine,” in Platonic
Ethics, Old and New (Ithaca, NY), 52–71.
Bordt, M. 2006. Platons Theologie (Munich).
Broadie, S. 2011. Nature and Divinity in Plato’s Timaeus (Cambridge).
Burnyeat, M. 1997. “First Words: A Valedictory Lecture,” Proceedings of the Cambridge
Philological Society 43, 1–20, reprinted in his Explorations in Ancient and Modern Philosophy
2 (Cambridge 2012), 305–25.
Burnyeat, M. 2004. “The Impiety of Socrates.” Ancient Philosophy 17, 1–12; repr. in T. Brickhouse
and N. Smith (eds.), The Trial and Execution of Socrates (Oxford 2002), 133–45, in
R. Kamtekar (ed.), Plato’s Euthyphro, Apology, and Crito: Critical Essays (Lanham, MD
2004), 210–28, and in M. Burnyeat, Explorations in Ancient and Modern Philosophy 2
(Cambridge 2012), 224–37.
Carone, G. 2005. Plato’s Cosmology and Its Ethical Dimensions (Cambridge).
Cherniss, H. 1954. “The Sources of Evil according to Plato,” Proceedings of the American
Philosophical Society 98, 23–30; repr. in G. Vlastos (ed.), Plato vol. 2 (Garden City, NY 1971),
244–58.
Jirsa, J. 2008. “Plato on Characteristics of God: Laws X, 887c5–899d3,” Rhizai 5, 265–85.
Johansen, T. 2004. Plato’s Natural Philosophy (Cambridge).
Karfík, F. 2004. Die Beseelung des Kosmos: Untersuchungen zur Kosmologie, Seelenlehre und
Theologie in Platons Phaidon und Timaios (Munich).
Lavecchia, S., 2006. Una via che conduce al divino. La “homoiosis theo” nella filosofia di Platone
(Milan).
644   David Sedley

Lennox, J. 1985. “Plato’s Unnatural Teleology,” in D. O’Meara (ed.) Platonic Investigations

(Washington, DC), 195–218. (Reprinted in Lennox, Aristotle’s Philosophy of Biology
(Cambridge, 2001, 280–302).
Mason, A. S. 2010. Plato (Durham, UK).
Mayhew, R. 2008. Plato, Laws 10. (Oxford).
Mayhew, R. 2010. “The theology of the Laws”, in C. Bobonich (ed.), Plato’s Laws: a Critical
Guide (Cambridge 2010), 197–216.
Menn, S. 1995. Plato on God as Nous (Carbondale and Edwardsville, IL).
Sedley, D. 2007. Creationism and Its Critics in Antiquity (Berkeley, CA).
Sedley, D. 2017. “Becoming Godlike,” in C. Bobonich (ed.), The Cambridge Companion to
Ancient Ethics (Cambridge), 319–37.
Van Riel, G. 2013. Plato’s Gods (Farnham, UK).
 chapter 27

Pl ato a n d A r istotl e
i n the Aca dem y

Christopher Shields

Even the more recent among the older thinkers found themselves at a
loss, lest it turn out according to them that the same thing should be at
the same time both one and many.
—Aristotle, Physics 185b25–27

1. Aristotle as a Source of Data

Regarding Plato’s Philosophy

We need not stray too far into baseless psychobiographical speculation to set aside
two competing, equally monodimensional treatments of Aristotle’s relation to Plato.
The first pictures Aristotle beginning his intellectual life as a meek and dutiful Platonist,
coming into his own as a philosopher only after the death of his master, some twenty
years beyond their earliest association.1 The second has him arriving in the Academy as
a fully formed Aristotelian, yet as a thinker too immature to grasp the subtlety and force
of his teacher’s philosophical accomplishments, with the result that he spent his time in
the school as an insufferable and captious critic of Plato.2 Still less is there reason to

1 So W. Jaeger characterizes Aristotle: “He had accepted Plato’s doctrines with his whole soul, and
the effort to discover his own relation to them occupied all his life, and is the clue to his development.
It is possible to discern a gradual progress, in the various stages of which we can clearly recognize the
unfolding of his own essential nature. . . . Just as tragedy attains its own special nature . . . ‘out of the
dithyramb’ by leading the latter through various forms, so Aristotle made himself out of the Platonic
philosophy” (Aristotle: Fundamentals of the History of His Development [Oxford: Oxford University
Press, 1934], 15).
2 An ancient tradition treats Aristotle as “the foal who kicked its mother” (Diogenes Laertius, Lives
of the Philosophers v 2). The ancient biographical evidence is collected in I. Düring, Aristotle in the
Ancient Biographical Tradition (Göteborg: Elanders, 1957). A still useful overview of the ancient
646   Christopher Shields

credit the bewildering contention, common to antiquity and modernity, that Aristotle
never developed to the point of grasping the rudiments of Platonic philosophy: “In the
first place,” says Burnet of Aristotle, “it is certain that he never understood the teaching
of the head of the Academy.”3 This is anything but certain. On the contrary, it is not
true and hence not even knowable; nor, indeed, is it even remotely credible. Rather, as
we should expect in the case of two philosophical geniuses interacting often with one
another for two decades,4 it is entirely likely that each understood much of the other,
that each benefited from the criticisms of the other, and that, consequently, each inevita-
bly learned much from the other.5
Exactly what they might have learned from each other is hard to say with anything
more than the confidence of informed conjecture. What we can do with a reasonable
assurance is to observe some obvious points of contact in the surviving writings of Plato
and Aristotle. Although he is never mentioned by name in Plato’s dialogues,6 Aristotle
offers important data concerning Platonic philosophy, mainly, though not exclusively,
through his characterizations and criticisms of Plato. For Aristotle discusses Plato and
his views frequently and in illuminating ways in his surviving corpus,7 and then again,
more continuously and fruitfully in his Peri Ideôn, or On Forms, which survives in a rea-
sonably intact version as a close paraphrase reproduced in a commentary on Aristotle’s
Metaphysics by Alexander of Aphrodisias, written in the late second or early third
century a.d.8 Aristotle’s criticisms are sometimes harsh and sometimes mild, and they are

accounts of Aristotle’s life in the Academy may be found in G. Grote, Aristotle (London, 1880), 1–26.
A much more sophisticated and engaging treatment of Aristotle as moving metaphysically closer to Plato
as he matures philosophically is offered by G. E. L. Owen, “The Platonism of Aristotle,” Proceedings of
the British Academy 51 (1966), 125–50. Although he enters many appropriate caveats and cautions,
Owen concludes: “It seems now possible to trace [Aristotle’s] progress from sharp and rather schematic
criticism of Plato to an avowed sympathy with Plato’s general metaphysical programme” (150).
3 J. Burnet, Platonism (Berkeley: University of California Press, 1928), 56. For another sort of dim
view of Aristotle’s interaction with Plato, see Cherniss (1944), critically discussed by Gerson (2014).
4 Their time together during these two decades would not have been uninterrupted. For a succinct
assessment of the evidence for Plato’s life and activities during the period in which Aristotle was a member
of the Academy, see Deborah Nails, The People of Plato (Indianapolis, Ind.: Hackett, 2002), 243–50.
5 This would explain Aristotle’s obvious reverence for Plato, whom he characterizes as “a man whom
the wicked have no place to praise: he alone, unsurpassed among mortals, has shown clearly by his own
life and by the pursuits of his writings that a man becomes happy and good simultaneously” (Frag. 650
R3, Frag. 673 R3, Olympiodorus, Commentarius in Gorgiam 41.9).
6 The Aristotle mentioned in the Parmenides is the son of Timocrates of Thorae (127a2, 136e7, 137c2;
cf. Seventh Letter 324b–d), who was to become one of the Thirty, not our Aristotle, son of Nicomachus,
the philosopher and member of Plato’s Academy.
7 Aristotle mentions Plato 54 times in his extant writings. He also mentions Socrates 143 times,
where very often—though the matter is permanently complicated—we may reliably treat his references
to Socrates as representations of authentic Platonic views. He also mentions by name 14 of Plato’s
dialogues, some such as the Apology and Euthydemus only once or twice, but others much more
frequently. Most often mentioned is the Timaeus, with 18.
8 The authorship of the Peri Ideôn is sometimes disputed; and even those who accept it as having a
genuinely Aristotelian provenance differ among themselves about the degree of closeness of the
paraphrase given by Alexander to an autograph by Aristotle. The fullest and most illuminating
treatment of this work is G. Fine, On Ideas (Oxford: Clarendon Press, 1993). Fine argues persuasively,
 Plato and Aristotle in the Academy   647

sometimes astute and other times curiously underdeveloped. In some cases, they are
simply obscure. Moreover, they are often not what they first seem. Indeed, as I argue
here, the process of becoming clear about some of Aristotle’s criticisms of Plato helps us
appreciate not only their real force but also the strength of Plato’s resilience in the face of
them. In this way, more than any other, it is possible to learn about Platonic philosophy
by studying Aristotle’s criticisms of Plato. After all, we do not have to suppose that
Aristotle is always right, is always fair, or is always disinterested, to regard him as a source
sufficiently familiar with Plato’s ideas to represent them in an informed manner. They
spring from the same tradition, and concern themselves with similar issues, even if they
pursue these issues in different idioms and with competing methods and conclusions.
Some of the most obvious points of contact between Plato and Aristotle are easy to
identify, and they cover the full range of the philosophical topics they each engage indi-
vidually. The most prominent criticisms are these:

1. Aristotle raises doubts about and rejects aspects of Plato’s theory of Forms.
2. Aristotle dismisses Plato’s soul-body dualism.
3. Aristotle expresses severe reservations about the tenability of Plato’s political
philosophy.9

Less obvious and more consequential, if also inevitably more obscure and less tractable,
are some methodological points of contact. We find in this area that:

4. Aristotle is dubious of Plato’s habitual and entrenched univocity assumption, to

the effect that core philosophical notions admit of single, non-disjunctive essence-
specifying accounts.10
5. Aristotle rejects Plato’s method of division as a technique for arriving at essence-
specifying definitions, whether univocal or not.11

with some caution, that the Peri Ideôn was indeed written by Aristotle, probably near the end of his
time in the Academy, and with the theory of Forms as it is advanced in Plato’s middle period as his
intended target.
9 As a general point, whenever we speak of Aristotle as criticizing Plato, we should always be
mindful of the implicit rider “in one phase of his development,” as applied to both philosophers. Thus,
more fully: “In one phase of his development, Aristotle is critical of Plato’s conception of the soul, in
one phase of his development regarding the nature of the soul.” A serviceable overview of some of the
main points of disagreement between Plato and Aristotle can be found in W. D. Ross, “The
Development of Aristotle’s Thought,” in I. Düring and G. E. L. Owen (eds.), Aristotle and Plato in the
Mid-Fourth Century (Göteborg: Elanders, 1960), 1–18.
10 For a treatment of Aristotle’s attitude toward univocity in both Plato’s and his own thought, see
C. Shields, Order in Multiplicity (Oxford: Clarendon Press, 1999). For a more elementary treatment of
the same topic, see Shields, “Learning about Plato from Aristotle,” in H. Benson (ed.), The Blackwell
Guide to Plato (Oxford: Blackwell, 2006), 403–41.
11 The subject of Aristotle’s criticisms of Platonic division are intriguingly complicated by the fact
that Aristotle is himself an unrepentant practitioner of the method. The topic thus provides fertile
ground for inquiries into Aristotle’s philosophical relationship to Plato. For good approaches to this
subject, see W. Cavini, “Naming and Argument: Diaretic Logic in Plato’s Statesman,” in C. Rowe (ed.),
648   Christopher Shields

In an effort to come to appreciate some otherwise easily overlooked problems in

Plato’s theory of Forms, problems whose interest and subtlety point to some engaging
aspects of their philosophical interaction, we will consider just one of Aristotle’s criti-
cisms of Plato, one that seems initially intended to land Plato in a straightforward and
inescapable contradiction. In the process of determining the actual force of Aristotle’s
criticism, we shall see first that it proves to be less straightforward than it initially
appears and then also that Plato need hardly be bowed before it. By understanding
Aristotle’s criticisms of his teacher in the context of their mutual association in the
Academy, we will thus also come to appreciate that Aristotle has much of consequence
to teach us about Plato.

2. Aristotle’s Criticisms
of Platonic Forms

Aristotle has no shortage of complaints about Plato’s Forms:

1. They are causally inert and so cannot explain change or generation (Meta. 991a8,
1033b26–28).
2. Postulating Forms offends theoretical economy (Phys. 259a8).
3. Forms, if ever they existed, would be epistemologically otiose (Meta. 991a12–14).
4. Introducing Forms as paradigms is empty metaphor (Meta. 991a20–23).
5. Forms cannot be essences if they are separated, since essences are intrinsic
­features of things (Meta. 991b1).
6. Forms are irrelevant to human conduct and so must be set aside from inquiries
into ethical virtue (EN 1096b32–4).

At his most caustic, Aristotle recommends a “good-bye to the Forms,” since “they are
jibber-jabber and even if they do exist they are wholly irrelevant” (APo. 83a32–34).
Different considerations motivate these different complaints, some more and some less
perspicuous, and some more and some less compelling. In general, all of these com-
plaints at least admit of rejoinders, in the sense that none of them purports to implicate
Plato in any immediate contradiction.
From this perspective, another one of Aristotle’s complaints evidently takes on a
special significance, since, understood one way, if cogent it entails that a feature of
Forms clearly accepted by Plato, their separation, results in a special absurdity, or even a
straightforward contradiction, to which there is no possible response beyond immediate
capitulation.12 In Metaphysics M 9, Aristotle complains that the successors of Socrates

Reading the Statesman (Sankt Augustin: Akademica Verlag, 1995), 123–38, and P. Pellegrin, “Division et
syllogisme chez Aristote,” Revue Philosophique de la France et de l’Etranger 171 (1981), 169–87.
12 This is, for example, how J. Annas understands the criticism: “The whole argument is designed
not, as before, to describe the theory of Forms, but rather to subject it to lethal criticism, by showing
 Plato and Aristotle in the Academy   649

went astray when they separated universals from particulars (Meta. 1086a31–b14). As
Aristotle recounts the situation, Socrates had first provided the impetus for seeking
the universal features shared by distinct particulars because he wanted to ascertain their
common nature: philosophers seek knowledge of what is captured in an essence-
specifying definition, since an essence-specifying definition states what holds univer-
sally and necessarily in its domain of investigation. If we wish to know what piety is, for
instance, then we isolate what is common to all and only instances of piety, uncover the
presence of that shared feature which makes all pious actions pious, and then put this
feature on display for the benefit of the discerning mind engaged in the project of philo-
sophical inquiry. Since what is laid bare in such an inquiry must be perfectly general,
this feature must also be something universal. So, for this reason, says Aristotle, Socrates
was right to attend to the universal, and commends him for doing so. For, “without the
universal, it is not possible to attain knowledge” (Meta. 1086b5–6). Socrates also receives
high marks from Aristotle for his admirable intellectual restraint, a virtue lacking, again
according to Aristotle, in his immediate successor. Though he sought adequate defini-
tions, “Nevertheless, Socrates surely never separated them from particulars; and in not
separating them, he thought rightly” (Meta. 1086b3–5). That he thought rightly, Aristotle
insists, can be appreciated by observing how those who do separate universals from
­particulars, the Platonists, go awry (Meta. 1086b5).
Aristotle thus suggests that intolerable results follow from the separation of Forms,
results evidently not attendant upon the mere postulation of universals, even if one
thinks of those universals as universal forms. So, from his perspective, there is nothing
wrong with the bare existence of universal forms: after all, Aristotle not only commends
Socrates in a general way but approves of his epistemic motives for accepting universals,
to the point of offering a highly technical and rigorous theory of scientific taxonomy and
inference of his own in which universals play a prominent and indispensable role.13
Indeed, the language of universals (ta katholou) belongs to Aristotle, but to neither
Socrates nor Plato.14 Moreover, in other contexts, Aristotle is altogether comfortable
with the existence of “common things” (koina), even when his primary dialectical
­purpose is precisely a refutation of Plato’s theory of Forms. Common things (koina) are
themselves universals, shared by many particulars. It is thus evidently not the universality
of Forms which earns Aristotle’s scorn.15 It is, rather, as he makes clear, separation and
its results (erga) which render Plato’s theory intolerable.16

that its very formulation involves a contradiction” (Aristotle’s Metaphysics Books M and N [Oxford:
Clarendon Press, 1976], 188).
13 See S. Mansion, Le Jugement d’Existence chez Aristote (Louvain: Éditions de l’Institut Supérieur de
Philosophie, 1946).
14 Though one does see clear precursors in Plato, for example, Meno 77a6.
15 I agree with M. Frede and G. Patzig, Aristoteles, “Metaphysik Z”: Text, Übersetzung und
Kommentar (Munich: Beck, 1988), vol. 1, 50–51, that attempts to distinguish koina and universals fail.
Though neither universals nor koina need to be regarded as separate, both are shared in the sense
sufficient for universality.
16 Indeed, in setting out the aporias of Meta. B, Aristotle suggests that the worries we have about
universals are really only worries about separated universals, as Platonic Forms are understood to be.
See 997b3–12, 999a19–22, 999b17–24.
650   Christopher Shields

What, though, is the absurdity Aristotle locates in separation of Forms? In Metaphysics

M 9, he foregrounds a single unacceptable consequence, though he does not argue for
this result in any detail, preferring instead to allude to fuller discussion elsewhere.17 That
result is this: by separating universals, the Platonists end up swallowing the view that
“universals and particulars are practically the same natures” (Meta. 1086b10–11), a
­formulation which softens Aristotle’s report earlier in the chapter that the Platonists “at
the same time make the Ideas, as substances, universal, and then again make them, as
­separate, belong to the class of particulars” (Meta. 1086a32–4). The harsher and the
softer complaints are related but nonetheless importantly distinct. Aristotle’s first com-
plaint appears to be categorial in character: Forms, he supposes, belong to two different
categories of being—namely, universals and particulars. This he supposes to be plainly
unacceptable. Unfortunately, Aristotle does not explain what is so bad about this alleged
consequence, perhaps because he takes as an obvious absurdity any commitment to the
existence of something which is at the same time both universal and particular. Still less
clear is why there should be a problem with the final thought of the chapter, that “univer-
sals and particulars are practically the same natures.” Indeed, it is not entirely clear even
what Aristotle means by this complaint.
Accordingly, it is not entirely clear why Plato should be unsettled by these results; nor
yet is it clear why, if the results are as problematic as Aristotle supposes, Plato is in fact
liable to them. So, there are at least four sorts of questions about Aristotle’s brief against
Plato’s theory of Forms in this passage. First, how does Aristotle argue—if he does so
argue—that Plato comes to be committed to the conclusion that Forms are both univer-
sals and particulars, or that, for Plato, universals and particulars are practically the same
natures? Second, what are the precise relations between these two claims? Does, for
example, one collapse into the other? Or is one meant to follow from the other? Or are
they simply distinct but related complaints? Third, however that may be, are there clear
expressions of the elements of Aristotle’s argument in the Platonic dialogues themselves,
or are we rather to understand Aristotle’s criticisms as rooted in principles not directly
endorsed by Plato (or not endorsed in the dialogues), so that he may or may not be
­constrained, upon reflection, to yield to Aristotle’s conclusions? Finally, there is a
straightforwardly philosophical question about the force of Aristotle’s complaints,
whatever the Platonic provenance of the view he means to assail: What is wrong with
allowing that Forms are both universals and particulars, or that universals and particu-
lars are practically the same natures? Again, pursuant to this last question is the prior
question of what Aristotle’s final complaint might mean.
The straightforwardly philosophical question should remain open for the time
being, at least in the sense that there might be any of a number of different things wrong
with the results Aristotle foists on Plato. Moreover, at this early stage of our inquiry, it
should be regarded as open in another direction as well: it may be a perfectly acceptable
consequence, one to which Plato can remain appropriately indifferent. That is, with
suitable understandings of the notions of universality and particularity, we may not be

17 For a brief presentation of the most prominent alternatives, see notes 22 and 23 of this chapter.
 Plato and Aristotle in the Academy   651

constrained to think of these categories as mutually exclusive.18 Perhaps when the

­relevant distinction is properly understood, Forms can be both universal and particular.
Similarly, perhaps we need not be troubled that universals and particulars are practically
the same natures. One might say, for instance, that being good and being right are nearly
the same sorts of attributes, even while insisting that they are nevertheless importantly
distinct features of our actions.
In the end, after sorting through these matters, it will turn out that Aristotle does
have a point on his side: there is something problematic in the way Plato understands
Forms vis-à-vis universals and particulars. Still, Aristotle is wrong if he thinks there is
an immediate argument to a self-contradictory conclusion to the effect that, necessarily,
every Form is at the same time both a universal and a particular—where these are
understood to be mutually exhaustive and exclusive categories of being. The important
objection present in Aristotle’s criticism will prove to have a rather different character.
His most telling point turns out to be both more interesting and less lethal than his initial
categorial concern—if, that is, we may understand his final remarks in Metaphysics xiii 9
as his final thought on the matter. On this alternative approach, Aristotle’s contention
turns out to be non-categorial: he argues that Plato’s conception of participation
as imitation leaves some things about the unity of sense particulars unexplained,
whether that unity is taken synchronically or diachronically. On this understanding, the
upshot of the criticism hardly spells the end to Plato’s theory of Forms. Instead, insofar
as they are legitimate, Aristotle’s concerns only force Plato, or a Platonist, to augment
and develop a theory that is otherwise overstressed by its own explanatory ambitions.
To reach this conclusion, we may proceed in two stages. First, we explore the precise
concern Aristotle has regarding particulars and universals in connection with Platonic
Forms. To this end we consider two versions of a standard categorial interpretation of
Aristotle’s criticism of Plato, both initially plausible as interpretations, but neither ultimately
at all forceful in terms of their critical purport. So, if he had only so much to say against
Plato, Aristotle’s criticisms would be wholly uncompelling. The best Aristotle could
hope for would be a kind of unsatisfactory stalemate, a result which would surely fall far
short of the sort of refutation of Plato’s theory of Forms he is usually thought to offer.
Once that is established, we turn to the second phase of our discussion. Upon closer
examination, it turns out that Aristotle also has another, deeper problem in view when
raising this objection to Plato. This deeper problem, unlike the one framed by the cate-
gorial version of his objection, locates Aristotle’s complaint not in some immediate category
error on Plato’s part but, rather, in the sorts of explanatory work Aristotle justifiably
expects Forms to discharge. While this deeper problem does have more traction, it
hardly constitutes a crippling objection to Plato’s theory of Forms. So, in the end, what

18 Thus, P. Strawson observes that relative to a logical subject criterion individuality might be
categorially unconstrained: “So anything whatever can appear as a logical subject, an individual. If we
define “being an individual” as “being able to appear as an individual,” then anything whatever is an
individual. So we have an endless variety of categories of individual other than particular—categories
indicated by such words as ‘quality,’ ‘property,’ ‘characteristic’ ” (Individuals [London: Methuen, 1959],
227). For Aristotle’s language of particularity, see Phys. 189b30, 195a32, PA 639b6, and DI 17a39–40.
652   Christopher Shields

remains viable in Aristotle’s anti-Platonic concern about the universality and particularity
of Forms does not reduce Plato’s theory to a category-based incompatibility. If this
is correct, then we may conclude that although Aristotle has put his finger on the pulse
of a genuine problem regarding the nature of Forms, his criticism fails to provide a
­neutral third party with any reason to relinquish a commitment to Forms—if at any rate
such a commitment is otherwise independently motivated. Accordingly, Aristotle’s crit-
icism, though pertinent, points in the end only to the need for further philosophical
work from Plato.

3. Aristotle’s Complaints
in Metaphysics M 9

Aristotle’s general critical argument begins rather surprisingly, by locating in Plato a

certain lack of philosophical imagination:

They at the same time make the Ideas, as substances, universals, and again, as
separate, belong as well to the class of particulars.19 Separation is the cause of the
resultant difficulties regarding Ideas. These things were shown to be problematic
earlier, because this cannot be. The reason why those who say that substances are
universal conjoin these things into the same is that they made substances not the
same as perceptibles. They thought that in the case of sensibles, particulars are in
flux and that none of them remains, whereas they thought of universals as beyond
(para) these and as being something else. Just as we said earlier, this is something
Socrates set in motion, because of his definitions, but even so he did not himself
separate them from particulars. And he thought rightly in not separating them. This
is clear from the results: for while without the universal it is not possible to attain
knowledge, separation is the cause of the difficulties which accrue concerning the
Ideas. They [Socrates’ successors] regarded it as necessary, if there are going to be
substances beyond (para) the sensible and flowing substances, that they be separate;
but they did not have others and instead selected the things predicated universally,
with the result that universals and particulars were practically the same sorts of
natures. (Meta. 1086a32–b11)

The argument here is not especially complex, though the ultimate conclusion seems
­surprising insofar as it weakens the bolder-sounding claim with which Aristotle begins.
He says first that the Platonists “make the Ideas, as substances, universals, and again, as

19 There is a textual problem here. I read hôs ousias with the ms. at 1086a33, rejecting Jaeger’s
seclusion, which is also accepted by W. D. Ross, Aristotle, Metaphysics (Oxford: Oxford University
Press, 1924), vol. 2, 462. Ross thinks it wrong to stress the substantiality of Forms in this context; but it
is appropriate for Aristotle to mention a connection between universality and substantiality, since (1)
he himself accepts such a connection in the Categories, and (2) he is here recounting what he takes to
be Plato’s motives in regarding Forms as universal.
 Plato and Aristotle in the Academy   653

separate, belong as well to the class of particulars” (Meta. 1086a32–4), only to conclude
ultimately, as we have seen, that they were thus saddled with the view that “universals
and particulars were practically the same sorts of natures” (Meta. 1086b10–11). The point
is not just that he says that they were practically (the normal meaning of schedon in this
context) the same sorts of natures but that he demurs from asserting directly that the
Ideas are both universals and particulars. Rather, he offers the somewhat opaque conclusion
that things spoken of universally and particularly are, for the Platonists, effectively the
same sorts of natures. That final conclusion is, in comparison with the clear and strident
claim with which he opens, both guarded and obscure.
So, there is a question about Aristotle’s ultimate objective in this passage. At first, his
argument seems to trace the particularity of Forms to their being substances. Since
­substances are separate, and separate things are particulars, it follows that Forms,
regarded as universals by the Platonists, are also particulars. In terms of their reasoning
process, the Platonists, as characterized by this approach, were simply at a loss: they
could not identify anything other than their own Forms to play the role of the objects of
knowledge. Since there do indeed need to be such objects, Plato was generally well moti-
vated, though, according to this line of thought, insufficiently resourceful. After rightly
recognizing the need for universal objects of knowledge, but seeing no alternatives to
their own separated Forms—which as separated must already be substances—the
Platonists end up committed to Forms as objects of knowledge as well as substances.
This, in turn, implicates them in a contradiction: substances are particular, and objects
of knowledge are universal.
Taken this way, laid out schematically, then, the argument of Metaphysics M 9 is this:

1. If knowledge is to be possible, there must be stable objects of knowledge, immune

to the flux suffered by sensibles.
2. Knowledge is possible.
3. So, there must be supersensible stable objects of knowledge, immune to the flux
suffered by sensibles.
4. As supersensible and stable, such objects might be: (i) in re, or (ii) ante rem.
5. If they are ante rem, the result will be that Forms are both universals and
particulars.
6. The Platonists, not recognizing 4(i) as an option, endorse ante rem objects of
knowledge.
7. If they both exist ante rem and are also objects of knowledge, then Platonic Forms
are both universals and particulars.
8. Hence, Platonic Forms are both universals and particulars.

This approach to Aristotle’s generative reconstruction of Plato’s motivations for

accepting separated Forms proceeds by taking seriously the diagnostic suggestion,
whether fair or unfair, that the Platonists saw no alternative to ante rem realism,
­evidently because they did not appreciate that universals might exist in rebus (Meta.
1087b9–10).
654   Christopher Shields

In reporting Plato’s motivations, Aristotle intimates that he himself accepts premises

1 and 2, and what they entail, which is premise 3, that there are supersensible stable
objects of knowledge. He parts company with Plato only at premise 4: he implies that he
recognizes the alternatives highlighted in premise 4, the very alternatives overlooked by
the insufficiently resourceful Platonists. According to his own self-presentation, on this
interpretation, Aristotle thereby positions himself to avoid the ensuing results, so that,
unlike the Platonists, he can offer forms which are universals but not separate and so
preserve the thought that universals and particulars are of radically different natures.
Presumably, his unstated conviction is that in rebus universals are dependent entities,
effectively his own preferred “common things” (koina), separate at most in thought or
definition (logô(i)), and so not separate simpliciter (haplôs). Since separation simplicter
belongs only to particulars, Aristotle avoids the consequences sketched for the Platonist.
This at any rate makes sense of Aristotle’s pointed insistence that the Platonists have
separation and its results (erga) to thank for their woes.
Several problems emerge when we consider the merits of this argument. In the first
instance, it is a bit difficult to take seriously the suggestion that Aristotle accuses the
Platonists of failing to recognize the prospect of in rebus universals. After all, Aristotle
himself, in this very passage, locates in Socrates precisely the wanted remedy: unseparated
common definitions which serve as the objects of knowledge. Still, perhaps one can
say on Aristotle’s behalf that while it is true that Socrates never separated his universals,
he never expressly refrained from separating them either. That is, it is compatible with
Socrates not having separated universals that he never so much as recognized as a rele-
vant alternative the nonseparation of Forms. For, as Aristotle himself suggests, Socrates
did not busy himself much with metaphysical questions; such inattention could well
result in his never considering either the general question of the nature of his universal
definitions or the specific question regarding their separation. When he then took his
metaphysical turn, Plato opted for separation without having first entertained the
­relevant alternatives. He saw that Forms could not be identical with sense particulars,
accepted their existence as immune to sensible flux and so in that way superior to sensi-
bles, and consequently invested them with ontological primacy and its attendant sepa-
ration. So, in principle, Aristotle might fairly accuse the Platonists of failing to discern
the relevant alternatives, while in the same breath crediting Socrates with having
refrained from going down the Platonist path.
Even so, there is a more serious problem, pertaining to premise 5, the premise that if
they are ante rem, Forms will be both universals and particulars. Here, in fact, there are
two distinct puzzles. The first, already mooted, concerns the force of the consequent,
while the second concerns the conceptual connection envisaged between being ante
rem and having a particular-like nature.
As for the first problem, the conclusion of Aristotle’s argument does not seem
fatal to the Platonist, since for all that has been established so far, something might be
both a universal and a particular.20 Aristotle must be assuming some views about the

20 Cf. note 18 of this chapter.

 Plato and Aristotle in the Academy   655

natures of particulars and universals and about the relation between separation and
­particularity—views which the Platonist may or may not be constrained to accept.
Perhaps their natures are not mutually exclusive, and perhaps something may be
­separate without being particular.
This leads into the second, more consequential form of oddness in the passage.
Aristotle does not conclude finally that Ideas will be both particulars and universals or
even that universals and particulars will have (echein) the same natures but, rather, that
they will be (einai) the same natures (tas autas phuseis einai; Meta. 1087b11). This is not
the point Aristotle is usually taken to be making in this passage;21 and it is moreover a
point which is harder to explicate than the simpler suggestion that Ideas are both uni-
versals and particulars. For the argument is not simply that since separation belongs to
substances and particularity to separate things, the Platonists, having made universal
Forms substances, must now also accept their particularity. Rather, the argument holds
that finding no other stable objects of knowledge than their own Forms, the Platonists
end up accepting as real and separate things those very Forms and that consequently,
according to Aristotle, they also end up regarding universals and particulars as being
effectively the same natures. This Aristotle supposes to be a problem for the Platonists.

4. The Source of Aristotle’s

Complaint: Two Questionable
Contentions

What then, precisely, is the final problem? Aristotle does not say but instead prefers to
refer back to an earlier discussion, which he fails to identify. Unfortunately, Aristotle’s
backward reference is a bit hard to pin down. The three most relevant passages, all
­proposed as the appropriate reference by at least one commentator or other,22 would

21 The Revised Oxford Translation skirts this issue, leaving a misleading impression: “So that it followed
that universals and individuals were almost the same sort of thing.” Annas has much the same: “With the
consequence that universals and particulars were almost the same sort of thing” (Metaphysics, 115).
22 Ross rightly observes that diêporêtai in Meta. 1086a34 “suggests a reference to Book B”
(Metaphysics, 462). Even so, it is difficult to know which argument in that book Aristotle might intend.
It certainly also remains open that Aristotle has another aporetic passage in mind, especially one
raising difficulties for the Platonists. J. Owens, The Doctrine of Being in the Aristotelian Metaphysics,
3rd ed. (Toronto: Pontifical Institute of Medieval Studies, 1978), agrees that the language suggests a
reference to Metaphysics B, but adds that it should not be understood as relying on Z 13–15. He thinks,
in fact, that “M 9–10 . . . presumes no knowledge in the ‘hearers’ of M 9, and does not positively refer to
ZH” (426). He, too, thinks that it is likely a reference to the last aporia of Metaphysics B, but focuses
exclusively on the epistemic horn that if principles are not universal they will not be knowable
(1003a15–17). That allusion would be peculiar, however, since that aporia holds quite generally and with
no special reference to those who postulate Forms. It is true that the next chapter, M 10, does allude to
the final aporia of Metaphysics B, at 1086b14–20. Owens does not make clear, however, why these are to
be treated as so closely linked.
656   Christopher Shields

seem to be Metaphysics B 6, where the fifteenth and final aporia of metaphysics is

retailed; Metaphysics M 4, where Aristotle attacks Platonic Forms in ways parallel to
Metaphysics M 9; and, finally, Metaphysics Z 13–15, where Aristotle sketches some of the
reasons that nothing spoken of universally is a substance and then draws out the conse-
quences of those reasons for the theory of Forms.
Of these, Metaphysics M 4 is the least plausible, since it contains no argument which is
even indirectly relevant to the claim that Platonic Forms are both universals and partic-
ulars or to universals and particulars being practically the same natures.23
Metaphysics B 6, by contrast, does seem to bear on this issue, at least orthogonally. There,
in the fifteenth aporia, Aristotle argues dilemmically, first, that no universal is a source
or principle (archê) (Meta. 1003a7–9) and then, second, that every principle (archê)
must be a universal (Meta. 1003a13–17). The argument for the first horn has it that:

1. Universals are common (koina) (1003a8–9).

2. Nothing common signifies some this (or, more loosely, any particular thing—­
tode ti; 1003a8–9).
3. So, no universal signifies any particular thing (tode ti).
4. Substance (ousia) signifies a particular thing (tode ti; 1003a9).
5. So, no universal is a substance (ousia).
6. The principles of things (archai) are substances (ousiai) (assumption).
7. So, no principle (archê) is universal.

This argument trades on some putative conceptual connections among substantiality,

particularity, and the nature of principles (archai).
The argument for the second horn, by contrast, finds its impetus in some epistemic
theses held in common by both Plato and Aristotle: knowledge (in the sense of epistêmê)
is in all cases of universals (1003a14–15; cf. Meta. 998a24–26, 998b4–6, 1013a14–17; APo.
87b28–88a17; Phys. 194b16–20); we do have knowledge of principles (archai)—or at least
such knowledge is possible; so, principles (archai) are universal. Since both horns can-
not be correct, something must give way.
Now, it is a matter of some dispute as to what gives way, in part because it is also a mat-
ter of dispute as to where Aristotle addresses the fifteenth aporia later in the Metaphysics,
if anywhere. This need not concern us here. Rather, for our purposes, it is important
only to observe that in mounting his anti-Platonic contention in the opening of our
­passage from Metaphysics M 9, Aristotle alludes to both horns of this dilemma.
Forms, as objects of knowledge, must be universal, as the second horn contends.

23 Annas, Metaphysics, along with H. Bonitz, Aristotelis Metaphysica, 2 vols. (Bonn:

Marcus, 1848, 1849), understands Aristotle’s complaint as relying on M 4. Annas is right that M 4
provides a backdrop of sorts to M 9; but it does not contain any argument with the conclusion to which
Aristotle here alludes: viz., that Forms are both universals and particulars. Ross is in my view right,
then, to suggest, as against Bonitz (and so prospectively as against Annas): “There is nothing in M 4,
5 . . . that quite suits the reference” (Metaphysics, vol. 2, 462).
 Plato and Aristotle in the Academy   657

But why must they be particular? Tracing out the pattern of inference in the first horn,
the argument seems to be:

1. Forms are principles (archai).

2. No principle (archê) is universal.
3. Whatever is not universal is particular.
4. So Forms must be particular.

Here, then, we have a rather different point than the direct claim, regularly and rightly
ascribed to Aristotle, that if Platonic Forms are substance (ousiai), they must be
particulars.24
This is a welcome result for Plato, since it is doubtful that this sort of anti-Platonic
argument will eventuate in anything more than an immediate and unilluminating stale-
mate. If Aristotle insists that all substances (ousiai) are particulars and Plato responds
that he rejects the motivating assumption of the inference, then no progress will have
been made in either direction. Rather, the case will have been aptly characterized by
Grote who, writing in the late nineteenth century, observed that such objections “are
founded upon Aristotle’s point of view, and would have failed to convince Plato.”25
Moreover, on Plato’s side is the ad hominem retort that Aristotle himself, at least in the
Categories, recognizes some substances, secondary substances, as universal, and even
allows degrees among substances by noting that the species is more a substance than the
genus in virtue of its being closer to the primary substance, better known than the genus,
and prior in predication to the genus, since the genus is predicated of the species, while
the species is not reciprocally predicated of the genus (Cat. 2a14–15, 2b8–14, 2b20–21).
Hence, at least at some periods in his career, Aristotle sees some point in according
the status of substance to at least these sorts of universals.26 So, there seems to be
no direct or easy inference from being substantial to being particular. At any rate, Plato
is hardly constrained to accept such an inference without some significant ancillary
argumentation.
Aristotle’s appeal to the status of principles (archai) may point to the needed
­argumentation, though the final aporia of Metaphysics B fails to provide it. Consequently,
though it is relevant to Aristotle’s contention that Forms are both universals and partic-
ulars, the final aporia of Metaphysics B fails to furnish a fully satisfactory back-reference
for Metaphysics M 9.

24 The various connections between form, separability, substance, and particularity in Plato and
Aristotle are well discussed by G. Fine, “Plato and Aristotle on Form and Substance,” Proceedings of the
Cambridge Philological Society (1983), 23–47.
25 Grote, Aristotle, 560.
26 This assumes that the species and genus of Cat. 5 are regarded not as abstract particulars, as sets
or aggregations of some kind, but as universals. This assumption is warranted by Aristotle’s treating the
species and genus as predicables (Cat. 2b32–35), together with his suggestion that whatever is by nature
predicated of many things is a universal (DI 17a38–40).
658   Christopher Shields

We may perhaps therefore turn to a second possible source for a fuller elaboration of
Aristotle’s complaint. Metaphysics Z 13–15 trace some of the consequences of Aristotle’s
own inquiry into substance (ousia) for Plato’s theory of Forms. Surprisingly, Z 15
­contains a direct and unargued ascription of both parts of the controversial thesis that
Forms are universals and particulars to the Platonists. First, Aristotle argues that no
Form can be defined, since “Form belongs to the class of particulars, as they say (hôs
phasi), and is separate” (Meta. 1040a8–9). Later in the same chapter Aristotle returns to
the epistemic basis for treating Forms as universals but adds a twist by arguing that
Forms must be universals, on the grounds that “every Idea is such as to be participated in
(methektê)” (Meta. 1040a26–27; cf. Meta. 990b28, 1079a25). This shows, according to
Aristotle, that Ideas cannot consist of other Ideas, since in that case that from which
Ideas would be constituted (presumably other Ideas) would equally need to be predi-
cated, under pains of their being simple, and so unknowable, in addition to their being
not such as to be participated in. If that is correct, then Ideas, as individuals, do not
admit of definitions.
Here, then, we read a rather different basis for the tension that Aristotle locates in
Plato’s views: Forms are accepted by Plato as the sorts of things which can be partici-
pated in, and so must be universals, and yet, “as they say,” Forms are also particulars.
Yet it is hard to know where they say that Forms are particulars. Plato does on occasion
refer to a Form as hekaston (e.g., Phaedo 78d3), where the language parallels in a non-
technical way Aristotle’s technical terminology for particulars, ta kath’ hekasta (Meta.
1086a34). Plato also on occasion uses singular referring terms when mentioning Forms
(Crat. 389d6; Parm. 133d8; Phd. 75b1, 75d2; Rep. 507b7, 597c9; Theaet. 146e9; Phil. 62a2).
Still, there is a perfectly unobjectionable way in which Forms are particulars, whether or
not they are also universals.
Let us call a deflationary particular anything which is a determinate subject of
predication.27 On that score, it is entirely possible to say, for example, that Beauty
Itself is a form, or that it is an abstract object, or that it is the most loved Form, and so
on. In these cases we are talking about that particular Form, that thing which is Beauty
Itself. Aristotle’s point cannot simply be that Forms are deflationary particulars as well as
universals, for there is nothing objectionable about their being both. Rather, his point
must be that Forms are robust particulars, those he contrasts with universals as early as
De Interpretatione 7. In that chapter, Aristotle clarifies his conception of the contrast by
explaining that “By universal I mean what is naturally predicated over (epi) more than
one thing, and by particular what is not” (DI 17a38–40; cf. Meta. 1039a1). Particulars thus
construed, robust particulars, cannot be over (epi) many things. Aristotle, at least in this
passage, presents universals and particulars as mutually exclusive categories of being.
Unfortunately, in this context Aristotle characterizes robust particulars only negatively:
they are never over (epi) many things.

27 So, a deflationary particular is equivalent to what Strawson calls an individual in the sense of a
logical subject. See note 18 of this chapter.
 Plato and Aristotle in the Academy   659

We know that Plato, in many dialogues, treats Forms as “one over many” (hen epi tôn
pollôn). Why, though, must he also treat them as robust particulars? According to one
common thought, Plato’s mimeticism commits him to the particularity of Forms. That
train of thought is this. Plato subscribes to mimeticism, the view that a particular x is
eponymously F only by imitating the Form F-ness. So, for example, we call Helen beauti-
ful because she somehow imitates the Form Beauty Itself. In general, since nothing
could be F, except by copying something which is itself F, the Form F-ness must itself
be F. For instance, since nothing white could be white by copying something which was
not white, the original, the Form, must also be white. Thus, we typically find scholars
maintaining that Plato’s mimeticism commits him to Forms which exemplify the
­properties copied by the particulars named after them.28
The general idea of such scholars is that, without resemblance, copying is impossible.
Of interest here is the initially plausible assumption that Platonic mimeticism, the view
that a particular x is F by imitating the Form F-ness, by itself commits Plato to a Form of
self-predication. The assumption is something more than merely initially plausible,
however, only if we assume additionally that mimeticism requires a symmetrical same-
ness of property instantiation. If we say, for example, that Rabelais resembles Augustine
in having a big nose, then we evidently do suppose that there is an attribute, having a big
nose, had by both Augustine and Rabelais. Indeed, it would be odd to say that they
resembled one another in this way if we were also prepared to deny that Augustine had
a big nose. This may suggest, then, that resemblance requires symmetry (if Rabelais
resembles Augustine with respect to F-ness then Augustine resembles Rabelais with
respect to this same property) and that this symmetry is unpacked only by supposing
that any such symmetry is underwritten by shared property instantiation. Taken
together, these theses do not commit Plato to the self-exemplification of Forms, since it
does not yet follow that Forms are themselves the properties they have; but it does
commit him to the view that sense-particulars and Forms share the properties in virtue
of which eponymous naming relations become possible.
It is clear, however, that scholars who argue this way are, as a conceptual matter, wrong
about resemblance; and they are, moreover, wrong about mimeticism and resemblance
in Plato’s middle and late periods. Plato seems perfectly aware that the reciprocity

28 A typical example is W. Runciman, “Plato’s Parmenides,” in R. E. Allen (ed.), Studies in Plato’s

Metaphysics (London: Routledge and Kegan Paul, 1965), 158: “If whiteness is white (which must follow
if white objects are white by resembling it) then whiteness is one of the class of white objects.”
W. D. Ross, Plato’s Theory of Ideas (Oxford: Oxford University Press, 1953), 89, offers a similar
judgment, when commenting on Cornford’s and Taylor’s earlier suggestions that copying is not mere
resemblance. They had argued, in Ross’s characterization, “that the relation of copy of original is not
one of likeness, since if A is like B, B is like A, but if A is a copy of B, B is not a copy of A.” Ross
responds: “But this defense itself fails. Grant that the relation is not merely one of likeness; it still
involves likeness, and likeness between two things involves some Form, some character, that they have
in common.” Ross is right to contend that copying is not the same relation as resembling (the latter, but
not the former, is symmetrical) but wrong to insist that copying involves resemblance, where that is
understood as shared property exemplification.
660   Christopher Shields

restriction fails, as does any notion of resemblance given in terms of shared property
exemplification. So, for example, Republic iii 395c–d, Plato restricts the kinds of
­imitation permitted to the guardians to those which imitate actions befitting of their
­station. They can imitate those who are courageous, temperate, pious, and free; but they
must avoid imitating the degenerate and slavish, “lest from enjoying the imitation,
they come to enjoy the reality.” His worry is that “imitations practiced from youth
become part of the nature and settle into habits of gesture, voice, and thought.” The
implication here seems to be that it is possible to imitate someone courageous without
being courageous. Perhaps this is no surprise: Homer, says Plato, imitates Chryses,
though neither Homer nor any verse of the Iliad is an old man supplicating the Achaeans
(Rep. 393d). Poetic imitation requires non-narrative projection, to be sure; but this
can be accomplished without anything’s exemplifying the very property the object
­represented is represented as exemplifying (cf. Cratylus 423c–d).
This last point is worth stressing, since the crucial point about mimeticism in this
context is not simply that the original and copy will exemplify some of the same properties.
For every set of two things share some property or other. Rather, the point is that para-
digmatism does not require that paradigms be paradigms by exemplifying the properties
for which they serve as paradigms.29 Consequently, nothing in Plato’s mimeticism
requires that he treat Forms as robust particulars. Therefore, if he thinks that Forms are
particulars, Plato does not indicate that this is so simply by embracing paradigmatism.
Accordingly, nothing about Plato’s mimeticism forces him to accept the robust particu-
larity of Forms.
So far, then, if we suppose that Aristotle’s objections in Metaphysics M 9 ultimately
rely on these sorts of considerations, we must conclude that he has failed to achieve his
end. We have found instead, by tracing out two versions of the anti-Platonic argument of
Metaphysics M 9, only that we arrive at a stalemate twice over. On the first approach,
Aristotle contends that, according to Plato, Forms are substances (ousiai) and so partic-
ulars, but also that they are over many things (epi tôn pollôn) and so universals. As we
have seen, Plato has an obvious response, that not every substance (ousia) is a particular,
a retort made all the more ready in view of the fact that at least in some stages of his
career Aristotle himself endorses the existence of substances (ousiai) which are univer-
sals (Cat. 2a14–15, 2b8–14, 2b20–21). On a second version of the same basic approach,
Platonic paradigmatism is ultimately the culprit, since it is supposed to entail that Forms
are universals and also particulars, because though they are themselves properties,
Forms must also exemplify the properties which their eponymous sense particulars
copy—namely, the properties they are. Otherwise, the objector contends, Forms could
not be the paradigms that Plato holds them to be. This inference relies on a false prem-
ise: that shared property exemplification is necessary for mimeticism. It also seems to
assume that the only subject capable of exemplifying a property is a robust particular,

29 This is a point well expressed by W. Prior, “The Concept of Paradeigma in Plato’s Theory of
Forms,” Apeiron 17 (1983), 33–42.
 Plato and Aristotle in the Academy   661

which is equally objectionable. So, here, too, it is perfectly appropriate for Plato simply
to shrug off the Aristotelian barrage. Taken together, Aristotle’s criticisms, so understood,
are wholly ineffectual.

