Wayne State University
Wayne State University Dissertations
1-1-2016
The Sophist In The Cave: Education Through
Names In Plato's Republic
Daniel Propson
Wayne State University,
Follow this and additional works at: https://digitalcommons.wayne.edu/oa_dissertations
Part of the Ancient History, Greek and Roman through Late Antiquity Commons, Linguistics
Commons, and the Philosophy Commons
Recommended Citation
Propson, Daniel, "The Sophist In The Cave: Education Through Names In Plato's Republic" (2016). Wayne State University
Dissertations. 1476.
https://digitalcommons.wayne.edu/oa_dissertations/1476
This Open Access Dissertation is brought to you for free and open access by DigitalCommons@WayneState. It has been accepted for inclusion in
Wayne State University Dissertations by an authorized administrator of DigitalCommons@WayneState.
THE SOPHIST IN THE CAVE
EDUCATION THROUGH NAMES IN PLATO’S REPUBLIC
by
DANIEL PROPSON
DISSERTATION
Submitted to the Graduate School
of Wayne State University,
Detroit, Michigan
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
2016
MAJOR: PHILOSOPHY
Approved By:
_________________________________________
Advisor
Date
_________________________________________
_________________________________________
_________________________________________
DEDICATION
To my five little philosophers.
ii
ACKNOWLEDGEMENTS
I am grateful to many, many people who have made this dissertation possible. My professors
at Wayne State have consistently challenged me to do my best work, and to them I am sincerely
grateful. I thank Lawrence Lombard for his endless curiosity and rigor, Bruce Russell for his
amiability and inquisitiveness, Susan Vineberg for her sound advice and her wit, Mike McKinsey for
his warmth and insight, Jonny Cottrell for his enthusiasm, and Greg Novack for his stories about
Sherlock and Moriarty. Phyllis Seals has been an indispensable source of practical assistance and
cheerfulness. My fellow current or erstwhile students David Baxter, Marcus Cooper, Dan Yeakel,
Jim Schwartz, Travis Figg, Jennifer Goossen, and Lucas Lockard have filled my life with interesting
conversations, and in no small way refined my thinking about Plato and philosophy. I also thank
the late Joseph Hinkins, who was the first graduate student I met at Wayne State, and whose passion
for philosophy inspired me.
After reading an early draft of Chapter I, Sean Stidd made a comment about Hilary Putnam
that transformed the chapter, and by extension the entire dissertation. Eric Hiddleston later helped
me refine my discussions of philosophy of language, so that I could engage with contemporary
practitioners of the discipline more effectively. Both Katherine Kim and John Corvino have been
extremely helpful to me throughout my time at Wayne State, and they certainly deserve my thanks.
Katherine is not just a great professor; she is a great friend, and her advice about how to work on a
dissertation has been invaluable. John is dedicated to the department and to his students, and he is
consistently available to help, even at a moment’s notice.
I have only Herb Granger to blame for my passion for Plato, and only Ken Walters to blame
for my passion for Greek. In only the third philosophy class I had ever taken, Professor Granger
iii
inspired my interest in Plato’s Republic, and patiently guided me through a graduate seminar that I
was hardly qualified to attend. His diligent instruction and his enthusiasm set the course for this
dissertation. As for Professor Walters, I began taking courses in Greek because they were a required
chore for my chosen profession. By the time he was done with me, I found myself utterly fascinated
by ancient Greek, both the language and the culture. Professor Walters deserves credit not only for
the hours he has spent poring over this manuscript, but also for any understanding I have of the
intellectual culture of ancient Greece.
My advisor, Josh Wilburn, appeared on the scene at exactly the right moment, and has never
ceased to amaze me with his diligence and insight. Josh has made me a much, much better
philosopher in the process. His patience with drafts that could be scattered and disorganized has
been impressive, and his concern for my professional future has been ever-present and
commendable. Thank you, Josh.
A number of friends and family members also deserve my thanks, for their unfailing love
and support. David Propson, Shelly Tyshka, Paul Propson, Mark Whitters, Josh Rock, Josh Pauley,
Mike Loveland, Dave O’Connor, Garrett Shireman, Brian Laba, Yvette Rock, Mike Miklosovic, and
Dan Smith: thank you! My gratitude also goes out to my father, whose perseverance through
adversity is the inspiration for my life, and my mother, who showed me what it means to love.
Finally, I would like to thank the crazy woman who encouraged her husband to follow his dream:
Priscilla Propson. Although I am surrounded by philosophers all day, you are the smartest person I
know, my dear, and certainly the greatest lover of wisdom I have ever known.
iv
TABLE OF CONTENTS
Acknowledgements _________________________________________________________________
iii
Introduction ______________________________________________________________________
2
Chapter One, “Education Through Names in the Cratylus” __________________________________
8
Chapter Two, “Onomastic Education in Plato’s Republic” ____________________________________ 44
Chapter Three, “Middle Virtue: Accurate ‘Poetry’ in Republic X” ______________________________
80
Chapter Four, “Language in the Cave” __________________________________________________ 104
Conclusion _______________________________________________________________________ 134
Bibliography .______________________________________________________________________ 138
Abstract __________________________________________________________________________ 145
Autobiographical Statement __________________________________________________________ 147
v
1
Be sure of it, my dear Crito: Incorrect speech is not only
discordant in its own right, but it also works mischief in our souls.