5. A Noncategorial Approach:
Forms as Archai

The shortcomings of these two versions of Aristotle’s criticism of Plato stem from a
­common source: they both understand Aristotle to be arguing that, as a categorial
­matter, any given Platonic Form must—impossibly—have the property of being both a
universal and a particular (“impossibly” because we are assuming robust particularity).
To be sure, Aristotle invites this sort of understanding when he suggests, for example,
that Forms are both suches and thises (Meta. 1038a34–1039a2), that the Third Man
results from treating something which signifies a quality or a relation as if it were an
individual (Soph. El. 178b36–179a10), or that the Platonists make Forms universals
even while treating them as belonging to the class of particulars (tôn kath hekaston,
Meta. 1086a33–34; cf. Meta. 1086b27). Still, in general, it seems hard to convict Plato of
holding explicitly or implicitly that the Forms are particulars in anything more than an
acceptably deflationary way. It equally seems difficult to understand mimeticism as
issuing in any kind of one-over-many hypothesis which may plausibly be construed
as entailing both universality and robust particularity for Forms. So, the categorial
approach leads to a dead end for Aristotle.
That dead end, however, does not convict Aristotle; it rather counsels rethinking the
genesis of his concern. In fact, if we focus on his preferred final language in Metaphysics
M 9, we recall that Aristotle does not ultimately draw the strong categorial conclusion
regularly ascribed to him. Returning again with greater care to the wording of the final
argument of Metaphysics M 9, we find a clue to a different direction of investigation. For
there at least Aristotle does not draw any direct categorial conclusion. Instead, he offers
a conclusion, muted in any case, for the thesis that universals and particulars are almost
the same sorts of natures. On closer inspection of Aristotle’s conclusion, we see directly
that this complaint, at least, is not that Forms will be universals and particulars but,
rather, on the contrary, that universals and particulars will be practically the same sorts
of natures.
In what sense will they be—not have—the same natures? In Physics ii 1, Aristotle
­identifies nature with “a principle (archê) and cause (aitia) of something’s being moved
and being at rest in that in which it belongs primarily, in itself, and not co-incidentally”
(Phys. 192b20–23). This he does in part because he distinguishes those things with
natures from those things which lack natures by arguing that everything which exists by
nature has an internal principle of motion and rest (archê tês kinêseôs kai staseôs; Phys.
662   Christopher Shields

192b13–14). Importantly, what it means to have a principle (archê) of motion is here

treated quite broadly, so that it encompasses not only locomotion but also growth and
diminution, along with alteration generally (ta men kata topon ta de kat’ auxêsin kai
phthisin, ta de kat’ alloiôsin; Phys. 192b14–15). So, if Aristotle expects a nature (phusis) to
be a principle (archê) and a principle to explain and ground a natural entity’s motion,
growth, and general alteration, it also follows that he looks to nature (phusis) to ground
and explain these same phenomena.30
This suggests a second, more promising sort of attack on Forms emanating from
Metaphysics M 9. On this approach, Aristotle need not rest with the unsustainable
­category-based refutation of Platonic Forms. Instead, he will offer a subtler, if less imme-
diately damaging objection. If we grant Plato mimeticism as a specification of participation
(methexis), then we have not compelled him to suppose that Forms, as paradeigmata,
must be robust particulars. Still, Plato is not yet out of the woods. For Aristotle is right to
insist that among the explananda of physics are the following: that there are unified enti-
ties which come into existence, grow, change, and die; that one organism may be like
another in many respects while nevertheless being numerically distinct from it; and,
generally, that individual organisms are unified systems of coordinated activity and
directionality. In brief, as students of nature and as metaphysicians, we need to discover
the principles (archai) of synchronic and diachronic unity.
If it turns out that Forms are virtually both universals and particulars as natures of
sensible substances, then if they are the principles (archai) of such substances, they will
fail to ground or explain the unity we observe in sensible substances. If Platonic Forms
must be universals, insofar as they are objects of knowledge and definition, then they
will explain, or help explain, what makes human beings be human beings—what it is that
Callias and Socrates have in common in virtue of which they are human. They will then
also explain what wisdom is—what all and only wise beings have in common in virtue of
which they qualify as wise. But then, if we understand that Alcibiades and Socrates are
not only wise but also both puckish and charming, while only one is musical and the
other is not, then we will need something beyond bare mimesis to explain why just these
mimetic tokens bundle together with those to constitute Alcibiades, and then again why
it is that he can remain numerically one and the same with the man whose youthful dark
hair has given way to stately blazing grey. In Aristotle’s terms, we need to know how
Forms, as universals, serve as the grounding principles (archai) of these facts. We need
to know, that is, why Platonic individuals are anything more than what Sellars aptly
termed “leaky bundles of abstract particulars,” mere congeries of particularized Form-
copies occurring together in indifferent regions of space.31
It will be by pushing discrete and evidently incompatible functional roles onto Forms
that Plato will end up with particulars and universals of virtually the same natures. If so
much is correct, then a better back reference for Aristotle’s underdeveloped suggestions

30 For an illuminating discussion of the character of principles (archai) in Metaphysics M and their
relation to Aristotle’s approach to change in the Physics, see Mueller (1987).
31 W. Sellars, “Raw Materials, Subjects and Substrata,” in E. McMullin (ed.), The Concept of Matter
(Notre Dame, Ind.: Notre Dame University Press, 1963), 263.
 Plato and Aristotle in the Academy   663

of Metaphysics M 9 will be not the last aporia but, rather, an earlier aporia of Metaphysics
B 4, the ninth aporia, which again deals with a worry about principles (archai), though
now of a different character: “if they are only one in form (eidos), nothing will be one in
number” (999b25–26). Aristotle’s dominant worry in this aporia seems to be that in
order to explain the unity of particulars, it is necessary for a principle to be one in number,
presumably for either one of two reasons. First, he may be assuming, as he had at
Metaphysics 993b23–31, and as Plato had in the Phaedo, that the aitia of something’s
being F must itself also be F, so that if a Form is a cause (aitia) of a sensible substance’s
being a unity, then it must itself be a unity, in the sense of being one in number. Second,
Aristotle seems to suggest that no given Platonic Form can do the job of bundling tropish
mirror images into genuine unities. Even the Platonic Form of Unity, as a universal,
seems to contribute just one more Form-copy to the bundle of otherwise non-united
and disparate Form copies. This is the point of Sellars’s colorful language: the bundle is
leaky in the sense that it is not properly even a bundle, since it has no internal principle
of structure or unification. If the Form of Unity adds just one more mimetic token to the
(putative) bundle, it is hard to appreciate how it could serve as a mechanism for unifying
or bundling the remaining mimetic tokens in the bundle. Later philosophers may wish
to speak of colocation, or coinstantiation, or coincidence in this connection, and their so
speaking may or may not be defensible. What is at issue in Aristotle’s criticism of Plato is
not whether any such gambit could ultimately succeed but, rather only that Plato owes
some such account. He needs to show how Forms ground both the synchronic and
diachronic unity of natural organisms. The bare resources of Forms and Form-copies
­present in his ontology do not point to any obvious way to discharge this obligation.
It is thus noteworthy that this noncategorial formulation of Aristotle’s objection does
not seek to force Plato into any immediate contradiction by treating Forms as both
universals and particulars. Instead, it treats Forms as overtaxed, from a functional or
explanatory point of view. Forms, as universals, permit genuine knowledge; but then,
again as universals, Forms fail to provide any principle of unity for sensibles. Hence,
they cannot be the principles (archai) Plato presumes them to be. Looked at this way, the
ninth aporia becomes immediately relevant to the argument of Metaphysics M 9. For
now Aristotle seems to have at his disposal the following series of inferences:

1. The principles (archai) of unified substances explain both their knowability and
their unity, taken both synchroncially and diachronically.
2. If they are the principles (archai) which ground (or explain) knowability, then
Forms are universal natures.
3. If they are the principles (archai) which ground (or explain) synchronic and
­diachronic unity, then Forms are particular natures.
4. Plato relies on Forms for both of these grounding (and explanatory) functions.
5. So Forms are both universal and particular natures.

In sum, if Forms are both the universal and particular natures of sensibles, then
universals and particulars will be virtually the same natures. Note, however, that
Forms may be both universals and particular natures without being both universals and
664   Christopher Shields

particulars: the nature of a relative will not itself be a relative—and the nature of a particular
need not itself be a particular. If humanity is Socrates’ nature, and Socrates is a sensible
particular, it does not follow that his nature is both a universal and a particular. Humanity
is, on the contrary, a universal and not a particular.
Thus, this argument has none of the force of the original categorial argument. It does
not seek to reduce Plato’s theory of Forms to an immediate or inescapable contradiction.
That, though, is a virtue: both versions of the categorial interpretation failed to convict
Plato of any egregious mistake or, indeed, of any mistake at all. By contrast, this noncat-
egorial version of Aristotle’s concern raises a legitimate worry and points to a bona fide
problem in Plato’s theory. If he overtaxes Forms by requiring them to function as both
universal and particular principles (archai), then Plato falls short of explaining either
knowability or unity, or both. Perhaps this is why, in Metaphysics M 9, Aristotle contends
that the successors of Socrates found no other principles beyond their own Forms. Their
doing so was understandable, but, oddly, if Aristotle is right, unacceptably economical.
Summarizing, then, it is natural to understand Aristotle’s complaints about Forms as
universals and particulars as proceeding in a categorial vein. If we understand him that
way, however, we uncover an argument sown with stridency but ultimately fruitless:
nothing about Plato’s treatment of Forms as substances (ousiai) or as paradigms (para-
deigmata) enjoins him to treat them as both universals and particulars. Plato is caught,
then, in no contradiction, and in no paradox of any kind, categorial or otherwise. Hence,
if this is the ultimate purport of his polemic, Aristotle will have missed his mark.
If, by contrast, we appreciate Aristotle’s aporematic interest in principles (archai) as
sources of both explanation and unification, then we see that he has a point, though not
one that prescribes any immediate rejection of the theory of Forms. Rather, when faced
with this criticism, we would want to look first to see whether Aristotle fares any better
in his own efforts to identify a principle (archê) which renders substance (ousia) both
one and knowable. For, after all, Aristotle himself faces the very tension he justifiably
highlights in Plato’s theory. Here, finally, the point is not simply a Platonic tu quoque. It is
rather that attention to the intricate intra-Academic dialectic that developed between
Plato and Aristotle serves to teach us something of value about each of them. The easy,
natural tendency to portray them as polar opposites entrenched in permanent philo-
sophical combat only serves to occlude this avenue of insight.

6. Conclusions

Scholars look to Aristotle as a source of data regarding Plato’s philosophy. To some, the
data Aristotle provides appear hopelessly tainted, because rife with polemic, habitually
unsympathetic, and even at times frostily caustic. A closer look at the dialectic of the
Academy suggests another vantage point from which to assess Aristotle’s contribution
to our understanding of Plato. From this angle, many of Aristotle’s criticisms prove more
multifaceted and less decisive than they may first present.
 Plato and Aristotle in the Academy   665

Indeed, although Aristotle’s writings contain many important and perfectly appropriate
criticisms of Plato’s theory of Forms, no one of them needs to be accepted by a Platonist
as unanswerable. Even the criticism which has been judged by scholars to be the most
directly devastating, because of its implicating Plato in an immediate categorial contra-
diction, proves upon closer inspection much less damaging than it first seems. It like-
wise proves to recommend a subtler and more consequential conclusion than what is
derivable from the strongly polemical version advanced in the categorial version of
Aristotle’s argument. Of course, Aristotle may, in fact, have a multiplicity of motives in
his worries about the particularity and universality of Forms; and it remains to be deter-
mined how his various worries may relate to one another. However they may, it emerges
upon investigation that Plato stands unconvicted of Aristotle’s categorial concerns about
Forms, at least as they are mounted in their most vigorous formulations.
Even so, it equally emerges upon investigation that among Aristotle’s concerns is
at least one fair worry about what threatens to be a false economy in Plato’s theory of
Forms: by introducing Forms as both objects of knowledge and as principles (archai)
of sense particulars, Plato saddles himself with a problem about the genesis of the unity
for sensibles—though, instructively, this is no less problem than that which surfaces
repeatedly in Aristotle’s own metaphysics in reference to his own positive treatment of
sensible substances. This suggests, then, not a linear Academic dialectic given in terms
of Aristotle objecting and Plato succumbing, or even an eristic contest given in terms of
Aristotle protesting and Plato retorting. Rather, Plato and Aristotle alike must grapple
with a fundamental problem of unity, the unity of complex particulars, which arises for
them in both its synchronic and diachronic guises.32 Although in their related ways,
both address this problem, neither Plato nor Aristotle emerges as the clear victor in
some arena of cleanly traceable intra-Academic dialectic; and yet each, it is fair to con-
clude, will have learned from the criticisms of the other—as we may yet learn from them
both. For the problem of how many things may also manage to be one thing finds no
more ready resolution today than it did in the days of Plato’s Academy.33

Bibliography
Ackrill, J. L. “In Defence of Platonic Division,” in J. L. Ackrill, Essays on Plato and Aristotle
(Oxford: Oxford University Press, 1997), 93–109.
Annas, J. Aristotle’s Metaphysics Books M and N (Oxford: Clarendon Press, 1976).
Bonitz, Hermann. Aristotelis Metaphysica, 2 vols. (Bonn: Marcus, 1848, 1849).
Burnet, John. Platonism (Berkeley: University of California Press, 1928).
Cavini, Walter. “Naming and Argument: Diaretitc Logic in Plato’s Statesman,” in C. Rowe
(ed.), Reading the Statesman (Sankt Augustin: Akademica Verlag, 1995), 123–38.

32 An appreciation of this more nuanced form of dialectical interaction is fully in evidence in Owen,
“Platonism.” Much of his landmark article is dedicated to showing, in effect, that we oversimplify the
complexity of Academic dialectic only at the cost of missing much of its real philosophical significance.
33 I thank Gail Fine for her clear and instructive comments on an earlier draft of this chapter. Her
generosity has improved the finished offering significantly.
666   Christopher Shields

Cherniss, Harold. Aristotle’s Criticism of Plato and the Academy, vol. 1 (Baltimore, MD: Johns
Hopkins University Press, 1944).
Düring, Ingemar. Aristotle in the Ancient Biographical Tradition (Göteborg: Elanders, 1957).
Fine, Gail. On Ideas (Oxford: Clarendon Press, 1993).
Fine, Gail. “Plato and Aristotle on Form and Substance,” Proceedings of the Cambridge
Philological Society 209 (1983), 23–47; reprinted in G. Fine, Plato on Knowledge and Forms
(Oxford: Oxford University Press, 2003), 397–425.
Frede, Michael, and Günther Patzig. Aristoteles, “Metaphysik Z”: Text, Übersetzung und
Kommentar (Munich: Beck, 1988).
Gerson, Lloyd P., “Harold Cherniss and the Study of Plato Today,” Journal of the History of
Philosophy, Volume 52, Number 3, July 2014, 397–409.
Graeser, Andreas, ed., Mathematics and Metaphysics in Aristotle (= Mathematik und Metaphysik
bei Aristotele: Akten des X. Symposium Aristotelicum) (Bern: Haupt, 1987).
Grote, George. Aristotle (London, 1880).
Jaeger, W. Aristotle: Fundamentals of the History of His Development (Oxford: Oxford
University Press, 1934); translated from the original German edition of 1923 by R. Robison.
Le Blond, J. Aristote, Philosophe de la vie: Le livre premier du traité sur les Parties des Animaux
(Paris: Aubier, Éditions Montaigne, 1945).
Le Blond, J. Logique et methóde chez Aristote (Paris: J. Vrin, 1939).
Lennox, James. Aristotle, On the Parts of Animals I–IV (Oxford: Clarendon Press, 2001).
Mansion, Suzanne. Le Jugement d’existence chez Aristote (Louvain: Éditions de l’Institut
Supérieur de Philosophie, 1946).
Mueller, Ian, “Aristotle’s Approach to Principles in Metaphysics M and N,” in Graeser, Andreas,
ed., Mathematics and Metaphysics in Aristotle (= Mathematik und Metaphysik bei Aristotele:
Akten des X. Symposium Aristotelicum) (Bern: Haupt, 1987), 241–259.
Nails, Deborah. The People of Plato (Indianapolis, Ind.: Hackett, 2002).
Owen, G. E. L. “The Platonism of Aristotle,” Proceedings of the British Academy 51 (1966),
125–50; reprinted in M. Nussbaum (ed.), Logic, Science, and Dialectic (Ithaca, N.Y.: Cornell
University Press, 1986), 200–21.
Owens, Joseph. The Doctrine of Being in the Aristotelian Metaphysics, 3rd ed. (Toronto:
Pontifical Institute of Medieval Studies, 1978).
Pellegrin, Pierre. “Division et syllogisme chez Aristote,” Revue Philosophique de la France et de
l’Etranger 171 (1981), 169–87.
Prior, William. “The Concept of Paradeigma in Plato’s Theory of Forms,” Apeiron 17 (1983),
33–42.
Ross, W. D. Aristotle, Metaphysics, 2 vols. (Oxford: Oxford University Press, 1924).
Ross, W. D. “The Development of Aristotle’s Thought,” in I. Düring and G. E. L. Owen (eds.),
Aristotle and Plato in the Mid-Fourth Century (Göteborg: Elanders, 1960), 1–18.
Ross, W. D. Plato’s Theory of Ideas (Oxford: Oxford University Press, 1953).
Runciman, W. “Plato’s Parmenides,” in R. E. Allen (ed.), Studies in Plato’s Metaphysics (London:
Routledge and Kegan Paul, 1965), 149–84.
Ryle, Gilbert. “Plato’s Parmenides II,” in R. E. Allen (ed.), Studies in Plato’s Metaphysics
(London: Routledge and Kegan Paul, 1965), 97–147; originally published in Mind 48 (1939),
302–25.
Ryle, Gilbert. Plato’s Progress (Cambridge: Cambridge University Press, 1966).
Sellars, Wilfrid. “Raw Materials, Subjects and Substrata,” in E. McMullin (ed.), The Concept of
Matter (Notre Dame, Ind.: Notre Dame University Press, 1963), 259–72.
 Plato and Aristotle in the Academy   667

Shields, Christopher. “Learning about Plato from Aristotle,” in H. Benson (ed.), The Blackwell
Guide to Plato (Oxford: Blackwell, 2006), 403–41.
Shields, Christopher. Order in Multiplicity (Oxford: Clarendon Press, 1999).
Stenzel, J. Plato’s Method of Dialectic, trans. D. J. Allan (Oxford, Oxford University Press, 1940).
Strawson, Peter. Individuals (London: Methuen, 1959).
Tarán, L. Speusippus of Athens (Leiden: Brill, 1981).
 chapter 28

Pl ato a n d Pl aton ism

Charles Brittain

1. Introduction

The attempt to understand and develop Plato’s philosophical views has a long history,
starting with Aristotle and Plato’s institutional successors in the Academy toward the
end of the fourth century b.c. But the development of a specifically Platonic philosophy
in the Academy or elsewhere was checked by the advent of the Hellenistic schools,
which advocated a more empirical approach to philosophical inquiry. As a result, the
idea that Plato’s dialogues already presented a well-defined, comprehensive, and essentially
correct philosophical system seems not to have arisen until the first century b.c.1 And it
was probably not until toward the beginning of the second century a.d. that a disparate
set of philosophers who identified themselves as “Platonists” conceived the project of
advocating and defending a specifically Platonic philosophy of this kind by systemati-
cally interpreting and explaining Plato’s texts.2 Over the next 500 years (c. 100–600 a.d.),
Platonist philosophers produced a huge corpus of philosophical work inspired by their
interpretations of Plato. The aim of this chapter is to introduce the reader to this
immensely varied and philosophically exciting—but, as yet, still largely unexplored—
tradition.3 The rest of this section gives some reasons why a modern student of Plato
might be interested in historical Platonism; Section 2 investigates the origins and evolution

1 I say “seems” because Antiochus of Ascalon, the originator of this idea in the first century b.c.,
claimed to be returning to the system set out by the early the Academics (see Cicero Academica
1.15–18). Antiochus’ historical claim is rejected by most scholars; but see D. Sedley, “The Origins of
Stoic God,” in D. Frede and A. Laks (eds.), Traditions of Theology (Leiden, 2002), 41–83.
2 The precise dating of the Platonist revival is disputed; see Section 2 and note 43 below.
3 A magnificent attempt to give a systematic presentation of Platonism through short excerpts with
commentaries is found in H. Dörrie and M. Baltes (trans.), Der Platonismus in der Antike, vols. 1–5
(Stuttgart, 1987–98). R. Sorabji, The Philosophy of the Commentators 200–600 a.d., 3 vols. (London,
2004), performs a similar service but emphasizes later Platonism owing to its primary focus on the
tradition of commentary on Aristotle.
670   Charles Brittain

of the Platonist movement; and Section 3 sketches its shifting epistemological foundations
and their relation to the Platonic dialogues.
The Platonist tradition is remarkably heterogeneous in comparison with other
ancient philosophical movements because it had neither a physical institution (a school)
to regulate membership, as the Hellenistic Stoics and Epicureans did, nor an explicit set
of its founder’s doctrines to regulate orthodoxy, as later Peripatetics did. The unity of the
tradition—that is, the sense in which the philosophers who identified themselves as
“Platonists” recognized their affiliation to a wider movement, and the basis for our
identification of “Platonism” as a single tradition—thus depends primarily on its partic-
ipants’ adherence to the project identified above: Platonists advocated and defended a
“specifically Platonic philosophy.” Given the openness of the Platonic dialogues to all
sorts of interpretation, however, this schema is too abstract. In practice, we can spell it
out as a threefold commitment to (a) the authoritative status of Plato’s work as contain-
ing in one way or another the correct philosophical doctrines; (b) a shared set of
assumptions about the inadequacy of empirical experience as a basis for understanding
the world, and about the existence and primacy of certain immaterial principles, includ-
ing “forms,” souls, and a transcendent god, that do explain it; and (c) an increasingly
keen interest in a range of religious practices and concerns.
The results of these commitments, when applied by Platonists to the interpretation of
Plato’s works, are likely to strike a modern student as rather distant from Plato’s text. It
turns out that (a) is not just an overly enthusiastic version of the principle of charity, but
becomes something close to a belief in Platonic infallibility; (b) leads the Platonists into
an ever-expanding dialectic of transcendence, producing increasingly complex hierar-
chies of metaphysical principles that seem more and more remote from Plato’s concerns;
and, in tandem with (b), (c) eventually introduces an overtly gnostic theory of “theurgy”
that seems at first sight to have much more in common with Christian soteriology than
with Plato’s Phaedrus or Symposium.4 It is thus a serious question whether a modern stu-
dent of Plato should be interested in Platonism, in a way that it is not in the case of later
Stoicism or Aristotelianism.5
One response to this question would be to argue that some of the Platonists were
nevertheless substantially right in their interpretations of Platonic texts (at least on cer-
tain central issues). But, while this view might be correct with respect to some details of
Platonic interpretation, it does not take into account a basic fact about Platonist exegesis
of Plato: the Platonists construed the interpretation of Plato as their primary function as
philosophers.6 Since their principal aim was to discover philosophical truths through the

4 See Section 2 below on (a); Sections 2 and 3 on (b); and note 62 on (c).
5 The two examples are different, however, in that students of Chrysippus cannot afford to ignore
late Stoics such as Epictetus or Hierocles because his works are lost, whereas students of Aristotle
merely lose out by not reading Alexander of Aphrodisias’ great exegetical work. So the Platonic case
parallels the second example, if it parallels either.
6 Proclus’ exegesis of the composition of the world soul in his Timaeus commentary is one case in
which a Platonist commentator is generally agreed to have solved a difficult question arising directly
from the Platonic text.
 Plato and Platonism   671

study of Plato rather than historical facts about Platonic arguments or texts, and given
that we are unlikely to subscribe ourselves to their rather implausible view about Plato’s
authoritative status ([a] above), it is probably a mistake to read their works as studies in
the history of philosophy.
A better case can be made, however, by considering the implications of some general
assumptions the Platonists tended to adopt about how to read Plato as a result of their
strange view about the status of the doctrines they found in his work. Three prominent
general assumptions are: (1) Plato’s dialogues (and other work) portray a consistent set
of doctrines; (2) his work presents a systematic philosophy; and (3) on most issues,
Plato’s views are best understood in the light of Aristotle’s development of them.7 We
don’t need to share any of these controversial assumptions, I suggest, to see how we
might benefit by reading the work of the philosophers who made them. At any rate,
I will give three reasons to think that we might.
The first is derived from the fact that most modern readers do not share assumption 1,
which is a “unitarian” view of the Platonic dialogues.8 One implication of adopting a
unitarian view is that the reader is compelled to consider and resolve the apparent
inconsistencies between various dialogues—for example, about the nature of the soul
and corresponding conception of virtue in the Phaedo and Republic. The Platonists offer
a range of solutions to such problems, which are often more plausible, and more philo-
sophically stimulating, than modern alternatives.9 The case above, for instance, which
modern scholarship has found particularly perplexing, was resolved by Plotinus by a
theory of “grades of virtue” reflecting the stage of self-awareness of the agent in her prog-
ress toward a correct understanding of one’s self as an immaterial and intellectual sub-
stance. On this account, the “civic virtues” of the Republic are preparatory for the higher
“cathartic virtues” of the Phaedo.10 The point here is not to argue that such Platonist
views are essentially correct but, rather, that engaging with them opens up a wealth of
philosophically rich and underexploited connections between the Platonic dialogues.11
A second reason is that the Platonist assumption that Plato is a systematic philosopher
(point 2 above) implies that we can find substantive theoretical answers to the philo-
sophical questions Plato raises. In the case of epistemology, for example, a Platonist

7 The third point was controversial in the second century a.d.; see Section 2, notes 51–53 below.
8 See Irwin’s discussion, chapter 3 of this volume.
9 A second case is the apparent inconsistency between the Phaedo, the Timaeus, or the Republic,
and the Phaedrus about the immortality of the non-rational soul. Proclus, for example, resolved this
controversial case by interpreting the two horses in the disembodied stages of Phaedrus myth in terms
of the circles of the same and the different that are constitutive of reason in the Timaeus, rather than
the two non-rational parts of the embodied soul in the Republic. His view is set out most clearly in his
student Hermeias’ Commentary on the Phaedrus.
10 Plotinus Ennead 1.2. See, e.g., J. Dillon, “An Ethic for the Late Antique Sage,” in L. Gerson (ed.),
The Cambridge Companion to Plotinus (Cambridge, 1996), 315–35; C. Brittain, “Attention Deficit in
Plotinus and Augustine,” Proceedings of the Boston Area Colloquium in Ancient Philosophy 18 (2003),
223–63; and at note 57 below.
11 In the case of Platonic ethics, Julia Annas has recently shown how using a Platonist text (Alcinous’
Handbook of Platonism) as a guide for reading Plato can yield illuminating philosophical results; see
J. Annas, Platonist Ethics Old and New (Ithaca, N.Y., 1999).
672   Charles Brittain

expects to discover not just how the apparently conflicting accounts of knowledge in
dialogues are consistent but a Platonic theory explaining the roles of perceptual experi-
ence and “recollection”—that is, our means of access to non-experiential knowledge—
in our acquisition of it.12 Some of the diverse results of this expectation are examined in
section 3, but we can note now that most Platonists found a developed epistemology and
theory of rationality in Plato’s Timaeus. Here, too, we do not need to accept the various
interpretations the Platonists offered to benefit from their recognition of the centrality
of this dialogue for Platonic epistemology.13 Another case in which the Platonist
assumption of systematicity led them to investigate issues in the dialogues that are usu-
ally ignored in modern philosophical scholarship concerns the topics of freedom,
self-determination, and divine providence. In the absence of a substantive body of mod-
ern scholarship on Plato’s views on these questions, the sophisticated work on them by
Plotinus, Porphyry, and Proclus is particularly illuminating.14 (The driving force in this
case was their further presupposition that Plato’s work contains substantive views about
all the central questions in philosophy; since these issues had become more salient in
the Hellenistic and early Imperial periods, the Platonists felt obliged to seek Platonic
responses to the theories of their Stoic and Peripatetic rivals; see section 2.)
The two reasons given so far might be construed uncharitably as the proposal to
exploit the vices of Platonist exegesis in pursuit of novel (in our context) and philosophi-
cally stimulating ways of reading Plato. The final reason, however, appeals to a significant
philosophical virtue of Platonist interpretations, viz. their sympathetic developments of
Platonic themes or ideas—often by appropriating and adapting Aristotle’s earlier, and
critical, reworking of them (point 3 above)—into sophisticated and original theories.15
The remarkable Platonist theories of the first principles, for example, constituted an
evolving effort to systematize and defend (by means of a causal theory) the apparently
scattered remarks of Plato on the interrelations between soul, intellect, the demiurge,

12 The Meno and Theaetetus appear to present a different view about the proper objects of
knowledge from the one found in the Republic or Sophist. The range of problems Platonists saw in
this field is set out in D. Sedley, “Three Platonist Interpretations of the Theaetetus” [“Three
Interpretations”], in C. Gill and M. McCabe (eds.), Form and Argument in Later Plato (Oxford, 1996),
79–103.
13 Modern readers have been slow to see this, perhaps because the Timaeus’ epistemology is hard to
fit within a developmentalist account (cf. G. E. L. Owen, “The Place of the Timaeus in Plato’s
Dialogues,” Classical Quarterly NS 3 [1953], 79–95). A notable non-Platonist exception is D. Frede, “The
Philosophical Economy of Plato’s Psychology: Rationality and the Common Concepts in the Timaeus,”
in M. Frede and G. Striker (eds.), Rationality in Greek Thought (Oxford, 1996), 29–56.
14 For Plotinus and Porphyry’s views on these issues, see W. Deuse, Untersuchungen zur
mittelplatonischen und neuplatonischen Seelenlehre [Seelenlehre] (Mainz, 1983). Proclus’ short treatises
on them have recently been translated into English in J. Opsomer and C. Steel, Proclus: On the
Existence of Evils (Ithaca, N.Y., 2003); Simplicius’ commentary on Epictetus (T. Brennan and C. Brittain
(trans.), Simplicius: On Epictetus’ Handbook (Ithaca, N.Y., 2002)) gives a useful introduction to late
Platonist views on these topics.
15 The Platonists, of course, denied that these were original theories rather than systematic
expositions of Plato’s sometimes obscure meaning. One reason for this was their assumption that they
had further information about Plato’s metaphysical principles from his (probably spurious) Letters, and
from various accounts of his “unwritten doctrines”; see Irwin, chapter 3 of this volume.
 Plato and Platonism   673

the Forms, the good, and the one.16 Their complex psychological theories attempted to
work out Plato’s suggestion that the soul is essentially a rational and immaterial sub-
stance that is independent of and prior to the material world, by explaining how such an
entity might interact with a body.17 And their elaborate hermeneutical theories were
developed in order to elaborate Plato’s diverse treatments of literary theory into a system
sufficiently complex to allow them to extract the last ounce of meaning from his texts.18
These sample cases are of enormous independent significance for the history of philoso-
phy, theology, and literature.19 But they are also directly relevant to the modern student
of Plato, in that they demonstrate some of the promising lines along which someone
sympathetic to Plato’s central metaphysical views might develop the insights she finds in
the Platonic texts.20

2. The Origins and Evolution

of Historical Platonism

Modern research on the origins of historical Platonism stems from the rejection of a
­traditional picture of a “Platonic School,” which was based on the false assumption that
Plato’s Academy in Athens was an enduring institution that ensured a more or less con-
tinuous transmission of Platonic teachings. The traditional picture distinguished five or
six discrete stages in the history of this “School”:

1. The “old” or “early” Academy, which lasted from 348 until 268 b.c.21
2. The “new” or “skeptical” Academy, from 268 to around 50 b.c.22

16 In Plotinus’ version, these Platonic principles are reduced to a theory of three hypostases (primary
ontological types or entities) of Soul, Intellect, and the One or the Good. The Forms are identified as
the thoughts that constitute Intellect (and Being), which is the Demiurge; the Good is identified with
the One. For Plotinus’ use of Aristotle in coming to these views, see note 58 below.
17 Some of the vast range of Platonist psychologies is reviewed in, e.g., Deuse, Seelenlehre, and
H. Blumenthal, Soul and Intellect: Studies in Plotinus and Later Neoplatonism (London, 1993).
18 R. Lamberton, Homer the Theologian (Berkeley, 1986), is a useful modern introduction to
Platonist hermeneutics; the anonymous Prolegomena to Plato is a less helpful ancient equivalent from
the sixth century a.d.
19 As the work of Origen and Augustine suggests, a good deal of Christian theology and
hermeneutics in the period from 150 to 600 a.d. is derived directly from Platonism. For a useful
overview of their relation, see C. Stead, Philosophy in Christian Antiquity (Cambridge, 1994).
20 A parallel history of the Stoic interpretations and adaptations of Plato would perform a similar
function for a reader who was unsympathetic to Plato’s dualism.
21 The most prominent old Academics were Speusippus (d. 338 b.c.), Xenocrates (d. 314 b.c.),
Crantor (d. 276 b.c.), and Polemo (d. 269 b.c.). An overview of the extant fragments of their work—
which was highly regarded by the Platonists—is given in J. Dillon, The Heirs of Plato: A Study of the Old
Academy, 347/274 b.c. (Oxford, 2003).
22 The most significant figures in the skeptical Academy were Arcesilaus (315–240 b.c.), Carneades
(214–128 b.c.), Philo of Larissa (159–84 b.c.), and Philo’s student Antiochus of Ascalon (c. 130–68 b.c.).
674   Charles Brittain

3. A “transitional” post-Academic period, from c. 50 b.c. to c. 70 a.d.23

4. A “middle” or early Platonist phase, from c. 70 to c. 230 a.d.24
5. A “Neoplatonist’’ or Plotinian phase of Platonism, from c. 230 to c. 300 a.d.25
6. A “late Neoplatonist” or late Platonist phase, from c. 300 to c. 600 a.d., including
6.1. a relatively unattested phase, from c. 300 to c. 400 a.d.;
6.2. the “Athenian School,” attested from c. 400 to its closure in 529 a.d.;
6.3. the “Alexandrian School,” attested from c. 435 to its closure in 611 a.d.26

(The italicized names are avoided in recent work on Platonism for reasons explained
below.)
Although the chronological framework of this traditional picture is generally
accepted, modern scholars reject the overall picture for three reasons.27 First, it conflates
the notions of a “school” as a physical institution and as a philosophical movement
(a “sect” or hairesis). But the Platonic Academy had ceased to exist as physical institution

Overviews of their work are given in C. Brittain, “Arcesilaus” (2005), J. Allen, “Carneades” (2004),
C. Brittain, “Philo of Larissa” (2006)—all in the Stanford Encyclopedia of Philosophy at http://plato.
stanford.edu/—and J. Barnes, “Antiochus of Ascalon” [“Antiochus”] in M. Griffin and J. Barnes (eds.),
Philosophia Togata (Oxford, 1989), 51–96.
23 Prominent figures in this period include Eudorus (late first century b.c.), Thrasyllus (d. 36 a.d.),
and Philo of Alexandria (d. 40–50 a.d.); see J. Dillon, The Middle Platonists [Middle Platonists]
(London, 1977), for an introductory account. (H. Tarrant, Thrasyllan Platonism [Ithaca, N.Y., 1993],
gives a controversial, maximalist interpretation of Thrasyllus’ work, but see notes 38–39 below).
24 Notable representatives include Plutarch of Chaeronea (c. 50–120 a.d.), Alcinous (c. 150–200
a.d.), Atticus (fl. 176 a.d.), Numenius (c. 150–200 a.d.), and Longinus (216–272 a.d.). G. Boys-Stones’
Platonist Philosophy 80 BCE to AD 250. An Introduction and Collection of Sources in Translation
(Cambridge, 2018) is set to be the standard doxographical collection and review for early Platonism in
English; more recent accounts of several important figures are given in W. Haase and H. Temporini
(eds.), Aufstieg und Niedergang der Römischen Welt [Aufstieg], Part 2, vols. 36.1–7 (Berlin, 1987–1994);
see note 52 below. H. Tarrant, Plato’s First Interpreters [First Interpreters] (Ithaca, N.Y., 2000), offers a
wide-ranging account focused on Platonic interpretation.
25 Plotinus (204–270 a.d.) and Porphyry (234–305 a.d.) are the key figures in this phase. The most
reliable modern translation of Plotinus’ work into English is L. Gerson (ed.), Plotinus The Enneads
(Cambridge, 2018). Three excellent introductions to Plotinus’ thought are E. Emilsson, Plotinus
(London, 2017), L. Gerson, Plotinus (London, 1994), and D. O’Meara, Plotinus: An Introduction to the
Enneads (Oxford, 1993). On Porphyry, see A. Smith, Porphyry’s Place in the Neoplatonic Tradition
[Porphyry’s Place] (The Hague, 1974), and the important new editions of two of his works by J. Barnes,
Porphyry: Introduction (Oxford, 2003), and L. Brisson (ed.), Porphyre: Sentences (Paris, 2005).
26 Iamblichus (c. 245–325 a.d.) is our principal representative of the “lost phase.” The leaders of the
Athenian school included Plutarch of Athens (d. 431 a.d.), Syrianus (d. 437 a.d.), Proclus (411–485
a.d.), and Damascius (465–540 a.d.). Notable teachers at the Alexandrian school included Ammonius
(c. 440–520 a.d.), Philoponus (c. 490–570 a.d.), Simplicius (c. 500–570 a.d.) and Olympiodorus
(c. 500–570 a.d.). R. Wallis, Neoplatonism (1972; 2nd ed. London, 1995), provides a rather out-of-date
account of late Platonism; but the historical circumstances of these philosophers and their work as
commentators on Aristotle is comprehensively reviewed in the essays in R. Sorabji (ed.), Aristotle
Transformed (Ithaca, N.Y., 1990).
27 See J. Glucker, Antiochus and the Late Academy (Göttingen, 1978), a groundbreaking account
of the actual history of the Platonic Academy and the discontinuous history of Platonism, which
investigates the history of the technical terms “Academic,” “Platonist,” “sect” (hairesis), “school-
succession” or “transmission” (diadoche), and “successor” (diadochos).
 Plato and Platonism   675

by the middle of the first century b.c.; thereafter, we know something about several
“private” Platonist teaching institutions—for example, Plotinus’ school in Rome,
Iamblichus’ in Chalcis in Syria, and the later Athenian and Alexandrian schools—but
there was no central “school.”28 Second, it conflates continued philosophical interest in
Plato’s thought outside an institutional framework with explicitly Platonic “movements”—
the skepticism of the “New Academics” or the Platonism of the self-identified
“Platonists”—by assuming that there was a continuous chain of Platonic “successors”
ensuring the transmission of teaching (diadoche) between stages 2 and 4.29 And third, it
conflates philosophical developments within the Platonist “movement” with chronologi-
cal periodizations and the emergence of particular teaching institutions in stages 4 to 6.30
The rejection of this traditional picture for a more accurate view of the various histor-
ical forms of Platonic studies allows us to replace ill-defined questions about the origins
of Platonism—such as ones framed in terms of the transmission of Platonic authority or
teachings or of an evolution toward a specific set of (usually Plotinian) metaphysical
doctrines—with a more determinate one about the philosophical movement of stages
4 to 6. Why was the development of an explicit “Platonism” delayed until the late first
century a.d. (stage 4)? A satisfying answer to this question should explain how a group
of philosophers came to see a Platonic philosophy as offering the solution to some
central difficulties in late Hellenistic philosophy (see Section 3 of this chapter). But we
might also expect to find a more strictly historical explanation for the fact that it was
only at this stage that a group of philosophers came to identify themselves as “Platonists,”
and our initial characterization of the movement (commitments a to c, p. 527) suggests
that this is likely to depend on the development of Plato’s status as an authoritative
philosopher (point a).

28 The primary sources for our knowledge of these schools are Porphyry’s Life of Plotinus, Eunapius’
Lives of the Sophists, Marinus’ Life of Proclus, and Damascius’ Life of Isidore (or Philosophical History);
see M. Edwards, Neoplatonic Saints: The Lives of Plotinus and Proclus by Their Students (Liverpool,
2000), and P. Athanassiadi, Damascius: The Philosophical History (Athens, 1999).
29 Some recent scholarship still assumes that we can trace at least the outlines of a continuous
transmission between private teachers and their students from Philo of Larissa or his student
Antiochus of Ascalon in stage 2 through to Eudorus in stage 3, and perhaps into stage 4; see, e.g.,
H. Tarrant, Scepticism or Platonism: The Philosophy of the Fourth Academy [Fourth Academy]
(Cambridge, 1985); Tarrant, First Interpreters; Dillon, Middle Platonists; and notes 35–36 below. But the
traditional institutional picture is not supported or implied by any ancient evidence.
30 “Middle Platonism” is used both as a chronological term, referring to Platonists prior to Plotinus
from period 4 (and sometimes their antecedents in period 3), and as a way of classifying doctrinal
divergences from Plotinus; as a result, the work of Longinus and Calcidius is chronologically
“Neoplatonist” but doctrinally “Middle Platonist,” while the reverse is true of Numenius. (The “middle”
and “neo-” classifications reflect the assumptions that the Platonism of period 4 was essentially a
revival of an original Platonism found in period 1, and that Plotinus’ work (in 5) marked a decisive
doctrinal shift from these earlier “Platonisms.”) The traditional view that the late Athenian and
Alexandrian Platonists defended radically distinct forms of Platonism rests on a misreading of the
evidence; see, e.g., I. Hadot, Le problème de néoplatonisme alexandrin: Hiéroclès et Simplicius (Paris,
1978). (The two schools were tightly connected, both socially and philosophically; Simplicius, for
instance, studied in both places.)
676   Charles Brittain

The problem of the delayed development of Platonism is an interesting one because

the general conditions for its emergence seem to have been in place from at least 50 b.c.
(stage 3). Notable among these conditions were three significant changes in late
Hellenistic philosophy. The first was the demise of the skeptical Academy by the middle
of the first century b.c., which released the interpretation of the Platonic corpus from
institutional constraints, and in particular from the skeptical interpretations of Plato
the Academics had advocated since the third century b.c.31 The second was the decen-
tralization of philosophy that ensued after the disruption of Athens during the first
Mithridatic war (89–84 b.c.) and the “pacification” of Greece by the Romans. This led
to a more pluralistic philosophical culture and paved the way for the development of a
tradition of written commentaries on works of recognized importance.32 And the third
was a widespread revival of interest in a set of metaphysical and (crudely) “otherworldly”
themes that had been abandoned or neglected by early Hellenistic philosophers but
were prominent in Plato: these included theories of transcendent principles and topics
such as divination, demonology, and the afterlife of the soul.33
The problem becomes more acute when we look at the changes to Plato’s status as a
philosophical authority that had already occurred by around the middle of the first
century b.c. One was that philosophers from various Hellenistic schools were now will-
ing to recognize Plato, as well as Aristotle and Pythagoras, as having the privileged status
of classical figures, whose views were worth considering even when they conflicted in
some ways with the orthodox doctrines of one’s own school. A second was that the death
throes of the skeptical Academy led to two explicit, though incompatible, appeals to the
Platonic tradition to validate the historical authenticity of entire philosophical positions.
In the case of Philo of Larissa, this meant an appeal to the continuity of the Academic

31 The philosophical demise of the skeptical Academy was caused by Philo of Larissa’s shift from
radical to mitigated skepticism in the 90s b.c., which triggered a revolt by his student Antiochus of
Ascalon, who left the Academy to set up a rival dogmatic “Old Academy.” This, in turn, led Philo to an
attempt to reposition the Academy as part of a critical—rather than skeptical—tradition; see Barnes,
“Antiochus,” and Brittain, “Philo of Larissa” (2006). Arcesilaus’ skeptical interpretations of Plato are
examined in J. Annas, “Plato the Sceptic,” Oxford Studies in Ancient Philosophy Supplementary Volume
(Oxford, 1992), 43–72, and Sedley, “Three Interpretations.” C. Brittain, Philo of Larissa [Philo] (Oxford,
2001), also covers the range of later skeptical Academic interpretations of Plato, Socrates, and the old
Academic philosophers (stage 1 previously described).
32 For a detailed analysis of these developments, and those mentioned immediately below, see
M. Frede, “Epilogue,” in K. Algra, J. Barnes, J. Mansfeld, and M. Schofield (eds.), The Cambridge History
of Hellenistic Philosophy (Cambridge, 1999), 771–97.
33 The renewed interest in metaphysics was partly due to the wider dissemination of Aristotle’s work
in the first century b.c., which rapidly generated a cross-party tradition of commentary and debate on
the Categories. Philosophical interest in Plato’s treatment of the whole set of themes is evident in Stoics
such as Posidonius (see, e.g., Galen PHP 4–5; Plutarch Proc. An. 1023b–d; and Cicero Div. 1.60–64),
and in the revival of philosophical and popular enthusiasm for Pythagoreanism. (These philosophical
changes seem to reflect a wider social change away from the optimistic sense that things could be
improved by natural and rational means here and now; this perhaps explains the success of the new
science of astrology in this period and the proliferation of soteriological sects. On this wider change,
see, e.g., E. Dodds, The Greeks and the Irrational [Greeks] [Berkeley, 1951], and P. Brown, The Making of
Late Antiquity [Cambridge, Mass., 1978].)
 Plato and Platonism   677

tradition—that is, to “the unity of the Academy,” including Socrates, Plato, the early
Academics, and his skeptical predecessors—to support his redefinition of Academic
philosophy as a nonskeptical tradition of critical inquiry. This was an unusual step for
an Academic (although claims that one’s views were consistent with a founding ortho-
doxy were a standard feature of the other Hellenistic schools).34 But it didn’t constitute
a return to Platonic philosophy because Philo used the thesis of the unity of the
Academy to defend his own empiricist views rather than a set of doctrines drawn from
Plato’s work.35
In the case of his rival Antiochus, however, it meant an appeal to the authoritative
consensus of the ongoing Platonic tradition—the old Academic and Peripatetic doctrines
that the skeptical Academics had misguidedly rejected—to validate the truth of the
dogmatic philosophical system that Plato himself had established. And since Antiochus’
position amounted to an explicit defense of a set of Platonic doctrines—a specifically
Platonic philosophy—it is at first sight hard to understand why it did not immediately
initiate a Platonist movement.36 But closer examination of his view provides two reasons
to explain this. The first is the content of the Platonic doctrines he advocated: Antiochus’
position was an empiricist one, and accordingly substituted an immanent Stoic god for
the transcendent metaphysical principles that were central to Platonism. The second is
his justification for this substitution: that is, his appeal to Platonic authority in the form
of the ongoing consensus of a broad tradition, which allowed him to abandon views
that had proven untenable and replace them with later, Stoic, alternatives.37 These two
Antiochian principles directly contradict the central theses of the Platonist movement.
The specific forms that direct appeals to Plato’s authority had taken in the early part of
the first century b.c. thus suggest one way to explain why the significant interest in
Platonic texts and themes over the next hundred years was not conceived as “Platonism”:
the philosophers involved had not developed an alternative Platonist ideology that was
capable of uniting their specific interest in Plato with their wider philosophical inter-
ests in transcendent metaphysical principles. Some prominent examples of these twin
interests in the period 50 b.c.–70 a.d. include Cicero’s translation of the Timaeus,

34 Philo’s appeal to the tradition contravened the extensive Academic critique of the use of authority
by dogmatic philosophers; see, e.g., Cicero Ac. 2.8 and DND 1.11; traces of this critique remain in Sextus
(e.g., PH 2.37–45) and Lucian’s Hermotimus. For general Hellenistic attitudes, see D. Sedley, “Plato’s
Auctoritas and the Rebirth of the Commentary Tradition” [“Plato’s Auctoritas”], in J. Barnes and
M. Griffin (eds.), Philosophia Togata II: Plato and Aristotle at Rome (Oxford, 1997), 110–29.
35 Pace Tarrant, Fourth Academy; see Brittain, Philo (2001), 129–265.
36 This seems clear for the reasons given below, as well as the almost complete silence about
Antiochus in the later tradition, which is broken only to criticize him as someone who attempted to
pollute Plato’s legacy with Stoicism (see, e.g., Sextus PH 1.235; Numenius fr.29 Des Places; and
Augustine Contra Academicos 3.41). It is denied, however, in, for example, Dillon, Middle Platonists.
37 Antiochus’ empiricism is clear from his wholesale adoption of a Stoic epistemology, attested in
Cicero Ac. 2. His views on Platonic metaphysics are more controversial, since the report in Cicero Ac. 1
gives both “old Academic” views and their “corrections” by the later tradition, such as Aristotle’s attack
on the theory of Forms, which leaves Antiochus’ own position unclear. But the section on “old
Academic” physics shows that he did agree with the Stoic “correction” that eliminated the Platonic view
that god and the soul were immaterial substances (Ac. 1.39).
678   Charles Brittain

Thrasyllus’ work on the Platonic dialogues, Eudorus’ work on “Pythagorean” (i.e., Platonic)
metaphysics, and Seneca’s discussion of Platonic principles in Letters 58 and 65.38 But it
is notable that none of these philosophers were considered by themselves or their con-
temporaries as primarily followers of Plato: their interest in Platonic metaphysics is
instead explained by their connections with established (late) Academic, Pythagorean,
or Stoic concerns.39
The additional element required for the development of an overtly Platonist ideology
was a theory that could justify an unqualified return to Plato at this late point in the
history of philosophy. That is, a theory was needed to explain why it made sense to
defend Plato’s views as a whole rather than defending some subset of them in light of the
intervening Hellenistic philosophical developments, which is what the various
Academics, Stoics, and neo-Pythagoreans of the first centuries b.c. and a.d. had done.
The theory that the Platonists seem to have adapted to serve this purpose drew on what
was by this time a popular notion that a common heritage of “ancient wisdom” was
encoded in the literature of the established cultures of the world.40 The point of the the-
ory was to suggest that Plato’s philosophy represented the final encoding of this “ancient
wisdom” into the form of philosophical writing (in Greek).41 The Platonists supported
this rather surprising claim by an argument from the history of post-Platonic philosophy,
which explained the chronic disagreements of the Hellenistic philosophers—and the
philosophical dead ends (empiricism, skepticism, etc.) they were driven to—as the
direct result of their dissensions from Platonic doctrine.42 An unqualified return to
Platonic wisdom was thus precisely what was required.

38 Eudorus’ interest in Plato is attested by his doxographical work on the creation of the world soul
in the Timaeus (Plutarch Proc. An. 1013b) and his neo-Pythagorean views on the first principles
(Simplicius In Phys. 188). Thrasyllus’ work on Plato’s dialogues is attested in Diogenes Laertius Lives of
the Philosophers 3.56–61.
39 Seneca was a Stoic. Thrasyllus was an astrologer known for his Pythagorean bent, who also edited
Democritus’ works. Cicero was a skeptical “Academic” who admired Plato; his Timaeus translation is
dedicated to the “Neopythagorean” Nigidius Figulus. Eudorus probably characterized himself as an
“Academic,” which usually meant a skeptical Academic at this time; but it is more likely that he saw
himself as affiliated to the Academic tradition as a whole—that is, a tradition that included the early,
nonskeptical Academics of period 1. It is unclear whether this would make him a late Antiochian
Academic—a view supported by the thoroughly Stoic ethical work ascribed to him in Stobaeus
Ec. 2.7—or a significant antecedent for Plutarch’s “Academic” Platonism (see notes 49–50 below).
40 The (Stoic) origin and evolution of this theory is examined in G. Boys-Stones, Post-Hellenistic
Philosophy (Oxford, 2001). Its Platonist proponents offer a range of recognized cultures, which varies
over time and depending on their aims. Prominent “sages” include the Egyptian Hermes, Persian
Zoroaster and Magi, Indian Brahmins, Jewish Moses, and Greek Homer, Musaeus, Orpheus, and
Pythagoras. The Chaldean oracles were added rather late to this list of ancient authorities—after they
had been forged in the mid-second-century a.d.
41 The exact nature of Plato’s relation to the ancient tradition is disputed. M. Frede, “Numenius,” in
W. Haase and H. Temporini (eds.), Aufstieg und Niedergang der Römischen Welt, Part 2, vol. 36.2
(Berlin, 1987), 1034–75, argues that Plato was regarded as the last of the ancients; Boys-Stones,
Post-Hellenistic Philosophy, sees Plato as the first to reconstruct the ancient wisdom correctly and in
philosophical Greek.
42 The theory is reconstructed and defended sympathetically in Boys-Stones, Post-Hellenistic
Philosophy, 99–150. As he shows, the general theory is present in Plutarch (fr. 157, 190 Sandbach),
 Plato and Platonism   679

Something like this view of the development of Plato’s status as a philosophical

authority seems to give the best available historical explanation for the evidence that the
Platonist movement did not take off until the end of the first century a.d, despite the
revival of interest in Plato 150 years earlier.43 At any rate, it is in the period from c. 70 to
c. 230 a.d. that a disparate group of philosophers accepting an explicitly Platonist ideology
set to work to explain and defend the doctrines they found in the Platonic corpus. One
result was a large and diverse mass of work, in the form of introductions to Plato, hand-
books on Platonic doctrines, Platonic lexica, commentaries on Platonic dialogues, and
essays on particular problems in Plato—of which only a few samples survive intact.44
Such works were as necessary then as now to explain textual, linguistic, and historical
difficulties in the dialogues; interpret their arguments, “myths,” and enigmatic (e.g.,
numerological) passages; elucidate their general intent; and understand how they fit
together to produce a coherent philosophical view.45 But since their authors were pri-
marily philosophers, their central purpose was not to offer interpretations of Plato’s
dialogues but to identify in them a systematic set of philosophical doctrines and defend
these against opposing views—that is, against both rival forms of Platonism and their
shared Stoic, Peripatetic, and skeptical philosophical opponents. As a result, the
Platonists of this period produced a second mass of more standard philosophical literature,

Numenius (fr. 1a Des Places), Celsus (Origen Against Celsus 1.14), Plotinus (e.g., Enn. 3.5.2), and
Porphyry (e.g., fr. 323–24 Smith). The related argument from the history of post-Platonic philosophy is
attested in Plutarch (Stoic. Rep. passim), Numenius (fr. 24), and Atticus (fr. 5 Des Places), among
others.
43 I should note, however, that a number of scholars, including Harold Tarrant and David Sedley,
prefer the view that a recognizable form of Platonism had already developed by the end of the first
century b.c., under the stimulus of Antiochus or Philo. The key disagreement between this view and
the one presented previously concerns the dating of the anonymous Commentary on the Theaetetus,
which attests to internal disagreements within a well-developed and explicitly Platonist movement, and
hence might support more “Platonist” interpretations of scantily known figures such as Eudorus or
Thrasyllus if it were datable to 25 b.c. or a.d. But the arguments for this early date, rather than 100 or
120 a.d., strike me as weak, since the controversies in Anon. are exactly those of the later period; see
D. Sedley and G. Bastianini (eds.), Commentarium in Platonis Theaetetum, Corpus dei papiri filosofici
greci e latini, vol. 3 (Florence, 1995), 227–562, and Brittain, Philo (2001), 249–54. It would also be strange
that later Platonists make no reference at all to this hypothetical early tradition of Platonism, beyond a
handful of citations of Eudorus and Thrasyllus that do not demonstrate its existence.
44 Some extant examples of the first two categories are Albinus’ Introduction to Plato, Alcinous’
Handbook, and Apuleius’ Plato’s Doctrines. Examples of the third and fourth categories are the
anonymous Commentary on the Theaetetus and Plutarch’s The Creation of the Soul in the Timaeus.
There is a considerable amount of evidence for the views of a number of early Platonists on
controversial issues in the Platonic dialogues for which later (fifth or sixth century a.d.) commentaries
are extant, viz. First Alcibiades, Cratylus, Gorgias, Parmenides, Phaedo, Phaedrus, Philebus, Republic,
and Timaeus.
45 The Platonist commentary tradition was preceded by an Aristotelian one, which may have taken
off in the late first century b.c.; see H. Gottschalk, “The Earliest Aristotelian Commentators,” in
R. Sorabji (ed.), Aristotle Transformed (Ithaca, N.Y., 1990), 55–81. The aims and scope of the early
Platonic commentaries—as manifested by our single example—are explained in Sedley, “Three
Interpretations,” and Sedley, “Plato’s Auctoritas.” On the later commentaries, see notes 63–64 below.
680   Charles Brittain

in the form of controversial works, treatises on specific philosophical problems, and

systematic treatments of metaphysics.46
Given the diversity of this material, we can notice here just two pervasive strands in it.
The first is that these Platonist texts share an overriding concern to justify the basic
assumptions of Platonism—that the world can only be understood by reference to a
set of immaterial principles inaccessible to empirical experience (commitments a to c,
p. 527)—against the criticism of their philosophical opponents. This is worth stressing
because the nature of our evidence for the period tends to obscure it. Owing to the
vagaries of textual survival, it is easy to forget that the “Hellenistic schools” continued to
flourish and produce significant work until at least the middle of the third century a.d.47
(This was publicly attested by the emperor Marcus Aurelius’ decision in 176/77 a.d. to
establish paid “chairs” for the four dominant “schools” of the period—that is, for the
Stoics and Epicureans, as well as for the Aristotelians and Platonists who had emerged
over the last century.) But even when we have corrected for our ignorance of the philo-
sophical context in which the Platonist work was produced, the form of the surviving
material—which is often fragmentary, or dedicated to religious or cultural topics that
may not strike us as obviously philosophical—can make it difficult to reconstruct the
underlying philosophical arguments.
The second strand is the struggle to define a narrower Platonist orthodoxy, which is
reflected in the variety of philosophical approaches the Platonists adopted for system-
atizing and defending the doctrines they found in the Platonic texts. Like modern readers
of Plato, the Platonists tended to read the dialogues in the light of their own philosophi-
cal preoccupations and training.48 In practice, this resulted in a series of more or less

46 See the excellent review by P. Donini, “Testi e commenti, manuali e insegnamento: la forma
sistematica e i metodi della filosofia in età postellenistica,” in W. Haase and H. Temporini (eds.),
Aufstieg und Niedergang der Römischen Welt, Part 2, vol. 36.7 (Berlin, 1994), 5027–100. Examples for the
first category include several extant works by Plutarch—Stoic Contradictions, Common Conceptions
(against the Stoics), and Against Colotes, and a summary of The Impossibility of Living Pleasantly
(against the Epicureans)—as well as the fragmentary remains of Numenius’ work, The Dissension of the
Academics from Plato, and of Atticus’ book, Against Aristotelian Interpretations of Plato. The second
category is represented by Plutarch’s extant Moral Virtue and pseudo-Plutarch’s On Fate; later extant
examples include most of Plotinus’ Enneads. There are no extant examples of the third category until
Porphyry’s Sententiae and Proclus’ Elements of Theology. (The extant material in these categories is thin
for the period prior to Plotinus, because it is only in the case of Plutarch that a substantial part of his
philosophical corpus survives intact.)
47 In the case of the Stoics, the texts that survive from this period are primarily those of “moralists”
such as Epictetus and Marcus Aurelius, whose books were valued by later Platonists and Christians for
their introductory ethical character, but innovative philosophical work was still being done; see Frede,
“Epilogue,” 779–81. Skeptical Academics had indeed largely disappeared by this time, but they were
replaced by Pyrrhonists; see C. Brittain, “ ‘Middle’ Platonists on Academic Scepticism” [“Scepticism”],
in R. Sorabji and R. Sharples (eds.), Greek and Roman Philosophy, 100 bc–200 ad, vol. 2 (London
2007), 297–315.
48 A possible exception to this is the “literal” interpretation of the creation of the world in time in
the Timaeus, advocated by Plutarch and Atticus, but rejected by the rest of the tradition. But it is not
clear that this is a “literal reading,” and it is likely that Atticus’ motivation for adopting it has some
connection to his opposition to Aristotle.
 Plato and Platonism   681

“Academic,” “Pythagorean,” “Stoic,” and “Aristotelian” versions of Platonism, depending

on the degree to which their proponents applied the refinements of these traditions to
their interpretations of Plato.49 At issue were both historical questions—such as the role
of Pythagoreanism in the formation of Plato’s doctrines and the kinds of skepticism
adopted by the “New Academics”—and philosophical questions about Platonic ethics,
psychology, and metaphysics.50 In ethics, Platonists agreed that the “end” was “likeness
to god,” but they disagreed about, for example, whether (Plato thought that) the human
good was limited to virtue—as the Stoics argued, and the “Socratic” dialogues may
suggest—or not—as Aristotle argued, and the later dialogues seem to allow.51 In psy-
chology, they agreed that the soul was essentially an immaterial and immortal sub-
stance, but they disagreed about, for example, its composition (in the Timaeus) and
whether its partition in some dialogues was compatible with an Aristotelian analysis
in terms of faculties.52 And in metaphysics, they agreed about the existence of Forms
and at least one divine intellect, but they disagreed radically about the number and kind
of ultimate principles.53

49 Plutarch regarded himself as an “Academic” (though the later tradition translated this as
“Platonist”); Numenius was often described as a “Pythagorean,” and we hear of a “Stoic and Platonist”
called Tryphon in Porphyry’s Life of Plotinus 17.
50 The debate about the degree of infidelity to Plato shown by the New Academics was fierce:
Numenius regarded them (and Antiochus) as complete renegades (fr. 24–28 Des Places); Plutarch
thought that their skepticism was largely compatible with Platonism (Platonic Questions 1); the
anonymous commentator on the Theaetetus claimed that “very few” of them were skeptical (cols.
54–55); in his exuberant early Platonist phase, Augustine suggested that they were all esoteric Platonists
(Against the Academics 3.37–42). A sympathetic account of the enduring New Academic trend in early
Platonism is given by J. Opsomer, In Search of the Truth: Academic Tendencies in Middle Platonism
(Brussels, 1998). Pythagoreanism was far less controversial, since Pythagoras’ influence on or
consonance with Plato was generally accepted; see, e.g., D. O’Meara, Pythagoras Revived (Oxford, 1989),
on the role of Pythagorean mathematical ideas in the development of Platonism.
51 The Stoic (or anti-Aristotelian) line is pursued vehemently by Atticus in fr. 2 Des Places. Plutarch
promotes an aggressively Aristotelian (or anti-Stoic) line in his treatise Moral Virtue; a more moderate
version is given in Alcinous Handbook 27–33.
52 On the composition of the soul, see, e.g., Plutarch, The Creation of the Soul in the Timaeus, and
Alcinous, Handbook 14; Atticus predictably rejects Aristotle’s contributions entirely (fr. 7). The remains
of an interesting earlier debate about the faculties of the soul can be seen in the fragments of Porphyry’s
The Faculties of the Soul. Fr. 252 (Smith) cites Numenius’ and Longinus’ conflicting views on the
(originally Stoic) faculty of assent and its relation to the Aristotelian faculty of representation
(phantasia); see L. Brisson and M. Patillon, “Longinus Platonicus Philosophus et Philologus,” in
W. Haase and H. Temporini (eds.), Aufstieg und Niedergang der Römischen Welt, Part 2, vol. 36.7
(Berlin, 1994), 5215–299, ad fr. 9 (pp. 5286–89), and M. Frede, “Numenius,” ad fr. 45 (pp.1070–74). But
this is just one example from a huge range of disputes.
53 These questions are the central focus of much work on the early Platonists; see, e.g., Dillon,
Middle Platonists. One of the most interesting debates concerned the relation between the Forms and
the divine Intellect—that is, whether the former were constitutive of the latter, or superior or
subordinate to it. (The Stoics took the Forms to be just thoughts of god, mistakenly hypostasized by
Plato. But if the Forms are the “model” and “perfect living thing” of Timaeus 29a and 31c, they should
be superior to at least the Demiurgic intellect.) See, e.g., Alcinous Handbook 9; Porphyry Life of
Plotinus 18 and 20; and Syrianus On Aristotle’s Metaphysics, pp. 104–07 (Kroll).
682   Charles Brittain

A good reason to take the work of Plotinus and his student Porphyry in the period
from c. 230 to 300 a.d. to mark a new stage in the history of Platonism is that their reso-
lutions to many of these questions became standard features of later Platonism.54 (A less
good though understandable reason is that we possess the entire works of Plotinus
and a substantial amount of Porphyry’s.) Like their predecessors, Plotinus and Porphyry
devoted a lot of attention to refuting the Platonists’ opponents (although by this stage they
include Christians and other gnostics, as well as Epicureans, Stoics, and Peripatetics).55
But they also succeeded in integrating enough Stoic and Aristotelian ideas into Platonism
to effect the eventual absorption of those rival traditions within it.56 Two central cases in
Plotinus are, first, his integration of Stoic ethics into a Platonic theory of virtue, by iden-
tifying its key theses—that rational virtue is the good and, accordingly, that irrational
emotions should be eliminated—as the principles of the second of three grades of
virtue.57 And second, his integration of Aristotelian theology into Platonic metaphysics,
by identifying the active intellect of De anima 3.5 with the unmoved mover and god
of Metaphysics 12 and both with the Platonic Demiurge, his own second principle.
(This meant taking the divine activity of Aristotle’s Intellect to be thinking—and so
constituting—the Platonic Forms; see Section 3 of this chapter.) In both cases, it is
notable that Plotinus regarded his predecessors as providing brilliant, but incomplete,
insights: the Stoic principles constitute only the second, rational, grade of virtue because
(Plato showed that) there is a higher virtue corresponding to our life as intellects; the
Aristotelian god is only the second principle because there must be a higher cause of
the unity of Intellect.58
Porphyry is known primarily as the disseminator of Plotinus’ doctrines through his
edition of the Enneads and in his own philosophical works. But the remarkable range of
his scholarship led him to two further acts of integration that had profound effects on
later Platonism. One was the salvaging of Aristotle’s Categories from Plotinus’ critique in
Enneads 6.1–3 by interpreting it as a work of “logic” rather than metaphysics and writing
an introduction and two separate commentaries on it. This allowed him to advocate a

54 Plotinus claimed that his views merely echoed those of his teacher in Alexandria, the shadowy
Ammonius Saccas, and Porphyry adverts explicitly to the use of Numenius’ and other philosophers’
works in Plotinus’ classes; see Porphyry’s Life of Plotinus 3 and 14.
55 See Section 3 of this chapter. Plotinus’ method in the Enneads is often to start with a critical
review of Epicurean, Stoic, and Peripatetic views on a topic before moving the argument to a higher
level that resolves the difficulties those views involve; see, e.g., Enn. 5.9 or 1.4. But he also directly
attacks, for example, Stoic materialism in Enn. 4.7, and the Aristotelian doctrine of the categories in
6.1–3, as well as the Gnostics in Enn. 2.9. Porphyry’s Against the Christians became a notorious work;
for controversial works against philosophical opponents, see, e.g., fr. 240–55 Smith.
56 See Frede, “Epilogue,” 793–97, citing Longinus’ lament at the state of non-Plotinian philosophy in
Porphyry’s Life of Plotinus 20.
57 See note 9 above.
58 The Aristotelian background of Plotinus’ doctrine of Intellect is reviewed in, e.g., S. Menn,
Descartes and Augustine (Cambridge, 1998), 73–129, and Gerson, Plotinus, part 1. Plotinus identified the
higher cause as the One of the Parmenides, which he took to be the same as the Good in Republic 509b,
by construing Plato’s phrase “beyond being” as meaning “above the Forms, i.e. above the divine
Intellect.”
 Plato and Platonism   683

strong version of the thesis of “the harmony of Plato and Aristotle”—that is, the claim
that, once properly understood, the two philosophers can be seen to be working
toward the same (Platonic) goal—which inspired the subsequent Platonist tradition of
sympathetic commentary on Aristotle (whose treatises were thereafter read as a propae-
deutic to the study of Plato).59 The second was his attempt to subjoin a set of religious
texts and practices to philosophy, as a way of offering a limited measure of ethical
progress to non-philosophers and counteracting the Christian alternative.60
Christianity became an increasingly significant force in the Roman empire in the
generations after Porphyry’s death in 305 a.d., as it was gradually adopted as the religion
of the imperial elite. In the end, this led to the dissolution of Platonism as an independ-
ent philosophical “movement” in the sixth century a.d., as its teaching was forbidden
and its institutions closed down. (Its ideas survived, however, in the new forms of
Christian and, later, Islamic philosophy.) But the 250 years between these events were
anything but a period of decline. Three notable innovations—all of them due in the first
instance to Iamblichus (c. 245–325 a.d.)—seem to characterize later Platonism.61
The central philosophical change was a major revision of Plotinian metaphysics,
involving the postulation of a new series of “unparticipated” principles to ground the
“participated” principles of Plotinus’ three hypostases, Soul, Intellect, and the One. The
causal theory driving this innovation led to the proliferation of mediating entities
between principles at different ontological levels, which are the subject of the increasingly
complex metaphysical and theological arguments in the extant later work of Proclus and
Damascius. A second change was the development of a systematic theory of “theurgy”—
that is, a theory of ritual or magical practices drawing on Porphyry’s first steps in this
direction—to bridge the gap the new metaphysics introduced between the increasingly
transcendent first principles and our limited capacity for intellection.62 The effect of this

59 Plotinus’ and Porphyry’s work on the Categories is examined in S. Strange, “Plotinus, Porphyry,
and the Neoplatonic Interpretation of the ‘Categories,’ ” in W. Haase and H. Temporini (eds.), Aufstieg
und Niedergang der Römischen Welt, Part 2, vol. 36.2 (Berlin, 1987), 954–74, and S. Ebbesen, “Pophyry’s
Legacy to Logic: A Reconstruction,” in R. Sorabji (ed.), Aristotle Transformed (Ithaca, N.Y., 1990),
141–71; see also Barnes, Porphyry: Introduction. The history of the harmony thesis has not been fully
studied; two recent and sympathetic (but controversial) studies are L. Gerson, Aristotle and Other
Platonists (Ithaca, N.Y., 2005), and G. Karamanolis, Plato and Aristotle in Agreement? (Oxford, 2006).
60 Porphyry’s religious philosophy is partially elucidated in his extant On Abstinence and Life of
Pythagoras. But we have fragments of numerous lost works, and one, The Return of the Soul, seems to
have contained a program describing various ethnic rituals as a means for moral improvement that
could constitute an alternative to the Christian “universal way”; see, e.g., fr. 283–302 Smith (which,
however, are largely drawn from Augustine’s polemic against Porphyry in City of God 10). See Smith,
Porphyry’s Place, Part 2.
61 For an overview of Iamblichus’ work, see J. Dillon, “Iamblichus of Chalcis (c. 240–325 a.d.),” in
W. Haase and H. Temporini (eds.), Aufstieg und Niedergang der Römischen Welt, Part 2, vol. 36.2
(Berlin, 1987), 862–909.
62 The theory is set out in Iamblichus’ De Mysteriis; see, e.g., Dodds, Greeks, 282–311, and the more
sympathetic treatment in Smith, Porphyry’s Place, Part 2. On the later Platonists’ rejection of Plotinus’
optimism about our intellectual abilities, see C. Steel, The Changing Self—A Study on the Soul in Later
Neoplatonism: Iamblichus, Damascius and Priscianus (Brussels, 1978), and J. Finnamore and J. Dillon,
Iamblichus De Anima (Leiden, 2002).
684   Charles Brittain

theory on the religious culture of the period was immense, as we can see from the failed
attempt of the emperor Julian to establish a theurgical Platonism as the state religion (in
360–63 a.d.). At first sight, its effect on Iamblichian philosophy looks severe, since it
seems to introduce a radical shift toward an unargued and philosophically uninteresting
mysticism. But its purpose was not to reject philosophical inquiry but to supplement it
with something like a theory of divine “special grace.” This may strike some modern
readers as unhelpful, but at bottom it is no more strange than some of the presupposi-
tions of Christian philosophy (at least in its ancient forms). The third change introduced
by Iamblichus was an intensified interest in commentary on Platonic dialogues, driven
by his search for deeper philosophical interpretations than his predecessors had offered.
Iamblichus’ innovations were enthusiastically embraced and refined by the remarkable
sequence of Platonists who studied and taught at the schools in Athens and Alexandria
from c. 400 to 529 and 611 a.d. (respectively). The massive extant output of these philos-
ophers defies any easy summary.63 But we can conclude this rapid survey by noting two
striking features of their work. The first is that it is no longer profoundly shaped by the
need to defend Platonism against “Hellenistic” materialism, empiricism, or skepticism.
This does not mean that it had no external opposition; but its (usually unspoken) philo-
sophical opponents were Christians who shared most of its central metaphysical pre-
suppositions. The result is a more hermetic or closed form of philosophy, which, like
much of the best philosophical work of our own time, is often inaccessible if one is not
already immersed in the tradition. The second is that a great deal of it is in the form of
philosophical commentary on Aristotle. The explanation for this is that the later
Platonists devised a set curriculum of works of increasing difficulty to provide the rigor-
ous training their advanced philosophical work presupposed. Since they accepted, in
varying degrees, Porphyry’s thesis of the harmony of Plato and Aristotle, but regarded
Aristotle’s work as primarily dealing with relatively easy topics, the curriculum included
the systematic study of his treatises as a preparation for understanding the more com-
plex metaphysical thought of Plato.64 It is perhaps unclear whether we should celebrate

63 For extended summaries of their thought, see note 3 above. The best starting point for reading the
later Platonists is Proclus’ Elements of Theology, which was translated with commentary by E. Dodds
(Oxford, 1933). The Platonist commentaries on Aristotle are being steadily translated into English in the
series The Greek Commentators on Aristotle, edited by Richard Sorabji. Some of the major
commentaries on Plato have been translated into modern English; see, e.g., L. Westerink, The Greek
Commentaries on Plato’s Phaedo (Amsterdam, 1976–1977); G. Morrow and J. Dillon, Proclus’
Commentary on Plato’s Parmenides (Princeton, N.J., 1987); and D. Baltzly, Proclus: Commentary on
Plato’s Timaeus, vol. 3 (Cambridge, 2007). There are also excellent French translations of Proclus’
Timaeus and Republic commentaries by A-J. Festugière and, more recently, a steady stream of Budé
editions of his works. The older English translations by Thomas Taylor are more reverent than helpful.
64 The late Platonist curriculum started with introductory ethical works, such as Epictetus’
Handbook or the Pythagorean Golden Verses—on which we have extant commentaries by Simplicius
and Hierocles, respectively (see note 14 above, and H. Schibli, Hierocles of Alexandria [Oxford, 2002]).
The second stage covered Aristotle, especially the “logical works” in the Organon, the Physics,
Metaphysics, and De anima (for all of which we still have extensive Platonist commentaries). The final
stage covered Plato and focused on metaphysics and theology (and especially the Timaeus and
Parmenides). The curriculum and its connection with the commentary tradition is described in detail
in L. Westerink, Anonymous Prolegomena to Platonic Philosophy (Amsterdam, 1962), and Westerink,
 Plato and Platonism   685

or regret the effects of this curricular decision on the history of medieval philosophy,
but this (along with major strands in Christian and Islamic philosophy) was the legacy
of Platonism.

3. Epistemological Foundations

In Section 1 of this chapter, I suggested that this broad philosophical tradition was
unified by its commitments to the authoritative status of Plato’s work and a set of shared
assumptions about its fundamental significance; in Section 2, I argued that the first
commitment is best understood as a reaction to the philosophical failures of the
Hellenistic schools. In this section, I sketch some of the ways in which Platonists elabo-
rated the shared assumptions of the second commitment and defended them against
their “Hellenistic” rivals.
The basic Platonist assumptions have been crudely characterized as two claims:
(1) the world is only intelligible through the structure and order imposed on it by
ontologically prior immaterial principles, and (2) knowledge accordingly presupposes
(non-empirical) cognitive access to these immaterial principles. The Platonists defended
these general claims both positively, by setting out a causal theory to justify (1) and an
epistemology to explain (2), and negatively, by arguing for the inadequacy of any mate-
rialist alternative to (1) or any empiricist alternative to (2). This section concentrates on
the second claim, however, since the defense of the first depends on the possibility of our
having access to the principles, and the arguments for (2) tell us something about the
nature of at least two of these principles: soul and intellect.65
Platonist epistemology depends on a basic set of positive claims about knowledge
derived from a (selective) reading of the Platonic dialogues.66 Three central theses are:

(a) The primary objects of knowledge are immaterial principles (starting with the
Forms).
(b) It is in principle possible to apprehend these objects (or at least the Forms)
directly through a nonrepresentational faculty of “intellection.”
(c) This possibility presupposes that embodied cognitive agents start off with a store
of non-empirical information—that is, some form of innate “knowledge.”

“The Alexandrian Commentators and the Introductions to Their Commentaries,” in R. Sorabji (ed.),
Aristotle Transformed (Ithaca, N.Y., 1990), 325–48.
65 On (1), see, e.g., L. Gerson, “Neoplatonism,” in C. Shields (ed.), The Blackwell Guide to Ancient
Philosophy (Oxford, 2003), 303–23. Proclus’ Elements of Theology is a clear exposition of the major later
Platonist arguments for (1); some notable earlier arguments are given by Numenius in On the Good
(fr.1–22 Des Places), Plotinus (see note 55 above), and Augustine (De Trinitate 9–10).
66 Platonist epistemology tended to start from the recollection doctrines of the Phaedo and
Pheadrus, rather than from the apparently belief-centered accounts of knowledge in the Meno and
Theaetetus, because the former lend themselves more readily to Platonist interpretations of the
metaphysical and psychological theories of Republic 5–7 and 10 and the Timaeus.
686   Charles Brittain

Since thesis (a) is a fundamental principle of Platonist metaphysics, and (b) amounts
to the claim that knowledge is possible, controversy within Platonism centered on (c).
The Platonic texts underlying this third thesis—the recollection argument in the Phaedo,
and its further elaboration, as the Platonists saw it, in the Phaedrus and Timaeus—raise a
number of general problems for any interpretation of Plato. Two particularly salient
questions concern the relations between perceptual (or “doxastic”) experience and
intellectual knowledge in Plato’s epistemology, and between the cognitive agent (the
soul) and its epistemic objects (the principles). Theories that address these questions
could, and Platonist theories did, take many forms, but a brief sketch of two general
responses and their reception within the tradition should suffice to indicate some of
the ways in which Platonism evolved.
The context for these Platonist theories, however, is not the relatively blank epistemo-
logical slate that Plato filled in with his theory of recollection but, rather, the developed
empiricist alternatives that the Hellenistic schools had devised in reaction to it. The
dominant alternative was the Stoic theory, which derived knowledge entirely from per-
ception, both directly (through the theory of the “cognitive” impression) and indirectly
via a process of natural concept formation based on secure perception.67 The latter proc-
ess provided a set of contentful “common” or basic conceptions—including concepts of
natural kinds, as well as logical notions such as consequence and inconsistency—that
the Stoics took to constitute our rationality.68 On their view, systematic knowledge of the
world could be acquired through a process of “articulating” the contentful conceptual
knowledge contained in the empirically acquired conceptions we apply in ordinary
experience, and using this as the basis for further empirical inquiry. The Stoics thus
denied the need for a Platonic theory of recollection—and, since they also construed the
basic principles of the world, including god and the soul, as material entities, they
rejected not just thesis (c), but also the existence of both transcendental objects of
knowledge (a) and an immaterial soul that might know them (b).
The first challenge for any Platonist theory of recollection (at least prior to late
Platonism, i.e., c. 300 a.d.) was accordingly to clear away the supposed confusions of
Stoic empiricism. The basis for this task was, unsurprisingly, the rather disparate set of
Platonic arguments designed to show the intrinsic deficiency of doxastic experience—
for instance, that the faculty of perception is conceptually impoverished in certain ways,
or that perceptual experience is fallible and restricted to the cognition of material

67 See, e.g., M. Frede, “Stoic Epistemology,” in K. Algra, J. Barnes, J. Mansfeld, and M. Schofield
(eds.), The Cambridge History of Hellenistic Philosophy (Cambridge, 1999), 295–322, and Frede, “The
Stoic Conception of Reason,” in K. Boudouris (ed.), Hellenistic Philosophy vol. 2 (Athens, 1994), 50–63.
Early Platonists were also concerned to argue against Peripatetic and Epicurean epistemologies (see
Plutarch fr. 215f Sandbach), but the surviving texts focus on the Stoics and their Academic rivals.
68 Some Stoic sources suggest that the theory depended on innate dispositions to form the
appropriate set of conceptions, which included moral concepts that may not be derivable from simple
perception in early childhood; see D. Scott, Recollection and Experience (Cambridge, 1995), 157–210,
and the articles in note 67 above.
 Plato and Platonism   687

objects or qualities.69 But this common stock of arguments could be mined to support a
wide range of positive theories, grounded in very distinct conceptions of perception and
the soul. The second challenge for Platonist theories of recollection was to deflate the
threat of skepticism, which had plagued the Stoics throughout the Hellenistic era and
remained a serious obstacle to any positive epistemology.70 In this case, however, the
Platonists shared a strategy, derived from their mutual acceptance of the thesis that
divergence from Platonic doctrine inevitably leads to intractable disagreements that can
only be solved by an unqualified return to Plato (see Section 2 of this chapter): they
argued that skepticism picks up on a fundamental problem of Stoic empiricism: viz., its
dependence on representations (phantasia) that yield only mediated or indirect appre-
hensions of their objects. The skeptics were correct to argue that such a theory can never
provide secure knowledge, but their own skepticism is parasitic on just this theory and
falls with it in the face of the Platonist alternative(s).71 As a result, Platonists were able to
supplement their general Platonic criticisms of doxastic experience with a battery of
more specific arguments drawn from the skeptical Academic (and, later, Pyrrhonist)
critique of Stoic epistemology.72
A number of early Platonists seem to have spelled out the Platonic doctrine of recol-
lection as a form of “cognitive dualism” modeled on the sharp distinction between dox-
astic and epistemic cognitive states in Timaeus 28–29.73 The heart of this view—which is

69 The primary Platonic sources for these criticisms of perception are the Phaedo and Theaetetus
(esp. 184–86), and Timaeus 28–29 and Republic 6–7. An early Platonist catalogue of such arguments is
given by Alcinous, Handbook ch. 25.3; individual elements are appealed to passim; see, e.g., Plutarch
Against Colotes 1116 a–b and 1118a–b.
70 Skepticism was a live option in the second century a.d., both in the form of the revived
Academicism of Favorinus (attested in Galen, On the Best Teaching Method) and in its new guise of
Pyrrhonism; see Brittain, “Scepticism.”
71 Traces of this strategy in early Platonism are discernible in Numenius and Alcinous; see Brittain,
“Scepticism,” and Boys-Stones, “Alcinous.” It becomes explicit in Plotinus; see Enn. 5.5 (and, e.g.,
Augustine Trin. 15.21), and D. O’Meara, “Scepticism and Ineffability in Plotinus” [“Scepticism”],
Phronesis 45.3 (2000), 240–51.
72 For the Academic criticisms of Stoic epistemology, see the articles cited in notes 22 and 31 above.
Their deployment by Platonists is attested in Plutarch’s controversial anti-Stoic works, Stoic
Contradictions and Common Conceptions (passim), and other early Platonist texts such as Plutarch fr.
215 (Sandbach), Anon. in Theaet., e.g., cols. 3.7–15 and 61, and Numenius fr. 24–8 (Des Places), as well
as in, e.g., Plotinus Enn. 4.6 and 4.7, Porphyry On the Capacities of the Soul (fr. 251–54 Smith) and
Augustine Against the Academics 3.26. These Academic arguments (or Academic developments of
Platonic criticisms) are often found in tandem with the Platonic criticisms cited in note 69 below.
73 Our sketchy evidence for the period makes it difficult to know how widely shared this sort of
theory was. I use Alcinous because recent work on this and related texts has made it possible to get a
fairly detailed picture of his theory. On the structure of Handbook ch. 4 and its relation to Platonic
texts, see D. Sedley, “Alcinous’ Epistemology,” in K. Algra, P. van der Horst, and D. Runia (eds.),
Polyhistor (Leiden, 1996), 300–12. On the Stoic and skeptic context that Alcinous aims to undermine,
see G. Boys-Stones, “Alcinous, Didaskalikos 4: In Defence of Dogmatism” [“Alcinous”], in M. Bonazzi
and V. Celluprica (eds.), L’eredità platonica: Studi sul platonismo da Arcesilao a Proclo (Naples, 2005),
203–34. Sedley, “Three Interpretations,” rightly contrasts Alcinous’ views and those of the anonymous
Commentary on the Theaetetus; but it seems to me that their differences are essentially over how to
interpret the dialogues rather than over the epistemological theory they find in it. (The term “cognitive
688   Charles Brittain

best attested in the sketch of Platonic epistemology in Alcinous’ Handbook, chapter 4

(from c. 150 a.d.)—is the thesis that our cognitive faculties flow into two autonomous
systems, based on distinct sources and objects. The doxastic system is based on the faculty
of perception and geared only for the cognition of material objects and states of affairs.
But, Alcinous argues, our ordinary cognitive lives—our abilities to make perceptual
judgments and to reason about empirical objects—are nonetheless fully explicable in
terms of evolved doxastic capacities derived from perception, memory, and a set of
empirical concepts they give rise to.74 The epistemic system, on this view, is thus not
required to explain ordinary cognition but, rather, only how philosophical knowledge is
possible: that is, how we can acquire the knowledge of immaterial objects (essences
and Forms) that is inaccessible to the doxastic system. Alcinous infers that its basis
must be some form of innate “knowledge” derived from our pre-incarnate intellection
of the Forms, as Plato suggested. But he elaborates this in terms of our innate posses-
sion of a second set of abstract “natural conceptions,” which give rise to our capacity
for epistemic reasoning. Philosophical inquiry, on this account, is a matter of actual-
izing or “recollecting” the (abstract and nonrepresentational) content latent in these
innate conceptions through a non-empirical process of “articulation,” described by
Alcinous as applying the methods of Platonic “dialectic”; when successful, it leads to
intellection of Forms.75
We can see even from this cursory sketch of cognitive dualism that it was conceived as
a theory of recollection that could replace the dominant Stoic epistemology of Alcinous’
time.76 The cognitive dualist preserves the Stoic view that our higher (epistemic) capaci-
ties are constituted by a set of contentful concepts, but it substitutes a Platonic innatist
explanation of their content for the Stoics’ empirical theory, which effectively severs the
connection between experience and philosophical knowledge in the Stoic theory. It is
also clear that this compromise is achieved at some cost. We can note two problems that
attracted the attention of later Platonists. The first is that the cognitive dualist’s rich
conception of “doxastic reason” seems to presuppose some complex rational capacities
that are not easily explicable by a Platonist without recourse to concepts derived from
“epistemic reason.”77 But once the cognitive dualist allows that our “natural conceptions”

dualism” is owed to Sedley, “Three Interpretations,” p. 91, who notes that Scott, Recollection, argues for a
version of this view as the correct interpretation of Plato’s theory in the Phaedo and elsewhere.)
74 Alcinous Handbook ch. 4.3–5. Alcinous’ theory of “doxastic reason” draws on the discussions of
perception and memory in the Theaetetus and Philebus; see Sedley, “Alcinous’ Epistemology.”
75 The process is called “articulation” in, e.g., Plutarch Platonic Question 1 and Anon. in Theaet. cols.
46–47. Alcinous’ description of “dialectic” and mathematics in chs. 5–7 covers a confusing mixture of
methods of argument employed by Plato. The core of it, however, is a version of the doctrine of
cognitive “ascent” suggested by the Line simile in Republic 6–7.
76 Owing to its form, Alcinous’ work does not make the polemical context explicit, but its anti-Stoic
intentions are clear from some arguments in the work and the context provided by his contemporaries;
see Boys-Stones, “Alcinous.” Similar but more palpably anti-Stoic views are found in, e.g., Plutarch fr.
215 (Sandbach); Anon. in Theaet., e.g., cols. 3.7–15 and 61; and Numenius fr. 24–8 (Des Places).
77 The clearest example is the case of moral reasoning, mentioned by Alcinous at the end of
chapter 4, which is naturally construed as involving the application of non-empirical moral concepts of
 Plato and Platonism   689

play a role in ordinary cognitive life, it is hard to see how he can maintain the firewall
between perception and intellection.78 The road is open for his critics to argue either
that such “natural concepts” are derived from perceptual experience by abstraction (the
Stoic model) or that all perceptual experience presupposes thick conceptualization
(Plotinus’ model). Cognitive dualism thus looks like an inefficient theory that unneces-
sarily undermines the unity of the soul.79
A second problem for cognitive dualism is that its explanation for the soul’s epistemic
potential in terms of innate but latent “natural conceptions” seems too weak to secure
knowledge. The cognitive dualist presupposes that once the process of recollection has
been triggered, the “articulation” of these natural conceptions is sufficient to culminate
in intellection of the Forms. But since the connection between the embodied soul and
the Forms is indirect—it depends on the partial grasp of the Forms (the natural concept)
being retained unconsciously and reactivated at a temporal distance—it is unclear that it
can yield unmediated knowledge. Most early Platonists seem to have assumed that their
position was immune to skepticism because the epistemic system was posited precisely
to escape the limited and mediated apprehension provided by doxastic representation
(phantasia).80 But later Platonists were less confident that the mere avoidance of percep-
tual or imaginative representation sufficed to avoid a mediated grasp of an object through
an “image” or “likeness”—that is, a derivative and incomplete form of knowledge.
The force of these objections to cognitive dualism is particularly evident in Plotinus’
radical revision of Platonist epistemology. The Enneads are primarily concerned with
the interrelations between the divine principles (Soul, Intellect, and the One or Good);
but Plotinus devoted a number of his psychological treatises to the elaboration of a
novel, and extremely complex, theory of perceptual experience.81 Plotinus rejected cog-
nitive dualism on the basis of an analysis distinguishing the physiological, sensory, and
judgmental aspects of perception. On his realist view, human perception as such is the

value. (Handbook 4.8 is confusing, because it seems to conflate the cognitive lives of ordinary and
perfect moral agents.)
78 In Handbook 25.5, Alcinous describes “irrational souls” as driven by perceptual representation
(phantasia) and deprived of reasoning, judgment, theorems, and universal apprehensions. In the
human case, it is not clear how doxastic reason can achieve, e.g., judgment and reasoning, on his view,
without actively applying “epistemic” concepts such as “truth” and “being” (see Theaet. 184–86).
79 Alcinous accepts the standard Platonist view that the soul is essentially an intellectual cognitive
agent constituted by something like the dualist’s “epistemic reason” (Handbook 14.1–2, cf. Plato, Timaeus
41c–42 and note 82 below). But the division of “reason” into two discrete systems, when combined with
the assumption that the soul’s nonintellectual faculties are not essential to it because they depend on
the body (Handbook 25.5; cf. Tim. 69–72), implies that ordinary cognitive life does not involve the soul’s
essential activity.
80 See note 71 above. Plutarch, however, may be an exception, since his “Academic Platonism”
sometimes drives him to allow that knowledge may not be possible (at least in this life); see his Platonic
Questions 1, and note 50 above.
81 See, e.g., Enn. 4.3–6. Excellent general accounts of Plotinus’ epistemology are provided in Gerson,
Plotinus, esp. 164–84, and E. Emilsson “Cognition and Its Object,” in L. Gerson (ed.), The Cambridge
Companion to Plotinus (Cambridge, 1996), 217–49. Emilsson’s Plotinus on Sense-Perception (Cambridge,
1988) is a brilliant reconstruction of the complexities of Plotinus’ theory.
690   Charles Brittain

direct apprehension of an object or quality, i.e., the judgment that it is there. But, while
this judgment is occasioned by a (psychological or nonmaterial) sensation triggered by
a physiological process, it is realized only through the activation of the “account” of the
object or quality—a logos or concept—that the agent already possesses. For if, as Plotinus
argues, the physiological processes involved in perception are in principle incapable of
transmitting cognitive information, these concepts cannot be generated by abstraction
from more primitive sensory inputs. Ordinary perception thus presupposes that the
soul, the cognitive agent, has the active use of the set of “natural conceptions” that
the cognitive dualist excluded from empirical experience: it is essentially structured
by “innate” concepts.
Plotinus’ rejection of Stoic empiricism is thus more radical than the cognitive dualist’s.
His interpretation of the theory of recollection takes it to apply to cognition quite
generally: all forms of thinking, including perception, presuppose that the soul actively
employs the set of innate “accounts” or concepts that makes them possible. (The theory
is based on Plotinus’ understanding of Phaedrus 249b–c and Timaeus 37a–c.) But
Plotinus’ theory of ordinary cognitive life is also closer to the Stoic model of rationality
than the cognitive dualist’s: since the (rational) soul is constituted by these immaterial
concepts, its existence depends on their continual activation—that is, thinking in
one form or another.82 A consequence of his positive re-evaluation of ordinary expe-
rience, however, was the downgrading of the philosophical achievement of conscious
“recollection”—the startling grasp on “equality itself ”—examined in the recollection
argument in Phaedo 72–78. Recollection in this sense is not the intellection of a Form,
as earlier Platonists had thought but, rather, a matter of “comparing new tokens with
old ones”: that is, becoming conscious of the content and interrelations of abstract
concepts the soul already has (or is).83 Plotinus accepted that “turning inside” in this
way was a necessary stage toward intellection; but he argued that an explicit or “artic-
ulated” conceptual grasp on an immaterial object was still an “image” or “likeness” of it,
albeit one at a higher and more abstract level than a merely perceptual or imaginative
representation. If the soul is able to get beyond the derivative knowledge provided by
the likenesses of Forms at the rational or conceptual level, it must have direct access to
the Forms themselves.
But Plotinus’ model for direct access to the Forms is the knowledge of the divine
Intellect. As noted in Section 2 of this chapter, Plotinus followed Aristotle in taking this
Intellect to be constituted by its intellection of itself, and he took the Forms to be the

82 Like Alcinous, Plotinus took the Timaeus account of the numerical composition of the (rational
and immortal) soul (Tim. 35a–37c and 41a–42d) to imply that it is just a set of logoi—“ratios” or
“accounts”—which were standardly construed as concepts of the Forms. For Platonists, however, the
existence or life of immaterial substances such as the soul consists in their activity.
83 See Enn. 5.3.2 and 4.3.25. Plotinus’ explicit treatments of recollection are unsystematic, however:
he doesn’t distinguish lexically between ordinary “recollection” (memory) and Platonic “recollection,”
and, since his various discussions of the latter take off from various Platonic texts, he doesn’t always
remember to reconcile them with his “downgraded” interpretation of the Phaedo-type. (Enn. 1.2.4 and
3.5.1, for instance, describe the ascent to an actual vision of the Forms via beauty in the Phaedrus and
Symposium as “recollection.”)
 Plato and Platonism   691

content of its thought, i.e., what it intellects, or itself.84 The metaphysical significance
of his identification of Forms (real being) with the Intellect is, of course, at the heart of
Plotinus’ Platonism. But our concern here is just with the crucial epistemological
implication that Plotinus argues for in Enn. 5.3: viz., that the knowledge that constitutes
both reality and the Intellect must be an eternal activity of self-knowing.85 A remarkable
feature of taking divine self-knowing as the paradigm of knowledge—and one that
Plotinus notes in Enn. 5.5.1–3—is that it eradicates any possibility of mediation or repre-
sentation between the cognitive subject and object (since the two are identical). But
once that possibility is removed, there is no further scope for skepticism: the Intellect’s
existence just is knowledge. (We might doubt the existence of the divine Intellect, of
course, but to do so is to undermine the intelligibility of the world, and hence the possi-
bility of thought itself, skeptical or otherwise.)
Thus, if the soul is able to realize non-derivative knowledge, and to do so requires
direct access to the Forms, the soul must be able to “conform to” or “become” Intellect.
And since Plotinus believes that the soul’s capacity for virtue demonstrates that it can
have genuine knowledge, he infers that it already has an intellectual capacity over and
above the “accounts” that make it a rational substance. But the only possible ground for
this capacity is the (active) Intellect itself, which is continually “writing” these accounts
on the soul (Enn. 5.3.4). Plotinus concludes that the soul is in fact always directly access-
ing the Forms, although we are not conscious of this: the soul does not entirely “descend”
from Intellect, its cause (Enn. 4.8.8).86
Plotinus’ theory of recollection met a mixed reaction in the later Platonist tradition.
The thesis that ordinary cognitive experience presupposes the soul’s prior possession of
at least some non-empirical structuring concepts was widely accepted.87 The later
Platonists’ rejection of cognitive dualism was qualified, however, by their develop-
ment of complex theories of empirical concept formation and a variety of extremely
rich and subtle analyses of perception, which remain largely unexplored in modern
scholarship.88 But the radical core of Plotinus’ interpretation of recollection—the thesis

84 For an introduction to Plotinus’ extraordinarily complex theory of intellect, see O’Meara,

Plotinus, 33–53, and Gerson, Plotinus, 42–78. A detailed reconstruction in English is forthcoming in
E. Emilsson’s Plotinus on the Intellect (Cambridge, 2007). On the background to Plotinus’ theory, see
notes 53 and 58 above.
85 The argument is examined in Emilsson’s “Cognition and Its Object” and I. Crystall, “Plotinus on
the Structure of Self-Intellection,” Phronesis 43 (1998), 264–87.
86 See S. Menn, “Plotinus on the Identity of Knowledge with Its Object,” Apeiron 34.3 (2001), 233–46.
As Menn shows, Plotinus thought that this heterodox doctrine was necessary to resolve the “greatest
difficulty” in Parmenides 133–34; he based his resolution on Sophist 238c.
87 See, e.g., Damascius’ analyses of the recollection argument in his Commentaries on the Phaedo
I.253–310 (esp. 1.274) and II.4–20 (esp. II.15). Similar views are expressed by Syrianus in On Aristotle’s
Metaphysics, 104–07 (Kroll), and by Proclus—see note 9 above.
88 See, however, the works cited in note 59 above on the role of empirical and non-empirical
concepts in the acquisition of natural language, e.g., and I. Mueller, “Mathematics and Philosophy in
Proclus’ Commentary on Book I of Euclid’s Elements,” in J. Pépin and H. Saffery (eds.), Proclus: Lecteur et
interprète des anciens (Paris, 1987), 305–18, on the role of abstraction in Platonist geometry. Work on
the later Platonists’ theories of perception has tended to focus on the diverse interpretations of
692   Charles Brittain

that it is the means of realizing the direct and necessary connection between the soul
and the divine Intellect—was flatly rejected by the great majority of later Platonists.89
Plotinus’ theory struck them as naively optimistic about the nature of the fallen (i.e.,
human) soul and its relation to the hierarchy of transcendent higher principles: the
immense gap between the two introduced by Iamblichian metaphysics was in principle
unbridgeable by unaided natural reason.90
A brief survey of Platonist epistemology can do no more than scratch the surface of
this massive body of sophisticated philosophical work by indicating some of the general
features of its evolution. But perhaps this is enough to suggest the value of the Platonist
tradition for a philosophical reader, both as a stimulus for producing rival interpreta-
tions of Plato and as a challenging philosophical movement in its own right.91

Bibliography
Allen, J. “Carneades,” Stanford Encyclopedia of Philosophy, http://plato.stanford.edu/entries/
carneades (2004).
Annas, J. “Plato the Sceptic,” Oxford Studies in Ancient Philosophy Supplementary Volume
(Oxford, 1992), 43–72.
Annas, J. Platonist Ethics Old and New (Ithaca, NY, 1999).
Athanassiadi, P. Damascius: The Philosophical History (Athens, 1999).
Baltzly, D. Proclus: Commentary on Plato’s Timaeus, vol. 3 (Cambridge, 2007).
Barnes, J. “Antiochus of Ascalon,” in M. Griffin and J. Barnes (eds.), Philosophia Togata
(Oxford, 1989), 51–96.
Barnes, J. Porphyry: Introduction (Oxford, 2003).
Blumenthal, H. Soul and Intellect: Studies in Plotinus and Later Neoplatonism (London, 1993).
Boys-Stones, G. “Alcinous, Didaskalikos 4: In Defence of Dogmatism,” in M. Bonazzi and
V. Celluprica (eds.), L’eredità platonica: Studi sul platonismo da Arcesilao a Proclo (Naples,
2005), 203–34.
Boys-Stones, G. Post-Hellenistic Philosophy (Oxford, 2001).