– Phaedo 115e1
1
Unless otherwise indicated, translations are my own.
2
INTRODUCTION
“What is the nature of the relationship between words and the objects they refer to?”
This question is surely central to philosophical inquiry, since errors of language inevitably
breed errors of investigation. Plato was one of the first people to ask the question. His answer, in
the Cratylus, has attracted a great deal of scholarly disapproval, and very little genuine philosophical
interest. Worse yet, scholars generally think that Plato’s views about philosophy of language were
confined to one (or, at best, two) dialogues, and that his philosophy of language played no particular
role in the metaphysical theories that we associate with “high Platonism.”
In this dissertation, I will contest all those common views. I will argue that philosophy of
language played a genuine role in Plato’s metaphysical doctrines in the Republic, and that these
doctrines are borne out by at least half a dozen other dialogues. I will argue that Plato’s views on
linguistic reference are, far from being obscure or outlandish, roughly comparable to certain widely
accepted modern views about linguistics, and reasonably plausible. I will argue that Plato’s cave
image dramatizes his philosophy of language – and his animosity against the sophists – in a
remarkable and previously unappreciated way.
I am not, however, making the claim that philosophy of language was somehow central to
the dialogues of Plato’s middle period. I do not think that theories of reference were on Plato’s
mind when he wrote the Republic, and I don’t think my thesis requires him to have been focused on
such a thing. I would merely say that Plato’s views about language informed his writing about other
topics, and that this connection can be traced in the dialogues. And perhaps this is just what we
would expect. Had Frege written about politics and propaganda, surely his views about language
would have influenced what he said. Why? Because language is the very lifeblood of political
3
information/misinformation, and – as Orwell recognized – the most powerful revolution proceeds
by controlling what things citizens are capable of referring to. The book 1984 is in the tradition of Plato’s
Republic, not (I hope) because the latter book does away with all freedom, but because 1984 carries
on the Republic’s tradition of media censorship. As I will argue, Plato believed that unjust states were
rife with linguistic inaccuracy about objects of the highest importance; citizens in such a state could
refer to objects like justice or temperance, but their notion of these words was radically at odds with
the genuine nature of such objects.
Plato’s solution to this sorry situation was censorship. According to Plato, the principle
guiding his censorship is truth (Rep. 490a).2 With a few exceptions, stories are to be censored
precisely insofar as they fail to conform to the truth (Rep. 377e). In Athens, however, it was a
common sophistical teaching that falsehood was impossible.3 Plato’s philosophy of language in the
Cratylus, the Sophist, the Statesman, and the Theaetetus was clearly meant to dispute this teaching as
thoroughly and decisively as possible. These dialogues, then, have at the very least the role of carving
out a place for the sort of falsity that must then be expunged from the Beautiful City. I see them as
doing something more. By promoting the notion that language proceeds from the lawgiver,
dialogues like the Cratylus and the Statesman make the claim that language is a political entity, thereby
suggesting that there is some connection between primary education and linguistic usage. This
suggestion, far from being disproven by the text of the Republic, is confirmed by it. The Republic is
not a dialogue about the philosophy of language, but it is highly dependent on dialogues that are.
2
Scholars may object that Plato is more moved by moral benefit than truth, in determining what stories are suitable to
tell. I do not think Plato would allow these two categories to be separated, however. From our perspective, Plato is
willing to broadcast lies that are morally advantageous. From his perspective, however, he is merely putting forward
factual inaccuracies that point to more fundamental truths (the Forms).
3
Gorgias expresses this view in his work “On What Is Not” (MXG 980a9-12) and Plato ascribes it to Protagoras at
Theaetetus 188d–189a.
4
My argument will proceed in the following way. In Chapter I, I will offer a reading of the
Cratylus. My goal there will be to understand what Plato meant by the claim that names can be false,
and I will argue that names are false just in case the embedded semantic content in the name is false
of the object the name refers to. The second chapter will delve into a number of books of the
Republic, and argue that Plato’s system of musical education there implicitly relies upon the notion
that names can be false, and that names are false just in case their semantic content is false of their
referent. In the third chapter, I will turn my attention to Book X of the Republic, and argue that its
castigation of imitative poetry leaves another type of poetry (literally “making,” ποίησις) unscathed:
the making of what I will call “intermediates.” The sort of intermediate Plato is most interested in
is “intermediate virtue.” I will argue that intermediate virtue is synonymous with demotic virtue,
and that the goal of proper musical education is the “making” of demotic virtue through the use of
true and accurate language. Finally, in the fourth chapter, I will argue that the picture of demotic
education I have drawn in the first three chapters can be fruitfully applied to the metaphysics of the
cave image. The puppets at the second level of the cave are, I will argue, the (conceptual) names
bestowed upon objects of value by the lawgiver. It is this sort of object that one must comprehend
before ascending to the real world, and the understanding of these names constitutes a genuine step
toward knowledge.