Aristotle’s theory in the De Anima commentary tradition; but see, e.g., P. Lautner, “Some Clarifications
on Proclus’ Fourfold Division of Sense-Perception in the Timaeus Commentary,” in M. Perkams and
R. Piccione (eds.), Proklos: Methode, Seelenlehre, Metaphysik (Leiden, 2006), 117–35.
89 The standard criticisms deriving from Iamblichus are recorded by Proclus, for example, in his
commentaries on the Parmenides (4.946–50 Cousin) and Timaeus (3.333–34 Diehl), and in his Elements
of Theology (x211 Dodds). As these contexts suggest, the criticisms were grounded on the apparent
incompatibility of Plotinus’ views with Plato’s claims about the effects of embodiment on the human
soul in Tim. 43a–44c and about the possibility of knowledge of the principles in Parm. 133–34.
90 See Section 2 and note 62 above. Note, however, that Plotinus himself had pointed the way toward
the invocation of superintellectual experiences through his views on the possibility of “experiencing”
the ultimate principle, the One or Good.
91 In the case of epistemology, the stimulus is provided by the Platonists’ focus on reconciling Plato’s
disparate arguments into unitary theories. The evolution sketched previously could be rephrased as a series
of such attempts, centered, respectively, on the recollection argument in the Phaedo (for the cognitive
dualists), the psychology of the Phaedrus, and the theory of rationality in the Timaeus (for Plotinus),
and all three with the doctrine of principles in the Parmenides (for later Platonists, such as Proclus).
 Plato and Platonism   693

Boys-Stones, G. Platonist Philosophy 80 bce to ad 250. An Introduction and Collection of

Sources in Translation (Cambridge, 2018).
Brennan, T., and C. Brittain. Simplicius: On Epictetus’ Handbook (Ithaca, NY, 2002).
Brisson, L. (ed.) Porphyre: Sentences (Paris, 2005).
Brisson, L., and M. Patillon. “Longinus Platonicus Philosophus et Philologus,” in W. Haase
and H. Temporini (eds.), Aufstieg und Niedergang der Römischen Welt, Part 2, vol. 36.7
(Berlin, 1994), 5215–299.
Brittain, C. “Arcesilaus,” Stanford Encyclopedia of Philosophy, http://plato.stanford.edu/entries/
arcesilaus (2005).
Brittain, C. “Attention Deficit in Plotinus and Augustine,” Proceedings of the Boston Area
Colloquium in Ancient Philosophy 18 (2003), 223–63.
Brittain, C. “ ‘Middle’ Platonists on Academic Scepticism,” in R. Sorabji and R. Sharples (eds.),
Greek and Roman Philosophy, 100 bc—200 ad, vol. 2 (London, 2007), 297–315.
Brittain, C. Philo of Larissa (Oxford, 2001).
Brittain, C. “Philo of Larissa,” Stanford Encyclopedia of Philosophy, http://plato.stanford.edu/
entries/philo-larissa/(2006).
Brown, P. The Making of Late Antiquity (Cambridge, MA, 1978).
Crystall, I. “Plotinus on the Structure of Self-Intellection,” Phronesis 43 (1998), 264–87.
Deuse, W. Untersuchungen zur mittelplatonischen und neuplatonischen Seelenlehre (Mainz, 1983).
Dillon, J. “An Ethic for the Late Antique Sage,” in L. Gerson (ed.), The Cambridge Companion
to Plotinus (Cambridge, 1996), 315–35.
Dillon, J. The Heirs of Plato: A Study of the Old Academy, 347/274 b.c. (Oxford, 2003).
Dillon, J. “Iamblichus of Chalcis (c. 240–325 a.d.),” in W. Haase and H. Temporini (eds.),
Aufstieg und Niedergang der Römischen Welt, Part 2, vol. 36.2 (Berlin, 1987), 862–909.
Dillon, J. The Middle Platonists (London, 1977).
Dodds, E. The Greeks and the Irrational (Berkeley, 1951).
Dodds, E. Proclus: The Elements of Theology (Oxford, 1933).
Donini, P. “Testi e commenti, manuali e insegnamento: la forma sistematica e i metodi della
filosofia in età postellenistica,” in W. Haase and H. Temporini (eds.), Aufstieg und Niedergang
der Römischen Welt, Part 2, vol. 36.7 (Berlin, 1994), 5027–100.
Dörrie, H., and M. Baltes (trans.) Der Platonismus in der Antike, vols. 1–5 (Stuttgart, 1987–1998).
Ebbesen, S. “Pophyry’s Legacy to Logic: A Reconstruction,” in R. Sorabji (ed.), Aristotle
Transformed (Ithaca, NY, 1990), 141–71.
Edwards, M. Neoplatonic Saints: The Lives of Plotinus and Proclus by Their Students (Liverpool,
2000).
Emilsson, E. “Cognition and Its Object,” in L. Gerson (ed.), The Cambridge Companion to
Plotinus (Cambridge, 1996), 217–49.
Emilsson, E. Plotinus on Sense-Perception (Cambridge, 1988).
Emilsson, E. Plotinus on the Intellect (Cambridge, 2007).
Emilsson, E. Plotinus (London, 2017).
Finnamore, J., and J. Dillon. Iamblichus De Anima (Leiden, 2002).
Frede, D. “The Philosophical Economy of Plato’s Psychology: Rationality and the Common
Concepts in the Timaeus,” in M. Frede and G. Striker (eds.), Rationality in Greek Thought
(Oxford, 1996), 29–56.
Frede, M. “Epilogue,” in K. Algra, J. Barnes, J. Mansfeld, and M. Schofield (eds.), The Cambridge
History of Hellenistic Philosophy (Cambridge, 1999), 771–97.
Frede, M. “Numenius,” in W. Haase and H. Temporini (eds.), Aufstieg und Niedergang der
Römischen Welt, Part 2, vol. 36.2 (Berlin, 1987), 1034–75.
694   Charles Brittain

Frede, M. “The Stoic Conception of Reason,” in K. Boudouris (ed.), Hellenistic Philosophy,

vol. 2 (Athens, 1994), 50–63.
Frede, M. “Stoic Epistemology,” in K. Algra, J. Barnes, J. Mansfeld, and M. Schofield (eds.), The
Cambridge History of Hellenistic Philosophy (Cambridge, 1999), 295–322.
George Boys-Stones. Platonist Philosophy 80 bce to ad 250. An Introduction and Collection of
Sources in Translation (Cambridge University Press, 2018).
Gerson, L. Aristotle and Other Platonists (Ithaca, NY, 2005).
Gerson, L. “Neoplatonism,” in C. Shields (ed.), The Blackwell Guide to Ancient Philosophy
(Oxford, 2003), 303–23.
Gerson, L. Plotinus (London, 1994).
Gerson, Lloyd P. (ed.), Plotinus the Enneads. (Translated by George Boys-Stones, John M.
Dillon, Lloyd P. Gerson, R.A.H. King, Andrew Smith and James Wilberding). (Cambridge
University Press 2018)/Eyjólfur Emillsson, Plotinus (Routledge 2017).
Glucker, J. Antiochus and the Late Academy (Göttingen, 1978).
Gottschalk, H. “The Earliest Aristotelian Commentators,” in R. Sorabji (ed.), Aristotle
Transformed (Ithaca, NY, 1990), 55–81.
Haase, W., and H. Temporini (eds.) Aufstieg und Niedergang der Römischen Welt, Part 2,
vols. 36.1–7 (Berlin, 1987–1994).
Hadot, I. Le problème de néoplatonisme alexandrin: Hiéroclès et Simplicius (Paris, 1978).
Karamanolis, G. Plato and Aristotle in Agreement? (Oxford, 2006).
Lamberton, R. Homer the Theologian (Berkeley, CA, 1986).
Lautner, P. “Some Clarifications on Proclus’ Fourfold Division of Sense-Perception in the
Timaeus Commentary,” in M. Perkams and R. Piccione (eds.), Proklos: Methode, Seelenlehre,
Metaphysik (Leiden, 2006), 117–35.
Menn, S. Descartes and Augustine (Cambridge, 1998).
Menn, S. “Plotinus on the Identity of Knowledge with Its Object,” Apeiron 34.3 (2001),
233–46.
Morrow, G., and J. Dillon. Proclus’ Commentary on Plato’s Parmenides (Princeton, NJ, 1987).
Mueller, I. “Mathematics and Philosophy in Proclus’ Commentary on Book I of Euclid’s
Elements,” in J. Pépin and H. Saffery (eds.), Proclus: Lecteur et interprète des anciens (Paris,
1987), 305–18.
O’Meara, D. Plotinus: An Introduction to the Enneads (Oxford, 1993).
O’Meara, D. Pythagoras Revived (Oxford, 1989).
O’Meara, D. “Scepticism and Ineffability in Plotinus,” Phronesis 45.3 (2000), 240–51.
Opsomer, J. In Search of the Truth: Academic Tendencies in Middle Platonism (Brussels, 1998).
Opsomer, J., and C. Steel. Proclus: On the Existence of Evils (Ithaca, NY, 2003).
Owen, G. E. L. “The Place of the Timaeus in Plato’s Dialogues,” Classical Quarterly NS 3 (1953),
79–95.
Schibli, H. Hierocles of Alexandria (Oxford, 2002).
Scott, D. Recollection and Experience (Cambridge, 1995).
Sedley, D. “Alcinous’ Epistemology,” in K. Algra, P. van der Horst, and D. Runia (eds.),
Polyhistor. Studies in the History & Historiography of Ancient Philosophy (Leiden, 1996),
300–12.
Sedley, D. “The Origins of Stoic God,” in D. Frede and A. Laks (eds.), Traditions of Theology
(Leiden, 2002), 41–83.
Sedley, D. “Plato’s Auctoritas and the Rebirth of the Commentary Tradition,” in J. Barnes and
M. Griffin (eds.), Philosophia Togata II: Plato and Aristotle at Rome (Oxford, 1997), 110–29.
 Plato and Platonism   695

Sedley, D. “Three Platonist Interpretations of the Theaetetus,” in C. Gill and M. McCabe (eds.),
Form and Argument in Later Plato (Oxford 1996), 79–103.
Sedley, D., and G. Bastianini (eds.), Commentarium in Platonis Theaetetum, Corpus dei papiri
filosofici greci e latini, vol. 3 (Florence, 1995), 227–562.
Smith, A. Porphyry’s Place in the Neoplatonic Tradition (The Hague, 1974).
Sorabji, R. (ed.) Aristotle Transformed (Ithaca, NY, 1990).
Sorabji, R. The Philosophy of the Commentators 200–600 ad, 3 vols. (London, 2004).
Stead, C. Philosophy in Christian Antiquity (Cambridge, 1994).
Steel, C. The Changing Self—A Study on the Soul in Later Neoplatonism: Iamblichus, Damascius
and Priscianus (Brussels, 1978).
Strange, S. “Plotinus, Porphyry, and the Neoplatonic Interpretation of the ‘Categories,’ ” in
W. Haase and H. Temporini (eds.), Aufstieg und Niedergang der Römischen Welt, Part 2,
vol. 36.2 (Berlin, 1987), 954–74.
Tarrant, H. Plato’s First Interpreters (Ithaca, NY, 2000).
Tarrant, H. Scepticism or Platonism: The Philosophy of the Fourth Academy (Cambridge, 1985).
Tarrant, H. Thrasyllan Platonism (Ithaca, NY, 1993).
Wallis, R. Neoplatonism (1972; 2nd ed. London, 1995).
Westerink, L. “The Alexandrian Commentators and the Introductions to Their Commentaries,”
in R. Sorabji (ed.), Aristotle Transformed (Ithaca, NY, 1990), 325–48.
Westerink, L. Anonymous Prolegomena to Platonic Philosophy (Amsterdam, 1962).
Westerink, L. The Greek Commentaries on Plato’s Phaedo (Amsterdam, 1976–77).
 Bibliography

This bibliography lists some works that might be useful to readers who wish to pursue the
study of Plato further. It is a revised version of the Bibliography to be found in the first
edition of this work. I have focused mainly on fairly recent work and on work in English. The
chapters contained in this volume are not listed here, nor have I listed all or only the sources
they cite; nor do I cover every topic the various chapters discuss. I refer the reader to the
­bibliographies at the end of each chapter for many further references. Though many works are
relevant in more than one section, I generally cite each work just once.

Abbreviations
CCP Kraut, (ed.) Cambridge Companion to Plato
EA Lee, Mourelatos, and Rorty (eds.), Exegesis and Argument
EPS Benson, (ed.) Essays on the Philosophy of Socrates
FA Gill and McCabe, (eds.) Form and Argument in Late Plato
LL Nussbaum and Schofield, (eds.) Language and Logos
LSD Owen, (ed.) Logic, Science and Dialectic
MAP Gentzler, (ed.) Method in Ancient Philosophy
PKF Fine, (ed.) Plato on Knowledge and Forms
PS Vlastos, Platonic Studies
SS Vlastos, Socratic Studies
SPM Allen, (ed.) Studies in Plato’s Metaphysics
Fine (ed.), Plato 1 Fine, G. (ed.) Plato 1: Metaphysics and Epistemology
Fine (ed.), Plato 2 Fine, G. (ed.) Plato 2: Ethics, Politics, Religion, and the Soul
Kamtekar (ed.), Plato’s EAC Kamtekar, R. (ed.) Plato’s Euthyphro, Apology and Crito
Vlastos (ed.), Plato 1 Vlastos, G. (ed.) Plato 1: Metaphysics and Epistemology
Vlastos (ed.), Plato 2 Vlastos, G. (ed.) Plato 2: Ethics, Politics, and Philosophy of Art
and Religion
Vlastos, Socrates Vlastos, G., Socrates, Ironist and Moral Philosopher

Texts and Translations

The Greek text of Plato is collected in five volumes published by Oxford University Press
(Oxford Classical Texts), edited by J. Burnet (1900–1907). These are gradually being replaced
with new editions. Two volumes have appeared so far: volume 1, edited by E. A. Duke,
W. F. Hicken, W. S. M. Nicoll, D. B. Robinson, and J. C. G. Strachan (1995); and The Republic,
edited by S. R. Slings (2003). Greek texts, with facing English translations, are available in the
Loeb library; the translations are of varying quality. The Budé texts are also worth consulting;
they contain a Greek text with facing French translations, and short notes. The Clarendon
698   Bibliography

Plato series provides generally accurate translations into English along with detailed notes.
The books in this series include:
Gallop, D. Phaedo (Oxford: Clarendon Press, 1975).
Gosling, J. C. B. Philebus (Oxford: Clarendon Press, 1975).
Irwin, T. H. Gorgias (Oxford: Clarendon Press, 1979).
McDowell, J. Theaetetus (Oxford: Clarendon Press, 1973).
Mayhew, R., Laws 10 (Oxford: Clarendon Press, 2008).
Meyer, S., Laws 1 and 2 (Oxford: Clarendon Press, 2015).
Taylor, C. C. W. Protagoras (Oxford: Clarendon Press, 1976/1991).
Translations (by various authors) of all of Plato’s surviving work, as well as of the spuria and
dubia, are in:
Cooper, John M., and Hutchinson, D. S. (eds.) Plato: Complete Works (Indianapolis, IN:
Hackett, 1997).
Some of these translations appear in separate volumes, with introductions, notes, and
bibliographies:
Frede, D. Plato: Philebus (Indianapolis, IN: Hackett, 1993).
Gill, M. L., and Ryan, P. Plato: Parmenides (Indianapolis, IN: Hackett, 1996).
Nehamas, A., and Woodruff, P. Plato: Phaedrus (Indianapolis, IN: Hackett, 1995).
Nehamas, A., and Woodruff, P. Plato: Symposium (Indianapolis, IN: Hackett, 1989).
Reeve, C. D. C. Plato: Cratylus (Indianapolis, IN: Hackett, 1998).
White, N. P. Plato: Sophist (Indianapolis, IN: Hackett, 1993).
Zeyl, D. Plato: Timaeus (Indianapolis, IN: Hackett, 1999).
Another widely used collection is:
Hamilton, E., and Cairns, H. (eds.) The Collected Dialogues of Plato including the Letters
(New York: Pantheon, 1961; Princeton, NJ: Princeton University Press, 1971).
Various journals are devoted to ancient philosophy. These include Ancient Philosophy, Apeiron,
Oxford Studies in Ancient Philosophy, and Phronesis.
The reader new to Plato might begin with:
Benson, H. (ed.) Essays on the Philosophy of Socrates (New York: Oxford University Press,
1992). Cited as EPS.
Fine, G. (ed.) Plato 1: Metaphysics and Epistemology (Oxford Readings in Philosophy series)
(Oxford: Oxford University Press, 1999). Cited as Fine (ed.), Plato 1.
Fine, G. (ed.) Plato 2: Ethics, Politics, Religion, and the Soul (Oxford Readings in Philosophy
series) (Oxford: Oxford University Press, 1999). Cited as Fine (ed.), Plato 2.
Irwin, T. H. Plato’s Ethics (New York: Oxford University Press, 1995).
Kraut, R. (ed.) Cambridge Companion to Plato (Cambridge: Cambridge University Press,
1992). Cited as CCP.
Kraut, R. Socrates and the State (Princeton, NJ: Princeton University Press, 1984).
Meinwald, C. Plato (New York: Routledge, 2016).
Taylor, C. C. W. Socrates (Oxford: Oxford University Press, 1998).
Vlastos, G. (ed.) The Philosophy of Socrates (Garden City, NY: Doubleday Anchor Books, 1971).
Cited as Vlastos (ed.), Philosophy of Socrates.
Vlastos, G. (ed.) Plato 1: Metaphysics and Epistemology (Garden City, NY: Doubleday Anchor
Books, 1971). Cited as Vlastos (ed.), Plato 1.
 Bibliography    699

Vlastos, G. (ed.) Plato 2: Ethics, Politics, and Philosophy of Art and Religion (Garden City, NY:
Doubleday Anchor Books, 1971). Cited as Vlastos (ed.), Plato 2.
Vlastos, G. Socrates, Ironist and Moral Philosopher (Ithaca, NY: Cornell University Press, 1991).
Cited as Vlastos, Socrates.
White, N. P. Plato on Knowledge and Reality (Indianapolis, IN: Hackett, 1976).
Williams, B. A. O. “Plato: The Invention of Philosophy,” in his The Sense of the Past (Princeton,
NJ: Princeton University Press, 2006), 148–86.
Readers interested in the chronology of Plato’s dialogues might look at:
Brandwood, L. The Chronology of Plato’s Dialogues (Cambridge: Cambridge University Press,
1990).
Brandwood, L., “Stylometry and Chronology,” in CCP, 90–120.
Cooper, J. “Introduction” to Plato: Complete Works, vii–xxvi.
Kahn, C. “On Platonic Chronology,” in J. Annas and C. Rowe (eds.), New Perspectives on Plato,
Modern and Ancient (Cambridge, MA: Harvard University Press, 2003), ch. 4.
Young, C. “Plato and Computer Dating,” Oxford Studies in Ancient Philosophy 12 (1994),
227–50.
A valuable source for information on the people mentioned in Plato’s dialogues is:
Nails, D. The People of Plato: A Prosopography of Plato and Other Socratics (Indianapolis, IN:
Hackett, 2002).
There is dispute about whether the dialogues that are generally taken to be early represent the
thought of the historical Socrates. For the view that they do, see:
Vlastos, G. Socrates, Ironist and Moral Philosopher (Ithaca, NY: Cornell University Press, 1991).
For a more skeptical assessment, see:
Kahn, C. Plato and the Socratic Dialogue: The Philosophical Use of a Literary Form (Cambridge:
Cambridge University Press, 1986).
Nehamas, A. “Voices of Silence: On Gregory Vlastos’s Socrates,” Arion, 3rd ser., 2 (1992),
156–86.
See also:
Taylor, C. C. W. Socrates (Oxford: Oxford University Press, 1998).
On the question of why Plato wrote dialogues, see:
Frede, M. “Plato’s Arguments and the Dialogue Form,” in J. Klagge and N. Smith (eds.),
Methods of Interpreting Plato and His Dialogues (Oxford Studies in Ancient Philosophy,
suppl. vol.) (Oxford: Clarendon Press, 1992), 201–19.
This topic is also touched on in:
Irwin, T. H. “Plato: The Intellectual Background,” in CCP, 51–89.
Kraut, R. “Introduction to the Study of Plato,” in CCP, 1–50.
For discussion of literary and dramatic aspects of the dialogues, and for Plato on art, see:
Annas, J. “Plato on the Triviality of Literature,” in J. Moravcsik and P. Temko (eds.), Plato on
Beauty, Wisdom and the Arts (Totowa, NJ: Rowman and Littlefield, 1982), 1–28.
Blondell, R. The Play of Character in Plato’s Dialogues (Cambridge: Cambridge University
Press, 2002).
700   Bibliography

Ferrari, G. “Plato and Poetry,” in G. Kennedy (ed.), Cambridge History of Literary Criticism
(Cambridge: Cambridge University Press, 1989), 92–148.
Gould, T. The Ancient Quarrel between Poetry and Philosophy (Princeton, NJ: Princeton
University Press, 1990).
Janaway, C. Images of Excellence: Plato’s Critique of the Arts (Oxford: Oxford University
Press, 1995).
Klagge, J., and Smith, N. (eds.) Methods of Interpreting Plato and His Dialogues (Oxford Studies
in Ancient Philosophy, suppl. vol.) (Oxford: Clarendon Press, 1992).
Moravcsik, J. “On Correcting the Poets,” Oxford Studies in Ancient Philosophy 4 (1986), 35–47.
Murdoch, I. The Fire and the Sun: Why Plato Banished the Artists (Oxford: Clarendon Press, 1977).
Nehamas, A. “Plato on Imitation and Poetry in Republic X,” in J. Moravcsik and P. Temko
(eds.), Plato on Beauty, Wisdom and the Arts (Totowa, NJ: Rowman and Littlefield, 1982),
47–78.
Nightingale, A. Genres in Dialogue: Plato and the Construct of Philosophy (Cambridge:
Cambridge University Press, 1995).
Rutherford, R. The Art of Plato: Ten Essays in Platonic Interpretation (Cambridge, MA: Harvard
University Press, 1995).
Urmson, J. “Plato and the Poets,” in J. Moravcsik and P. Temko (eds.), Plato on Beauty, Wisdom
and the Arts (Totowa, NJ: Rowman and Littlefield, 1982), 125–36.
Woodruff, P. “What Could Go Wrong with Inspiration? Why Plato’s Poets Fail,” in J. Moravcsik
and P. Temko (eds.), Plato on Beauty, Wisdom and the Arts (Totowa, NJ: Rowman and
Littlefield, 1982), 137–50.

Background and Context

Barnes, J. Early Greek Philosophy (Harmondsworth, UK: Penguin, 1987).
Barnes, J. The Presocratic Philosophers, rev. ed. (London: Routledge and Kegan Paul, 1982).
Bett, R. “The Sophists and Relativism,” Phronesis 34 (1989), 139–69.
Boardman, J., Griffin, J., and Murray, O. (eds.) The Oxford History of the Classical World
(Oxford: Oxford University Press, 1991).
Curd, P. K. The Legacy of Parmenides: Eleatic Monism and Later Presocratic Thought (Princeton,
NJ: Princeton University Press, 1997).
Dodds, E. R. The Greeks and the Irrational (Berkeley: University of California Press, 1951).
Dover, K. Greek Popular Morality in the Time of Plato and Aristotle (Oxford: Blackwell, 1974).
Ehrenberg, V. From Solon to Socrates: Greek History and Civilization during the Sixth and Fifth
Centuries B.C., 2nd ed. (London: Methuen, 1989).
Field, G. C. Plato and His Contemporaries: A Study in Fourth-Century Life and Thought
(London: Methuen, 1930).
Gallop, D. Parmenides (Toronto: University of Toronto Press, 1991).
Grote, G. A History of Greece, 6th ed., 12 vols. (London: Dent, 1872).
Guthrie, W. K. C. A History of Greek Philosophy 1: The Earlier Presocratics and the Pythagoreans
(Cambridge: Cambridge University Press, 1962).
Guthrie, W. K. C. A History of Greek Philosophy 2: The Presocratic Tradition from Parmenides
to Democritcus (Cambridge: Cambridge University Press, 1965).
Guthrie, W. K. C. A History of Greek Philosophy 3: Fifth-Century Enlightenment (Cambridge:
Cambridge University Press, 1969).
Horky, P. Plato and Pythagoreanism (Oxford: Oxford University Press, 2013).
 Bibliography    701

Hornblower, S. The Greek World 479–323 BC, 3rd ed. (London: Routledge and Kegan Paul,
2002).
Hornblower, S., and Spawforth, A. (eds.) The Oxford Classical Dictionary, 3rd. ed. (Oxford:
Oxford University Press, 1996).
Huffman, C. Philolaus of Croton: Pythagorean and Presocratic (Cambridge: Cambridge
University Press, 1993).
Hussey, E. The Presocratics (London: Duckworth, 1972).
Hussey, E. “The Beginnings of Epistemology: From Homer to Philolaus,” in S. Everson (ed.),
Epistemology (Cambridge: Cambridge University Press, 1990), 11–38.
Irwin, T. H. Classical Thought (Oxford: Oxford University Press, 1989).
Irwin, T. H. “Plato: The Intellectual Background,” in CCP, 51–89.
Kahn, C. The Art and Thought of Heraclitus (Cambridge: Cambridge University Press, 1979).
Kahn, C. Pythagoras and Pythagoreans: A Brief History (Indianapolis, IN: Hackett, 2001).
Kerferd, G. The Sophistic Movement (Cambridge: Cambridge University Press, 1981).
Kirk, G. S., Raven, J. E., and Schofield, M. The Presocratic Philosophers: A Critical History with
a Selection of Texts, 2nd. ed. (Cambridge: Cambridge University Press, 1983).
Lloyd, G. E. R. Hippocratic Writings (Harmondsworth, UK: Penguin, 1978).
Lloyd, G. E. R. Magic, Reason and Experience: Studies in the Origin and Development of Greek
Science (Cambridge: Cambridge University Press, 1979).
Lloyd, G. E. R. The Revolutions of Wisdom (Berkeley: University of California Press, 1997).
Long, A. A. (ed.) The Cambridge Companion to Early Greek Philosophy (Cambridge: Cambridge
University Press, 1999).
Morgan, M. “Plato and Greek Religion,” in CCP, 227–47.
Morgan, M. Platonic Piety: Philosophy and Ritual in Fourth Century Athens (New Haven, CT:
Yale University Press, 1990).
Palmer, J. A. Plato’s Reception of Parmenides (Oxford: Clarendon Press, 1999).
Sedley, D. Creationism and Its Critics in Antiquity (Berkeley: University of California Press,
2013).
Taylor, C. C. W. The Atomists Leucippus and Democritus: Fragments (Toronto: University of
Toronto Press, 1999).
The rest of this bibliography lists, first, books that cover a fairly broad range of topics and/or
dialogues, then various other works, listed under the name of the dialogue or topic with which
they are primarily concerned. Material already mentioned previously is generally not men-
tioned again.

General Books
Allen, R. E. (ed.) Studies in Plato’s Metaphysics (London: Routledge and Kegan Paul, 1965).
Cited as SPM.
Barney, R., Brennan, T., Brittain, C. (eds.) Plato and the Divided Self (Cambridge: Cambridge
University Press, 2012).
Benson, H. (ed.) A Companion to Plato (Oxford: Blackwell, 2006).
Bobonich, C. Plato’s Utopia Recast: His Later Ethics and Politics (Oxford: Clarendon Press,
2002).
Crombie, I. M. An Examination of Plato’s Doctrines, 2 vols. (London: Routledge and Kegan
Paul, 1963).
Dancy, R. Plato’s Introduction of Forms (Cambridge: Cambridge University Press, 2004).
702   Bibliography

Denyer, N. Language, Thought and Falsehood in Ancient Greek Philosophy (London: Routledge
and Kegan Paul, 1991).
Fine, G. Plato on Knowledge and Forms: Selected Essays (Oxford, Clarendon Press, 2003).
Cited as PKF.
Gentzler, J. (ed.) Method in Ancient Philosophy (Oxford: Clarendon Press, 1998). Cited as MAP.
Gill, C., and McCabe, M. M. (eds.) Form and Argument in Late Plato (Oxford: Clarendon
Press, 1996). Cited as FA.
Gill, M. L. Philosophos: Plato’s Missing Dialogue (Oxford: Oxford University Press, 2012).
Gosling, J. C. B. Plato (London: Routledge and Kegan Paul, 1973).
Gosling, J. C. B., and Taylor, C. C. W. The Greeks on Pleasure (Oxford: Clarendon Press, 1982).
Grote, G. Plato and the Other Companions of Socrates, new ed., 4 vols. (London: John Murray,
1888).
Grube, G. M. A. Plato’s Thought, with a new introduction, bibliographic essay, and bibliography
by D. Zeyl (Indianapolis, IN: Hackett, 1980).
Guthrie, W. K. C. A History of Greek Philosophy 4: Plato the Man and His Earlier Dialogues
(Cambridge: Cambridge University Press, 1975).
Guthrie, W. K. C. A History of Greek Philosophy 5: Later Plato and the Academy (Cambridge:
Cambridge University Press, 1979).
Hardie, W. F. R. A Study in Plato (Oxford: Clarendon Press, 1936).
Harte, V. Plato on Parts and Wholes: The Metaphysics of Structure (Oxford: Clarendon Press,
2002).
Irwin, T. H. Plato’s Moral Theory: The Early and Middle Dialogues (Oxford: Clarendon Press,
1977).
Irwin, T. H. Plato’s Ethics (Oxford: Clarendon Press, 1995).
Kahn, C. Plato and the Socratic Dialogue: The Philosophical Use of a Literary Form (Cambridge:
Cambridge University Press, 1996).
Kahn, C. Plato and the Post-Socratic Dialogue (Cambridge: Cambridge University Press, 2013).
Lee, E. N., Mourelatos, A. D. P., and Rorty, R. M. (eds.), Exegesis and Argument: Studies in
Greek Philosophy Presented to Gregory Vlastos (Assen, The Netherlands: Van Gorcum, 1973).
Cited as EA.
McCabe, M. M. Plato and His Predecessors (Cambridge: Cambridge University Press, 2000).
McCabe, M. M. Plato’s Individuals (Princeton, NJ: Princeton University Press, 1994).
Nussbaum, M., and Schofield, M. (eds.) Language and Logos: Studies in Ancient Greek
Philosophy Presented to G. E. L. Owen (Cambridge: Cambridge University Press, 1982).
Cited as LL.
Owen, G. E. L. Logic, Science and Dialectic: Collected Papers in Greek Philosophy, ed.
M. Nussbaum (Ithaca, NY: Cornell University Press, 1986). Cited as LSD.
Penner, T. The Ascent from Nominalism: Some Existence Arguments in Plato’s Middle Dialogues
(Dordrecht, The Netherlands: Reidel, 1987).
Prior, W. Unity and Development in Plato’s Metaphysics (London: Croom Helm, 1985).
Robinson, R. Plato’s Earlier Dialectic, 2nd ed. (Oxford: Clarendon Press, 1953).
Ross, W. D. Plato’s Theory of Ideas (Oxford: Clarendon Press, 1951).
Rowe, C. Plato and the Art of Philosophical Writing (Cambridge: Cambridge University Press,
2007).
Russell, D. Plato on Pleasure and the Good Life (Oxford: Clarendon Press, 2005).
Scott, D. Recollection and Experience: Plato’s Theory of Learning and Its Successors (Cambridge:
Cambridge University Press, 1995).
 Bibliography    703

Silverman, A. The Dialectic of Essence: A Study of Plato’s Metaphysics (Princeton, NJ: Princeton
University Press, 2003).
Vasiliou, I. Aiming at Virtue (Cambridge: Cambridge University Press, 2008).
Vlastos, G. Platonic Studies, 2nd. ed. (Princeton, NJ: Princeton University Press, 1981). Cited as PS.
Vlastos, G. Studies in Greek Philosophy, 2 vols., ed. D. Graham (Princeton, NJ: Princeton University
Press, 1993, 1995).
Vogt, K. Belief and Truth: A Skeptic Reading of Plato (Oxford: Oxford University Press, 2012).

The Early Dialogues: General

Allen, R. E. Plato’s Euthyphro and the Earlier Theory of Forms (London: Routledge and Kegan
Paul, 1970).
Beversluis, J. Cross-Examining Socrates: A Defense of the Interlocutors in Plato’s Early Dialogues
(Cambridge: Cambridge University Press, 2000).
Brickhouse, N., and Smith, T. Plato’s Socrates (Oxford: Oxford University Press, 1994).
Brickhouse, Thomas C., and Smith, Nicholas D. (2010) Socratic Moral Psychology (Cambridge:
Cambridge University Press).
Guthrie, W. K. C. Socrates (Cambridge: Cambridge University Press, 1971). Originally pub-
lished as Part 2 of Guthrie, A History of Greek Philosophy 3 (Cambridge: Cambridge
University Press, 1969).
Judson, L. and Karasmanis, V. (eds.), Remembering Socrates (Oxford: Clarendon Press,
2006).
Kamtekar, R. (ed.) Plato’s Euthyphro, Apology and Crito: Critical Essays (Lanham, MD:
Rowman and Littlefield, 2004). Cited as Kamtekar (ed.), Plato’s EAC.
Morrison, D. (ed.) Cambridge Companion to Socrates (Cambridge: Cambridge University
Press, 2011).
Penner, T., and Rowe, C. Plato’s Lysis (Cambridge: Cambridge University Press, 2005).
Reeve, C. D. C. Socrates in the Apology: An Essay on Plato’s Apology of Socrates (Indianapolis,
IN: Hackett, 1989).
Santas, G. Socrates: Philosophy in Plato’s Early Dialogues (London: Routledge and Kegan
Paul, 1979).
Stokes, M. Dialectic in Action: An Examination of Plato’s Crito (Swansea, UK: Classical Press
of Wales, 2005).
Stokes, M. Plato: Apology (Warminster, UK: Aris and Phillips, 1997).
Stokes, M. Plato’s Socratic Conversations: Drama and Dialectic in Three Dialogues (London:
Athlone Press, 1986). (Focuses on the Laches, Protagoras, and Symposium.)
Vlastos, G. Socratic Studies, ed. M. F. Burnyeat (Cambridge: Cambridge University Press, 1994).
Cited as SS.

The Early Dialogues: Epistemology, Metaphysics, and Method

Benson, H. “Misunderstanding the ‘What Is F-ness?’ Question,” Archiv für Geschichte der
Philosophie 72 (1990), 125–42. Also in EPS, 123–36.
Benson, H. “The Priority of Definition and the Socratic Elenchos,” Oxford Studies in Ancient
Philosophy 8 (1990), 19–65.
Benson, H. “The Problem of the Elenchus Revisited,” Ancient Philosophy 7 (1987), 67–85.
Benson, H. Socratic Wisdom (New York: Oxford University Press, 2000).
704   Bibliography

Beversluis, J. “Does Socrates Commit the Socratic Fallacy?,” American Philosophical Quarterly
24 (1987), 211–23. Also in EPS, 107–22.
Burnyeat, M. F. “Examples in Epistemology,” Philosophy 52 (1977), 381–98.
Crombie, I. M. “Socratic Definition,” Paideia 5 (1976), Special Plato Issue, 80–102. Also
in J. Day (ed.), Plato’s Meno in Focus (London: Routledge and Kegan Paul, 1994),
172–207.
Fine, G. ‘Does Socrates Claim to Know That He Knows Nothing?’, Oxford Studies in Ancient
Philosophy 35 (2008), 49–88.
Geach, P. T. “Plato’s Euthyphro: Analysis and Commentary,” Monist 50 (1966), 369–82. Also in
Kamtekar (ed.), Plato’s EAC, 23–34.
Kraut, R. “Comments on Gregory Vlastos, “The Socratic Elenchus,” Oxford Studies in Ancient
Philosophy 1 (1983), 59–70.
Lesher, J. H. “Socrates’ Disavowal of Knowledge,” Journal of the History of Philosophy 25 (1984),
275–88.
Mackenzie, M. M. “The Virtues of Socratic Ignorance,” Classical Quarterly 38 (1988),
331–50.
Nehamas, A. “Confusing Universals and Particulars in Plato’s Early Dialogues,” Review of
Metaphysics 29 (1975), 287–306.
Prior, J. “Plato and the ‘Socratic Fallacy,”’ Phronesis 43 (1998), 97–113.
Vlastos, G. “Is the Socratic Fallacy Socratic?,” Ancient Philosophy 10 (1990), 1–16. Revised
version in SS, 67–86.
Vlastos, G. “Socrates’ Disavowal of Knowledge,” Philosophical Quarterly 35 (1985), 1–31. Also
in Fine (ed.), Plato 1, 64–92.
Vlastos, G. “The Socratic Elenchus,” Oxford Studies in Ancient Philosophy 1 (1983), 27–58. Also
in Fine (ed.), Plato 1, 36–63.
Vlastos, G. “What Did Socrates Understand by His “What Is F?” Question?,” in PS, 410–17.
Wolfsdorf, D. “Socrates’ Avowals of Knowledge,” Phronesis 49 (2004), 75–142.
Woodruff, P. “Expert Knowledge in the Apology and Laches: What a General Needs to Know,”
Proceedings of the Boston Area Colloquium in Ancient Philosophy 3 (1987), 79–115.
Woodruff, P. “Plato’s Early Theory of Knowledge,” in S. Everson (ed.), Epistemology (Cambridge:
Cambridge University Press, 1989), 60–84. Also in EPS, 86–106.

The Early Dialogues: Ethics, Politics, Religion and The Soul

Annas, J. “Virtue as a Skill,” International Journal of Philosophical Studies 3 (1995), 227–43.
Annas, J. “Virtue as the Use of Other Goods,” Apeiron 26 (1993), 53–66.
Bostock, D. “The Interpretation of Plato’s Crito,” Phronesis 35 (1990), 1–20. Also in Kamtekar
(ed.), Plato’s EAC, 210–28.
Brickhouse, T., and Smith, N. “Socrates on Goods, Virtue and Happiness,” Oxford Studies in
Ancient Philosophy 5 (1987), 1–28.
Brickhouse, T., and Smith, N. Socrates on Trial (Princeton, NJ: Princeton University Press,
1989).
Brown, L. “Did Socrates Agree to Obey the Laws of Athens?,” in Judson and Karasmanis
(eds.), Remembering Socrates, 72–87.
Burnyeat, M. F. “The Impiety of Socrates,” Ancient Philosophy 17 (1997), 1–12. Also in Kamtekar
(ed.), Plato’s EAC, 150–62.
Burnyeat, M. F. “Virtues in Action,” in Vlastos (ed.), Philosophy of Socrates, 209–34.
 Bibliography    705

Cohen, S. M. “Socrates and the Definition of Piety,” Journal of the History of Philosophy 9 (1971),
1–13. Also in Vlastos (ed.), Philosophy of Socrates, 158–86, and in Kamtekar (ed.), Plato’s
EAC, 35–48.
Devereux, D. “Courage and Wisdom in Plato’s Laches,” Journal of the History of Philosophy 15
(1977), 129–41.
Devereux, D. “Socrates’ Kantian Conception of Virtue,” Journal of the History of Philosophy 33
(1995), 341–408.
Devereux, D. “The Unity of the Virtues in Plato’s Protagoras and Laches,” Philosophical Review
101 (1992), 765–89.
Evans, M. ‘Lessons of Euthyphro 10a-11b’, Oxford Studies in Ancient Philosophy 42 (2012),
1–38.
Harte, V. “Conflicting Values in Plato’s Crito,” Archiv für Geschichte der Philosophie 81 (1999),
117–47. Also in Kamtekar (ed.), Plato’s EAC, 229–46.
Irwin, T. H. “Coercion and Objectivity in Plato’s Dialectic,” Revue Internationale de Philosophie
40 (1986), 47–74.
Irwin, T. H. “Socrates the Epicurean?,” Illinois Classical Studies 11 (1986), 85–112. Also in EPS,
198–219.
Judson, L. “Carried Away in the Euthyphro,” in D. Charles (ed.), Definition in Greek Philosophy
(Oxford: Oxford University Press, 2010), 31–61.
Kamtekar, R., “Socrates on the Attribution of Conative Attitudes,” Archiv für Geschichte der
Philosophie 86 (2006), 127–62.
Kamtekar, R. Plato’s Moral Psychology: Intellectualism, the Divided Soul, and the Desire for
Good (Oxford: Oxford University Press, 2017).
Mackenzie, M. M. Plato on Punishment (Berkeley: University of California Press, 1981). (Also
discusses this topic in other periods of Plato’s thought.)
McPherran, M. The Religion of Socrates (University Park: Pennsylvania State University Press,
1996).
Moss, J. “Hedonism and the Divided Soul in Plato’s Protagoras,” Archiv für Geschichte der
Philosophie 96 (2014), 285–319.
Penner, T. “Desire and Power in Socrates,” Apeiron 24 (1991), 147–201.
Penner, T. “Knowledge vs. True Belief in the Socratic Psychology of Action,” Apeiron 26 (1996),
199–230.
Penner, T. “Socrates on the Strength of Knowledge: Protagoras 351B–375E,” Archiv für
Geschichte der Philosophie 79 (1997), 117–49.
Penner, T. “Socrates on Virtue and Motivation,” in E. N. Lee, A. D. P. Mourelatos, and
R. M. Rorty (eds.), Exegesis and Argument: Studies in Greek Philosophy Presented to Gregory
Vlastos (Assen: Van Gorcum, 1973), 133–51. Volume cited as EA.
Penner, T. “The Unity of Virtue,” Philosophical Review 82 (1973), pp. 35–68. Also in EPS, 162–84,
and in Fine (ed.), Plato 2, 78–104.
Rudebusch, G. Socrates, Pleasure, and Value (New York: Oxford University Press, 1999).
Santas, G. “Plato’s Protagoras and Explanations of Weakness,” Philosophical Review 75 (1966),
3–33. Also in Vlastos (ed.), Philosophy of Socrates, 264–98.
Santas, G. “Socrates at Work on Virtue and Knowledge in Plato’s Laches,” Review of Metaphysics
22 (1969), 443–60.
Santas, G. “The Socratic Paradoxes,” Philosophical Review 73 (1964), 147–64.
Segvic, H. “No One Errs Willingly: The Meaning of Socratic Intellectualism,” Oxford Studies
in Ancient Philosophy 19 (2000), 1–45.
706   Bibliography

Sharvy, R. “Euthyphro 9d–11b: Analysis and Definition in Plato and Others,” Nous 6 (1972),
119–37.
Shaw, C. Plato’s Anti-hedonism and the Protagoras (Cambridge: Cambridge University Press,
2015).
Smith, N., and Woodruff, P. (eds.). Reason and Religion in Socratic Philosophy (Oxford: Oxford
University Press, 2000).
Taylor, C. C. W. “The End of the Euthyphro,” Phronesis 27 (1982), 109–18.
Taylor, C. C. W. “Socratic Ethics,” in B. S. Gower and M. C. Stokes (eds.), Socratic Questions
(London: Routledge and Kegan Paul, 1992), 137–52.
Vlastos, G. Introduction to Plato: Protagoras (Indianapolis, IN: Bobbs Merrill, 1965).
Vlastos, G. “Socrates on Acrasia,” Phoenix 23 (1969), 71–88. Also in G. Vlastos (ed.), Studies in
Greek Philosophy, vol. 2 (Princeton, NJ: Princeton University Press, 1995), 43–59.
Vlastos, G. “Socratic Piety,” in Vlastos, Socrates, 157–78. Also in Fine (ed.), Plato 2, 56–77, and
in Kamtekar (ed.), Plato’s EAC, 49–71.
Vlastos, G. “The Unity of the Virtues in the Protagoras,” Review of Metaphysics 25 (1972), 415–58.
Revised version with additional notes in PS, 2nd ed., 221–69, 418–23.
Woolf, R. (2000), “Callicles and Socrates: Psychic (Dis)harmony in the Gorgias,” Oxford
Studies in Ancient Philosophy 18: 1–40.
Woozley, A. D. Law and Obedience: The Arguments of Plato’s Crito (London: Duckworth, 1979).
Woozley, A. D. “Socrates on Disobeying the Law,” in Vlastos (ed.), Philosophy of Socrates,
229–318.
Young, G. “Socrates and Obedience,” Phronesis 19 (1974), 1–29.
Zeyl, D. “Socrates and Hedonism,” Phronesis 25 (1980), 250–69.
Zeyl, D. “Socratic Virtue and Happiness,” Archiv für Geschichte der Philosophie 64 (1982), 225–38.

Meno
Benson, H. Clitophon’s Challenge: Dialectic in Plato’s Meno, Phaedo, and Republic (Oxford:
Oxford University Press, 2015).
Bluck, R. S. Plato’s Meno (Cambridge: Cambridge University Press, 1964).
Charles, D. “Types of Definition in Plato’s Meno,” in Judson and Karasmanis (eds.), Remembering
Socrates, 110–28.
Charles, D. “The Paradox in the Meno and Aristotle’s Attempts to Resolve It,” in Charles (ed.),
Definition in Greek Philosophy (Oxford, 2010), 115–50.
Day, J. (ed.) Plato’s Meno in Focus (London: Routledge and Kegan Paul, 1994). Cited as Day
(ed.), Meno in Focus.
Devereux, D. “Nature and Teaching in Plato’s Meno,” Phronesis 23 (1978), 118–126.
Dimas, P. “True Belief in the Meno,” Oxford Studies in Ancient Philosophy 14 (1996), 1–32.
Fine, G. “Inquiry in the Meno,” in CCP, 200–226. Also in PKF, 44–65.
Fine, G. “Knowledge and True Belief in the Meno,” Oxford Studies in Ancient Philosophy 27
(2004), 41–81.
Fine, G. The Possibility of Inquiry (Oxford: Oxford University Press, 2014).
Franklin, L. “The Structure of Dialectic in the Meno,” Phronesis 46 (2001), 413–39.
Gentzler, J. “Recollection and ‘The Problem of the Socratic Elenchus,’ ” Proceedings of the
Boston Area Colloquium in Ancient Philosophy 10 (1996), 257–95.
Judson, “Hypothesis in Plato’s Meno,” Philosophical Inquiry 41 (2017), 29–39.
McCabe, M. M. “Escaping One’s Own Notice Knowing: Meno’s Paradox Again,”’ Proceedings
of the Aristotelian Society, 109 (2009), 233–56.
 Bibliography    707

Moravcsik, J. M. E. “Learning as Recollection,” in Vlastos (ed.), Plato 1, 53–69.

Nehamas, A. “Meno’s Paradox and Socrates as a Teacher,” Oxford Studies in Ancient Philosophy
3 (1985), 1–30. Also in EPS, 298–316, and in Day (ed.), Meno in Focus, 221–48.
Nehamas, A. “Socratic Intellectualism,” Proceedings of the Boston Area Colloquium in
Ancient Philosophy 2 (1986), 275–316.
Perin, C. “Knowledge, Stability, and Virtue in the Meno,” Ancient Philosophy 32 (2012), 15–34.
Scott, D. Plato’s Meno (Cambridge: Cambridge University Press, 2006).
Sharples, R. W. Plato: Meno (Warminster, UK: Aris and Phillips, 1985).
Thompson, E. S. The Meno of Plato (London: Macmillan, 1901).
Vlastos, G. “Anamnêsis in the Meno,” Dialogue 4 (1965), 143–67.
Vlastos, G. “Elenchus and Mathematics,” American Journal of Philology 109 (1988), 362–96.
Also in his Socrates, 107–31, and in EPS, 137–61.
White, N. P. “Inquiry,” Review of Metaphysics 28 (1974), 289–310.
Wilkes, K. “Conclusions in the Meno,” Archiv für Geschichte der Philosophie 61 (1970), 143–53.
Also in Day (ed.), Meno in Focus, 208–20.

Cratylus
Ackrill, J. L. “Language and Reality in Plato’s Cratylus,” in his Essays on Plato and Aristotle
(Oxford: Clarendon Press, 1994), 33–52. Also in Fine (ed.), Plato 1, 125–42.
Adelmollo, F. The Cratylus of Plato (Cambridge: Cambridge University Press, 2011).
Annas, J. “Knowledge and Language: The Theaetetus and the Cratylus.” See under Theaetetus.
Barney, R. Names and Nature in Plato’s Cratylus (London: Routledge and Kegan Paul, 2001).
Barney, R. “Plato on Conventionalism,” Phronesis 42 (1997), 143–61.
Baxter, T. The Cratylus: Plato’s Critique of Naming (Leiden: E. J. Brill, 1992).
Fine, G. “Plato on Naming,” Philosophical Quarterly 27 (1977), 289–301. Also in PKF, 117–31.
Kahn, C. “Language and Ontology in the Cratylus,” in EA, 152–76.
Keller, S. “An Interpretation of Plato’s Cratylus,” Phronesis 45 (2000), 284–305.
Ketchum, R. “Names, Forms, and Conventionalism: Cratylus 383–95,” Phronesis 24 (1979),
133–47.
Kretzmann, N. “Plato on the Correctness of Names,” American Philosophical Quarterly 8
(1971), 126–38.
Mackenzie, M. M. “Putting the Cratylus in Its Place,” Classical Quarterly NS 36 (1986),
124–50.
Schofield, M. “The Dénouement of the Cratylus,” in LL, 61–81.
Sedley, D. Plato’s Cratylus (Cambridge: Cambridge University Press, 2003).
Silverman, A. “Plato’s Cratylus: The Naming of Nature and the Nature of Naming,” Oxford
Studies in Ancient Philosophy 10 (1992), 25–72.
Spellman, L. “Naming and Knowing: The Cratylus on Images,” History of Philosophy Quarterly
10 (1993), 197–210.
Williams, B. A. O. “Cratylus” Theory of Names and Its Refutation,” in LL, 83–93.

Phaedo and the Theory of Forms

Ackrill, J. L. “Anamnêsis in the Phaedo,” in his Essays on Plato and Aristotle, 13–32. Also in EA,
177–95.
Annas, J. “Aristotle on Inefficient Causes,” Philosophical Quarterly 32 (1982), 311–26.
Bailey, D. “Platonic Causes Revisited,” Journal of the History of Philosophy 52 (2014), 15–32.
708   Bibliography

Bolton, R. “Plato’s Distinction between Being and Becoming.” See under Theaetetus.
Bostock, D. Plato’s Phaedo (Oxford: Clarendon Press, 1986).
Butler, T. “A Riveting Argument in Favor of Asceticism in the Phaedo,” History of Philosophy
Quarterly 29 (2012), 103–33.
Cherniss, H. F. “The Philosophical Economy of the Theory of Ideas,” American Journal of
Philology 57 (1936), 445–56. Also in SPM, 1–12, and in Vlastos (ed.), Plato 1, 16–27.
Devereux, D. “Separation and Immanence in Plato’s Theory of Forms,” Oxford Studies in
Ancient Philosophy 12 (1994), 63–90. Also in Fine (ed.), Plato 1, 192–214.
Dimas, P. “Recollecting Forms in the Phaedo,” Phronesis 48 (2003), 175–214.
Fine, G. “Forms as Causes: Plato and Aristotle,” in A. Graeser (ed.), Mathematics and
Metaphysics in Aristotle (Bern: Haupt, 1986), 69–112. Also in PKF, 350–96.
Fine, G. “The One over Many,” Philosophical Review 89 (1980), 197–240.
Fine, G. On Ideas: Aristotle’s Criticism of Plato’s Theory of Forms (Oxford: Clarendon Press,
1993).
Fine, G. “Separation,” Oxford Studies in Ancient Philosophy 2 (1984), 31–87. Also in PKF,
252–300.
Fine, G. “The ‘Two Worlds’ Theory in the Phaedo,” British Journal for the History of Philosophy
24 (2016), 557–72.
Fine, G. “Plato on the Grades of Perception: Theaetetus 184–6 and the Phaedo,” Oxford Studies
in Ancient Philosophy 53 (2017), 65–109.
Franklin, L. “Recollection and Philosophical Reflection in Plato’s Phaedo,” Phronesis 50 (2005),
289–314.
Frede, D. “The Final Proof of the Immortality of the Soul,” Phronesis 23 (1978), 27–41.
Gallop, D. “Plato’s Cyclical Argument Recycled,” Phronesis 27 (1982), 207–22.
Gentzler, J. “Sumphônein in Plato’s Phaedo,” Phronesis 36 (1991), 265–76.
Gosling, J. C. B. “Similarity in Phaedo 73b seq.,” Phronesis 10 (1965), 151–61.
Hartman, E. “Predication and Immortality in Plato’s Phaedo,” Archiv für Geschichte der
Philosophie 54 (1972), 215–28.
Irwin, T. H. “Plato’s Heracleiteanism,” Philosophical Quarterly 27 (1977), 1–13.
Kelsey, S. “Recollection in the Phaedo,” Proceedings of the Boston Area Colloquium in Ancient
Philosophy 16 (2000), 91–121.
Keyt, D. “The Fallacies in Phaedo 102a–107b,” Phronesis 8 (1963), 167–72.
Kirwan, C. “Plato and Relativity,” Phronesis 19 (1974), 112–29.
Lovibond, S. “Plato’s Theory of Mind,” in S. Everson (ed.), Psychology (Cambridge: Cambridge
University Press, 1991), 35–55.
Malcolm, J. Plato on the Self-Predication of Forms: Early and Middle Dialogues (Oxford: Clarendon
Press, 1991).
Mates, B. “Identity and Predication in Plato,” Phronesis 24 (1979), 211–29.
Matthews, G., and Blackson, T. “Causes in the Phaedo,” Synthèse 79 (1989), 581–91.
Matthews, G., and Cohen, S. M. “The One and the Many,” Review of Metaphysics 21 (1968),
630–55.
Miller, T. “Socrates’ Warning against Misology,” Phronesis 60 (2015), 145–79.
Mills, K. W. “Plato’s Phaedo 74,” Phronesis 2 (1957), 128–74; and 3 (1958), 40–58.
Nehamas, A. “Plato on the Imperfection of the Sensible World,” American Philosophical
Quarterly 12 (1975), 105–17. Also in Fine (ed.), Plato 1, 171–91.
Nehamas, A. “Predication and Forms of Opposites in the Phaedo,” Review of Metaphysics 26
(1973), 461–91.
 Bibliography    709

Nehamas, A. “Self-Predication and Plato’s Theory of Forms,” American Philosophical Quarterly

16 (1979), 93–103.
Owen, G. E. L. “A Proof in the Peri Ideôn,” Journal of Hellenic Studies 77 (1957), 103–11. Also in
SPM, 243–312, and in LSD, 165–79.
Rowe, C. J. Plato: Phaedo (Cambridge: Cambridge University Press, 1995).
Scott, D. “Platonic Anamnêsis Revisited,” Classical Quarterly NS 37 (1987), 346–66.
Sedley, D. “Equal Sticks and Stones,” in D. Scott (ed.), Maieusis: Studies on Greek Philosophy in
Honour of M. F. Burnyeat (Oxford: Oxford University Press 2007), 68–86.
Sedley, D. “Form-Particular Resemblance in Plato’s Phaedo,” Proceedings of the Aristotelian
Society 106 (2006), 309–25.
Sedley, D. “Platonic Causes,” Phronesis 43 (1998), 114–32.
Sedley, D. “Teleology and Myth in Plato’s Phaedo,” Proceedings of the Boston Area Colloquium
in Ancient Philosophy 5 (1991), 359–83.
Taylor, C. C. W. “The Arguments in the Phaedo concerning the Thesis That the Soul Is a
Harmonia,” in J. Anton and A. Preus (eds.), Essays in Ancient Greek Philosophy, vol. 2 (Albany:
State University of New York, 1983), 217–31.
Vlastos, G. “Degrees of Reality in Plato,” in R. Bambrough (ed.), New Essays on Plato and
Aristotle (London: Routledge and Kegan Paul, 1965), 1–18. Also in PS, 58–75.
Vlastos, G. “A Metaphysical Paradox,” Proceedings and Addresses of the American Philosophical
Association 39 (1966), 5–19. Also in PS, 43–57.
Vlastos, G. “Reasons and Causes in the Phaedo,” Philosophical Review 78 (1969), 291–325. Also
in Vlastos (ed.), Plato 1, 132–66, and in PS, 76–110.
White, N. P. “Forms and Sensibles: Phaedo 74bc,” Philosophical Topics 15 (1987), 197–214.
White, N. P. “Perceptual and Objective Properties in Plato,” Apeiron 22 (1989), 45–65.
White, N. P. “Plato’s Metaphysical Epistemology,” in CCP, 277–310.
Woolf, R. “Misology and Truth,” Proceedings of the Boston Area Colloquium in Ancient Philosophy
23 (2007), 1–16.

Republic: General
Adam, J. The Republic of Plato, 2 vols. (Cambridge: Cambridge University Press, 1902).
Annas, J. An Introduction to Plato’s Republic (Oxford: Clarendon Press, 1981).
Cross, R. C., and Woozley, A. D. Plato’s Republic: A Philosophical Commentary (London:
Macmillan, 1964).
Ferrari, G. F. R. Cambridge Companion to Plato’s Republic (Cambridge: Cambridge University
Press, 2007).
Halliwell, S. Plato: Republic 5 (Warminster, UK: Aris and Phillips, 1993).
Halliwell, S. Plato: Republic 10 (Warminster, UK: Aris and Phillips, 1988).
Joseph, H. W. B. Knowledge and the Good in Plato’s Republic (Oxford: Clarendon Press, 1948).
Kraut, R. (ed.) Plato’s Republic: Critical Essays (Lanham, MD: Rowman and Littlefield, 1997).
McPherran, M. Plato’s Republic: A Critical Guide (Cambridge: Cambridge University Press,
2010).
Murphy, N. R. The Interpretation of Plato’s Republic (Oxford: Clarendon Press, 1951).
Ostenfeld, E. (ed.) Essays on Plato’s Republic (Aarhus, Denmark: Aarhus University Press, 1998).
Reeve, C. D. C. Philosopher-Kings: The Argument of Plato’s Republic (Princeton, NJ: Princeton
University Press, 1988).
Santas, G. (ed.) The Blackwell Guide to Plato’s Republic (Oxford: Blackwell, 2006).
710   Bibliography

Scott, D. Levels of Argument: A Comparative Study of Plato’s Republic and Aristotle’s Nicomachean
Ethics (Oxford: Oxford University Press, 2015).
White, N. P. A Companion to Plato’s Republic (Indianapolis, Ind.: Hackett, 1979).

Republic: Metaphysics and Epistemology

These works focus on topics particular to the Republic; more general discussions of the Theory
of Forms were listed previously in the section on the Phaedo and the Theory of Forms.
Burnyeat, M. F. “Platonism and Mathematics: A Prelude to Discussion,” in A. Graeser (ed.),
Mathematics and Metaphysics in Aristotle (Bern: Haupt, 1987), 213–40.
Fine, G. “Knowledge and Belief in Republic V,” Archiv für Geschichte der Philosophie 60 (1978),
121–39. Also in PKF, 65–84.
Fine, G. “Knowledge and Belief in Republic V–VII,” in S. Everson (ed.), Epistemology (Cambridge:
Cambridge University Press, 1990), 85–115. Also in Fine (ed.), Plato 1, 215–46, and in PKF,
85–116.
Fogelin, R. “Three Platonic Analogies,” Philosophical Review 80 (1971), 371–82.
Gentzler, J. “How to Know the Good: The Moral Epistemology of Plato’s Republic,” Philosophical
Review 114 (2005), 469–96.
Gosling, J. C. B. “Doxa and Dunamis in Republic V,” Phronesis 13 (1968), 119–30.
Gosling, J. C. B. “Republic V: ta polla kala,” Phronesis 5 (1960), 116–28.
Harte, V. “Language in the Cave,” in D. Scott (ed.), Maieusis: Studies in Honour of M. F. Burnyeat
(Oxford: Oxford University Press, 2008), 195–215.
Harte, V. “Knowing and Believing in Republic 5,” in V. Harte and R. Woolf (eds.), Rethinking
Ancient Philosophy: Old Chestnut and Sacred Cows (Cambridge: Cambridge University Press,
2017), 141–62.
Karasmanis, V. “Plato’s Republic: The Line and the Cave,” Apeiron 21 (1988), 147–71.
Moss, J. “Plato’s Appearance/Assent Account of Belief,” Proceedings of the Aristotelian Society
114 (2014), 213–38.
Santas, G. “The Form of the Good in Plato’s Republic,” Philosophical Inquiry (1980), 374–403;
abbreviated version in Fine (ed.), Plato 1, 247–74.
Schwab, W. “Understanding epistêmê in Plato’s Republic,” Oxford Studies in Ancient Philosophy
51 (2016), 41–85.
Scott, D. “From Painters to Poets: Plato’s Methods in Republic 10,” Proceedings of the Aristotelian
Society 116 (2016), 289–309.
Scott, D. “Knowledge, True Belief and Poetry in Republic 10,” in Harte and Woolf (eds.),
Rereading Ancient Philosophy, 163–80.
Smith, N. “Plato on Knowledge as a Power,” Journal of the History of Philosophy 38 (2000),
145–68.
Smith, N. “Plato on the Power of Ignorance,” R. Kamtekar (ed.), Virtue and Happiness: Essays
in Honour of Julia Annas. Published in Oxford Studies in Ancient Philosophy Supplementary
Volume, 2012), 51–73.
Strang, C. “Plato’s Analogy of the Cave,” Oxford Studies in Ancient Philosophy 4 (1986),
19–34.
Taylor, C. C. W. “Plato and the Mathematicians,” Philosophical Quarterly 17 (1967), 193–203.
Wilson, J. S. “The Contents of the Cave,” Canadian Journal of Philosophy (suppl.) 2 (1976),
111–24.
Woolf, R. “Truth as a Value in Plato’s Republic,” Phronesis 54 (2009), 9–39.
 Bibliography    711

Republic: Ethics, Psychology, and Religion

Annas, J. “Plato and Common Morality,” Classical Quarterly NS 28 (1978), 437–51.
Annas, J. “Politics and Ethics in Plato’s Republic,” in O. Höffe (ed.), Politeia (Berlin: Akademie
Verlag, 1997), 141–60. Volume cited as Höffe (ed.).
Bobonich, C. “Akrasia and Agency in Plato’s Laws and Republic,” Archiv für Geschichte der
Philosophie 72 (1994), 3–36.
Brown, E. “Justice and Compulsion for Plato’s Philosopher-Rulers,” Ancient Philosophy 20
(2000), 1–17.
Brown, E. “Minding the Gap in Plato’s Republic,” Philosophical Studies 117 (2004), 275–302.
Brown, E. “The Unity of the Soul in Plato’s Republic, in Barney et al. (eds.), Plato and the
Divided Soul, 53–74.
Burnyeat, M. F. “Plato on Why Mathematics Is Good for the Soul,” Proceedings of the British
Academy 103 (2000), 1–81.
Cooper, J. “The Psychology of Justice in Plato,” American Philosophical Quarterly 14 (1977),
151–57.
Cooper, J. “Plato’s Theory of Human Motivation,” History of Philosophy Quarterly 1 (1984),
3–21. Also in Fine (ed.), Plato 2, 186–206.
Dahl, N. “Plato’s Defence of Justice,” Philosophy and Phenomenological Research 51 (1991),
809–34. Also in Fine (ed.), Plato 2, 207–34.
Demos, R. “A Fallacy in Plato’s Republic?,” Philosophical Review 73 (1964), 395–98. Also in
Vlastos (ed.), Plato 2, 52–56.
Everson, S. “The Incoherence of Thrasymachus,” Oxford Studies in Ancient Philosophy 16
(1998), 88–131.
Gill, C. “Plato and the Education of Character,” Archiv für Geschichte der Philosophie 67 (1985),
1–26.
Heinaman, R. “Plato’s Division of Goods in the Republic,” Phronesis 47 (2002), 309–35.
Höffe, O. (ed.) Politeia (Berlin: Akademie Verlag, 1997). Cited as Höffe (ed.).
Irwin, T. H. “The Parts of the Soul and the Cardinal Virtues,” in Höffe (ed.), 119–39.
Kahn, C. H. “Plato’s Theory of Desire,” Review of Metaphysics 41 (1987), 77–103.
Kenny, A. J. P. “Mental Health in Plato’s Republic,” Proceedings of the British Academy 55 (1969),
229–53. Also in Kenny, The Anatomy of the Soul (Oxford: Blackwell, 1973), 1–27.
Kerferd, G. B. “The Doctrine of Thrasymachus in Plato’s Republic,” Durham University Journal
9 (1947), 19–27.
Kirwan, C. “Glaucon’s Challenge,” Phronesis 10 (1965), 162–73.
Kraut, R. “The Defense of Justice in Plato’s Republic,” in CCP, 311–37.
Kraut, R. “Plato’s Comparison of Just and Unjust Lives,” in Höffe (ed.), 271–90.
Kraut, R. “Reason and Justice in the Republic,” in EA, 331–37.
Kraut, R. “Return to the Cave: Republic 519–521,” in Fine (ed.), Plato 2, 235–54.
Lorenz, H. The Brute Within (Oxford: Clarendon Press, 2006).
Lorenz, H. “Desire and Reason in Plato’s Republic,” Oxford Studies in Ancient Philosophy 27
(2004), 83–116.
Mahoney, T. “Do Plato’s Philosopher-Rulers Sacrifice Self-Interest to Justice?,” Phronesis 37
(1992), 265–82.
Moss, J. “Appearances and Calculations: Plato’s Division of the Soul,” Oxford Studies in Ancient
Philosophy 34 (2008), 35–68.
Moss, J. “Pleasure and Illusion in Plato,” Philosophy and Phenomenological Research 72 (2006),
503–35.
712   Bibliography

Nicholson, P. P. “Unravelling Thrasymachus’ Argument in the Republic,” Phronesis 19 (1974),

210–32.
Penner, T. “Plato and Davidson: Parts of the Soul and Weakness of the Will,” Canadian Journal
of Philosophy (suppl.) 16 (1990), 35–74.
Penner, T. “Thought and Desire in Plato,” in Vlastos (ed.), Plato 2, 96–118.
Price, A. “Plato and Freud,” in C. Gill (ed.), The Person and the Human Mind: Issues in Ancient
and Modern Philosophy (Oxford: Clarendon Press, 1990), 247–70.
Reeve, C. D. C. “Socrates Meets Thrasymachus,” Archiv für Geschichte der Philosophie 67 (1985),
246–65.
Robinson, R. “Plato’s Separation of Reason and Desire,” Phronesis 16 (1971), 38–48.
Sachs, D. “A Fallacy in Plato’s Republic,” Philosophical Review 72 (1963), 141–58. Also in Vlastos
(ed.), Plato 2, 35–51.
Santas, G. “Two Theories of the Good in Plato’s Republic,” Archiv für Geschichte der Philosophie
57 (1980), 223–45.
Scott, D. “Metaphysics and the Defence of Justice in the Republic,” Proceedings of the Boston
Area Colloquium in Ancient Philosophy 16 (2000), 1–20.
Scott, D. “Platonic Pessimism and Moral Education,” Oxford Studies in Ancient Philosophy 17
(1999), 15–36.
Scott, D. “Plato’s Critique of the Democratic Character,” Phronesis 45 (2000), 19–37.
Stalley, R. “Plato’s Arguments for the Division of the Reasoning and Appetitive Elements
within the Soul,” Phronesis 20 (1975), 110–28.
Taylor, C. C. W. “The Role of Women in Plato’s Republic,” Oxford Studies in Ancient Philosophy,
suppl. vol. (2013), 75–87.
Vlastos, G. “Justice and Happiness in the Republic,” in Vlastos (ed.), Plato 2, 66–95. Also in PS,
111–39.
Waterlow [Broadie], S. “The Good of Others in Plato’s Republic,” Proceedings of the Aristotelian
Society 72 (1972–73), 19–36.
Wedgwood, R. “The Coherence of Thrasymachus,” Oxford Studies in Ancient Philosophy 53
(2017), 33–63.
White, N. P. “The Classification of Goods in Plato’s Republic,” Journal of the History of Philosophy
22 (1984), 393–421.
White, N. P. “Happiness and External Contingencies in Plato’s Republic,” in W. C. Starr
and R. C. Taylor (eds.), Moral Philosophy (Milwaukee: Marquette University Press,
1989), 1–21.
White, N. P. “The Ruler’s Choice,” Archiv für Geschichte der Philosophie 68 (1986), 22–46.
Wolfsdorf, D. “Pleasure and Truth in Republic 9,” Classical Quarterly 63 (2013), 110–38.
Woods, M. J. “Plato’s Division of the Soul,” Proceedings of the British Academy 73 (1987),
23–47.
Young, C. “Polemarchus’ and Thrasymachus’ Definitions of Justice,” Philosophical Inquiry 2
(1980), 404–19.

Republic: Politics and the Ideally Just State

Many of the works listed in the previous two sections are also relevant here.
Ackrill, J. L. “What’s Wrong with Plato’s Republic?,” in his Essays on Plato and Aristotle, 2nd ed.
(Oxford: Clarendon Press, 2001), 230–52.
 Bibliography    713

Annas, J. “Plato’s Republic and Feminism,” Philosophy 51 (1976), 307–21. Also in Fine (ed.), Plato 2,
265–79.
Brown, L. “How Totalitarian Is Plato’s Republic?,” in E. Ostenfeld (ed.), Essays on Plato’s Republic
(Aarhus, Denmark: Aarhus University Press, 1998), 13–27.
Burnyeat, M. F. “Culture and Society in Plato’s Republic,” Tanner Lectures on Human Values 20
(Salt Lake City: University of Utah Press, 1999), 215–324.
Burnyeat, M. F. “Utopia and Fantasy: The Practicability of Plato’s Ideally Just City,” in J. Hopkins
and A. Savile (eds.), Psychoanalysis, Mind and Art: Perspectives on Richard Wollheim
(Oxford: Blackwell, 1992), 175–87. Also in Fine (ed.), Plato 2, 279–308.
Ferrari, G. City and Soul in Plato’s Republic (Sankt Augustin: Akademia Verlag, 2003; Chicago:
University of Chicago Press, 2005).
Kraut, R. “Egoism, Love, and Political Office,” Philosophical Review 82 (1973), 330–44.
Neu, J. “Plato’s Analogy of State and Individual: The Republic and the Organic Theory of the
State,” Philosophy 46 (1971), 238–54.
Popper, K. R. The Open Society and Its Enemies, vol. 1: The Spell of Plato (Princeton, NJ:
Princeton University Press, 1971; originally published London: Routledge and Kegan
Paul, 1945).
Schofield, M. Plato: Political Philosophy (Oxford: Oxford University Press, 2006).
Taylor, C. C. W. “Plato’s Totalitarianism,” Polis 5 (1986), 4–29. Also in Fine (ed.), Plato 2, 280–96.
Tuana, N. (ed.). Feminist Interpretations of Plato (University Park: Pennsylvania State
University Press, 1994).
Vlastos, G. “Does Slavery Exist in Plato’s Republic?,” Classical Philology 63 (1968), 291–95. Also
in PS, 140–63.
Vlastos, G. “The Theory of Social Justice in the Polis in Plato’s Republic,” in H. North (ed.),
Interpretations of Plato (Leiden: Brill, 1977), 1–40.
Vlastos, G. “Was Plato a Feminist?,” Times Literary Supplement 4, 485 (March 17, 1989), 276,
288–89. Also in Vlastos, Studies in Greek Philosophy, vol. 2, 133–43.
Williams, B. A. O. “The Analogy of City and Soul in Plato’s Republic,” in EA, 196–206. Also in
Fine (ed.), Plato 2, 255–64.

Symposium, Phaedrus, and Platonic Love

Bett, R. “Immortality and the Nature of the Soul in the Phaedrus,” Phronesis 31 (1986), 1–26.
Also in Fine (ed.), Plato 2, 425–50.
Bury, R. G. The Symposium of Plato, 2nd ed. (Cambridge: Cambridge University Press, 1932).
Destrée, P. and Giannopoulou, G. Plato’s Symposium: a Critical Guide (Cambridge: Cambridge
University Press, 2017).
Dover, K. J. Plato: Symposium (Cambridge: Cambridge University Press, 1980).
Ferrari, G. Listening to the Cicadas: A Study of Plato’s Phaedrus (Cambridge: Cambridge
University Press, 1987).
Ferrari, G. “Platonic Love,” in CCP, 248–276.
Ferrari, G. “The Struggle in the Soul: Plato, Phaedrus 253c7–255a1,” Ancient Philosophy 5 (1985),
1–10.
Frede, D. “Out of the Cave: What Socrates Learnt from Diotima,” in R. Rosen and J. Farrell (eds.),
Nomodeiktes: Greek Studies in Honor of Martin Oswald (Ann Arbor: University of Michigan
Press, 1993), 397–422.
Hackforth, R. “Immortality in Plato’s Symposium,” Classical Review 64 (1950), 42–45.
714   Bibliography

Hackforth, R. Plato’s Phaedrus (Cambridge: Cambridge University Press, 1952).

Heath, M. “The Unity of Plato’s Phaedrus,” Oxford Studies in Ancient Philosophy 7 (1987),
150–73.
Heath, M. “The Unity of the Phaedrus: A Postscript,” Oxford Studies in Ancient Philosophy 7
(1989), 189–91.
Hunter, R. Plato’s Symposium (Oxford: Oxford University Press, 2004).
Kosman, A. “Platonic Love,” in W. H. Werkmeister (ed.), Facets of Plato’s Philosophy (Assen:
van Gorcum, 1976), 53–69.
Lesher, J., Nails, D., and Sheffield, F. (eds.) Plato’s Symposium: Issues in Interpretation and
Reception (Cambridge, MA: Harvard University Press, 2007).
Mahoney, T. “Is Socratic Eros in the Symposium Egoistic?,” Apeiron 29 (1996), 1–18.
Moravcsik, J. M. E. “Reason and Eros in the Ascent-Passage of the Symposium,” in J. Anton and
G. Kustas (eds.), Essays on Ancient Greek Philosophy (Albany: State University of New York
Press, 1971), 285–302.
Nussbaum, M. “The Speech of Alcibiades: A Reading of the Symposium,” in her Fragility of
Goodness (Cambridge: Cambridge University Press, 1986), 165–200.
Price, A. Love and Friendship in Plato and Aristotle (Oxford: Clarendon Press, 1989; 2nd ed. 1997).
Price, A. “Loving Persons Platonically,” Phronesis 26 (1981), 25–34.
Rossetti, L. (ed.) Understanding the Phaedrus: Proceedings of the Second Symposium Platonicum
(Sankt Augustin: Academia Verlag, 1992).
Rowe, C. J. “The Argument and Structure of Plato’s Phaedrus,” Proceedings of the Cambridge
Philological Society 32 (1986), 106–25.
Rowe, C. J. Plato: Phaedrus (Warminster, UK: Aris and Phillips, 1986).
Rowe, C. J. Plato: Symposium (Warminster, UK: Aris and Phillips, 1998).
Rowe, C. J. “The Unity of Plato’s Phaedrus: A Reply to Heath,” Oxford Studies in Ancient
Philosophy 7 (1989), 175–88.
Santas, G. Plato and Freud: Two Theories of Love (Oxford: Blackwell, 1988).
Scott, D. “Socrates and Alcibiades in Plato’s Symposium,” Hermathena 168 (2000), 25–37.
Sheffield, F. Plato’s Symposium: The Ethics of Desire (Oxford: Oxford University Press, 2006).
Sheffield, F. “Psychic Pregnancy and Platonic Epistemology,” Oxford Studies in Ancient
Philosophy 20 (2001), 1–35.
Vlastos, G. “The Individual as Object of Love in Plato,” in his PS, 3–34. Also in Fine (ed.), Plato
2, 137–63.
Wardy, R. “The Unity of Opposites in Plato’s Symposium,” Oxford Studies in Ancient Philosophy
23 (2002), 1–61.

Parmenides
Allen, R. E. Plato’s Parmenides: Translation and Analysis (Minneapolis: University of Minnesota
Press, 1983).
Cohen, S. M. “The Logic of the Third Man,” Philosophical Review 80 (1971), 448–75. Also in
Fine (ed.), Plato 1, 275–97.
Cornford, F. M. Plato and Parmenides (London: Routledge and Kegan Paul, 1939).
Fine, G. “Owen, Aristotle, and the Third Man,” Phronesis 27 (1982), 13–33.
Geach, P. “The Third Man Again,” Philosophical Review 65 (1956), 72–82. Also in SPM, 265–77.
McCabe, M. M. “Unity in the Parmenides: The Unity of the Parmenides,” in FA, 5–47.
Meinwald, C. “Good-bye to the Third Man,” in CCP, 365–96.
 Bibliography    715

Meinwald, C. Plato’s Parmenides (New York: Oxford University Press, 1991).

Mignucci, M. “Plato’s Third Man Arguments in the Parmenides,” Archiv für Geschichte der
Philosophie 72 (1990), 143–81.
Miller, M. Plato’s Parmenides (Princeton, NJ: Princeton University Press, 1986).
Moravcsik, J. M. E. “The “Third Man” Argument and Plato’s Theory of Forms,” Phronesis 8
(1963), 50–62.
Owen, G. E. L. “Notes on Ryle’s Plato,” in LSD, 85–103. Also in Fine (ed.), Plato 1, 298–319.
Peterson, S. “A Reasonable Self-Predication Premise for the Third Man Argument,” Philosophical
Review 82 (1973), 451–70.
Peterson, S. “The Greatest Difficulty for Plato’s Theory of Forms: The Unknowability Argument
of Parmenides 133c–134,” Archiv für Geschichte der Philosophie 63 (1981), 1–16.
Rickless, S. Plato’s Forms in Transition (Cambridge: Cambridge University Press, 2007).
Ryle, G. “Plato’s Parmenides,” in SPM, 85–103.
Sayre, K. Parmenides’ Lesson: Translation and Explication of Plato’s “Parmenides” (Notre Dame,
IN: University of Notre Dame Press, 1996).
Schofield, M. “Likeness and Likenesses in the Parmenides,” in FA, 49–77.
Sellars, W. “Vlastos and the “Third Man,” Philosophical Review 64 (1955), 405–37.
Strang, C. “Plato and the Third Man,” Proceedings of the Aristotelian Society 37 (suppl.) (1963),
147–64. Also in Vlastos (ed.), Plato 1, 184–200.
Vlastos, G. “The Third Man Argument in the Parmenides,” Philosophical Review 63 (1954),
319–49. Also in SPM, 231–63.
Vlastos, G. “Plato’s “Third Man” Argument (Parm. 132a1–b2): Text and Logic,” Philosophical
Quarterly 19 (1969), 289–81. Also in PS, 342–65.

Theaetetus
Ackrill, J. L. “Plato on False Belief: Theaetetus 187–200,” Monist 50 (1966), 383–402. Also in
Ackrill, Essays on Plato and Aristotle, 53–71.
Annas, J. “Knowledge and Language: The Theaetetus and the Cratylus,” in LL, 95–114.
Benson, H. “Why Is There a Discussion of False Belief in the Theaetetus?,” Journal of the History
of Philosophy 30 (1992), 171–99.
Bolton, R. “Plato’s Distinction between Being and Becoming,” Review of Metaphysics 29 (1975),
66–95.
Bostock, D. Plato’s Theaetetus (Oxford: Clarendon Press, 1988).
Broadie, S. “The Knowledge Unacknowledged in the Theaetetus,” Oxford Studies in Ancient
Philosophy 51 (2016), 87–117.
Burnyeat, M. F. “Conflicting Appearances,” Proceedings of the British Academy 65 (1979),
69–111.
Burnyeat, M. F. “Examples in Epistemology: Socrates, Theaetetus, and G. E. Moore,” Philosophy
52 (1977), 381–98.
Burnyeat, M. F. “Idealism and Greek Philosophy: What Descartes Saw and Berkeley Missed,”
Philosophical Review 90 (1982), 3–40.
Burnyeat, M. F. “The Materials and Sources of Plato’s Dream,” Phronesis 15 (1970), 101–22.
Burnyeat, M. F. “Plato on the Grammar of Perceiving,” Classical Quarterly NS 26 (1976),
29–51.
Burnyeat, M. F. “Protagoras and Self-Refutation in Later Greek Philosophy,” Philosophical
Review 85 (1976), 44–69.
716   Bibliography

Burnyeat, M. F. “Protagoras and Self-Refutation in Plato’s Theaetetus,” Philosophical Review 85

(1976), 172–95. Also in S. Everson (ed.), Epistemology (Cambridge: Cambridge University
Press, 1989), 39–59.
Burnyeat, M. F. “Socrates and the Jury: Paradoxes in Plato’s Distinction between Knowledge
and True Belief,” Aristotelian Society 54 (suppl.) (1980), 177–91. (Reply by J. Barnes,
193–206.)
Burnyeat, M. F. “Socratic Midwifery, Platonic Inspiration,” Bulletin of the Institute of Classical
Studies 24 (1977), 7–16. Also in EPS, 53–65.
Burnyeat, M. F. The Theaetetus of Plato (translated by M. J. Levett with revisions by Burnyeat
and introductory essay by Burnyeat) (Indianapolis, IN: Hackett, 1990).
Chappell, T. Reading Plato’s Theaetetus (Sankt Augustin: Academia Verlag, 2004; and
Indianapolis, IN: Hackett, 2005).
Cooper, J. “Plato on Sense-Perception and Knowledge (Theaetetus 184–6),” Phronesis 15 (1970),
123–46. Also in Fine (ed.), Plato 1, 355–76.
Cooper, J. Plato’s Theaetetus (New York: Garland, 1990).
Cornford, F. M. Plato’s Theory of Knowledge: The Theaetetus and Sophist (London: Routledge
and Kegan Paul, 1935).
Crivelli, P. “The Argument from Knowing and Not Knowing in Plato’s Theaetetus (187e5–188c8),”
Proceedings of the Aristotelian Society, n.s. 96 (1995/96), 177–96.
Crivelli, P. “Allodoxia,” Archiv für Geschichte der Philosophie 80 (1998), 1–29.
Dancy, R. “Theaetetus’ First Baby: 151e–160e,” Philosophical Topics 15 (1987), 61–108.
Day, J. “The Theory of Perception in Plato’s Theaetetus 152–183,” Oxford Studies in Ancient
Philosophy 15 (1997), 51–80.
Fine, G. “Conflicting Appearances: Theaetetus 153d–154b,” in FA, 105–33. Also in PKF, 160–83.
Fine, G. “False Belief in the Theaetetus,” Phronesis 24 (1979), 70–80. Also in PKF, 213–24.
Fine, G. “Knowledge and Logos in the Theaetetus,” Philosophical Review 88 (1979), 366–97.
Also in PKF, 225–51.
Fine, G. “Plato on Perception,” Oxford Studies in Ancient Philosophy (suppl.) (1988), 15–28.
Fine, G. “Protagorean Relativisms,” Proceedings of the Boston Area Colloquium in Ancient
Philosophy 10 (1996), 211–43. Also in PKF, 132–59.
Frede, D. “The Soul’s Silent Dialogue: A Non-aporetic Reading of the Theaetetus,” Proceedings
of the Cambridge Philological Society 215 (1989), 20–49.
Frede, M. “Observations on Perception in Plato’s Later Dialogues,” in his Essays in Ancient
Philosophy (Oxford: Clarendon Press, 1987), 3–8. Also in Fine (ed.), Plato 1, 377–83.
Holland, A. J. “An Argument in Plato’s Theaetetus: 184–6,” Philosophical Quarterly 23 (1973), 97–116.
Kanayama, Y. “Perceiving, Considering, and Attaining Being (Theaetetus 184–186),” Oxford
Studies in Ancient Philosophy 5 (1987), 29–81.
Kerferd, G. B. “Plato’s Account of the Relativism of Protagoras,” Durham University Journal 42
(1949), 20–26.
Ketchum, R. “Plato and Protagorean Relativism,” Oxford Studies in Ancient Philosophy 10 (1992),
73–105.
Lee, E. N. “ ‘Hoist on His Own Petard,’ ” in EA, 225–61.
Lesher, J. H. “Gnôsis” and Epistêmê in Socrates’ Dream in the Theaetetus,” Journal of Hellenic
Studies 89 (1969), 72–78.
Lewis, F. “Foul Play in Plato’s Aviary,” in EA, 262–84.
Lewis, F. “Two Paradoxes in the Theaetetus,” in J. M. E. Moravcsik (ed.), Patterns in Plato’s
Thought (Dordrecht: Reidel, 1973), 123–49.
 Bibliography    717

Matthen, M. “Perception, Relativism and Truth: Reflections on Plato’s Theaetetus 152–60,”

Dialogue 24 (1985), 33–58.
McDowell, J. “Identity Mistakes: Plato and the Logical Atomists,” Proceedings of the Aristotelian
Society 70 (1969–70), 161–80. Also in Fine (ed.), Plato 1, 384–97.
Modrak, D. “Perception and Judgment in the Theaetetus,” Phronesis 26 (1981), 35–54.
Morrow, G. “Plato and the Mathematicians: An Interpretation of Socrates’ Dream in the
Theaetetus (201e–206c),” Philosophical Review 79 (1970), 309–33.
Nehamas, A. “Epistêmê and Logos in Plato’s Later Thought,” Archiv für Geschichte der Philosophie
66 (1984), 11–36.
Rudebusch, G. “Plato on Sense and Reference,” Mind 104 (1985), 526–37.
Ryle, G. “Letters and Syllables in Plato,” Philosophical Review 69 (1960), 431–51.
Ryle, G. “Logical Atomism in Plato’s Theaetetus,” Phronesis 35 (1990), 21–46.
Sedley, D. The Midwife of Platonism: Text and Subtext in Plato’s Theaetetus (Oxford: Clarendon
Press, 2004).
Sedley, D. “Three Platonist Interpretations of the Theaetetus,” in FA, 79–103.
Silverman, A. “Flux and Language in the Theaetetus,” Oxford Studies in Ancient Philosophy 18
(2000), 109–52.
Silverman, A. “Plato on Perception and ‘Commons,’ ” Classical Quarterly NS 40 (1990), 148–75.
(Also discusses the Timaeus.)
Thaler, N. “Plato on the Importance of ‘This’ and ‘That’: The Theory of Flux and Its Refutation
in the Theatetus,” Oxford Studies in Ancient Philosophy 45 (2013), 1–42.
Tigner, S. “The ‘Exquisite’ Argument at Theaetetus 171a,” Mnemosyne 24 (1971), 366–69.
Waterfield, R. Plato: Theaetetus (Harmondsworth: Penguin, 1987).
Waterlow, S. “Protagoras and Inconsistency,” Archiv für Geschichte der Philosophie 59 (1977),
19–36.
Williams, B. A. O. “Introduction,” Plato’s Theaetetus, trans. M. J. Levett, rev. M. F. Burnyeat
(Indianapolis, IN: Hackett, 1992).
Williams, C. J. F. “Referential Opacity and False Belief in the Theaetetus,” Philosophical
Quarterly 22 (1972), 289–302.
Woolf, R. “Plato and the Norms of Thought,” Mind 122 (2013), 171–216.

Timaeus
Broadie, S. Nature and Divinity in Plato’s Timaeus (Cambridge: Cambridge University Press,
2012).
Bryan, J. Likeness and Likelihood in the Presocratics and Plato (Cambridge: Cambridge University
Press, 2012).
Burnyeat, M. F. “Eikôs Muthos,” Rhizai 2 (2005), 143–66.
Calvo, T., and Brisson, L. (eds.) Interpreting the Timaeus and Critias (Proceedings of the 4th
International Platonic Symposium) (Sankt Augustin: Academia Verlag, 1997).
Cherniss, H. F. “A Much Misread Passage in Plato’s Timaeus,” American Journal of Philology 75
(1954), 113–30.
Cherniss, H. F. “The Relation of the Timaeus to Plato’s Later Dialogues,” American Journal of
Philology 78 (1957), 225–66. Also in SPM, 339–78.
Cornford, F. M. Plato’s Cosmology: The Timaeus of Plato (London: Routledge and Kegan Paul,
1937).
Driscoll, J. “The Platonic Ancestry of Primary Substance,” Phronesis 24 (1979), 253–69.
718   Bibliography

Frede, D. “The Philosophical Economy of Plato’s Psychology: Rationality and Common

Concepts in the Timaeus,” in M. Frede and G. Striker (eds.), Rationality in Greek Thought
(Oxford: Clarendon Press, 1996), 1–34.
Frede, M. “Being and Becoming in Plato,” Oxford Studies in Ancient Philosophy (suppl.) (1988),
37–52.
Gill, M. “Matter and Flux in Plato’s Timaeus,” Phronesis 32 (1987), 34–53.
Johansen, T. K. “Body, Soul and Tripartition in Plato’s Timaeus,” Oxford Studies in Ancient
Philosophy 19 (2000), 87–111.
Johansen, T. K. Plato’s Natural Philosophy: A Study of the Timaeus-Critias (Cambridge:
Cambridge University Press, 2004).
Keyt, D. “The Mad Craftsman of the Timaeus,” Philosophical Review 80 (1971), 230–35.
Kung, J. “Mathematics and Virtue in Plato’s Timaeus,” in J. Anton and A. Preus (eds.), Essays in
Ancient Greek Philosophy, vol. 3 (Albany: State University of New York Press, 1989), 303–39.
Kung, J. “Why the Receptacle Is Not a Mirror,” Archiv für Geschichte der Philosophie 70 (1988),
167–78.
Lennox, J. “Plato’s Unnatural Teleology,” in D. O’Meara (ed.), Platonic Investigations (Washington,
DC: Catholic University Press of America, 1985), 195–218.
Mason, A. S. “Immortality in the Timaeus,” Phronesis 39 (1994), 90–97.
Mills, K. W. “Some Aspects of Plato’s Theory of Forms: Timaeus 49c ff.,” Phronesis 13 (1968),
145–70.
Mohr, R. The Platonic Cosmology (Leiden: Brill, 1985).
Mohr, R. and Sattler, B. (eds.), One Book, the Whole Universe: Plato’s Timaeus Today (Las
Vegas: Parmenides, 2010).
Morrow, G. “The Demiurge in Politics: The Timaeus and the Laws,” Proceedings and Addresses
of the American Philosophical Association 27 (1953–54), 5–23.
Morrow, G. “Necessity and Persuasion in Plato’s Timaeus,” Philosophical Review 59 (1950),
147–64. Also in SPM, 421–37.
Osborne, C. “Topography in the Timaeus,” Proceedings of the Cambridge Philological Society 34
(1988), 104–14.
Owen, G. E. L. “The Place of the Timaeus in Plato’s Dialogues,” Classical Quarterly NS 3 (1953),
79–95. Also in SPM, 313–38, and in LSD, 65–84.
Owen, G. E. L. “Plato and Parmenides on the Timeless Present,” Monist 50 (1966), 317–40.
Also in LSD, 27–44.
Patterson, R. “The Unique Worlds of the Timaeus,” Phoenix 35 (1981), 105–19.
Reed, N. H. “Plato on Flux, Perception and Language,” Proceedings of the Cambridge
Philological Society 18 (1972), 65–77.
Reydam-Schils, G. J. (ed.) Plato’s Timaeus as Cultural Icon (Notre Dame, IN: University of
Notre Dame Press, 2003).
Robinson, J. “The Tripartite Soul in the Timaeus,” Phronesis 35 (1990), 103–10.
Sedley, D. Creationism and Its Critics in Antiquity (Berkeley: University of California Press,
2007).
Sedley, D. “The Ideal of Godlikeness,” in Fine (ed.), Plato 2, 309–28.
Silverman, A. “Timaean Particulars,” Classical Quarterly NS 42 (1992), 87–113.
Strange, S. K. “The Double Explanation in the Timaeus,” Ancient Philosophy 5 (1985), 25–39.
Also in Fine (ed.), Plato 1, 397–415.
Taylor, A. E. A Commentary on Plato’s Timaeus (Oxford: Clarendon Press, 1928).
Vlastos, G. “Creation in the Timaeus: Is It a Fiction?,” in SPM, 401–19.
 Bibliography    719

Vlastos, G. “The Disorderly Motion in the Timaeus,” Classical Quarterly 33 (1939), 71–83. Also
in SPM, 379–99.
Vlastos, G. Plato’s Universe (Seattle: University of Washington Press, 1975).
Zeyl, D. “Plato and Talk of a World in Flux: Timaeus 49a6–50b5,” Harvard Studies in Classical
Philology 79 (1975), 125–48.

Sophist
Ackrill, J. L. “Plato and the Copula: Sophist 251–59,” Journal of Hellenic Studies 77 (1957),
1–6. Also in Ackrill, Essays on Plato and Aristotle, 80–92, and in Vlastos (ed.), Plato 1,
210–22.
Ackrill, J. L. “Sumplokê Eidôn,” Bulletin of the Institute of Classical Studies 2 (1955), 31–35. Also
in Ackrill, Essays on Plato and Aristotle, 72–79, and in SPM, 199–206, and in Vlastos (ed.),
Plato 1, 201–09.
Bluck, R. S. Plato’s Sophist: A Commentary (Manchester: Manchester University Press, 1975).
Bostock, D. “Plato on “Is-Not” (Sophist, 254–9),” Oxford Studies in Ancient Philosophy 2 (1984),
89–119.
Brown, L. “Being in the Sophist: A Syntactical Inquiry,” Oxford Studies in Ancient Philosophy 4
(1986), 49–70. Also in Fine (ed.), Plato 1, 455–78.
Brown, L. “Definition and Division in the Sophist,” in D. Charles (ed.), Ancient Theories of
Definition (Oxford: Clarendon Press, 2010), 151–71.
Brown, L. “Innovation and Continuity: The Battle of Gods and Giants, Sophist 245–259,” in
MAP, 181–207.
Campbell, L. The Sophistes and Politicus of Plato (Oxford: Clarendon Press, 1867).
Crivelli, P. Plato’s Account of Falsehood: A Study of the Sophist (Cambridge: Cambridge
University Press, 2012).
Frede, M. “The Literary Form of the Sophist,” in FA, 132–51.
Frede, M. “Plato’s Sophist on False Statements,” in CCP, 397–424.
Frede, M. Prädikation und Existenzaussage (Göttingen: Vandenhoeck and Ruprecht, 1967).
Hestir, B. Plato on the Metaphysical Foundation of Meaning and Truth (Cambridge: Cambridge
University Press, 2016).
Heinaman, R. “Being in the Sophist,” Archiv für Geschichte der Philosophie 65 (1983), 1–17.
Heinaman, R. “Communion of Forms,” Proceedings of the Aristotelian Society 83 (1982–83),
175–90.
Heinaman, R. “Once More: Being in the Sophist,” Archiv für Geschichte der Philosophie 68
(1986), 121–25.
Heinaman, R. “Self-Predication in the Sophist,” Phronesis 26 (1981), 55–66.
Ketchum, R. “Participation and Predication in the Sophist 251–60,” Phronesis 23 (1978), 42–62.
Keyt, D. “Plato on Falsity: Sophist 263b,” in EA, 285–305.
Keyt, D. “Plato’s Paradox That the Immutable Is Unknowable,” Philosophical Quarterly 19
(1969), 1–14.
Kostman, J. “False Logos and Not-Being in Plato’s Sophist,” in J. M. E. Moravcsik (ed.), Patterns
in Plato’s Thought (Dordrecht: Reidel, 1973), 192–212.
Lee, E. N. “Plato on Negation and Not-Being in the Sophist,” Philosophical Review 81 (1972),
267–304.
Lewis, F. “Did Plato Discover the estin of Identity?,” California Studies in Classical Antiquity 8
(1975), 113–43.
720   Bibliography

Lewis, F. “Plato on ‘Not,’ ” California Studies in Classical Antiquity 9 (1976), 89–115.

Malcolm, J. “Plato’s Analysis of to on and to mê on in the Sophist,” Phronesis 12 (1967),
130–46.
Malcolm, J. “Remarks on an Incomplete Rendering of Being in the Sophist,” Archiv für Geschichte
der Philosophie 67 (1985), 162–65.
McDowell, J. “Falsehood and Not-Being in Plato’s Sophist,” in LL, 115–34.
McPherran, M. “Plato’s Reply to the ‘Worst Difficulty’ Argument of the Parmenides: Sophist
248a–249d,” Archiv für Geschichte der Philosophie 68 (1986), 233–52.
Moravcsik, J. M. E. “Being and Meaning in the Sophist,” Acta Philosophica Fennica 14 (1962),
23–78.
Moravcsik, J. M. E. “Sumplokê eidôn and the Genesis of Logos,” Archiv für Geschichte der
Philosophie 42 (1960), 117–29.
Nehamas, A. “Participation and Predication in Plato’s Later Thought,” Review of Metaphysics
36 (1982), 343–74.
Notomi, N. The Unity of Plato’s Sophist (Cambridge: Cambridge University Press, 1999).
Owen, G. E. L. “Plato on Not-Being,” in Vlastos (ed.), Plato 1, 223–67. Also in Fine (ed.), Plato
1, 416–54.
Pelletier, F. J. Parmenides, Plato, and the Semantics of Not-Being (Chicago: Chicago University
Press, 1992).
Prior, W. “Plato’s Analysis of Being and Not-Being in the Sophist,” Southern Journal of Philosophy
18 (1980), 199–211.
Reeve, C. D. C. “Motion, Rest, and Dialectic in the Sophist,” Archiv für Geschichte der Philosophie
67 (1985), 47–64.
Roberts, J. “The Problem about Being in the Sophist,” History of Philosophy Quarterly 3 (1986),
229–43.
van Eck, J. “Falsity without Negative Predication,” Phronesis 40 (1995), 20–47.
van Eck, J. “Plato’s Logical Insights: Sophist 254d–257a,” Ancient Philosophy 20 (2000),
53–79.
Vlastos, G. “An Ambiguity in the Sophist,” in PS, 270–322.
Wiggins, D. “Sentence Meaning, Negation, and Plato’s Problems of Not-Being,” in Vlastos
(ed.), Plato 1, 268–303.

Statesman
Ackrill, J. L. “In Defense of Platonic Division,” in O. P. Wood and G. Pitcher (eds.), Ryle: A
Collection of Critical Essays (Garden City, NY: Doubleday Anchor Books, 1970), 373–92.
Also in Ackrill, Essays on Plato and Aristotle, 93–109.
Bobonich, C. “The Virtues of Ordinary People in Plato’s Statesman,” in C. J. Rowe (ed.), Reading
the Statesman (Sankt Augustin: Academia Verlag, 1995), 313–20.
Cooper, J. “Plato’s Statesman and Politics,” Proceedings of the Boston Area Colloquium in
Ancient Philosophy 13 (1999), 71–104. Reprinted in his Reason and Emotion: Essays on
Ancient Philosophy (Princeton: Princeton University Press, 1999), 165–91.
Gill, C. “Plato and Politics: The Critias and Politicus,” Phronesis 24 (1979), 148–67.
Gill, C. “Rethinking Constitutionalism in Statesman 291–303,” in C. J. Rowe (ed.), Reading the
Statesman (Sankt Augustin: Academia Verlag, 1995), 292–305.
Lane, M. Method and Politics in Plato’s Statesman (Cambridge: Cambridge University Press,
1997).
 Bibliography    721

Moravcsik, J. M. E. “The Anatomy of Plato’s Divisions,” in EA, 324–48.

Moravcsik, J. M. E. “Plato’s Method of Division,” in J. M. E. Moravcsik (ed.), Patterns in Plato’s
Thought (Dordrecht: Reidel, 1973), 158–80.
Owen, G. E. L. “Plato on the Undepictable,” in EA 349–61. Also in LSD, 138–47.
Rowe, C. J. (ed. and trans.) Plato: Statesman (Warminster: Aris and Phillips, 1995).
Rowe, C. J. “The Politicus: Structure and Form,” in FA, 153–78.
Rowe, C. J. (ed.) Reading the Statesman: Proceedings of the Third Symposium Platonicum (Sankt
Augustin: Academia Verlag, 1995).
Skemp, J. B. Plato’s Statesman (London: Routledge and Kegan Paul, 1952; 2nd ed. Bristol:
Bristol Classical Press, 1987).

Philebus
Bury, R. G. The Philebus of Plato (Cambridge: Cambridge University Press, 1897; reprinted by
Arno Press in 1973).
Carpenter, A. “Ranking Knowledge in Plato’s Philebus,” Phronesis 60 (2015), 180–205.
Cooper, J. “Plato’s Theory of Human Good,” Journal of Philosophy 74 (1977), 714–30. Also in
Fine (ed.), Plato 2, 329–44.
Cooper, N. “Pleasure and the Good in Plato’s Philebus,” Philosophical Quarterly 18 (1968),
12–15.
Dancy, R. “The One, the Many, and the Forms: Philebus 15b1–8,” Ancient Philosophy 4 (1984),
160–93.
Dybikowski, J. “False Pleasure and the Philebus,” Phronesis 15 (1970), 147–65.
Dybikowski, J. “Mixed and False Pleasures in the Philebus: A Reply,” Philosophical Quarterly
20 (1970), 244–47.
Fletcher, E. “The Divine Method and the Disunity of Pleasure in the Philebus,” Journal of the
History of Philosophy 55 (2017), 179–208.
Fletcher, E. “Plato on Pure Pleasure and the Best Life,” Phronesis 59 (2014), 113–42.
Frede, D. “Disintegration and Restoration: Pleasure and Pain in Plato’s Philebus,” in CCP,
425–63.
Frede, D. “Rumpelstiltskin’s Pleasures: True and False Pleasures in Plato’s Philebus,” Phronesis
30 (1985), 151–80.
Gosling, J. C. B. “False Pleasures: Philebus 35c–41b,” Phronesis 4 (1959), 44–54.
Gosling, J. C. B. “Father Kenny on False Pleasures,” Phronesis 6 (1961), 41–45.
Hackforth, R. Plato’s Examination of Pleasure (Cambridge: Cambridge University Press,
1945).
Hampton, C. Pleasure, Knowledge, and Being: An Analysis of Plato’s Philebus (Albany: State
University of New York Press, 1990).
Harte, V. “The Philebus on Pleasure: The Good, the Bad and the False,” Proceedings of the
Aristotelian Society 104 (2003–2004), 111–28.
Harte, V. “Desire, Memory and the Authority of Soul: Plato Philebus 35CD,” Oxford Studies in
Ancient Philosophy 46 (2014), 33–72.
Harte, V. “The Life of Protarchus’ Choosing: Plato Philebus 20b–22c,”’ in Mi-Kyoung Lee (ed.)
Strategies of Argument: Essays in Ancient Ethics, Epistemology and Logic. (Oxford: Oxford
University Press, 2014), 3–20.
Harte, V. “Some ‘Value of Knowledge” Problems in Plato’s Philebus,” Proceedings of the Aristotelian
Society Supplementary Volume 92 (2018).
722   Bibliography

Harte, V. “Plato’s Philebus on the Value of Idle Pleasure,” in David Brink, Susan Sauvé Meyer,
and Christopher Shields (eds.), Virtue, Happiness, and Knowledge: Themes from the Work of
Gail Fine and Terence Irwin (Oxford: Oxford University Press, 2018).
Kenny, A. “False Pleasures in the Philebus: A Reply to Mr. Gosling,” Phronesis 5 (1960), 45–52.
Meinwald, C. “One/Many Problems,” Phronesis 41 (1996), 95–103.
Meinwald, C. “Plato’s Pythagoreanism,” Ancient Philosophy 22 (2002), 87–101.
Meinwald, C. “Prometheus” Bounds,” MAP, 165–80.
Moravcsik, J. M. E. “Forms, Nature and the Good in the Philebus,” Phronesis 24 (1979), 81–104.
Penner, T. “False Anticipatory Pleasures: Philebus 36a3–41a6,” Phronesis 15 (1970), 166–78.
Sayre, K. “The Philebus and the Good,” Proceedings of the Boston Area Colloquium in Ancient
Philosophy 2 (1987), 45–71. (Reply by P. Mitsis, 72–78.)
Striker, G. Peras und Apeiron (Göttingen: Vandenhoeck and Ruprecht, 1970).
Taylor, A. E. Philebus and Epinomis (London: T. Nelson, 1956).
Thomas, C. “Plato’s Prometheanism,” Oxford Studies in Ancient Philosophy 31 (2006), 23–32.
Waterfield, R. “The Place of the Philebus in Plato’s Dialogues,” Phronesis 25 (1980), 270–305.
Waterfield, R. Plato: Philebus (Harmondsworth: Penguin, 1982).

Laws
Annas, J. Virtue and Law in Plato and Beyond (Oxford: Oxford University Press, 2017). Also
discusses the Republic.
Bobonich, C. “Persuasion, Compulsion, and Freedom in Plato’s Laws,” Classical Quarterly 41
(1991), 365–88. Also in Fine (ed.), Plato 2, 373–403.
Bobonich, C. “Plato’s Theory of Goods in the Laws and Philebus,” Proceedings of the Boston
Area Colloquium in Ancient Philosophy 11 (1997), 101–39. (Reply by J. Gentzler, 140–51.)
Bobonich, C. “Reading the Laws,” in FA, 249–82.
Bobonich, C. Plato’s Laws: a Critical Guide (Cambridge: Cambridge University Press, 2010).
England, E. B. The Laws of Plato, 2 vols. (New York: Arno Press, 1976). Originally pub. 1921.
Laks, A. “The Laws” in C. Rowe and M. Schofield (eds.), The Cambridge History of Greek and
Roman Political Thought (Cambridge: Cambridge University Press, 2000), 258–92.
Menn, S. Plato on God as Nous (Carbondale: Southern Illinois University Press, 1995).
Morrow, G. Plato’s Cretan City (Princeton, NJ: Princeton University Press, 1993). Orig. pub.
1960.
Meyer, S. “Plato on the Law” in H. Benson (ed.), A Companion to Plato, 373–87. Also discusses
other dialogues.
Meyer, S. “Self-Mastery and Self-Rule in Plato’s Laws,” in Brink, Meyer, and Shields (eds.),
Virtue, Happiness, and Knowledge.
Roberts, J. “Plato on the Causes of Wrongdoing in the Laws,” Ancient Philosophy 7 (1987), 23–37.
Saunders, T. “Plato’s Later Political Thought,” in CCP, 464–92.
Saunders, T. Plato’s Penal Code: Tradition, Controversy, and Reform in Greek Penology (Oxford:
Clarendon Press, 1991).
Saunders, T. “The Socratic Paradoxes in Plato’s Laws,” Hermes 96 (1968), 421–34.
Saunders, T., and Brisson, L. Bibliography on Plato’s Laws (Sankt Augustin: Academia Verlag,
2000).
Schofield, M. “Injury, Injustice, and the Involuntary in the Laws,” Oxford Studies in Ancient
Philosophy, suppl. vol. (2012), 103–14.
Stalley, R. E. An Introduction to Plato’s Laws (Indianapolis, IN: Hackett, 1983).
 Bibliography    723

Warren, J. “Comparing Lives in Plato, Laws 5,” Phronesis LVIII 4 (2013), 319–46.
Wilburn, J. “Akrasia and Self-Rule in Plato’s Laws,” Oxford Studies in Ancient Philosophy XLIII
(2012), 25–53.
Wilburn, J. “Moral Education and the Spirited Part of the Soul in Plato’s Laws,” Oxford Studies
in Ancient Philosophy XL (2013), 63–102.

Aristotle’s Criticism of Plato

Annas, J. “Aristotle on Substance, Accident and Plato’s Forms,” Phronesis 22 (1977), 146–60.
Annas, J. Aristotle’s Metaphysics M and N (Oxford: Clarendon Press, 1976).
Annas, J. “Forms and First Principles,” Phronesis 19 (1974), 257–83.
Cherniss, H. F. Aristotle’s Criticism of Plato and the Academy (Baltimore: Johns Hopkins
University Press, 1944; repr. New York: Russell and Russell, 1962).
Cherniss, H. F. The Riddle of the Early Academy (Berkeley: University of California Press,
1945).
Devereux, D. “The Primacy of Ousia: Aristotle’s Debt to Plato,” in D. O’Meara (ed.), Platonic
Investigations (Washington, DC: Catholic University of America Press, 1985), 226–32.
Düring, I., and Owen, G. E. L. (eds.) Aristotle and Plato in the Mid-Fourth Century (Göteborg:
Elanders Boktryckeri Aktiebolag, 1960).
Fine, G. On Ideas: Aristotle’s Criticism of Plato’s Theory of Forms (Oxford: Clarendon Press,
1993).
Fine, G. “Plato and Aristotle on Form and Substance,” Proceedings of the Cambridge Philological
Society 209 (1983), 23–47. Also in PKF, 397–425.
Matthews, G. B., and Cohen S. M. “The One and the Many,” Review of Metaphysics 21 (1968),
630–55.
Owen, G. E. L. “Dialectic and Eristic in the Treatment of Forms,” in LSD, 221–38. Reprinted
from Owen (ed.), Aristotle on Dialectic: The Topics (Oxford: Oxford University Press, 1968),
103–25.
Owen, G. E. L. “Logic and Metaphysics in Some Early Works of Aristotle,” in LSD, 180–99.
Reprinted from Düring and Owen (eds.), Aristotle and Plato in the Mid-Fourth Century
(Göteborg: Elanders Boktryckeri Aktiebolag, 1960), 163–99.
Owen, G. E. L. “The Platonism of Aristotle,” in LSD, 200–20. Reprinted from Proceedings of
the British Academy 51 (1966), 125–50.
Owen, G. E. L. “A Proof in the Peri Ideôn,” in LSD, 165–79. Reprinted from Journal of Hellenic
Studies 77 (1957), 103–11.
Shields, C. Order in Multiplicity: Homonymy in the Philosophy of Aristotle (Oxford: Clarendon
Press, 1999).
Shields, C. “Learning about Plato from Aristotle” in H. Benson (ed.), A Companion to Plato,
403–41.

Later Platonism
Annas, J. Platonic Ethics, Old and New (Ithaca, NY: Cornell University Press, 1999).
Baltzly, D. Proclus: Commentary on Plato’s Timaeus, vol. 3 (Cambridge: Cambridge University
Press, 2007).
Baltzly, D. Proclus: Commentary on Plato’s Republic (Cambridge: Cambridge University Press,
2018).
724   Bibliography

Barnes, J. “Antiochus of Ascalon,” in M. Griffin and J. Barnes (eds.), Philosophia Togata 1

(Oxford: Clarendon Press, 1989), 51–96.
Barnes, J. Porphyry: Introduction (Oxford: Clarendon Press, 2003).
Blumenthal, H. Soul and Intellect: Studies in Plotinus and Later Neoplatonsim (Aldershot:
Variorum, 1993).
Boys-Stones, G. Post-Hellenistic Philosophy: A Study of Its Development from the Stoics to
Origen (Oxford: Oxford University Press, 2001).
Boys-Stone, G. R. Platonist Philosophy 80 BC to AD 250: An Introduction and Collection of
Sources in Translation (Cambridge: Cambridge University Press, 2018).
Boys-Stone, G. R. and Rowe, C. J. The Circle of Socrates: Readings in the First-Generation
Socratics (Indianapolis, IN: Hackett, 2013).
Brennan, T., and Brittain, C. Simplicius: On Epictetus’ Handbook (Ithaca, NY: Cornell
University Press, 2002).
Brittain, C. Philo of Larissa: The Last of the Academic Sceptics (Oxford: Oxford University
Press, 2001).
Dillon, J. Alcinous: The Handbook of Platonism (Oxford: Clarendon Press, 1996).
Dillon, J. The Heirs of Plato: A Study of the Old Academy, 347–274 B.C. (Oxford: Clarendon
Press, 2003).
Dillon, J. The Middle Platonists: A Study of Platonism, 80 B.C. to A.D. 220 (London: Duckworth,
1977).
Dodds, E. R. Proclus. The Elements of Theology, 2nd ed. (Oxford: Clarendon Press, 1963).
Emilsson, E. Plotinus on Sense-Perception (Cambridge: Cambridge University Press, 1988).
Emilsson, E. Plotinus on the Intellect (Cambridge: Cambridge University Press, 2007).
Emilsson, E. Plotinus (New York: Routledge, 2017).
Gerson, L. Aristotle and Other Platonists (Ithaca, NY: Cornell University Press, 2005).
Gerson, L. From Plato to Platonism (Ithaca, NY: Cornell University Press, 2013).
Gerson, L. (ed.) The Cambridge Companion to Plotinus (Cambridge: Cambridge University
Press, 1996).
Gerson, L. “Neoplatonism,” in C. Shields (ed.), The Blackwell Guide to Ancient Philosophy
(Oxford: Blackwell, 2003), 303–23.
Gerson, L. Plotinus (London: Routledge and Kegan Paul, 1994).
Gertz, S. Death and Immortality in Late Neoplationism (Leiden: Brill, 2011).
Glucker, J. Antiochus and the Late Academy (Göttingen: Vandenhoeck and Ruprecht, 1978).
Karamanolis, G. Plato and Aristotle in Agreement? (Oxford: Clarendon Press, 2006).
Morrow, G., and Dillon, J. Proclus’ Commentary on Plato’s Parmenides (Princeton, NJ:
Princeton University Press, 1987).
Jackson, R., Lycos, K., and Tarrant, H., Olympiodorus: Commentary on Plato’s Gorgias (Leiden
and Boston: Brill, 1998).
O’Meara, D. Plotinus: An Introduction to the Enneads (Oxford: Clarendon Press, 1993).
Runia, D. and Share, M. Proclus: Commentary on Plato’s Timaeus, vol. 2 (Cambridge:
Cambridge University Press, 2008).
Sedley, D. “Alcinous’ Epistemology,” in K. Algra, P. van der Horts, and D. Runia (eds.),
Polyhistor (Leiden: E. J. Brill, 1996), 300–12.
Smith, A. Porphyry’s Place in the Neoplatonic Tradition (The Hague: Nijhoff, 1974).
Sorabji, R. (ed.) Aristotle Transformed (Ithaca, NY: Cornell University Press, 1990).
Sorabji, R. The Philosophy of the Commentators 200–600 AD, 3 vols. (London: Duckworth, 2004).
Stead, C. Philosophy in Christian Antiquity (Cambridge: Cambridge University Press, 1994).
 Bibliography    725

Steel, C. The Changing Self: A Study on the Soul in Later Neoplatonism (Brussels: Paleis der
Academiën, 1978).
Tarrant, H. Proclus: Commentary on Plato’s Timaeus, vol. 1 (Cambridge: Cambridge University
Press, 2007).
Tuominen, M. The Ancient Commentators on Plato and Aristotle (Stocksfield: Acumen, 2009).
Wallis, R. Neoplatonism (London: Duckworth, 1972; 2nd ed., 1995).
Westerink, L. Anonymous Prolegomena to Platonic Philosophy (Amsterdam: North-Holland,
1962).
Westerink, L. The Greek Commentaries on Plato’s Phaedo, 2 vols. (Amsterdam: North-Holland,
1976–1977).
 Index Locorum

Aristotle Magna Moralia

1182a15–30: 404
Categories 1200b34–37: 414n15
1a24–25: 17
2a14–15: 657, 660 Metaphysics
2b20–21: 657, 660 1.1: 198
2b32–35: 657n26 1.6: 15
2b8–14: 657, 660 5.1: 287n1
13b14–19: 469n38 5.23: 17
7.13: 29
De anima 12.8: 638
402a21–22: 287 13.4: 15
406b24–28: 520 13.9: 15
408b12–15: 515 987a29b–7: 404
408b30–31: 520 987a32–b10: 57
417b20: 297n35 987b18–988a17: 349n39
433a15–20: 298n40 987b9–10: 481
988b6–8: 298n40
De caelo 990b17: 239
279b32–280a2: 291n14 990b28: 658
991a12–14: 648
De interpretatione 991a12–20: 61
16b11–15: 469n38 991a20–23: 648
17a1–2: 503 991a8: 648
17a20–24: 503 991b1: 648
17a38–40: 651n18, 657n26, 658 993b23–31: 663
17a38–b1: 15 997b3–12: 649n16
17a8–9: 503 998a24–26: 656
998b4–6: 656
Eudemian Ethics 999a19–22: 649n16
1216b3–10: 404 999b17–23: 649n16
1219a11–18: 425n31 999b25–26: 663
1249a21–b4: 377 1003a13–17: 656
2.2–3: 377 1003a14–15: 656
1003a15–17: 655n22
Generation and Corruption 1003a7–9: 656
324b13–18: 298n40 1010a7–15: 58
778b1: 297n35 1013a14–17: 656
728   index locorum

Metaphysics (Continued) Parts of Animals

1013a16–17: 291n16 639b6: 651n18
1038a34–1039a2: 661 639b–640a: 297n37
1039a1: 658
103b26–28: 648 Physics
1040a26–27: 658 4.3: 17
1040a8–9: 657 185b25–27: 645
1079a13: 239 189b30: 651n18
1079a25: 658 192b13–14: 661
1086a31–b14: 649 192b14–15: 662
1086a32–34: 653 192b20–23: 661
1086a32–b11: 652 194b16–20: 656
1086a33: 652n19 194b29–32: 297n35
1086a33–34: 661 194b30: 298n38
1086a34: 655n22, 658 195a32: 651n18
1086b10–11: 35, 650, 653 195b21–27: 354
1086b27: 661 202a9–12: 354
1086b32–34: 650 209b11–16: 61n46
1086b3–5: 649 240a28–b7: 519
1086b5: 649 259a8: 648
1086b5–6: 649
1087b11: 655 Poetics
1087b9–10: 653 1447b11: 41
1178b12–32: 398 1456b8–18: 607
1447b13–16: 292n17
Nicomachean Ethics
2.3: 377 Politics
2.5: 377 1264b26: 359n1
6.12–13: 377 1265a6–8: 622
1094a1–22: 343n19 1305a11–15: 606
1095b14–22: 341n15
1096b32–34: 648 Posterior Analytics
1097a15–1097b6: 343n19 1.9–10: 287n2
1097b14–21: 341n14 2.19: 198
1102b16–18: 520 100b10: 277n32
1105a26–b5: 408n7 72b19–24: 277n32
1107a1–2: 377 83a32–34: 648
1110b18–24: 418n21 87b28–88a17: 656
1113a15–31: 417n20 99b20: 277n32
1140b21–25: 408n8
1140b33: 277n32 Rhetoric
1141a9–12: 408n8 1387b23–25: 300
1144b28–30: 404 1407b6–9: 607
1145b23–24: 414n15
1171a10–12: 552 Sophistical Refutations
1177a12–18: 341n15 22: 29
1177b26–1178a8: 341n15 165b20–21: 607
 index locorum   729

178b27–28: 503 23a–b: 99

178b36–179a10: 661 23b3–4: 123, 123n14
183b7–8: 404 23b7: 121n5
226b1–231b8: 405 24a: 381
24c–25c: 607
Topics 24d–28a: 101
113a24–32: 240n27 24e–25a: 610
137b3–13: 240n27 26b: 106
141b26: 281n39 26d–e: 77
143b25–32: 240n27 28a–33c: 42
147a5–11: 240n27 28b–d: 611
148a14–22: 240n27 29a: 399
159a15–35: 231n3 29a5–b5: 5
29b: 3, 380n5, 381, 429, 611
29b1–7: 387
Plato 29b6–7: 121–23, 122n13, 577n6
29d–30b: 537
First Alcibiades [?] 29d3–5: 136
106b: 231n4 29d7–30b4: 419
118a: 585n25 29d–e: 576, 605
29e1–3: 124
Apology 29e3–30a3: 406
18a7–19a7: 120 29e5–30a3: 124
18a7–20c3: 405 30a9–b2: 124
19b4–c1: 120 30a–b: 399, 534
19d: 611 30b2–4: 363n12
19d–20c: 430 30c7–d6: 426
20a2–b9: 338 30e6–31a2: 576
20a–c: 430 31a9–c3: 415n17
20c: 611 31b: 605
20c1–2: 120 31c4–32a3: 576, 578
20d6–e3: 5 32a4–e1: 136
21–34: 97 32b8–d1: 586n28
21a–22e: 379 32e–33a: 576
21b: 419 33a–b: 611
21b4–5: 120–21 33a–c: 42n4
21b6: 120 34a: 41
21b–e: 392 36: 97
21c8–d1: 121n5 36b–c: 576
21d: 3, 380, 429–30 36c2–d1: 123n17
21d1–7: 120 36c3–d1: 419
21d2–8: 5 36c3–d10: 419
21d4–5: 122 36c4–8: 576
22b–c: 610 36d: 583
22e7–23a5: 405 36d–e: 576
23a3–5: 121 37b: 429
23a6–b4: 121 37b2–8: 123n14
730   index locorum

Apology (Continued) Cratylus

37b2–c2: 395 383a1–384e2: 487
37b7–8: 122 383a5–b2: 487
37e5–38a1: 120 383b2–3: 486
38a: 201n60 384a4–5: 487
38a1–7: 404, 419 384a–c: 44
38a5–6: 119 384d1–5: 486
39a6–b6: 426 385a1–391a1: 487–88
40c–e: 399–400, 589n33 385b3–d1: 89n79
41b5–7: 124 386d: 115
41b5–c4: 340n12 387c5: 89n79
41b7–c4: 419 389a5–390a10: 488
41c–e: 535 389d6: 658
41e1–42a1: 123n17 390b1–d8: 489
390c: 231n3
Charmides 390d9–391b3: 489
153e–154d: 569 391a: 96
154d: 96 391b4–5: 489
154e–155a: 231n4 391c10–421c2: 489
155c–e: 96 393d3–4: 489
158: 122n57 399a–b: 326n44
158d8–159a4: 406 399c1: 486
159b–d: 388 404b: 187n10
159c–160b: 104n54 414a8: 486
159e: 579 421c3–427d3: 489–90
160b3: 104n54 421c5: 486
160b7: 104n54 421d–e: 326n44
160b–c: 388 422a1–e1: 495
160e: 579 423b9–424a1: 488
161a8–9: 424 423c–d: 660
162b9–10: 121n5 424c5–d4: 488n17
163a–d: 44 425a2–3: 495n35
164c–d: 104 426d3–e6: 492n25
164d: 105, 108 427d4–440e7: 490–92
165b–c: 380n5, 381 429c6–430a5: 493, 497
166: 98 430a6–431c3: 497
166c–d: 201n60 430a7–8: 497n39
166d: 96 430d3: 497n40
167e1–5: 417 430d6: 497n40
169c: 104 430e1: 497n39
170d–171c: 581n15 431b4: 497n39
171d5–e5: 581 433e8: 486
171e7–172a5: 420n25 437d10: 72n14
173c7–d5: 420n25 439b–440e: 58
175b3–4: 488n19
176a6–b4: 121n5 Critias
176a–b: 7n16, 394n33 106a4: 291n15
191a: 388 108a–b: 85n57
 index locorum   731

Crito 271e–272b: 202n65

44b5–c5: 132 272c–d: 96
45c6–7: 132 272e: 96
45d8–46a2: 132 273a: 96
46b3–6: 132, 137 275–276: 104
46b3–c3: 577, 582 275a–c: 231n4
46d: 201n60 275e: 104n53
46d1–2: 133 278d1–3: 421
474a–b: 578 278e: 8n17
47a–48b: 577 278e–282a: 577n5, 592n43
47c11: 578n10 278e–282d: 532, 534
47c8–48a4: 426 278e3–281b4: 420
47d–48a: 154n23, 535 280c3–d1: 422
47d–e: 588n33 280c3–d7: 420n26
47e4–6: 420n26 280d4–6: 425n32
48–49: 7 281: 99, 112
48b: 18n51 281a2–6: 422
48b10–d6: 133 281c: 366, 369n24
48b2–4: 133 282d8–283b3: 421
48b2–8: 420n25 283–288: 105
48b4–5: 133 283a1–4: 426n35
48b4–6: 419 283a–288a: 493n28
48b7–8: 133 283e7–184c6: 333n66
48d–e: 577 284b1–c6: 493
49a–d: 535 285: 99
49b9–c11: 133 285–286: 333n66
49c10–d5: 578 285c4–6: 580
49c11–d5: 133 285d–288a: 112
49c7–8: 133n45 285e9–286c9: 270n16
49e6–50a3: 135, 137 287a–b: 104
49e6–7: 134 288a: 105
49e9–50a2: 134 288d5–292e5: 421
508e–509a: 578 289b7–c4: 488n20
50a2–3: 134 290: 106, 110
50a4–5: 121n5, 134–35, 137 290c: 115n80
50a6–51c5: 135 290e1: 78
50d5: 137 291a: 112n72
50e1: 137 291b4–7: 422, 422n28
51c6–53a8: 135 291c4–9: 424
51e7–52a3: 136 291d: 110
52a6: 137 292b1–e5: 421
52b: 101n34 292b6–c1: 422n28
53a9–54d2: 135 292e: 110
293: 98n17
Euthydemus 293a: 96
271: 93 293d–296d: 115
271–272: 101n35 296e3–297a2: 123n14
271a: 54 298b: 96
732   index locorum

Euthydemus (Continued) 10e10–11a3: 126, 126n23

303: 106 10e2–11a6: 127
304b–307c: 54 10e2–5: 125
10e6–9: 125
Euthyphro 11a: 384
2–4: 93 11a3–6: 127
2a–5d: 94 11a5–b1: 129
3d5–9: 415n17 11b: 111
4a–d: 96 11b6–7: 131
4d9–e1: 124 11b6–8: 128
4e4–8: 124 11b6–e1: 134n50
4e–5a: 96, 431 11e–15e: 94
4e9–5a2: 125 11e7–12d11: 128
5a: 393 12e6–8: 128
5a–c: 96 13d8–9: 128
5c4–d4: 6 14b–c: 393
5c9: 125 14c: 111
5d1: 15 14c6: 128
5d–11b: 94 15b1–6: 128
5d7: 236 15c: 104, 111
5d8–6e2: 6 15c11–d8: 124
5d8–e2: 125 15c1–3: 128
6c–e: 383 15c8–9: 128
6d: 111, 239 15c–e: 96
6d10–11: 15, 125, 129 15d–16a: 96
6d11: 5, 129n36, 459 15d4–8: 7n16, 162n3
6d86–8: 6 15e–16a: 94
6d9–e1: 6
6d–e: 431 Gorgias
6e1: 5 447a: 141n2
6e3–6: 398 448c: 43
6e4: 5 448e6–7: 498
8a: 111 449a: 606
9c: 111 449a–d: 145
9d: 97n13 449b–c: 142n3
9e1–3: 6 452e: 145, 606
9e4–7: 170n47 454c–455a: 611
10–11: 105 454e: 608
10–11b5: 6 454e–455a: 145
10a4: 127 456b–c: 608
10a5–c13: 126 457: 104
10c10–13: 127 457a–458a: 201n60
10c10–14: 125 458a: 611
10d1–e1: 127 459c8–460a4: 510
10d4: 127 459e: 608
10d6–8: 125, 127 460a5–c6: 426
10d9–11: 125 460b7–461a7: 510
 index locorum   733

460c3: 513 476a–477a: 153

461a4–7: 513 476a–e: 582
461b: 151, 608 477b: 153
461b1: 339n7 478b8: 582n16
461b–462e: 146n10 478d–480d: 583
462b: 43 479b3–c4: 426
463a: 54 480a6–b2: 426
464b2–c3: 407n6 481b5: 41
464b2–c5: 422 482: 96, 106
464b–465d: 146 482b–c: 104
464c–465c: 608n7 482c–e: 151n19
464d: 311 483b–484a: 607
465a: 611n13 484a: 231n4
465b–c: 610 485a–e: 540
465d: 146 485e–486b: 46
466a–468e: 580 486c: 231n4
466a–471d: 53 486e: 96
466b4: 151 489: 98, 104
466b–c: 151 491a–b: 155
466b–d: 145 491b2–4: 514
466c1: 151 491b–492a: 511–12
466d–480d: 63 491b–c: 104
466d6–e2: 511 491d4–e1: 512
466d8–e2: 151 491d7–e1: 423
466d–e: 611 491d–e: 155
466e13–467a10: 415n18 492a: 155
467c5–481b5: 415 492a2: 154
467e: 151 492c4–494a5: 425
468a: 238 492d: 156
468b1–6: 510–11 492e–495a: 543
468b1–8: 513, 528n15 493a: 156, 221, 612
468b2: 151 493a1–b3: 423
468c5–7: 151 493a3–4: 155, 511
469a–475e: 579 493a5–493c4: 512
469b–c: 152 494a–c: 154n25
470c–471a: 535 494c–d: 154
470e: 152, 605 494d4: 154
470e4–11: 419 494e7: 154
471a: 153 495a–b: 148n15
471e1–472c: 142 495a–c: 144
471e–472c: 535 495d: 143
472b8: 495n35 495d–499d: 573
472d–48a: 535 495e: 611
474b–c: 611 497d8–e3: 424
474c–475e: 153 498a–499b: 154n26
475b8–c3: 153n22 499b–e: 148n15
475d8–9: 153n22 499e7–500a3: 524n32
734   index locorum

Gorgias (Continued) 517a4–6: 407n6

49c6–e9: 417 517b2–c2: 416, 423, 423n29
500e–501b: 538 517b–c: 609
501c: 293 517c5–518a7: 424
501d–502e: 610 519c2–520e11: 405
502e–503b: 577 519c3–d1: 405
503c6–d3: 416 520b2–3: 422
503d6–504e4: 422 521c7–d3: 123n14
504a–b: 444 521d–522c: 581
504b7–c9: 422–23 521d6–8: 178n90, 407n6
504c5–9: 426n35 521d–e1: 576–77
504c5–d3: 426 521e–522c: 46
504d: 155 523a: 106
504d1–e4: 423–24 523a–527a: 142
504d5–e4: 407n6, 423n29 524b: 399
504e6–505b1: 425 524d: 399
505a2–4: 420n26 525a3–6: 426, 512
505a6–10: 416, 514 527c6–e5: 426n35
505b1–2: 416
505e: 390 Hippias Major
506c: 156 282b1–e8: 405
506c–507c: 144 283c3–5: 405
506c5–e4: 422–24, 425n32 284b–d: 576n4
506e1–507a2: 424 285b–e: 44
507a5–c5: 423n29 285c–d: 607n4
507b8–c5: 419, 420n25 286c: 108
507b–c: 157 289a8–289b3: 484
507c–508a: 539 289a–c: 443
507c–509c: 157 291a: 104
507c8–d6: 425n32 304c1–e1: 294
507c9–d1: 426n35 304c–e: 95
507d: 156 304d–e: 431, 589n33
507e–508a: 52 304e3–9: 294
507e6–508a4: 426 Hippias Minor
508a8–b3: 419 365d6–369a2: 406
509a: 380n5 368b–c: 607n4
509c–511a: 52 372b–e: 380n5
509c–e: 580 372c8–376b6: 408
509d7–e2: 423n29 372d3–e1: 408
509e: 9, 154 375d7–376b6: 406
509e2–7: 415n18, 418 376b4–6: 408
512a2–5: 420n26 376b8–c6: 408
513e8–514a3: 577
515a1–d1: 423 Ion
515b–519a: 53 531a–b: 611n12
515b–522b: 597n48 531e–532a: 610
515c–4517a6: 406 535e–536a: 610
516d: 611 542b1–4: 121n5
 index locorum   735

Laches 630b–c: 367–68

181: 96 630c: 360
186e–187c: 232n8 630c1–6: 360, 364
187e–188a: 232n9 630c4–d1: 365
187e–188b: 538 630c8–d: 361, 364
190a6–c2: 162n3 630d: 372
190b–c: 394n33 630d5–6: 360n7
192a–b: 432 630d–631a: 360
192b9–c1: 129n37 630e1–2: 360–61, 364, 372
192c: 579 630e2–3: 364–65
192c2–d11: 8n17 630e–631a: 375
193e: 105 631a3–4: 360n7
194a: 98 631a5: 364–65, 372
194d–198b: 366n15 631b: 372
195e–196a: 48 631b–d: 362
196b–d: 393 631b3–4: 372–73
196c–d: 392 631b3–d6: 361
196e: 7n16 631b6: 362
196e1–197c4: 6n13 631b6–7: 362n11
197a: 392–93 631b7–631d2: 363
198d–199e: 147n13 631b8: 363n12
198e–199a: 48 631b–d: 362, 364–65, 367n19, 373n31, 374,
199d4–e1: 409, 414 376, 533–34, 592n43, 597n47
199d–e: 8n18 631c5–6: 367
199e: 393 631c5–d1: 365
200: 98 631c6: 376
200e: 380n5 631c7: 367n21, 370
201c2: 339n7 631c7–8: 367
631d5–6: 362
Laws 631d6: 363
624a1–5: 372 631d6–632c2: 361
625a6–7: 359 631e4–632a1: 361
625c6–8: 360 631e–632a: 374
626e: 373 632c2–7: 361
627c–628d: 373 632c5–6: 362n11
628a6–10: 360n7 632c–d: 374n33
628c6–7: 360n7 632d–e: 372
628c9–d3: 373 632e: 367
628d8–e5: 360n7 632e6: 360n7
628e: 372 6335–d6: 366n17
629a–631a: 372 633a8–9: 375
629a–b: 368 633b: 372
629e1–630b3: 364 633b–c: 373
630a–631a: 364, 367, 367n19, 371 633c8–d3: 366n17
630a7–b2: 360, 364 633d: 373
630b: 369 633d1–3: 372
630b2–3: 360, 364–65 634a: 366n17
630b6–7: 366 634b7–c4: 372
736   index locorum

Laws (Continued) 687e: 370

635e–636a: 372 687e7–9: 362n11
636c6: 571 688a: 360
637b6–638b9: 372 688a1–b1: 360n7
638c1–4: 360n7 688a5: 360n7
643e6: 364 688b: 374n33
644b–645c: 377n36 688b2–3: 362n11
644b6–8: 373 688b7: 362n11
644b6–c2: 373 688e4: 369
644c–645a: 373 689a5–c3: 369–70
644d: 542 689a–e: 369–70, 374
645b2–3: 373 689b: 370, 600
645b–d: 624 689c2: 369
645d–649d: 622 689d4–e1: 362n11, 370
647c7–649d2: 373 691a7: 362n11
652a–660c: 544 691c–d: 600
653a: 374n33 692a: 368
653a–c: 377n36 693b5–c6: 360n7
653b: 623 693b–c: 368
653b–c: 375n34 693c2–6: 370
653d: 623–24 694c–696a: 368
653e–654a: 624 696a: 375
656c: 624 696b: 367
659d–661d: 374 696b10–11: 366
660a–661a: 533–34 696b2–4: 363
660d–661a: 368 696b2–5: 368
660e–661c: 363 696b6–e1: 368
660e–661e: 592n43 696b–c: 368, 374
661a2: 366 696c: 376
662b–664b: 624 696c5–8: 370
662d–64c: 600 696c8–10: 370, 374, 376
664a: 612, 624 696e1: 369
664c: 623 697b: 364, 372
664e: 624 697b2–c2: 363
664e–665a: 624 705d: 360
665a: 623 705d3–706a4: 360n7
665c–666c: 622 709e–710b: 368
666b: 623 709e8–710a1: 368
666c: 624 710a3–b2: 368
666e–667a: 372n28 710a5: 369
667a: 369 710a5–6: 369–70
667a1–2: 364 710b1: 369
667b: 623 710c: 369n24
667c–669c: 624 713c–d: 600
670d–e: 374n33 713e8–714a2: 362
671b–e: 622, 624 714b8–c1: 360n7
681b: 374 718e–723d: 360
 index locorum   737

719e–723c: 623 860e6: 359

720c–e: 591n42 870a7–c1: 363
722b5–c2: 597 870a–b: 364, 372
727b: 363–64 870b2–6: 363
728c–d: 364 875a–d: 600
728c–e: 363 879b–882c: 63
731b3–4: 374n32 886a: 635
732d–743e: 600 890e–891a: 598
733b–d: 544 893b–894c: 635
739a3–740a2: 596 895a–896a: 624
739b8–c3: 597n47 895a–b: 636
739c1: 597n47 895c–896c: 636
743c: 597n47 896c–897b: 637
753b: 599n51 896d5–e7: 639
760b–761e: 599n51 896e1: 639n11
765d–e: 605 897b–898c: 637
773c–d: 374 898c–899c: 637
77e: 591n42 899b5–8: 635
781e–782a: 291n15 900d1–2: 364n14
785b: 599n51 900d5–e2: 364n14
787d: 624 900d7: 371n25
788d–793e: 623 900d–e: 367–68
790c–d: 624 902a–b: 600
791a: 624 903b–d: 601
791c: 367, 375 906a–b: 601
801d: 624 906b2: 371n25
803c–804c: 542 908c–d: 633
803d: 605 913a6: 362n11
806c5–7: 597n47 935e5: 362n11
808d–e: 624 936b: 368
809e–810b: 623 961a–c: 359
811c–d: 623 962b1–5: 360n7
811c–e: 623 962d6–e9: 360n7
812b–c: 623 963a: 360
812c–817e: 624 963a1–b4: 360n7
812d–e: 623 963a6–9: 366
813d–e: 623 963a8: 362n11, 376
814e–816e: 623 963a8–9: 376
817e–818a: 598 963a–965e: 375
819b–822c: 623 963c12: 362n11
820c4–5: 598 963c–968a: 360
836c: 571 963c–e: 367
838a–841a: 544 963e1–8: 366, 369
838a–841e: 541 963e6–7: 362n11
838e–839b: 571 964b: 368
846d–847a: 597n47 964b6: 362n11
857c2–e5: 597–8 965b–968c: 623
738   index locorum

Laws (Continued) 74d5–e11: 482

965d: 367 75c5–7: 166n20
965d2: 362n11 75c–d: 202n65
966b: 455n21 76d: 432
967e: 455n21 77b3–5: 162n6
968a–c: 359 77b–78b: 152n21
77c1–2: 511
Letter VII [?] 77d7–e4: 418
324a: 48 78c4–d3: 161n1
324b–d: 646n6 78c6–7: 162n6
327b–328d: 64 78d2–e6: 161
341: 103n41 79a3–6: 161
342d8: 500n48 79e7–80a4: 170n49
79e7–80b4: 170n48
Lysis 80: 104, 111
204a: 231n4 80a: 98, 108
204b–c: 96 80a4–7: 170n49
207: 106n57 80b: 95
209c–e: 576n4 80b1–2: 170n49
210a9–b4: 581 80b4: 168
211b: 231n4 80d: 382, 432
212a1–7: 7n16 80d5–8: 11, 162, 168, 169n37
223b: 431 80d6–7: 173n65
223b3: 339n7 80d6–8: 432
223b3–8: 162n3 80d7–8: 171
80e: 433
Menexenus 80e1–2: 169
235e–236c: 53 80e1–5: 11
80e2–5: 168
Meno 80e4–5: 169n37
70a1–4: 161 81: 115
71a1–b8: 498 81a10–e3: 163n8
71a4–b8: 11 81a11: 171n51
71a6: 168 81a–86c: 186, 193
71b: 381–82, 431 81a–b: 156n29
71b1–8: 162 81a–d: 190n21
71b1–c4: 406 81b: 431
71b2–9: 165, 166n22 81b–86b: 613
71b3: 168 81c9–d3: 167
71e1–72a5: 162 81c–d: 382, 434
71e2–5: 161 81d4–6: 170, 178n89
72c7–8: 483 81e3–4: 170n45
72c8: 239 81e4–85d5: 163
72e–77a: 438 81e6–82a3: 178n89
73a7–b1: 161 82b–87c: 55
73d9–10: 161 82e: 169n41
74a4–6: 161 82e4–6: 178n89
74b2–77a2: 162 84a2–b1: 170n49
 index locorum   739

84a3–b1: 175 94b4–6: 406

84b6–7: 170n49 94b4–8: 406
84b9–c1: 170n49 94e: 46
84b9–d1: 170n46 96b6–8: 178
84c–d: 169n41 96d–98d: 588n33
85b–d: 436 97a: 13
85c10–d1: 12 97a9–b3: 166n22
85c2: 12 97a–b: 436
85c6–7: 176n80 97b5: 163n10
85c6–d11: 175 97b9: 163n10
85c7: 163n10 97c4: 163n10
85c9–10: 175n75 97c9: 163n10
85c9–d1: 164, 177n83 97d2: 163n10
85c–d: 219, 397 97d6–98b6: 134n50
85d6–7: 198n51 97e6: 163n10
85d6–86c4: 163n8 97e6–98a8: 163–64
85d–86c: 199n54 97e–98a: 611n13, 613n17
85d8–d1: 176n80 98a: 13, 270, 273, 396, 436
86a7: 163n10 98a4–5: 178n89
86a–b: 436 98a8: 163n10
86b6–9: 436 98b1–5: 177
86b6–c2: 171n51 98b2: 163n10
86c6–100c2: 175 98b7: 163n10
86d3–100c2: 163 98c10: 163n10
86d3–87b5: 163n9 98c1–2: 163n10
86d8–e1: 498 98d1: 163n10
87b3: 498 99a1–2: 163n10
87b5–c9: 176 99a1–5: 177n83
87b6–89c4: 163 99a5: 163n10
87b6–c1: 170n45, 178n89 99c1–d9: 177n83
87c1–96b10: 11 99e4–100a7: 176n78
87c–89a: 577n5, 592n43 99e4–100b: 407n6
87d: 579 99e6: 177n83
87d8–e1: 424 100a: 190n29, 589n33
88a–b: 366, 369n24 100b4–7: 162
89c5–100b4: 163
89d3–96c10: 406n5 Parmenides
89d6–e3: 178n87 126b8: 231
90b3–e8: 406 126c: 231
90e10–91c5: 405 126e5: 232
91a1–6: 161n1 127–129: 93n2
91a1–b5: 406 127–136: 113
91c–92e: 607 127a2: 646n6
91e: 45 127a–b: 101n37
92d–93a: 610 127c: 95
93a5–94e2: 406 127c5: 233
93a–94e: 609 127c–e: 77
93d9–10: 178n88 127e: 233
740   index locorum

Parmenides (Continued) 130e5–131a2: 458n10, 481

127e11–12: 233 131b5–6: 242
127e8: 233 131b–c: 238
127e8–10: 232 131e11: 242
128a7: 233 132a: 241
128b1–2: 248 132a10–b2: 241
128b4–6: 248 132a1–4: 239
128d–e: 78 132a1–b2: 476
128e6–130a2: 240 132a3: 251
128e3–4: 248 132a6–8: 240
128e6: 234 132a–b: 29
128e6–129a1: 236 132b3–4: 242
128e6–129a3: 233, 248 133–134: 691n86, 692n89
128e6–129a6: 238 133a11–b2: 244
1297–130e1: 240 133a8–9: 244
129a: 233 133b–134e: 29
129a1: 237 133b3–4: 243–44
129a1–2: 236–37 133c10–d4: 244
129b: 232, 240 133c8–d5: 481
129b1–2: 236–37 133d7–134a2: 253
129b1–3: 236 133d7–e3: 244
129b6–c1: 236 133d8: 658
129b–e: 233 133e3: 244
129c1: 232 134b9–12: 245
129c2–3: 236 134c: 245
129d: 115 134d9–e7: 245
129d5: 236 134e10–135a3: 245
129d6–e1: 234, 237 135b1–4: 247
129d6–e4: 236 135b5–c1: 235, 248
129d7: 237 135b5–c2: 482n4
129d7–8: 237, 244 135b5–c3: 246
129e1: 237 135c–137b: 232
129e–130a: 232 135c9–d1: 245–46, 253
129e3: 237 135c9–d2: 240
129e5–a2: 236 135d: 95
130a2: 234, 237, 240 135d5: 232
130a7: 235 135e1–3: 247
130b: 237 135e2–4: 247
130b1–e3: 458, 482 135e8–136a2: 247
130b2: 234 135e9: 256
130b2–5: 234 136b3–4: 247
130b7–8: 236 136b6–c4: 247
130c1: 237 136b7–8: 256
130d: 237 136c5–6: 246, 253
130d3–4: 458 136e1–3: 246, 253
130e: 95 136e–137a: 95
130e–131e: 238 136e7: 646n6
 index locorum   741

136e8–137b4: 248 66c6–7: 588n32

137: 104 66d3–e4: 588n32
137b: 106n58 67b: 586n27
137b1: 232 67b9–10: 588n32
137c–165e: 249 67e4: 199
137c2: 646n6 68a7–b3: 588n32
137c4: 248n51 68b: 586n27, 588n33
137d: 250 68b8–c3: 369
139d1: 250 68b9–10: 588n32
142d9–143a3: 257 68b–c: 586, 587n29, 587n30
145b: 250n58 68c–69b: 200
146c: 251 68d–69b: 224
148a–d: 251 69a5–c2: 369
149e8–11: 235n16 69a–d: 541, 588n33
152a–e: 94 69b3–4: 586n28
158b1–c6: 254 69b7–8: 587
159b: 250n58 69c3–6: 587
160b2–3: 248 69d2: 199
166c: 232 69e6–72e1: 14
166c1: 256 70a5: 190
166c1–4: 232 70b3–4: 190
166e5: 248 70c–72d: 185–86
72–78: 690
Phaedo 72e: 115
57a: 183 72e3–78b3: 14
58d–59b: 184 72e–73a: 192
59a: 199 72e–76d: 613
59b: 41, 101 72e–78a: 186
60c–61b: 184 73a–b: 193
60d2: 199 73b–76e: 458n11
60e–61a: 199 73c–76c: 193–94
61b–69e: 184–185 74a9–12: 248n51
61d: 190 74a–c: 58
64a: 541 74b: 115, 586n27
64a4: 199 74b7–c3: 468
64b9: 199 74b8: 474n52
64c: 189 74b–c: 267
64e2: 199 74c2: 468n33
64e4–6: 192 74d4–8: 467
65–67: 16 751b: 658
65c–e: 586n27 752d2: 658
65d13: 456n3 75c10–d5: 461
65d4–e5: 461 75d: 234
65d–66a: 58, 196 75d10–11: 197
66a3: 456n3 75d9–10: 197
66b2: 199 75e2–3: 197
66b–e: 586n27 75e5–6: 198
742   index locorum

Phaedo (Continued) 84a–b: 200

76: 14 84c: 201
76a–c: 197 84c–91e: 187
76b: 32, 445n21 84d: 106
76b5–6: 273n23 84e–85b: 203n66
76d7–9: 234 85c–d: 201
76d–e: 196 85e3–86d3: 514
76e1–2: 198 86b: 14, 588n33
77a–b: 199 87a: 199
77b–e: 516 87a–88b: 14n41
77e: 108n67 88a: 98
78–79: 469 88a–b: 191n25
78a: 200 88c: 110, 201
78b4–84b8: 14 88d: 200
78b–84b: 186–87 89a: 200
78c10–79a5: 465 89d: 202
78c1–5: 516 89d–e: 202
78c7: 186n8 90–91: 106
78c–79e: 58 90a–d: 202
78c–e: 59 90b: 202
78d10–e2: 481 90b–d: 609
78d10–e4: 474 90c: 202
78d3: 658 90d–e: 202
79a6–7: 465 91c: 201, 203
79e: 190 91e–92a: 199
80a: 191, 200 92a–101e: 187–88
80d5–e1: 516 92a7–93a9: 515
80e: 200 92c: 105
80e6: 199 92d: 199
81b4–5: 196 92d–e: 196
81b–83e: 225 94b7–d2: 515
81b–c: 587n30 94b8–d6: 514
81c–d: 191 94c10–e2: 516
81d–e: 191 95d4–102a9: 16
82a11–b1: 369 96: 105
82a12–13: 587n30 96–106: 457, 459
82a–b: 588n33 96a–101e: 234
82a–d: 587 96c–e: 16
82b7: 587n31 97b8–98c2: 342
82b–d: 200 97c: 77
82c3: 199 97c1–2: 376
82d–83e: 586, 587n29 97c–99d: 16
82e: 191 97d–99d2: 273n23
82e–83d: 200 99b–101d: 16
83b6: 200 99c–100a: 201n61
83c: 202n63 99c9–d1: 16
83d: 191 99d–100a: 202
 index locorum   743

100b: 234 253d–255e: 569

100c2–7: 476–77 253d–256e: 541
100c9–d8: 129n36 255e–256a: 573
102–107: 105 256a–c: 566–67
102a: 110 256b–e: 572
102a10–107b10: 14 259b–d: 200n57
102a–b: 105n55 262c: 56
102b–105b: 462 265c8–266c1: 366
102b–105c: 16 265d–266c: 451n34
102b11: 458n10 265e1–2: 345n23
102b1–2: 481 265e1–3: 484
102b3–5: 471 266a: 567
102b3–6: 466, 468 266b3–5: 482n4
102d–103c: 462 266d–269c: 607
102d–106e: 188 270c: 605
103c13: 463 274e–275b: 193n29
104d1–7: 463n28 275d–e: 102
105b: 477n62 276c–277a: 620
105c1–6: 439 277b5–8: 355
107a: 105n55 277b–c: 605
107b: 201 277c: 571
107c–118a: 189 278d–e: 72n14
114d: 202 278e–279b: 56
115: 14
115d: 192 Philebus
117c–d: 203n67 11b1: 338
118a: 42 11b4–c3: 340
11b7: 340n13
Phaedrus 11b9: 340
230d: 101n34 11d4–6: 340
244a–245a: 619 12–22: 106
245a: 610 12d5–e4: 345
245c–257b: 195n36 13a8: 340
245c–e: 624 13b7: 340
246a–250c: 517 13c6–19b8: 337
246a–e: 541 13e4–6: 340, 340n13
246c: 553 14c7–15c3: 341
246d–250e: 613 16a4–5: 338
247c2: 241n31 16b–18d: 451n34
247c3: 470 16b4–18d2: 341
247c–e: 241 16c2–3: 355n54
249b5–c3: 482n4 17a4: 355
249b–c: 690 18b–d: 279
249d–251e: 569 18e8–22c4: 341
251a1: 572 19: 104
252a: 573 19b5: 338
252c–253c: 634 19c4–5: 338
744   index locorum

Philebus (Continued) 51b9–c1: 352

19c4–d6: 340 51c1–2: 353
19c5: 339n7 51c1–d9: 352
19d6–e5: 338–39 51d1: 352
19e–20a: 113 51d7: 352n45
20: 98 51e–52a: 545
20a1–8: 339 51e7–52a1: 353
20b–23a: 532n3 52c2–3: 353n47
20c–21a: 61 53c: 61
20d–21a: 113 53c4–55a3: 343, 354
21e3–4: 338 55b1–c1: 343
22ec5–e3: 342 55c4–59d9: 343
23a: 108 55d5–55c6: 355
23b5–9: 337 56d1–57e5: 355
23b6–9: 342 57e6–59d9: 355
23c9–12: 346 58a7–8: 338
26a4: 346 59d10–64a6: 344
26a8: 347n34 59d4–5: 357n55
26b5–7: 346 60a–61a: 532n3
26e6–7: 354 60a7–b4: 340
27e5–7: 342 60b4: 340n13
28a1–31a3: 342 61a4: 356
28b7–10: 339 61b11: 356
28c–30e: 542 62a2: 658
28c6–7: 342 62b8–9: 344
31a5–6: 342 63b2–4: 356
31b2: 350 63c8: 356
31b2–53c3: 343 64a7–65a5: 344
31c2–32b4: 353 66a4: 344
31c2–d10: 343 66a6–b3: 347n34
33e3: 197n48 66c9: 344
34d10–35d6: 517 67b: 100
36b: 351n43 67b11–13: 339
36c6: 351
37b10–c2: 498 Protagoras
37e5–40d10: 351 309a–310a: 141n1
38c: 107n64 309c–310a: 94
38c5–e8: 503 310a2: 339n7
41a7–50e2: 351 310d–e: 144
41b: 201n60 311a1–2: 338
41c–d: 351n43 311b: 112n72
42c–47d: 543 312a: 98
43b1–c6: 343 313–314: 99
44b–e: 61 313c3–4: 121n5
45d–e: 573 313e–314a: 581n15
47c–d: 351n43 314b: 607
50e5–52c2: 343–44, 352 314b8–e2: 338
 index locorum   745

314d: 141–42 335c7–d5: 330n8

314e–315a: 141n2 336c: 231n4, 235
314e–316b: 141n2 337a–c: 44
315b9: 176n78 338b3: 330n8
315c8: 176n78 338e–348e: 142
315c8–d4: 338 338e–349a: 147
316d–317a: 609 339: 98n17
317c: 141n2 339a: 607, 610
317c7–d7: 406 339d–342a: 44
317d5–e2: 339n8 339e: 78
318d–e: 44 340b–d: 610
318e–319a: 44, 145, 606, 607n4 347b: 607n4
318e4–319a7: 406 348b: 231n4
318e5–319a2: 161 348b–c: 104
319a3–b1: 405 349d: 147
319b3–d7: 406 349e–351a: 147
319d–320b: 609 349e–360e: 609
319d7–320b5: 406 351: 106
320b4–5: 405 351a–e: 147
320c–328d: 142 352: 148
322c–d: 614 352a8–c7: 410
322d–323c: 219 352b–353a: 609
323b–333e: 609 352b3–8: 508
323c5–d7: 177n83 352b3–c2: 409
323d–327c: 609n9 352c: 150
324a–c: 145n8 352c4–6: 507
324d–328c: 145 352e5–353a6: 411
325c–328a: 609 352e6–353a1: 8n21
325e–326a: 147n12 353: 148
326e6–327c3: 178n88 353–358: 105
327c1–d4: 408n8 353a: 105n56
327c–d: 614 353c–354e: 9, 282
328b–c: 144 353c4–354e2: 410
329a–b: 103n41 353c9–d4: 412
329b–d: 146 355a5–d3: 410
329d–333e: 235 355c–e: 148
331c–333c: 609 356a5–7: 412
331c–d: 97n13 356a–c: 149
332a–333b: 94 356c4–357b3: 412
333–334: 98 356d4–e2: 412
333d: 78 356d7–e2: 508
334–335: 106 356d–e: 539
334c–335d: 147 357b5–6: 508
334c–d: 104 357c6–d3: 412
335a: 142n3 357c6–e2: 411
335c: 78, 104 357e: 150
335c3: 339n7 357e2: 510
746   index locorum

Protagoras (Continued) 366e: 212n6

357e4–8: 405 367c: 536
358a: 105n56 367e: 211
358a–b: 148n15 368c–369b: 590n38
358b6–d2: 508 368d–369a: 208
358c1–3: 510, 512 370a–b: 219
358c1–5: 10, 411 374d–376c: 219
358c3–5: 412 374e–376d: 369n24, 374n32
358c–d: 8n17 376c–398b: 208
358d–e: 44 376c–d: 615
360d: 366n15 377a11–b8: 526
361a3–b7: 406 377a12: 288n6
362a1–3: 339n8 377a–383a: 616–17
377a–b: 612, 615
Republic 379a–e: 627
327a: 114 387d–e: 536n7
331c: 443 389b–c: 597n48
331e–332d: 42 390d–391e: 617–18
335b: 78 390d7–391a2: 526
336–337: 98 392d–395c: 618
338c–339a: 607 393d: 660
338d–e: 615 395c: 615
342c: 293 395c–d: 660
343: 98 400a4–c5: 348n36
343a: 100 400d: 611
343c: 19, 207 401b: 618
344a: 207 401c–402a: 623
344d: 78 402c: 616
344e: 535 403c: 615
345b: 99 405a: 592n42
346a: 97 410–411: 612
350c–d: 78 410b: 616
350d: 104 412b–414b: 219
351a5–6: 123n14 412e: 612
353d11–354a2: 420n25 413b–c: 612
354a10: 114n78 414b–415d: 219, 616
354c1: 135 416d–417b: 209n3
357a: 106 419a: 25b67
358: 108n67 419a–421c: 597n47
358a1–3: 425n32 420b: 218
358b: 211–12, 212n6 420b–421c: 25, 217
360e–362c: 536 421b–c: 590
362d–364c: 614 422e–423d: 218
364e: 633, 633n6 423e–424a: 208
365e: 633n6 423e7–424a3: 597n47
366c–d: 19n53 426a2: 495n35
366d–367e: 535 427d: 536n7
 index locorum   747

427d1–7: 425n32 442c9–d2: 520n7

427e–434c: 590n38 442c–d: 540
429b–c: 572 443c–444a: 540
429c–430b: 366n15, 612 443c9–e2: 346n32
429c–435c: 590 443e: 212
429c7–430b9: 413n14 444c1–d1: 426n35
430a–c: 616 444c8–d10: 426n35
430c3: 590n38 444e–445a: 536n7
430e3–4: 346n32 445a–b: 208
431d1: 554n2 449c: 208
433: 100 451c–457c: 209
434d: 208 454a: 107n65
435d: 216, 616 454a–c: 202n65
435d9–436a3: 526 456d: 590n35, 591n42
435d9–436b2: 521–22 461b–c: 572
435e: 22n62 461c–464d: 593
435e–441c: 21, 554n2 462a: 218
436b–441c: 618n21 472b–e: 536n7
436b9–10: 518 473c–d: 209
436d1–2: 518 473c–e: 440
436d4–436e2: 518 473d: 23
437a3–8: 518 473d7–439b1: 511
437b1–c5: 519 474b–480a: 59
437e4: 234 474b–c: 223
438a: 21n59 474c1–3: 591n42
438d7: 520n7 474d–e: 561
439b3–c8: 519–20 475b–484a: 234
439b3–d2: 525 475e: 227
439b8–c1: 518 475e–476e: 462
439c5–d2: 521 475e–480a: 220, 261, 440
439d1–8: 521 475e–502a: 222
439d6: 568 476–480: 106, 108n68
439d6–440a4: 413n14 476a1: 248n51
439d7–8: 523–24 476a–480a: 458n11, 535, 579
439e1–4: 523 476a5: 345n23
439e5–440a7: 522–23 476a–541b: 214n11
440a9–b4: 523 476b: 225, 591n42, 592
441a7–b2: 526 476b–c: 225
441b5: 514 476c: 223
441c1–2: 524 476d: 108
441c2: 520n7 476e: 108
441c4–7: 521 476e–477a: 58
441d11–444a2: 590n38 476e–480a: 591n42
441d–443e: 540 476e7–477a1: 440
442a7–b1: 521 477a2–7: 441
442b10–c5: 590n38 477a–b: 441
442b–c: 366n15 477b11–478a4: 442
748   index locorum

Republic (Continued) 506b–e: 215

477b5–8: 442 506c: 24, 444
477c6–d5: 442 506e3: 300n44
478b5–c2: 493 506e–509d: 444
478d: 443 507b7: 658
478e–479e: 267 508d: 112
478e7–479c5: 484 508d–583b: 208n2
478e7–479d10: 473 509a–b: 444
478e7–479d5: 468 509b: 682n58
479a: 441 509d–511e: 444
479a–b: 195 50e1–5: 527
479a–c: 23n65 510b2–9: 288
479d3–5: 443 510c: 445n21
479e: 225 510c–511d: 445
479e–480a: 443 511b7: 288
484a: 223 511b–c: 107n65
484d: 23 511d: 89n79
485b: 59 514a–517a: 189, 223, 444
485d: 214 514a–517b: 209
487a: 214, 220 514a–540c: 615
487b: 222 515b7–10: 589n34
488a–b: 220 516c4–d7: 592
488c: 220 516e–517a: 593
488c–495c: 220–21 516e–517e: 46
490a–495c: 221 517d–e: 591n42
491a–495b: 64 518b–c: 224
492a: 45 518c–d: 613
492a–c: 608 518d10: 89n78
492c: 221 518d–519a: 613
493a–c: 608 518d9–519a1: 426n35
493e: 225 518d–e: 625
494a: 220 519c–521c: 594
494c–495b: 221 519e: 25
495a–496a: 227 519e–520a: 25, 218
498b8–c4: 425n32 520b7–c4: 593
499a: 202n65 520c: 24, 440n16
500c: 214, 618 520c1–5: 593
501d: 215 520c–d: 591n42
501e6–502a2: 591n42 520e: 215
504a–540c: 222 521c–534e: 288
504a–b: 616 521d: 226
504b: 216 521d–525b: 616
505a: 444 522a: 616
505d11–e4: 418 522a4–9: 589, 591n42
505d5–9: 418 522b2–531d4: 24n69, 349
505e1–2: 528n14 522e: 226
506b–507a: 42n3 523–525: 107n64
 index locorum   749

523a–524c: 225 558c–562a: 210

523a–525b: 234 558d–559c: 213
523b: 462 559b: 213
523b–524d: 621 560e–561b: 213
523c–524d: 614 561c6–d4: 522
523d4–5: 464 562a–569c: 210
523d5–6: 462 571a–572b: 572
524a6–10: 462 571a–580c: 210
524b: 444 571b–572b: 567–68
524c3–d1: 462 571c3–d3: 524
525a–540a: 615, 621–22 571d1–4: 520n7
525c: 226 573b9–d5: 569
526d: 226 573c: 569
527c: 226 574b–c: 568
529c–531c: 55 579e: 213
530b: 226 5807b1–4: 554n2
530d: 52 580a: 210, 213
531c: 226 580a–c: 591
531c1–4: 3478n35 580c: 536n7
531c–d: 444 580c–581b: 618n21
531e: 32, 445n21 580d–588a: 210
532–533: 107n65 580d6–7: 520n7
532a1–2: 288 580e2–581d3: 525–26
532a6–7: 355 581b6–11: 521
533b1–3: 355 581b8: 520n7
533b–c: 445n21 583b–588a: 214n11
534b: 32 584b: 545
534b3–4: 355 584c–585a: 545
534b3–7: 273n23 585b–586d: 543
534b–c: 445 585b–586e: 622
534d8–535a1: 107 586c–587a: 540
537d–539d: 227 586e–587a: 543
538c–539c: 621 588b–589b: 541
538c–539d: 202n62 588b–591b: 540
540b: 215 588b–591d: 536
545a–580a: 554n2 588b–592b: 535
545c–548d: 210 58b–592b: 210
548b: 200n57 590a–591c: 591n42
548d–550c: 210 590c–d: 546
550d–553a: 210 591d1–3: 346n32
553a–555a: 210 595a–600c: 619–20
553d2–4: 524n12 596a10–b10: 488
554d–e: 213n8 596a5–7: 463
554e1: 554n2 596a–597e: 234
555b–558c: 210 596a6–8: 482
557a–558c: 221 596a7–9: 458n10
557a–564a: 47 596b1–2: 463
750   index locorum

Republic (Continued) 240e1: 334n69

597c9: 658 240e–241: 333
597d–598b: 293 243d–244b: 315n19
601d1–e3: 488n20, 489 244c: 104
602a–b: 620 250a8–d3: 315n19
602e–606d: 618–19 250c6–7: 322
603a1–2: 520n7 251–252: 312
604a10–b2: 520 251a5–6: 313
604d4–5: 524n12 251b3: 314
605c–d: 98 251b8–c1: 314
605e: 99 251c–261b: 451
606b1: 618n21 252b9–10: 314
606b7–8: 618n21 252d: 106n59
607a: 619 252d2–11: 317n21
608c: 210 253e: 85n58
611a10–612a: 517 255a4–255b7: 322, 322n33
612e–613b: 542 255a–b: 317n21
613a–b: 536n7 255c12–13: 322n33
617e: 640 255c–d: 321n31
255e–256e: 315–17, 323
Sophist 256a10–b4: 312, 317, 319
216: 93n2, 97 256b: 106n59
217a: 309 256b3–4: 318n22
217c–d: 106n58 256b6–10: 319, 319n28
217c–e: 310n3 256b6–8: 317n21
217e: 80 256e6–7: 322n32
217e1: 339n7 256e6–8: 500n47
218b7–c1: 492 257b: 331n60
219a–232a: 451n34 257b10–c3: 493
219b4–6: 292n19 257b1–c3: 330
231c–e: 310n4 257b6–7: 331n60
232–233b: 231n4 257b6–c2: 326n44
232d: 607 257b–c: 331–32
234b5–d1: 493 258d: 325n38
236b7: 327n50 259d9–264b10: 481
236d5–241b4: 493 260–264: 333–34
236d5–264b10: 493 260a7–8: 494
236d9–264b8: 309 260b10–c4: 325
236e: 114 260b3–4: 494
236e4: 310–11 260b5–261c10: 494
237b7–e7: 334n68 260c3–4: 334, 334n68
237d–264b: 311–12 260d1–2: 334, 334n68
238c: 691n86 261: 100
239a8: 311n10 261a–b: 325
239b4: 311n10 261c: 325
239b9: 311n10 261d1–262e10: 494–96
239d1: 311n10 261d2–4: 496n37
240d9: 333 261d7: 325
 index locorum   751

261e5: 494n30 269d–e: 642n15

262a2: 326n41 273b–d: 642n15
262a3: 494n30 279a–311c: 451n35
262a3–5: 494n31 281: 104n54
262a6: 494n30 284b–e: 62
262a6–8: 494n32 285: 104
262b10–c1: 494n32 287b10–c5: 482
262b2: 325 296a–297b: 597n48
262b5–6: 494n31 299e: 106
262c9: 495n34 306a–11a: 374
262d2–6: 326, 326n42 306a–b: 366, 369n24
262d4: 495n34 307d–308b: 595
262d9: 494n30 307e: 369
262e11–263b13: 499 308e–309a: 595
262e11–263d5: 494 309c: 374
262e13–14: 494n31–32, 497 309c5–7: 595
262e13–263a11: 496 309e–310a: 595
262e9–10: 498 311a: 595
263: 331n60 311b9–c1: 595
263a12–b3: 498 311c5–6: 595
263a6: 334n70
263b: 334 Symposium
263b11–12: 322n32, 499–500n47, 172b: 339n7
500n48 172c1: 339n7
263b4–11: 327 174–175: 96
263b4–5: 329n53 175: 99
263b7: 328, 331 175c–d: 570
263c1–12: 496 177d7–8: 123n16
263c7: 334n70 177e: 96
263d: 332 178d–179a: 559
263d1: 332n60 180d–181d: 567
263d1–4: 327, 327n48–49, 329–30 186a–c: 567
263d6–264b5: 494 189d–193: 559
263e: 107n64 194d: 232n8
263e10–264a3: 503 197b: 554
263e4–5: 503 199e–201c: 533
264b–268d: 451n35 201a4–5: 561
265a: 310n4 202a: 455n21
265b9–10: 292n19 204d: 104
268c: 310n4 204d5: 554n4
204e1–205a3: 425
Statesman 204e–205a: 532
257a: 85n58 204e–209e: 554–59
258b–268d: 451n35 205a: 554n4
262a3–263b12: 482 205a1–3: 555
262a5–263b11: 484 205b: 292n18
262a–e: 30n77 205d2: 554n4
269–274: 114n76 206a11–12: 554, 556, 567
752   index locorum

Symposium (Continued) 152c: 263n5, 264, 270

206b7–8: 555 152c–183c: 282
206c4–5: 556 152d3–6: 498
207d3–4: 300n44 152d–e: 265
208a4: 197n48 153d–160e: 265
208a6–b2: 300n44 156: 100
209b–c: 563, 566 156a–157c: 265
210a–212b: 553 157e–160e: 266
210a–c: 561–62 160c: 608
210c5–6: 561n6 161b9–163a3: 487n14
210d4: 560 166c9–167d5: 487n14
210d5–6: 562 167a6–8: 493
210e–211b: 195 167e–168c: 202n65
210e2–211b5: 484 168a–b: 202n62
210e4–5: 560 169e–171d: 268n12
210e6–211a5: 466 170: 106
211a: 59 172–177: 111
211a–b: 58 172–178: 97n16
211c8–d1: 560 172c: 46
211d: 563 176a: 641
211e–212a: 561 176a–b: 542
212a3–4: 563 178–179: 105
213c–d: 569 180a6–7: 495n35
216: 96 181–183: 31
217a–219e: 569 181c–183c: 263, 266
217b: 96 182d–e: 267
217c1–2: 495n35 184–186: 31, 268, 282, 687n69, 689n78
219e: 101n34 184b–186a: 447
184b–186e: 263
Theaetetus 184d–e: 268n13
143a: 109 185a6–7: 447
143a–b: 77 185b7–c2: 447
144b–c: 79 186a10–b1: 447
145c–151d: 262 186b11–c1: 448n26
146b: 231n4 186b–c: 269, 447–48
146e9: 658 186c7: 447n25
147a: 236 186d2–5: 448
147c: 236 187a–b: 449
148e–151d: 262n4 187b: 263
149a4: 123n16 187b–c: 269
150d4: 339n7 187c7–200c7: 493
151–171: 105 187d2–5: 447
151a3: 339n7 188a–c: 270
151e: 262, 264, 607 188c–189b: 270
151e1–3: 447 188c5–8: 272n20
152a: 264, 608 188c9–189b9: 493
152b2–9: 487 189–190: 107n64
 index locorum   753

189b: 332 27c5–30c1: 290

189b–191a: 270 27d–28a: 439n16
189e–190a: 107 27d5–28a4: 469
189e6–190a6: 503 27d5–28b2: 290
190e5–196d2: 497n41 28–29: 687n69
191a–196c: 270–71 28a3: 290n13
195b–196c: 272 28a6–b2: 488
196a4–7: 503 28b1: 291
197a–20d: 270 28b2–29b2: 290
197b: 197n48 28b2–c1: 290
200b–c: 113 28b6: 291n15
200d–201c: 263 28b6–7: 289n7
201a–c: 449 28c: 629
201a–d: 272–73 28c1–2: 291
201b: 443 28c2: 292n20
201c–d: 263, 273 28c2–29b2: 488
201d–203c: 263 28c2–5: 291
201d–203d: 273 28c5–29b1: 293
201e–202b: 274 29a1–3: 291n15
202c: 455n21 29b1–c3: 294
202e: 275 29b2–3: 288
203b: 275 29b9–c1: 295
203c–205e: 282 29b–c: 289
203d–206b: 275 29b–d: 632
203e–205e: 263 29d2: 439n16
206a–c: 263, 282 29d5: 296
206b–e: 450 29d7–30a2: 296–97
206c–207b: 278 29e1–2: 301n48
206c–208b: 278 29e–30a: 627
206c–210a: 263, 278 29e–31b: 630
207b–208b: 278 29e4: 289n7, 295n29
208c–210a: 450 30a6–b6: 524n12
208c–210c: 279–80 30a6–c1: 299
208d–210a: 272 30c–31b: 294
210b–d: 262 32c1: 292n20
210d: 47, 95 33a6–b1: 524n12
33b: 629
Timaeus 34a7: 292n20
17a: 114n77 34b: 630
17a5: 339n7 34b9: 292n20
17b–19a: 85n57 35a–37c: 690n82
20: 93n2 378a: 628
21a: 114 37a1: 299
26b–c: 633 37a2: 292n20
26e: 114n78 37a2–c5: 503
27a5–6: 288 37a–c: 690
27a–b: 85n57 37b4–8: 440n16
754   index locorum

Timaeus (Continued) 51b7–d3: 463n28

37c–38b: 299 51d–52a: 439n16
37c–38c: 292n24 51e: 455n21
380a–b: 628 51e5: 273n23
38b6: 292n20 51e6–52b5: 469n40
38b6–9: 300n45 52a4–7: 481
38c4: 292n20 52a9–b1: 302
38e6: 292n20 52e4: 303
39e–42e: 640 52e–53a: 303
39e7–9: 299n41 53a7: 302
40b–c: 633 53b: 642
40c–41a: 631 53b1–6: 302
40d–41a: 631–32 53b2: 302
41a–42d: 690n82 53b4–5: 303
41a5: 292n20 53c5–6: 304
41a7–b6: 300n45 53c–68d: 304
41a–d: 634 53c–d: 303
41c–42: 689n79 53d: 296n33
41c–d: 292 53d4–7: 293n25, 296
42d: 292 54b4–5: 304
43a–44c: 692n89 54b–c: 303
46a2–c6: 301n50 55b4: 292n20
46b1: 301n50 55e: 303
46c–e: 641n13 55e4: 296n32
46d1: 297n34 56c3–7: 304
46d–e: 301 57c: 303
46e6: 297n34 59c–d: 289
47e2–48e1: 301 61b6: 297n34
47e3–53c3: 290 61d3: 296n32
47e–48a: 641n13 63e8: 297n34
47e–68d: 303 68e4: 292n20
48a: 304, 641–42 68e–69b: 304
48a2: 299n41 69–72: 689n79
48a6–7: 301 69c4: 292n20
48a7–b3: 301n49 69c5–d4: 517
48b3–5: 302 70e–71e: 599n52
48b7–c6: 295 75a–b: 303, 641, 641n13
48c2–d4: 289n8 76c6: 297n34
48d2–3: 295 76d6: 297n34
48e6: 296n32 77b3–6: 524
49a6–50a4: 302n54 77b–c: 599n52
49b–50b: 302 80a1: 297n34
49b8–c1: 303 87e5: 297n34
49d4: 303 90a–d: 542
50b1–2: 302 91d: 292n23
50c7–d2: 469n30 91d1–5: 292n21–22
50d–51b: 302 92c8–9: 300n45
 index locorum   755

Plotinus 4.6: 687n72

4.7: 687n72
Enneads 4.8.8: 691
1.2.4: 690n83 5.3.2: 690n83
1.2: 671n10 5.3.4: 691
1.4: 682n55 5.3: 691
2.9: 682n55 5.5.1–3: 691
3.5.1: 690n83 5.5: 687n71
3.5.2: 679n42 5.9: 682n55
4.3–6: 689n81 6.1–3: 682, 682n55
 Index Nominum

Ackrill, J. 195n32–33, 196n43, 216n16, 312–14, Barton, J. 43n8

312n12, 321, 324, 324n36, 438n14, 487n15 Bastianini, G. 679n43
Adalier, G. 262n3, 270n17, 272n20–21, Bedu-Addo, J. 195n32
275n28, 277n34 Benson, H. 4–5, 5n10, 7, 104n52, 122n12,
Adam, J. 348n36 130n41, 132n44, 135n51, 167n24, 169n40,
Ademollo, F. 492n26 170n49, 173n61, 176n80, 177n82, 188n14,
Algra, K. 456n1 383n8, 386, 386n14, 388, 388n17, 391–92,
Allen, J. 674n22 391n31, 392n32, 394, 431n5, 432n6, 577n6,
Allen, R. 95n9, 129n36, 234n15, 398–99, 578n7–8, 581n15
399n42 Berges, S. 576n2
Alline, H. 72n14, 81n34, 82n35 Berman, S. 153n22
Anagnostopoulos, M. 527n13 Betegh, G. 185n6, 189n16, 190n22
Annas, J. 2, 9–10, 10n28, 11n30, 16n30, 18n50, Beversluis, J. 100n24, 387n15
26n70, 34n90, 60n45, 72n15, 83n46, 97n15, Blank, D. 44n11
106n63, 165n18, 189n16, 207n1, 217n18–19, Bloch, E. 96n10
279n36, 420n24, 457n7, 458n11, 462n24–25, Blondell, R. 95n8, 101n33
463n26, 524n11, 532n3, 535n4, 536n7, 542n14, Bluck, R. 87n67
576n2, 598n49, 620n26, 628n2, 648n12, Blumenthal, H. 673n17
655n21, 656n23, 671n11, 676n31 Bobonich, C. 2, 21n58, 27–28, 353n49, 359n2,
Apelt, O. 500n48 360n5–6, 361n8, 362n10, 363, 363n13,
Apolloni, D. 186n9, 241n28 366n16–17, 367n18, 369n23, 375n34, 419n23,
Armstrong, D. 472, 472n48 518n5, 524n11, 540n10, 576n3, 582n19,
Armstrong, J. 542n14 584n23, 586n26, 587n29, 587n31, 590n36–37,
Athanassiadi, P. 675n28 591n41, 592n43, 593n44, 595n46, 598n49–50,
599n51–52, 623n32, 624n33
Bae, E. 245n48 Bolton, R. 298n40
Bailey, D. 188n14, 239n24, 241n30, 241n34, Bonitz, H. 656n23
255n86 Bordt, M. 627n1
Baltes, M. 291n14–15, 300n43, 669n3 Borsche, T. 488n20
Baltzly, D. 684n63 Bostock, D. 32n87, 136n55, 164n13, 167n26,
Barker, A. 345n27, 348n35, 349n39, 355, 355n53 172, 172n60, 186n7, 186n9, 188n15, 195n32–33,
Barker, E. 576n2 195n37, 196n42–43, 272n20, 277n33, 313n13,
Barnes, J. 78n29, 83n43, 84n52, 95n7, 122n10, 319n27, 320n29, 436n10, 439n14–15, 448n28,
164n11, 165n17, 167n27, 168n31, 186n7, 449n31, 450n33, 456n1, 458n10, 482n5, 484n9
273n22, 445n20, 449n32, 674n22, 674n25, Boter, G. 86n61, 87n66–67, 88n71
676n31, 683n59 Bouchet, C. 54n34
Barney, R. 44n9, 405n3 Boys-Stones, G. 674n24, 678n40–42, 687n71,
Bartels, M. 359n1, 364n14, 373n29 687n73, 688n76
 index nominum   757

Bradley, F. 324, 324n36 Cohen, S. 6n12, 30n80, 239n23, 240n26,

Brandwood, L. 72n16, 76n22 242n35–37, 243, 243n41, 256n89, 257n91
Brennan, T. 672n14 Cooper, J. 26n70, 31n85, 69n5, 97n5, 98n19,
Brickhouse, T. 104n52, 122n10, 124n20, 100n27, 120n3, 156n29–30, 215n12, 268n13,
128n32, 131n42, 136n57, 386n13, 387n15, 269n15, 341n15, 379n1, 426n34, 447n24,
389n23, 409n11, 413n13, 419n23, 420n26, 461n20, 507n1, 513n2, 524n10, 524n10,
576n2, 582n17, 582n19 576n2, 595n46, 597n48, 606n3
Brisson, L. 229, 360n5, 362n10, 599n51, Cornford, F. 14n42, 85n57, 94n5, 262n3,
674n25, 681n52 272n20, 291n15, 294n27, 313n15, 319,
Brittain, C. 36, 86n62, 456n6, 671n10, 672n14, 319n28, 659n28
674n22, 676n31, 677n35, 679n43, 680n47, Correia, F. 126n25
687n70–71 Coxon, A. 59n44
Broadie, S. 277n34, 289n10, 300n47, Crivelli, P. 2, 17, 34, 311n9, 315n18, 315n20, 317,
475n53, 628n3 317n21, 318n25, 320–23, 322n33–34, 323n35,
Brown, E. 215n14, 594n45 325n40, 329n52, 330n56, 331n60, 332n60–61,
Brown, L. 2, 25n68, 33–34, 33n88, 218n21, 333n65, 458n10, 493n29, 500n50, 501n54,
310n5, 311n9, 321n31, 325n38, 438n13, 501n53 502n57, 502n59, 503n62
Brown, P. 676n33 Crombie, I. 162n5, 166n20, 276n31
Brumbaugh, R. 86n61 Crooks, J. 189n17
Brunschwig, J. 232n9 Cross, R. 207n1, 217n17
Brunt, P. 62n52 Crystal, I. 59n44, 691n85
Bryan, J. 294n28 Curd, P. 59n44
Burkert, W. 51n24–26, 60n45
Burnet, J. 69n1, 88n74, 337n2, 606n3, Dancy, R. 56n39, 193n30, 195n32, 196n44,
646, 646n3 265n8, 345n27
Burnyeat, M. 24n66, 27n73, 31n83, 31n85, Day, J. 176n79, 177n83
32n86–87, 70n7–8, 99n21, 100n27, 101n32, Denyer, N. 145n9, 188n15, 314n17, 324, 324n37,
103n44, 106n62, 112n72, 164n11–12, 165n14, 326n44, 472n47, 486n12, 495n33
165n17–8, 166n19, 170n50, 226n29, 243n43, Derrida, J. 103n42
262n4, 263n6, 265n9, 268n13, 272n20, Des Places, E. 361n9, 677n36, 679n42,
273n22, 273n24, 274n25–26, 275n27, 276n29, 681n50–51, 685n65, 687n72, 688n76
277n34, 279n35–36, 280n37, 288n4–5, Destrée, P. 552n1, 620n27
295n30–31, 326n46, 327, 337n1, 349n39, Deuse, W. 672n14, 673n17
444n19, 447n24, 449n32, 463n26, 493n28, Devereux, D. 2, 3n6, 8–10, 8n20, 124n20,
517n4, 520n8, 610n10, 621n30, 634n8, 635n9 142n4, 178n89, 409n11, 413n13, 414n16,
Butler, T. 586n27 471n43, 532n3
Diès, A. 347n33, 350n40, 352n45, 361n9
Carlini, A. 86n60, 87n67 Dillon, J. 55n35, 56n39, 78n28, 82n37, 83n46,
Carone, G. 156n30, 527n13, 634n7, 639n12 89n76, 612n16, 671n10, 673n21, 674n23,
Castagnoli, L. 2, 4n7, 12n35, 14, 14n40 675n29, 677n36, 681n53, 683n61–62,
Cavini, W. 647n11 684n63
Chappell, T. 262n3, 274n25, 277n34 Dimas, P. 166n24, 169n38, 172n56, 173n61,
Charles, D. 162n5, 174n69, 432n6 176n80, 195n32
Cherniss, H. 55n35, 61n46, 74n18, 460n18, Dodds, E. 46n16, 49n21, 53n29, 86n64, 87n67,
638n10, 639n12, 646n3 88n74, 676n33, 683n62, 684n63, 692n89
Clay, D. 99n22, 100n29 Donini, P. 377n37, 680n46
Code, A. 290n13, 469n41 Donnellan, K. 491n24
758   index nominum

Dorion, L.-A. 143n6, 155n28, 404n2 477n61, 483n7, 484n8, 484n10, 485n11,
Dörrie, H. 82n35, 669n3 577n6, 586n27, 613n17, 622n31, 646n8,
Dorter, K. 183n2 651n18, 657n24
Dover, K. 71n10 Finnamore, J. 683n62
Doyle, J. 121n5, 144n7 Fletcher, E. 341n15, 343n18, 344n22, 346n28,
Duke, E. 69n1, 311n7 349n37, 351n41, 353n49
Duncombe, M. 245n48, 503n61 Folch, M. 598n50
Dunn, F. 43n8 Ford, A. 44n12
Düring, I. 83n45, 645n2 Forster, M. 123n15, 178n89
Frances, B. 252n69
Ebbesen, S. 683n59 Franklin, L. 195n32, 613n17
Ebert, T. 200n58 Frede, D. 188n15, 339n6, 339n9, 343n18,
Ebrey, D. 183n2 346n28, 347n34, 349n38, 351n41–42,
Edmonds, R. 185n6 352n45–46, 439n16, 672n13
Edwards, M. 675n28 Frede, M. 31n85, 70n7–8, 100n30, 231n3,
Emilsson, E. 674n25, 689n81, 691n84–85 269n14, 290n13, 297n34, 310n3, 311n7,
Enoch, D. 581n15 312n11, 315n18, 317, 320–23, 321n30–31,
Eucken, C. 53n31 322n32–33, 325n40, 326, 326n42, 326n45–46,
Evans, M. 126n25, 127n29, 128n31, 327n47, 327n50, 328n51, 334n70, 345n23,
343n18, 354n51 457n9, 469, 469n41, 478, 496n38, 498n43,
499n45, 500n49, 501n51, 524n10, 649n15,
Ferejohn, M. 224n27, 330n56, 414n16, 676n32, 678n41, 680n47, 681n52, 682n56,
501n53 686n67
Ferrari, F. 229, 300n43 Frege, G. 312n11, 313, 318–19, 497
Ferrari, G. 22n62, 101n34, 103n42, 228, Furley, D. 45n14, 436n10
541n12, 552n1 Furley, W. 47n18
Festugière, A. 71n13, 684n63
Field, G. 41n1 Gadamer, H. 100n29
Fine, G. 3n4, 5n9, 12n33, 14n39, 16n46, 17n47, Gallop, D. 94n4, 105n55, 183n2, 186n7, 186n9,
23n63, 31n83, 32n86–87, 35n92, 60n45, 94n4, 188n15, 190n18, 195n32, 195n36, 196n41,
121n8, 122n10–11, 129n38, 162n5, 164n11–13, 199n55, 201n61, 202n63, 456n1
165n15, 166–68, 166n19–23, 167n26–27, Gauthier, D. 217n19
167n29–30, 168n31–32, 168n34–35, 169n38, Geach, P. 130–31, 131n42, 165, 165n15–16,
169n40, 169n43, 170n49, 171n53, 172n57, 173, 239n23, 241n32, 242n36, 387, 388n16,
173n64, 173n67, 174n68–69, 175n73, 175n75, 472n47
176n81, 177n83, 178n85, 193n30, 195n32, Gentzler, J. 173n62, 175n75, 176n80
196n44, 197, 197n46–47, 198n50, 198n53, Gergel, T. 612n16
201n59, 223n24, 234n14–15, 237n20, 240n26, Gerson, L. 456n1, 646n3, 674n25, 682n58,
241n33, 242n36, 243n38–39, 243n41, 252n68, 683n59, 685n65, 689n81, 691n84
254n78, 261n1, 265n9, 270n17, 271n18, Gettier, E. 164n13, 437n12
272n20, 273n24, 276n30, 277n33–34, Giannopoulou, Z. 552n1
279n35, 280n38, 288n3, 296n33, 298, Gifford, M. 48n19
298n40, 382n6, 397–98, 397n37, 398n41, Gill, C. 79n33, 96n10, 100n24, 100n29, 456n1
430n2, 436n9–10, 437–38n12, 438n13, Gill, M. 30n79, 85n58, 231n2, 233n10, 234n15,
439n15, 441n17, 446n23, 450n33, 458n11, 236n18, 238n21, 239n24, 242n37, 243n43,
458n13, 463n27, 468n36, 471, 471n44–44, 248, 248n50–51, 249n54, 250n57–58, 250n61,
472n47, 473n49, 474n50, 476–77, 476n59, 251n63–64, 254n74, 255n86, 256n90, 330n56
 index nominum   759

Giovanelli, P. 54n34 Huffman, C. 49n20–21, 51n26, 52n27,

Glucker, J. 674n27 187n11
Gomez-Lobo, A. 426n34 Hunter, R. 84n53, 101n38, 552n1
Gonzalez, F. 261n1 Hutchinson, D. 69n5, 120n3, 606n3
Gosling, J. 11n30, 195n38, 313n13, 314n16,
318n24, 346n30, 347n34, 473n49 Irwin, T. 1n2, 3, 6n12, 8n22, 9n24, 10n28–29,
Gottschalk, H. 679n45 18n50, 35n91, 58n42, 70n7, 73n17, 76n23,
Gould, T. 619n22 123n15, 144n6, 155n28, 162n7, 164n13,
Griffith, M. 43n7 166n22, 170n49, 173n64, 176n80, 184n4,
Griffith, T. 53n30, 141, 143n5, 155n28, 193n28, 195n37, 207n1, 234n12, 282n41,
361n9, 362n10 299n41, 341n15, 359n1, 366n17, 367n19,
Griswold, C. 552n1 371n27, 385, 385n9, 390–91, 390n25–27,
Grote, G. 19n55, 44, 75n20, 82n35, 646n2, 391n28–30, 393, 398n40, 403n1, 404n2,
657, 657n25 408n9, 410n12, 419n23, 420n24, 423n29,
Grube, G. 121n5, 209n4, 621n29 426n34, 430n4, 459n16, 466n30, 467n32,
Gulley, N. 166n19, 172n56, 173n62, 193n28, 473n49, 474n51, 484n10, 485n11, 520n6,
196n43, 198n52–53, 385 528n14, 532n3, 535n4, 536n7, 552n1, 575n1,
Guthrie, W. 43n7, 44n9, 44n11, 45, 45n14–15, 576n2–3, 591n41, 607n6, 608n8, 615n18,
49n21, 52n28, 53n29–30, 55n35–36, 56n39, 671n8, 672n15
59n43, 60n45, 61n46, 71n11 Isnardi Parente, M. 56n39

Hackforth, R. 72n14, 183n2, 299n42, 347n34, Jachmann, G. 83n43

351n41, 352–53, 357n55 Jaeger, W. 645n1, 652n19
Hadot, I. 675n30 Janaway, C. 605n1
Hadot, P. 537n8 Jirsa, J. 634n7
Halliwell, S. 95n8, 98n18, 99n21, 109n69, Johansen, T. 18, 191n23, 290n11, 295n31,
292n17, 620n27 298n40, 301n51–52, 302n55, 405n3,
Halperin, D. 564n7 628n3–4
Hare, R. 122n10 Johnson, C. 576n2, 580n12
Harold, J. 619n23 Johnson, R. 54n34
Harte, V. 2, 14–16, 23n65, 135n53, 137, Jouanna, J. 43n5
137n58–59, 195n37, 224n27, 276n29, 302n56, Judson, L. 2, 7n16, 11, 13, 13n37–38, 14n39,
345n24, 375n35, 456n1, 586n26, 587n29, 126n23, 127n29, 128n31, 163n9, 174n72,
589n34 175n76, 176n79, 179n91
Hatzistavrou, A. 581n15
Havelock, E. 224n26 Kagan, D. 48n19
Heckel, M. 621n29 Kahn, C. 41n1, 51n24, 53n30, 57n40, 58n41,
Heidel, W. 51n25 72n16, 76n24, 77n26, 97n15, 101n36, 143n5,
Hermann, F. 234n13, 300n47, 143–44n6, 193n28, 231n1, 236n18, 237n19,
620n27 242n37, 313n14, 338n3, 359n1, 380, 404n2,
Hestir, B. 329n52, 500n49 457n8, 459n17, 460n19, 590n38, 597n47,
Hicken, W. 311n7 613n17
Hobbs, A. 522n9 Kamen, D. 189n17
Hoekstra, M. 326n41 Kamtekar, R. 2, 27, 189n16, 209n5, 418n22,
Hook, L. van 53n31 540n11, 591n41, 598n50, 615n19, 616n20,
Horky, P. 52n27 624n33
Hornblower, S. 48n19, 64n56, 77n25 Kanayama, Y. 188n14
760   index nominum

Karamanolis, G. 683n59 Makin, S. 476n56

Karasmanis, V. 162n5, 163n9 Malcolm, J. 241n28
Karfik, F. 300n42, 627n1 Mansfeld, J. 85n56
Kelsey, S. 196n43, 475n53, 587n29 Mansion, S. 649n13
Kennedy, G. 53n31 Manuwald, B. 147n11
Kerferd, G. 44n9, 53n29 Marquez, X. 599n51
Keuls, E. 327n48, 328, 329n52, 499n45, Martin, M. 172n55
500n50, 620n25 Martinelli Tempesta, S. 87n67
Kidd, I. 84n51 Mason, A. 301n52, 641n13
Kingsley, P. 189n16 Mates, B. 313n14
Kirk, G. 50n23, 51n24, 51n26, 57n40, Matthews, G. 2, 5, 5n11, 7, 11n31, 12, 123n14,
58n41, 59n44 130n41, 131n43, 386n13, 577n6, 586n26,
Klosko, G. 599n51 611n15
Knorr, W. 51n25, 55n37 Mayhew, R. 359n3, 634n7
Konstan, D. 300n46 McCabe, M. 4, 42n2, 79n33, 96n12, 97n13,
Krämer, H. 103n43 102n40, 107n64, 168n34, 169n38,
Kraut, R. 2, 12n33, 20, 22n61, 26n71, 97n15, 173n63, 173n65, 456n1, 458n12.
120n2, 121n7, 133n45, 136n56–57, 213n8, See also Mackenzie, M.
215n13–14, 218n20, 389n23, 406n4, 576n2, McCoy, J. 193n28
578n9, 585n24, 587n30 McDowell, J. 172n55, 262n2, 272n20, 328n51,
Kretzmann, N. 488n16, 488n19 333–34, 333n63, 333n66–67, 334n68–70,
Kripke, S. 395, 395n34 449n30
McGinn, M. 172n55
Laks, A. 360n5, 362n10, 576n2, 598n49 McPartland, K. 577n6, 578n8
Lamberton, R. 673n18 McPherran, M. 128n32, 132n44, 228
Lane, M. 134n48, 576n2 McTighe, K. 418n21
Lang, P. 344n21 Meadows, K. 362n10
Lautner, P. 692n88 Meinwald, C. 30n81, 34–35, 241n33, 243n40,
Lavecchia, S. 628n2 245n47, 247n49, 248n50–51, 250–55,
Lee, M. 2, 31–32, 31n83, 44n9, 264n7, 265n9 250n57–59, 251n62–64, 251n66–67,
Lennox, J. 297n34, 348n36, 641n14 252n68–69, 253n71–72, 254n74–77, 255n83,
Lesher, J. 552n1 255n86, 256n90, 338n3, 345n24, 345n27,
Lewis, F. 272n20, 313n14, 318n25, 319n27 346n29, 346n31, 476n57
Livingstone, N. 53n31 Mekler, S. 82n39
Lloyd, G. 43n5–7, 163n9 Menn, S. 163n9, 276n31, 300n42, 456n1,
Lockwood, M. 313n14, 318n26 606n2, 629n5, 682n58, 691n86
Long, A.A. 456n4 Metzger, B. 77n25
Long, A.G. 93n3, 142n3, 157n32, 183n2 Meyer, S. 2, 28
Lorenz, H. 2, 8n21, 14, 21, 21n58, 191n26, Meyer, S. 363n13, 373n29–30
299n41, 540n10, 554n2, 590n37, 599n52, Migliori, M. 232n5
624n33 Millar, A. 178n85
Ludwig, W. 69n5 Miller, D. 302n53
Lynch, J. 50n22, 53n31, 54n32, 62n48 Miller, M. 251n63, 345n24, 596n47
Lynch, T. 229 Miller, T. 202n64
Mirhady, D. 53n31
Mabbott, J. 332 Moravcsik, J. 167n28–29, 168n35, 169,
Mackenzie, M. 154n24 169n42, 172, 172n56, 172n59, 173n65,
Madison, L. 189n17 314n16, 347n33
 index nominum   761

Moreschini, C. 232n5 Pasquali, G. 86n61, 89n75

Morgan, K. 106n63 Patillon, M. 681n52
Morgan, M. 456n1 Patzig, G. 649n15
Morrow, G. 62–63, 62n50, 63n53, 71n12, Peacock, H. 243n43, 249n55, 255, 255n84
72n14, 301n51, 359n4, 360n5, 684n63 Peipers, D. 501n51
Moss, J. 150n17, 154n27, 514n3, 582n17, Pellegrin, P. 648n11
619n24 Pelletier, F. 252n69–70, 253n71, 256, 257n91,
Most, G. 189n17 493n29
Motte, A. 459n14 Penner, T. 21n57, 129n39, 134n13, 147n11,
Mueller, I. 297n34, 662n30, 691n88 239n23, 241n29, 255n88, 256n89, 398n40,
Muniz, F. 345n27 409n11, 413n13–14, 414n16, 432n6, 459n16,
552n1, 610n12
Nails, D. 4n8, 41n1, 70n6, 101n35, 184n3, Peponi, A.-E. 598n50
232n7, 339n5, 552n1, 564n7, 646n4 Perin, C. 164n11, 178n85
Natorp, P. 456n5 Perl, E. 300n43
Nehamas, A. 96n12, 97n14, 98n18, Peterson, S. 17, 28–29, 30n81, 232n6, 234n13,
99n21, 162n5, 162n7, 165n18, 166n19, 239n23, 241n29, 241n32, 242n36, 245n48,
167n25, 167n30, 169n36, 169n40, 248n50, 251n63, 251n65, 253n73, 254n77,
169n42, 170n49, 175n74–75, 176n80, 255n87, 457n8, 476n57, 476n59
279n36, 383n8, 467n31–32, 561n6, Pitkin, H. 582n16
620n25, 621n28 Polansky, R. 389n23
Nettleship, R. 207n1 Popper, K. 25, 216–18, 216n16, 217n17,
Neu, J. 22n61, 217n19 217n19, 517n4
Nicoll, W. 311n7 Pradeau, J.-F. 360n5, 362n10
Nightingale, A. 101n33, 224n26, 598n49 Prauscello, L. 598n50
Nisbett, R. 582n19 Press, G. 99n22
Norlin, G. 53n31 Price, A. 582n19
Notomi, N. 229, 309n1, 310n6, 311 Price, H. 332
Nuchelmans, G. 496n36 Price, S. 47n18
Nussbaum, M. 95n8, 552n1 Prichard, H. 19n52
Priest, G. 231n1
O’Brien, M. 43n7, 147n11 Prior, W. 129n36, 130n41, 660n29
O’Meara, D. 674n25, 681n50, Putnam, H. 395, 395n35
687n71, 691n84
Obdrzalek, S. 554n3 Raven, J. 50n23, 51n24, 51n26,
Ober, J. 47n17 57n40, 59n44
Opsomer, J. 672n14, 681n50 Reale, G. 103n43
Ostwald, M. 50n22, 53n31, 54n32, 62n48 Reeve, C. 209n4, 419n23, 487n15, 490n21,
Owen, G. 59n44, 74n18, 97n15, 249n53, 311n9, 621n29
318, 318n26, 334n68, 456n1, 457n8, 460n18, Reshotko, N. 419n23
501n51, 502n60, 646n2, 655n22, 665n32, Reynolds, L. 77n25, 86n63, 86n65
672n13 Rhodes, P. 47n18, 48n19, 64n58
Owens, J. 655n22 Richardson, H. 149n16
Rickless, S. 231n2, 233n10, 234n15, 237n20,
Palmer, J. 59n44, 456n2 239n24, 240n25, 241n33–34, 243n42–43,
Pangle, T. 576n2 248, 250n57–58, 251n63, 252n68–69, 254n77,
Papillon, T. 53n31 255–57, 255n82
Parfit, D. 217n19 Riginos, A. 55n37, 72n14
762   index nominum

Robinson, D. 89n77, 311n7, 499n45–46 167n26, 169n38, 178n85, 193n28, 195n32, 196,
Robinson, R. 188n14, 276n31, 491n23 196n42, 196n45, 216n15, 225n28, 229, 382n6,
Roochnik, D. 95n8 438n13, 438–39n14, 578n7, 587n29, 599n52,
Rorty, R. 104n50 613n17, 619n22, 621n30, 686n68, 688n73
Rosen, G. 126n25 Sedley, D. 2, 6, 28, 57n40, 59n44, 72n14,
Ross, D. 60n45, 467n32, 482n2, 501n51, 89n79, 97n15, 98n20, 100n28, 147n11,
647n9, 652n19, 655n22, 656n23, 659n28 154n24, 183n2, 184n3, 187n11, 188n12–13,
Ross, L. 582n19 188n15, 189n16, 190n20, 194n31, 195n34–35,
Rowe, C. 72n15, 89n77, 97n15, 156n30, 183n2, 195n37, 201n61, 234n12, 241n33, 254n81,
186n8, 241n31, 289n10, 552n1, 576n2 262n4, 268n13, 270n16, 275n28, 277n34,
Rudebusch, G. 345n27, 496n36 280n38, 291n15, 292n24, 298, 298n39,
Runciman, W. 659n28 305n57, 326n44, 327n48, 333n64, 359n3,
Runia, D. 305n57 457n9, 468n34, 468n37, 476n56–57, 483n6,
Russell, B. 255–56, 274, 498n42 486n13, 487n15, 488n18–19, 490n22, 542n14,
Russell, D. 144n6, 150n18, 545n16 594n45, 628n2–3, 641n14, 669n1, 672n12,
Rutherford, R. 101n33 676n31, 677n34, 679n43, 679n43, 679n45,
Rutten, C. 459n14 687–88n73, 688n74
Ryan, P. 231n2 Segvic, H. 162n7
Ryle, G. 79n32, 82n38, 249n52, 332, 332n62, Sellars, W. 662–63, 662n31
457n8, 517n4 Shackleton Bailey, D. 82n39
Sharma, R. 483n6
Sachs, D. 19n55, 212n7 Sharples, R. 166n23, 167n25, 169n42, 177n83
Samaras, T. 576n2, 598n49, 599n51 Shaw, J. 150n18
Santana, A. 578n7–8 Sheffield, F. 552n1, 566n8
Santas, G. 9n26, 105n56, 162n7, 229, 409n11 Shields, C. 15, 17n48, 35–36, 647n10
Saunders, T. 62–63, 62n51, 63n54, 576n2 Shorey, P. 75n19, 254n79
Sayre, K. 93n1, 234n11, 241n34, 247n49, Silverman, A. 439n14, 446n23, 473n49
249n56, 251n63, 254n77, 275n28, 349n39, Singpurwalla, R. 409n11, 413n13
501n52 Slings, S. 69n1, 87n69, 88n72, 89n78–79
Schaffer, J. 126n25 Smith, A. 674n25, 679n42, 681n52, 682n55,
Scheppers, F. 326n41 683n60, 683n62, 687n72
Schibli, H. 684n64 Smith, J. 463, 464n29, 483n6
Schiefsky, M. 43n8 Smith, N. 104n52, 122n10, 124n20, 128n32,
Schnieder, B. 126n25 131n42, 132n44, 136n57, 386n13, 387n15,
Schofield, M. 3, 50n23, 51n24, 51n26, 53n30, 389n23, 409n11, 413n13, 419n23, 420n26,
55n38, 57n40, 59n44, 60n45, 61n47, 62n49, 576n2, 582n17, 582n19
62n52, 64n59, 89n79, 96n11, 143n5, 243n43, Solmsen, F. 81n34, 84n54
360n5–6, 361n9, 362n10, 373n29, 405n3, Sommers, F. 313n14
488n19, 490n21, 492n25, 576n2, 598n49, Somville, P. 459n14
607n5 Sorabji, R. 456n1, 669n3, 674n26, 684n63
Schöpsdau, K. 360n5 Spawforth, A. 77n25
Schwab, W. 14n39, 164n11, 165n17–18, Stead, C. 673n19
166n19–20, 168n31, 168n33 Steel, C. 672n14, 683n62
Scolnicov, S. 103n43, 231n1, 234n11, 234n14, Stokes, M. 95n8, 100n30
250n57, 250n60 Strachan, J. 311n7
Scott, D. 2, 12n33, 12n35, 18, 20, 25, 70n7, Strang, C. 239n23
163n9, 164n11–12, 166n19, 166n21, 166n23, Strange, S. 683n59
 index nominum   763

Strauss, L. 103n44 414n16, 419n23, 430n3, 432n6, 436n10,

Strawson, P. 651n18, 658n27 459n16, 471n44, 476n57, 532n3, 566n8,
Striker, G. 346n30 578n7, 611n14–15
Szaif, J. 329n54, 330n56, 332n61, Vogt, K. 344n21, 357n57
334n70
Szlezak, T. 103n43 Wallis, R. 674n26
Wardy, R. 53n29
Tarán, L. 56n39, 83n43 Warren, J. 2, 3n6, 8, 10n28, 149n16, 184n5
Tarnopolsky, C. 154n27 Waterfield, R. 561n6
Tarrant, H. 84n55, 674n23–24, 675n29, Waterlow, S. 265n9
677n35, 679n43 Weiss, R. 137n58–59, 586n28
Taylor, A. 84n51, 249n54, 275n28, 659n28 Wells, R. 86n61
Taylor, C. 2, 5, 10n28, 11n30–31, 12, 13n38, 14, West, M. 185n6
17n47, 23–24, 23n64, 31n82, 31n85, 44n9, Westerink, I. 69n3, 338n4, 684n63–64
45n13, 122n10, 148n15, 161n2, 167n26–27, White, N. 26n70, 167n25, 169n40, 173n61,
168n33, 169n40, 174n69, 176n78, 216n16, 207n1, 212n6, 272n20, 484n9, 503n63,
281n40, 282, 310n6, 386n13, 409n11, 410n12, 532n3, 576n2
432n6, 439n15, 458n11, 577n6, 590n39, Wilberding, J. 591n41
613n17, 621n30, 622n31 Wilburn, J. 410n12
Taylor, T. 684n63 Wilkes, K. 176n79, 177n83, 178n90
Thaler, N. 277n34 Williams, B. 22n62, 104n51, 486n13, 490n21,
Thomas, R. 77n26 614n18
Thompson, W. 54n33, 426n33 Williams, C. 272n20
Too, Y. 44n12, 53n31 Williams, T. 196n45
Trivigno, F. 147n12 Wilson, N. 77n25, 86n63, 86n65
Tuozzo, T. 343n18 Wittgenstein, L. 171n53, 274
Wolfsdorf, D. 578n7
van Eck, J. 312n12, 318n22, 329n55, 331n60, Woodruff, P. 122n10, 147n12, 386n13, 539n9,
501n51 561n6
Van Riel, G. 629n5 Woolf, R. 121n5, 157n31, 191n23, 202n62,
Vasiliou, I. 586n27, 587n30, 591n40–42 587n29
Vegetti, M. 229 Woozley, A. 207n1, 217n17
Vlastos, G. 5n9, 7n15, 10n29, 16n46, 20n56,
29n76, 49n21, 64n57, 96n12, 97n15, 104n52, Yi, B. 245n48
122, 122n9, 129n38, 147n11, 152n20, 153n22, Yunis, H. 77n26
188n13, 213n10, 217n19, 234n15, 237n20,
239n23, 241n28, 242n36, 254n80, 255n85, Zalta, E. 252n69–70, 253n71, 256, 257n91
292n24, 312, 312n12, 318n23, 320, 320n29, Zeyl, D. 41n1, 293n26, 295, 299, 302n54,
382n6, 383n7, 385–86, 385n10–11, 388–90, 302n56, 410n12
389n19–22, 389n24, 393, 400, 400n43, Zuntz, G. 50n23
 General Index

Academy 49–50, 50n22, 54–55, 55n35, 60–62, Antiochus 669n1, 673n22, 675n29,
78, 78n28, 82–85, 404, 645–46, 645–46n2, 676n31, 677, 677nn36–37, 679n43,
673–75, 674n27 681n50
account Antipater 535n4
and knowledge 32, 166n19, 263, 270, Antisthenes 41, 55, 55n38, 184n3, 275n27
280–81, 445–46, 449–51, 451–53, 537–38 Anytus 46, 120, 161, 607
compositional 450 apeiron, the unlimited 340–50 passim, 349n39
of properties 438 aporia 81, 98, 99–101, 108, 170n49, 244,
See also definition; logos 311–12, 403–04, 421–24 passim, 450,
actions 649n16, 654–56 passim, 663
and the good 527, 528n15 appearance 412–14, 508, 539, 582n19,
and verbs 494–95 618–20, 621n29
involuntary 417–18, 418n21 approximation view 467–68, 468n35, 476n60.
versus objects 494n32 See also compresence of opposites
voluntary 407–08 Arcesilaus 83–84, 86n60, 673n22, 676n31
Agathon 553–55, 569–70 archê, principle 287–89, 290–91, 291n16, 294,
aitias logismos 13, 17, 32, 163–65, 164n11, 167, 296, 301, 656–57
170, 177, 270, 273, 396, 443, 448n27, 449 Archelaus of Macedonia 153
akrasia Archytas 49–52, 49n21, 52n27
and appearance 143, 150, 413–14 Aristophanes of Athens 52, 106n60, 559–60,
and belief 8, 411–14 564–67
and ignorance 10, 150, 411, 509–10, 512 Aristophanes of Byzantium 84–85
and knowledge 8, 150, 411–14 Aristotle
and pleasure 10 and philia, friendship 552
and Protagoras 8–10, 143, 148, 150, 409–14, and Plato 35–36, 56–61, 61n46, 72n14,
507–10, 512–13 78n29, 80–81, 83, 83n46, 298n40, 349n39,
common view 409–12 359n1, 376–77, 444n20, 503, 596, 622–23,
kinds of 8n20 636, 645–47, 645n1, 645–46n2, 646n7
meaning of 409n10 on causation 297–98, 298n38, 300, 354
Socratic denial of 8–10, 409, 507–10 on definition 281
Alcibiades 553, 559, 567–70 on envy 300
Alcinous 687–90 passim on first principles 277n32, 287
analogy on forms 15, 35–36, 240n27, 349n39,
of the cave. See cave analogy 397–99, 458n13, 459, 471, 648–65
ship of state 220–21 on language 503
soul and state 25, 217, 546 on motion 636
statesman and doctor 597 on Socrates 73, 76–77, 398, 403–04, 646n7,
with health 153–54, 156, 423, 426 648–49, 654
with perceptual judgments 412 on the soul 507, 509, 515, 520
Anaxagoras 77, 188, 342–43, 354, 376 Aristoxenus 51, 61n46, 444n20
 general index   765

assets 420–22, 424. See also external goods final 297–98, 298n40, 376
authoritarianism 211, 216, 544 in Phaedo 16–17, 187, 467–77
auto kath’ hauto 233–35, 244, 321n30, 322n33 intelligent 301–02, 305
autonomy 25–27, 214, 582 of evil 627, 637, 638–42
auxiliaries 22, 25, 208, 219, 413n14, 589–91, See also forms, and causal responsibility
591n40, 591n42, 592n43, 593, 595, 597n47, cave analogy 23–24, 23n63, 46, 189, 209,
598, 600 222–23, 223n25, 225–27, 444, 470, 591n42,
592–94, 592n43, 613
beauty 560–63 Cephalus 48, 63, 79, 231–32, 340, 572
being change 15–16, 33, 59, 72, 195, 266–68, 283, 297,
and falsehood paradox 230 303, 311, 315–240 passim, 317n21, 318n22,
and knowledge 440–41, 455–56 322n34, 436n10, 451, 465–66, 469, 636–37,
and names 489–90 648, 652–54, 662, 662n30. See also
See also einai, predication Heracliteanism
belief, doxa choice of lives 211–12
and akrasia 411–14, 508–10 Christianity 670, 673n19, 680n47, 682–84,
and appearance 412–13 683n60
and consistency 392–93 Chrysippus 83
and falsehood paradox 32, 493–94, Cicero 84, 677–78, 678n39
502–03 classification 451, 453, 488
and knowledge 163, 175–79, 435–38, Cleinias
442–43, 449–50 in Euthydemus 110, 419–21, 533
and recollection 435–36 in Laws 635, 639
and the elenchus 389 coercion 580–83, 587, 591, 597, 600
instability of 413 compresence of opposites 15–17, 23, 23n65,
true belief 7, 7n16, 12–13, 18, 31–32, 163, 171, 29, 29n74, 195, 466–70, 467n32,
171n52–53, 173–79 passim, 175n74, 474–75, 568n35. See also
176n80–81, 177n82–83, 269–70, 374, approximation view
374n33, 385, 389, 397, 413, 430, 450, 567, consistency 104–05, 112
579–84 passim, 611–12, 620. See also convention, nomos 45, 632
knowledge, and true belief cosmos 287–305 passim, 355, 542, 548, 598,
benefit 146, 152, 390, 420–21, 424, 444, 629–30, 637
533–34, 563, 570–71, 576–601 passim, courage, andreia
613, 617 and knowledge 392–93, 538
boulê, deliberation 578. See also practical and shame 154
deliberation and the virtues 147, 147n13, 150, 360–61,
364–65, 393
Callias 338, 339n5, 339n8, 662 as self-control, sôphrosunê 372–73
Callicles 46, 46n16, 53n29, 98, 106, 142–44, raw 366, 368–69, 371–72, 374
148n15, 151n19, 154–58 passim, 415n18, refined 366–67, 371–72, 374–75
417, 423, 512–14, 564n7 Cyrenaics 9n23
capacity 442–43
cause definition
and aitia, aition, and/or aitios 16–17, 17n46, and virtue 162–63, 577
290–91, 294, 296–98, 297n34, 343, Aristotle’s conception of 281
436n10, 457n9, 640, 661 by genus and species 281, 451
and necessity 301–05 in Parmenides 236, 241, 245, 250–51
causal principles 630 nominal 129, 398
efficient 297–98, 298n38, 298n40, 354 priority of definition 162
766   general index

definition (Continued ) dialogues

real 129, 398 and fiction 101–03, 109–10
Delphic oracle 120, 123, 132–33, 199n57, and oral performance 77–80
379, 429 and Plato’s philosophy 80–81
demiurge 292–93, 299–300, 629–31, 640, 672, and Platonism 680–81, 681n50
673n16, 682. See also god and scepticism 81, 100, 676n31
democracy composition of 71, 77–80, 97n15, 101–03
Athenian 47–48, 53, 64, 221 developmental view of 2–3, 81, 97n15,
critique of 219–22 142–43, 254–55, 457–58, 531–32, 599
democratic man 213–14 dialogue form 4, 77–81, 93–94–95, 97,
instability of 221–22 100–03, 105–07, 109–14, 143,
pure 221 337–40, 455
Democritus 456, 456n4, 471, 531, dramatic detail of 93–94
678n39 dubious 70
dependency thesis 577n5, 582n18, 584n23, interpretation of 95–96, 97–101, 142–43,
592n43 200–01
desire late 3, 60–62, 72–74, 340, 345, 375n35,
and akrasia 412, 415, 509 493n27, 594–601
and Charmides 417 order of 1–4, 71–77, 142–43
and erôs 553–54 publication of 77–78, 82–84
and Gorgias 143, 150–57, 415–18 spurious 69–70, 69n3, 69n5, 84n50, 85–86,
and mental conflict 514 86n60
and motivation 148, 510–11 transitional 426
and pleasure 511–13, 524–25, 543–44 unitarian view of 3, 671
and Protagoras 146–50, 409–14, 415–16 Socratic 41, 41n1, 74, 77, 111, 162, 184n4,
and the appetitive part 521, 523–25 262, 337, 405, 426–27, 429, 459, 531–32,
and the good 142, 146–57 passim, 415–18, 575n1, 576–85 passim, 601, 681
510–14, 527–28 Dion (brother-in-law of Dionysius I) 49,
and the rational part 524, 541 52, 64
and wanting 553–54 Dionysius I (tyrant of Syracuse) 49, 52, 64
appetitive, epithumia 21, 155, 191, 215, Diotima 59, 106, 466, 551–73 passim
416n19, 417, 423, 521, 523, 526–27, 554, 591, disavowal of knowledge
619n22, 624n33 and Apology 4, 121–23, 380–81, 385, 387,
bodily 514 429–30
good-dependent 8, 8n22, 21 and Charmides 381
good-independent 21, 22n8, 143, 143n9, and Crito 137–38
150–52, 155–57 and definitional knowledge 384
non-rational 3, 3n6, 28, 73, 213–16 passim, and kalon kagathon, fine and good 120,
582n19, 600, 619 123, 379–81, 386, 429–30
rational, boulêsis 8, 151–52, 416n19, 417, and limited knowledge 386–87
417n20 and Meno 382–83
spirited 21, 215, 527 and paradox 121–22
versus aversion 519–21 and Plato’s views 429
developmental view. See dialogues, and political philosophy 577–78
developmental view of and priority of definition 4–5, 130–31
dialectic 107–09, 202n65, 209, 231, 231n3, and Socrates 4–5, 7, 121–23, 380–87
348–50, 355, 622, 688n75. See also and systematic knowledge 430n3
education, dialectical and Vlastos 122, 385–86, 390, 430n3
 general index   767

interpretations of 4–5, 122–23 elenchus

moderate interpretation of 384, 387 and Apology 380, 392
two-senses interpretation 5, 385 and Benson’s view 391–92
See also knowledge and Euthyphro 124–32 passim
divided line, image of 23, 23n63, 222, 288, and Gorgias 390
296n33, 444–45, 591n42 and Irwin’s view 390–91
doctrinal interpretation 80–81, 81n34, 115. and knowledge 384, 389–90
See also dialogues and Laches 388
dream theory 32, 263, 274–78, 274n26, and Meno 162, 396–97
282–83 and political philosophy 578–79
dualism and priority of definition 129–31, 387–88
cognitive 687–89, 687–88n73 and Vlastos’s view 388–89
cosmic 639 problem of 6–7, 11, 13, 389–92, 396–97, 611
soul-body 85, 190–91, 399–400, 647 standard 388–89
elitism 546–47
education epistemic deference 581–82
and character 546 epistemological holism. See knowledge,
and dancing 623–24 and holism
and pleasure 544 eristic 169n40, 202–203, 202n65
and Socrates 42, 148, 157, 606 erôs
and stories 616–19 and beauty 554–55
and the polis 18, 579 and moderation, sôphrosunê 558
and the soul 613–14, 623–25 and sex 555–56
dialectical 209, 226–27, 547, 613, 615–16, and the good 554–57
620–22 and the tyrannical man 567–70
in Athens 606–07, 609–10 as desire 553–54
in Republic 18, 614–16. See also education, bad form of 567–70
of the guardians contrasted with philia 551–53
mathematical 24, 209, 211, 222, goal of 568
226, 226n29, 288, 444, 547, 583, intellectual 214–15
589–90, 591n42, 598, 613, 615–16, self-involvement of 557–59
620–22 essence 5, 15, 322, 398, 414, 490, 498, 647–49
musical 27, 220, 363, 373, 374n33, 589–90, ethics
595, 598n50, 605, 606n2, 610, 610n11, 612, and happiness 531–33
615–17, 620–24 and metaphysics 547–48
of the guardians 24, 208–09, 220, 222, 226, and Socrates 7–11, 403
226n29, 374n32, 615–16 ethical rationalism 111–12
Plato’s concern with 27, 605 Plato’s 18, 531–33
eidos etymology 489–90
and piety 5, 125, 129, 383–84 eudaimonia, happiness
as form 15n43 and assets 420–22, 424
as kind 35 and ethics 531–33
auto to eidos 125, 383 and the good 419–25, 533–35
meaning of 233–34, 345n23 and Gorgias 146, 539–40
See also forms and pleasure 10–11, 34, 147, 543–44
einai, to be 23, 23n64, 33n88, 312–13, 320–22, and Protagoras 146–47, 539
441, 447n25, 448n28, 456. See also and Socrates 8–9, 425–27
being, predication and the political art 146, 421–24
768   general index

eudaimonia, happiness (Continued ) and Phaedo 1, 14–18, 186, 192–99, 457,

and virtue 18, 142, 419–25 passim, 533–36, 461–62, 465–66, 467–68, 471, 474,
538–39, 543, 545–46 476, 484
and wisdom 421–25 and Phaedrus 1, 195n36, 196, 241, 241n31,
common view of 421, 532–33 345n23, 460, 470, 690n83
contemplative view 34, 590–91, 630 and philosophers 214–16, 586, 653–55
in Euthydemus 419–22, 532 and Platonism 685, 688–91 passim
in Republic 18, 25–27 and predicative expressions 482–85
of the city 217–18 and properties 439
psychological eudaimonism 9, 21, 425n32 and recollection 195, 438
rational eudaimonism 9, 18, 21, 26, 580 and Republic 1, 23–24, 462–64, 473–74
theory of 585 and self-predication 29–30, 29n74,
evaluative concepts 224 240–44, 251–53, 255n86, 315n18,
examined life 119–21, 134–35 476–77, 659
expertise 430 and sensibles 15–17, 24, 195
explanation and separation 17, 234–37, 244, 398, 471,
and forms 438–39 648–50, 652–55, 658
and inquiry 105 and Socrates 397–400
in Phaedo 187, 438–39 and statements 500
necessary and sufficient conditions and Symposium 1, 466
of 384, 395 and Theaetetus 261–62, 263–64, 269
teleological 439, 444 and Timaeus 2n3, 469–70
external goods 124, 154, 535. See also assets apprehension of 220, 223–25, 685
as condition of knowledge 261–62,
fallacy of irrelevance 19–20, 212–13, 212n7 269, 273
false judgement 270–72, 272n20, 493, 503 as meanings 17, 458n10, 485
falsehood paradox 493, 496–98, 502–503 as objects of inquiry 452
family 209, 551, 557, 568, 589, 593 as paradigms 293–94, 648, 660
first principles 56, 277, 277n32, 288, 672, as particulars 15, 35–36, 648–65 passim
678n38, 683. See also archê, principle as principles of the cosmos 293
forms as principles, archai 661–64
and Aristotle. See Aristotle, on forms as standards 484–85
and causal responsibility 457–58, 457n9, as theoretical entities 457
476, 627 as universals 15, 35–36, 397–98, 400,
and change 465–66, 469 431–32, 472–74, 472n47, 648–65 passim
and development of Plato’s characterization of 465–70
thought 457–58 criticisms of 2, 35–36, 35n92, 648–61.
and Euthyphro 398–99 See also Aristotle, on forms; third
and explanation 15–16, 187, 439 man argument
and genus and species 35, 250, 281, 345, form of animal 640
349, 482, 640 form of beauty 1, 16, 194, 223, 225, 457, 466,
and opposites 462 476–77, 552, 559–65 passim, 613
and Parmenides 2, 17, 28–30, 60, 231–57 form of equal 1, 194–96, 461, 467
passim, 458, 482 form of likeness 233–38 passim, 248
and participation 16, 29, 35, 57, 129n36, 238, form of the good 18, 20, 24, 194, 209, 216,
319, 347n33, 472n46, 562, 589–90n34, 651, 222, 226, 276–77, 288–89, 296, 444–45,
658, 662, 683 453, 553, 594, 621–22
 general index   769

form of the just 30n77, 194, 209, 586 and Protagoras 142
form of unity 445 and Republic 418
immanence of 17, 61, 399, 471, 477 and the cosmos 293, 298–301
linguistic dimension of 17, 481–85, 482n4 and the Promethean method 348
location of 470–72 and virtue 390, 419–25
motivation for 16, 58–59, 397–99 as order and proportion 24, 444
order of 214–15 classification of goods 533–34
perfection of 467–69 divine goods 361–64, 375–76
scope of 29, 458, 460–64, 481–85 hierarchy of goods 361–67, 375–76
theory of 2, 14, 17, 28–29, 59–60, 73–75, human goods 361–64, 375–76
77, 113, 129n36, 234, 253–56, 263, 283, Plato’s lecture on 61n46, 444, 444n20
397–99, 457–58, 472–73, 481–85, 647–51, supreme 376–77, 444–45n20
664–65 Gorgias 145, 151, 407n6, 510, 605–08,
four-fold division 34, 342, 344, 346–49, 608nn7–8, 612–13
354, 356 grounding relation 126–28, 129
guardians 22, 24–27, 25n67, 208, 209n3, 216,
genus 281, 345, 345n25, 451, 453, 482, 657, 219–20, 222, 226, 361, 369n24, 374n32,
657n26 541, 597n47, 615–16, 618, 623, 660
Glaucon’s challenge 19–20, 106, 207–08, Gyges 19, 211, 614n18
211–12, 212n6, 215, 535–36, 614
gnosticism 670, 682, 682n55 health 133, 151, 153–54, 185, 210, 213, 346, 356,
god 361–64, 420, 422–23, 461, 533–35
as cause of good 627, 642 hedonism 9–11, 34, 61, 106, 114, 143, 143–44n6,
as ethical paradigm (homoiosis theôi) 542, 148n15, 150, 282, 282n41, 410n12, 584n23.
628, 638 See also pleasure
as nous, intellect 6, 299, 299n41, Heracliteanism 18, 31, 56–60, 265–67,
299–300n42, 629, 641–42 282–83, 446
created gods 629, 631–32 hiatus 56, 60, 75, 75n21
in Timaeus 298–300, 302–05, 627–34 Hippias 44, 106n57, 176n78, 394, 408,
passim 538, 607n4
shape of 628–30 historical Socrates 41, 73, 76–77, 96–97, 97n15,
world-god 629–30, 637 183–84, 184n4, 201, 379–80, 403–04.
See also demiurge See also Aristotle, on Socrates
good fortune 151, 420 Homer 78, 98, 210, 514–15, 592, 605, 617, 660
good, agathon homosexuality 541, 544, 564, 571–72
actual good 142, 146, 150, 152, 415–18
and desire 142, 146–57 passim, 415–18, Iamblichus 674n26, 675, 683–84, 683n61,
510–14, 527–28 692n89
and erôs 554 imitation
and explanation 446 and education 618, 624, 660
and Gorgias 142, 539–40 and names 489–91
and happiness 419–25, 533–35 criticism of 99, 458n13, 620, 660
and knowledge 340 mimeticism 35, 651, 662
and order 444 infallibilism 264–66, 268, 270, 385–86
and Philebus 340–45 inquiry 42, 99, 104, 147, 169–70, 196, 437, 452,
and pleasure 147, 337–45, 350–54, 531, 611, 614, 677, 688. See also Meno’s
410, 423 paradox, and problem of inquiry
770   general index

intertextuality 114–15 and recollection 381–83, 437n11

Isocrates 44, 52–56, 53n31, 54n34, 64, 606, and Socrates 379–80, 396
607n5, 621n30 and specification 431–33
and teaching 176–77
justice, dikaiosunê and testimony 169–70
and art 407–09 and the soul 436
as convention 45 and true belief 13, 31–32, 163–65, 171n52–53,
as knowledge 510–11 173–79 passim, 269–73, 385, 413, 430–38
as psychic order 22, 208, 213–14, 423–27 passim, 443, 446, 449–50, 452
civil 207–208 and virtue. See virtue, as knowledge
conventional 19–20, 212–16, 213n8 as expertise 508
defense of 19, 153, 207, 211–13 as justified-true-belief 13–14, 122, 164–68,
in Republic 18, 536n6 164n13, 174, 396, 437–38n12, 577n6
individual 22, 207, 216 as primary 431–32
metaphysical defense of 20, 214–16 as technê, craft 42
platonic 19, 212–13, 224, 426 as true belief with an account 32, 163–64,
psychological defense of 20, 213–14 263, 270, 272–83 passim, 396, 446, 449,
in Laws 364–65, 374–75 611n13, 613n17
as understanding 164–68
kalon, fine by acquaintance 24, 271n18, 441, 622
as admirable 147, 153, 374 coherentist view 446
as difficult 394 definitional 58, 130–31, 135n52, 384, 387–88,
meaning of 344n20 393–96, 398, 577
kinds derivative 450
and concepts 686 empirical 13, 198, 452
and the Promethean method 346–47. ethical 578
See also Promethean method exact 435–36
communion of kinds 33, 311, 315–24 experiential 434, 449
passim, 322n33 expert 5, 430–31
greatest kinds 35, 311, 315, 451 hierarchical 281, 445
knowledge of 430–32 in Meno 163–68, 175–79, 432–38
natural 394–95 in Republic 440–46, 452
knowledge in Theaetetus 261–83 passim, 446–51
accurate 355, 393 inferential 434–35, 434n8
and akrasia 150, 409–14 innate 12, 197–98, 685, 688
and asymmetry 263, 273–74, 276, 283 integration of 112
and classification 451, 453 latent 382–83, 397
and elenchus 390 limited 386–87
and first principles 287 manifest 12, 382–83, 397
and forms 16, 24, 245, 261, 269, 281, 440, 443 necessary and sufficient conditions
and foundationalism 445 394–95
and holism 24, 32, 32n87, 446, 584 of forms 17, 24, 220, 440, 443, 461,
and motivation 508–509 586, 588n33
and perception 16, 31, 262–63, 264–69, of good and bad 133, 508–09
282–83, 446–49 perceptual 437, 437n11, 443–44n18
and priority of definition 5, 5n10, 11, prenatal 12, 13n36, 175n75, 186,
128–31, 387–88, 393–94, 396, 431n5 197, 197n47
 general index   771

propositional 24, 441, 447–49 Meno’s paradox

senses of 384–86 and the elenchus 170n49, 396–97
stability of 413, 436–37 and problem of discovery 12, 162, 168–70,
sufficiency of 150 171–72, 381–83, 432–34
supremacy of 508–09 and problem of inquiry 12, 162, 168–70,
systematic 452 382, 396, 432–34
and recollection 12–13, 163, 171–75, 382–83,
language 396–97, 434–38
and Cratylus 481, 485–92 recognitional response 13, 171–75
and forms 481–85 true-belief response 12–13, 173–75
and Sophist 492–503 method
late-learners 33, 310–15, 313n15, 315n18–19, of division 34, 104n54, 279, 309, 345,
321–24, 322n34, 335 345n24, 348, 351, 355, 451, 492, 494, 647,
laws 647n11
and obedience 7 of hypothesis 110n70, 163, 174–75, 179,
and persuasion 360–61, 597–98 187–88, 188n14, 196, 201, 247–49, 256,
and teleology 360 282, 408, 609
in Crito 135–38 mind 34, 271–72, 338, 342, 344, 447.
rule of law 360 See also nous, intellect
Leontius 413n14, 522–23 money 525–26
logical atomism 274 motion, kinêsis
logos 32–33, 120, 120n1, 125, 128, 131–34 and the principle of opposites 518–20
passim, 137–38, 185–86, 188–89, 199–203 divine 636
passim, 231, 263, 273–83 passim, 290, kinds of 635–36
294–95, 312–13, 315, 325–28, 333–35, self-motion 636
362n11, 366, 369–70, 445–46, 445n22, motivation
449, 494, 498, 537–38, 561, 627, 654, and desire 148, 510–11
690–91, 690n82 and knowledge 143, 508–10
love 20, 551–73. See also erôs; philia, and the good 511–12
friendship non-rational 582n19, 584, 592.
See also desire, non-rational
macro-micro account 22, 22n62 Plato’s theory of 507
Magnesia 359–60, 635, 638 music 51, 184, 199, 347–48, 355
manuscripts 86–89, 638–39
Marcus Aurelius 680, 680n47 names
marriage 564–65 and classification 488
mathematics 24, 35, 44, 49–51, 51n25, 55, and conventionalism 486–89,
55n37, 211, 222, 226, 226n29, 288, 349, 486n13
355–56, 444–45, 452, 548, 548n18, 583, and Cratylus 485–92
589–90, 615–16, 631, 688n75 and definite descriptions 491, 491n24
matter 641–42 and imitation 489–90
meaning 492–503. See also forms, and naturalism 487–89, 490–92
as meanings and nature 487–89
means-end reasoning 21n58, 151, and reference 491–92, 495–98
524–26 and stating 32–34, 275n27, 314–15
medicine 42–43, 145–46, 293, 347, 419, and the legislator’s art 488–89
422–23, 597–98 correctness of 486
772   general index

names (Continued ) and reason 464

primary and derivative 489–90 and sensory awareness 268n13, 268–69
usage 486 and summoning 462–64, 621
versus statements 502–03 and the senses 264, 268, 268n13, 447
necessity 301–05, 636, 641 perceptual experience 470–71, 672,
Nicias 48, 98, 109, 393, 538 686, 689
nocturnal council 360, 360n5, 375, perceptual illusion 508
599n51 perfection 585, 588–89, 594, 600–01
non-identity assumption 30, 591n42 Pericles
non-self-explanation principle 242–43, and virtue 406–07
252–53 funeral speech 47
nous, intellect 277, 294, 299n41, 340n13, philia, friendship 551–52
342–43, 355–56, 357n55, 361–62, 362n11, Philo of Larissa 675n29, 676–77, 676n31
367, 376. See also god, as nous philosophers
number 621. See also mathematics and the forms 196–97, 214–16
as rulers 216, 222–23, 359–60
one over many assumption 15, 30, 30n79, philosophy
239–40, 242–43, 246, 252, 252n70, and Phaedo 199–203
255–56, 623, 659, 661 as charm 186, 200
one/many problem 28, 341, 345, 348 as inquiry 42, 104
onoma, name 494–95, 496n37. as preparation for death 184–85, 191, 200
See also names definition of 223–25
ontology 455–56, 492–503 pretenders to 226–27
organic theory of the Republic 25, 217–18 practice of 42
piety 128, 383–84
pain 343, 350–53, 544–45. See also pleasure Platonic Corpus 69–70, 82–86
painting 69n5, 102, 353, 463, 620, 620n26 Platonic division 343, 345, 345n24, 349, 351,
paradeigma 129n36, 384, 398, 662, 664. 355, 647n11. See also method, of division
See also forms, as paradigms Platonism
Parmenides 58–59, 59n44, 294n28, 456, and orthodoxy 680–81
456n2, 456n4, 471 and recollection 686–92
participation. See forms, and participation commitments of 36, 670, 675, 680, 685
particulars delayed development of 676–79
and forms 15, 35–36, 61, 195n37, epistemology 685–92 passim
224, 243, 347n33, 400, 443, 463–65, heterogeneity of 669–70
472–78, 472n47, 482–85, 589–90n34, interpretive method 670–71, 672n15
650–65 passim Middle Platonism 674, 674n24, 675n30
and knowledge 224, 272–73, 281, 283, traditional picture of 673–75, 675n30
443, 586n26 pleasure
peras, limit 340–50 passim and akrasia 148, 410–12
perceptibles 465–70 and desire 511–13, 524–25, 543–44
perception, aisthêsis and education 544
and change 465–66 and genus-species 349
and knowledge 31, 262–63, 264–69, and harmony 343
282, 446–49 and rhetoric 608n7, 610
and propositional content 31, 268–69, and sophistry 610
271n19, 447–49 and the apeiron 348–50
 general index   773

and the appetitive part 524–26 basic and non-basic 483–84

and the good 147–48, 338, 343, 349, essentialist 322–24
423, 539 ordinary 322–24
and virtue 544 versus identity sentences 33, 33n88,
calculation of 148–49 271n18, 275, 312–15, 317, 320–21,
false 350–51, 351n42, 353n48 323–24, 326
in Philebus 340–57 passim principle of differentiation of capacities 442
in Protagoras 10–11, 142–43, 147, 539 principle of opposites 518–21
mixed and unmixed 350–52, 545 principle of relatives 511, 521
nature of 545 principle of specialization 25, 219
plasticity of our capacity for 544 priority of definition. See knowledge,
sexual 154, 351 and priority of definition
Plotinus Proclus 670n6, 671n9, 672, 672n14, 674n26,
and Aristotelian theology 682 683, 684n63
and forms 690–91 production 421–22
and Stoic ethics 682 Promethean method 341–50, 347n33, 350
and virtue 671 passim, 354–56
epistemology 690–92 propagation 557–59
method of 682n55 properties
school of 675 and forms 15–17, 224–25, 398, 400, 431–32,
poetry 69, 99, 208, 210, 463, 526, 607n4, 610, 438–39, 462, 464, 477–78, 484n10,
616–20 passim, 619n24, 620nn26–27 659–60
political art and summoning 462–64, 621n29
and citizenship 28, 595 apprehension of 447
and happiness 421–24 propositions 441–42
and Socrates 576–77 pros heauto 250–54, 251n63, 252n70, 253n72
and the good 421 pros ta alla 250–53, 251n63
and virtue 422–24 Protagoras
political philosophy and measure doctrine 31, 45, 264–65,
and Gorgias 423–24 607, 609n9
and love 562–63 and perception 446
and Plato’s views 18, 27, 47, 575, 595, 601 and relativism 31, 31n83, 265–66, 487
and Socrates 576–85 and teaching 144–45, 150, 161, 606–610
and the elenchus 578–79 Protarchus 80, 98, 104, 106, 108,
and the late dialogues 594–601 338–57 passim
and the middle dialogues 210–11, 586–94 psychic harmony 19–20, 155–57, 157n31, 208,
and the Socratic dialogues 576–85 212, 216, 423–25. See also soul, psuchê
Porphyry 672, 672n14, 674n25, 682–84, psychological conflict
682n54, 683nn59–60 and akrasia 143, 410, 412–13, 415, 417
Posidonius 84, 676n33 and belief 413
practical deliberation 133–35, 137 and knowledge 148, 413–14
predication and sexual attraction 563–64
and “is,” 23, 33, 312–13 and Socratic intellectualism 150
and false statements 100, 325–35 passim, and Socratic intellectualism 515
500–502 and tripartion of the soul 519–21
and forms 481–85 in Odyssey 514–15
and principle of opposites 519 psychological egoism 415n17
774   general index

punishment 145n8, 153, 154n24, 156, sight-lovers 179, 223, 225–27, 442–43
535, 544, 579, 582–83, 582n17, 628, Simonides 147, 147n12, 610
633, 638 skepticism 81, 100, 100n28, 675, 676n31, 681,
Pythagoras 44, 50–52, 676, 678n40, 681n50 681n50, 687, 687n70, 689, 691
Pythagoreanism Socrates
and music 348 as historical figure. See historical Socrates
and Orphism 185, 189, 190n22 intellectual autobiography of 16, 187, 198,
and Plato 49–52, 56 200n58
characteristics of 49 Socratic fallacy 387–88
in Phaedo 189–90, 200n58 Socratic intellectualism 143, 150, 162, 509–10,
512–14, 515. See also akrasia; knowledge;
rationalization 413–14 motivation; virtue
realism 456 Socratic paradoxes 8–11, 154, 405
reason sophists
and the good life 120, 340 and Gorgias 141
as a power 613 and philosophy 310
function of 540–41, 599 and Protagoras 44–45, 141
receptacle 302–04, 302n55, 469n40, 641 and Socrates 405
recollection and virtue 406–09
and a priori inquiry 435 definition of 44, 310, 405
and experience 435 Plato’s view of 44–45, 310, 607–610,
and forms 438 607n6
and immortality of the soul 14, 186, soul, psuchê
192–93, 434 and Apology 399–400
and Meno’s paradox. See Meno’s paradox, and education 613–14, 623–25
and recollection and Gorgias 143, 155–57, 399–400, 423–25,
and perceptual knowledge 198, 510–14
436–37, 437n11 and motivation 507
and serial reminding 434–35 and Phaedo 14, 189–92, 514–17
regress 30, 166n20, 195n36, 241, 252, 256, 445, and Protagoras 155, 507–10
450, 452–53, 638. See also third man and recollection 434, 436–38
argument and Republic 517–27
resemblance 659–60, 659n28 and self-motion 636
rhetoric 145–46, 151, 606–08, 611 and Socrates 399–400
and the affinity argument 14, 186–87, 190,
secret doctrine 263, 265–67 199n56, 200–01, 516
seduction 570–72 and the body 190–91, 515, 637
self-control, sôphrosunê 96, 155–57, 368–75 and the final argument 14, 105, 188, 192n27,
passim, 388, 403, 423–24, 512–13, 196, 199n56
538, 558 and the opposites argument 14, 185–86, 201
self-predication. See forms, and and the recollection argument 14, 185,
self-predication 192–99, 195n32, 201
sêmeion, distinguishing mark 280–81, 494n30 and the self 191–92
Seneca 678, 678n39 and virtue 422–25
senses 198, 264. See also perception, aisthêsis, appetitive part, epithumêtikon 21, 155, 210,
and the senses 423–24, 523–24, 526, 540–41, 554, 554n2,
sensibles 15–17, 35, 447. See also perceptibles 568, 590–91, 599n52, 612, 618–19, 618n21,
sexuality 572–73 621n29, 622, 671n9
 general index   775

as charioteer 195n36, 517, 541, 567, 571 Stilpo 314

as prisoner 184, 190–91, 225 Stoicism 83–84, 85n59, 282, 305n57, 357n56,
as ruler 191, 637 420n24, 535, 535n4, 538–39, 542–43,
beauty of 560, 562 547–48, 670, 673n20, 676n33, 677–78,
death as separation of 185, 189 677n36–37, 678n39–40, 680–82, 680n47,
harmonia theory of 187, 190, 200, 514–15 681nn51–53, 686–88, 690
immortality of 14, 49, 95, 105, 110, 171, succession 15–16
174n71, 191–93, 199n54, 200–01, 434, sun, image of 23–24, 23n63, 187–78, 222–23,
514–17 444–45
non-rational, alogiston, part 524
parts of 21, 155, 212, 423, 514–17, 518–28 technê, craft
passim, 540–41, 612, 622 and knowledge 42–44
pregnancy of 556–59 and Socrates 405
pursuit of the good 527–28 and the Promethean method 341,
rational part, logistikon 21, 21n58, 208n2, 346, 349
210, 214, 423–24, 521–23, 527, 541, 554n2, and the sophists 145, 405
590, 591n40, 618 as a capacity for opposites 407
spirited part, thumoeides 21, 210, 515, 522, craft analogy 10, 537–38
524, 540–41, 554n2, 557, 566, 590–91, development of the idea of 539n9
590n38, 606n2, 612, 618–19, 618n21, in Sophist 310
621n29, 622, 671n9 of measurement 149–50, 508, 539
tripartition of 155, 191, 299n41, 515, 519–22, opposed to empeiria 146
540n10, 556 teleology 16, 18, 24, 115, 188, 360, 361, 376–77,
world-soul 631, 637–38 439–40, 444, 446, 622
species 345–48, 451, 453. See also eidos; tetralogies 69–70, 84–86, 141n1
genus theology
speech-acts 487–88, 496–97 and monotheism 635, 643
Speusippus 56, 56n39, 60–61, 60n45, 82, in Republic 627
673n21 Plato’s 629
stability (political) 584, 588, 593–94, 600 scientific 632, 634–35, 638, 643
statements theological arguments 6, 635, 636–37, 642
account of 33, 325–27 traditional 632–34, 638, 643
and falsehood paradox 230, 451–61 passim third man argument 29–30, 29n76, 239,
and forms 500 242–43, 246, 251–53, 476, 661
as truth bearers 496n36, 498 thirst 511–12
extentionalist interpretation of 501–02 thought, dianoia 296n33, 503
incompatibility interpretation of 328–30, Thrasyllus 84–85, 84n55, 85n56, 141n1,
332, 501 674n23, 678n38–39
incompatibility range interpretation Thrasymachus 19, 78, 98, 207, 212, 615
of 330–32, 335, 501–02 truth 124, 498, 616–17
late-learners’ view of 312–15 two worlds theory 17, 23–24, 23n65,
Oxford interpretation of 327n50, 328–31, 261n1
501–02 tyrant
primary and derivative 495–98, 496n36 in Gorgias 145, 151, 153, 415–18, 511
Stilpo’s view of 314–15 in Republic 207, 210, 211–12
versus names 33–34, 274, 312, 314–15,
492–503 passim uniqueness assumption 29–30, 30n79,
Stephanus 69n4 240n25, 242n35, 243, 256–57, 257n91
776   general index

universals 15, 35–36, 621n28, 648–65 passim. teachability of 11, 142, 150, 161, 163, 178–79,
See also forms, as universals 406–08, 609, 609n9
utilitarianism 25n69, 217, 217n19 unity of 28, 146–47, 147n11, 150, 364, 375,
414, 538
verbs 494–98
vice 210, 300 wax block 271
virtue “what is F?” question
analogy with health 153–54, 156, 423, 426 and being, ousia 384, 490
and akrasia 409–410 and definitional knowledge 5, 15, 128–29,
and education 609 236, 384, 388, 431n5
and Euthydemus 537 and Euthyphro 5–6, 128–29, 383–84
and Gorgias 422–24 and forms 5, 15, 461
and happiness 419–25, 543, 576 and manifest knowledge 383
and Laches 538 and Meno 162, 382
and pleasure 545–46 and necessary and sufficient conditions
and reason 540–41 5–6, 384
and recollection 383 and the elenchus 5, 162, 393, 397
and Socrates 123–24, 403–09 and the problem of discovery 433
and the good 423–24 and virtue 7–8, 146–47, 162
and the ideal state 546–47 aporetic reading of 395–96
and the sophists 405–09 whole-part account 22, 22n62
as a craft, technê 10, 150, 405–09, 537–39 wisdom, sophia, phronêsis
as knowledge 8, 150, 175–79, 404, 409 and assets 420–22, 424
as order among parts 422–25 and happiness 419–25
as psychic order 423–25 and harmony 370
as true belief 176–77, 179 and self-control 512–13
dominance of 545 and Socrates 5, 120–21,
entirety of 360, 364–65, 374–75 123, 429
ignorance of 381–82 as an art, technê 421–22, 424
necessity of 536 as divine 362, 367
of the soul 423–25 human wisdom 121, 124
parts of 146–47, 150, 161, 374–75 priority of 376
Socratic theory of 404
sufficiency of 8, 10, 10n29, 18n50, 55n38, Xenophon 41, 44, 509, 606n2
420n26, 534–36, 535n4, 547, 580n11,
584n23 Zeno of Citium 83