A Few Comments
I should note from the start that my comments about philosophy of language here are not
presented (for the most part) in the lexicon of the modern discipline, and that this is done quite
intentionally. Ancient philosophers did not conceptualize language in the way that we do, and
forcing their concepts into our terms will only sometimes be illuminating. As a result, I have tried
5
to approach Plato’s philosophy using Plato’s terms. At the same time, my hope is that – especially
in the case of more obscure terms, like mimesis – I have offered something that can be recognized as
engaging with the types of problems we grapple with in contemporary philosophy.
My treatment of the Cratylus, as well, requires a word of explanation in advance. I will only
deal briefly and tangentially with the question many scholars consider central to any study of the
Cratylus: “Why does Plato spend so much time making his peculiar and implausible etymologies?”
Since I think we lack evidence to decide that question, I intentionally skirt the issue in the text. I
do propose one minimal claim about the etymologies – that they are partially intended to ridicule
sophists – but this minimal claim is such a commonplace that it hardly needs a defense. I believe
that we can pull a number of significant philosophical claims out of the Cratylus without solving its
central interpretive puzzle, and I focus my attention on those claims.
Finally, let me make a comment about the role of the historical Socrates in my dissertation.
Unless otherwise indicated, I use the name “Socrates” to denote the Platonic character, not the
Athenian philosopher. There are doubtless contributions that Socrates the Athenian made to
Plato’s philosophy of language, but it is almost impossible to know what these contributions were.
The passages I discuss from the Euthydemus (likely an early dialogue) may be the clearest clue we have
about Socrates’s genuine attitude toward language: if they are any indication, he was aware of the
dangers of ambiguity and open to the notion that names admit of correctness and incorrectness.
It may be that Socrates’s most significant stamp on Plato’s philosophy of language, however,
came at the moment of his trial, when (if Plato can be trusted) Socrates contrasted the “language
spoken here [in the law courts]” (τῆς ἐνθάδε λέξεως, Apol. 17d) with both ordinary language and
the language of philosophy. If we look closely at that claim, we can see a sort of anticipation of the
6
idea that the poets and the sophists are at a “third remove” (Rep. 599d) from the truth, since they
do not even speak in the ordinary fashion, which is regarded as superior to the fashion of the
sophists. Moreover, Socrates’s speech embodies the failure of philosophical language to convince
the hostile masses in the cave (Rep. 517a; cf. Gorg. 486a-b), and his opponents’ speeches embody the
capacity for sophists, rhetoricians, and poets to manipulate moral terms in a way that dupes ordinary
people into believing falsely and acting wrongly. This failure, and the resultant death of Socrates,
may have played a role in forming Plato’s conviction that the public square must not be controlled
by sophists and orators, and thus his determination that philosophers must take a central role in
both government and media.
We can see, then, that Plato’s opposition to sophistical forms of communication was no
small aspect of his psychology. Sophistry had been the death of Socrates, and Plato feared that it
would be the death of Athens. Unlike Socrates, however, it would appear that Plato decided to fight
fire with fire. Instead of declaring rhetoric utterly worthless, Plato (in the Phaedrus) forges a place
for a reformed τέχνη of rhetoric. And, as I will argue here, he also makes a place for arational but
non-sophistical rhetorical persuasion in the Republic and the Laws.
Toward the end of the Apology, Socrates warns the Athenians that the fruit of his death for
Athens will be
a reckoning … far more unpleasant than the manner of my death. Critical questioners (οἱ
ἐλέγχοντες) will approach more of you, though I have heretofore held them back. And,
since they are young, they will make even more difficulties for you and make you even more
perturbed. (39c-d)
One can hardly help thinking that Plato had himself in mind, and his scathing condemnation of
Athens, in the cave image, surely was meant as a repudiation of such fools as would put to death a
man of incomparable wisdom. We can hardly doubt that Plato would have failed entirely in his
7
criticism, however, if he had not himself been a master of rhetorical persuasion – a method that
Socrates himself intentionally eschewed. This dissertation aims to consider what sort of thing a
purified rhetoric might be, and how such rhetoric might be used to educate the populace. Whether
such a departure from his own methods would impress or horrify the historical Socrates is anybody’s
guess.
8
CHAPTER 1: EDUCATION THROUGH NAMES IN THE CRATYLUS
One of the most striking aspects of Plato’s philosophy of language is his notion that words
can be true or false. In the Cratylus, Socrates proposes that the assignment of appropriate names (τὰ
προσήκοντα) results in “true” names, whereas the assignment of incorrect (μὴ ὀρθῶς) names
results in “false” names (431a-b; 430d). To the modern ear, this sounds like a category error. Truth
and falsity are properties of propositions, not of words. The notion that words admit of falsity
sounds at best eccentric and at worst nonsensical. One might hope, then, that Plato’s theory of the
falsity of words might be quarantined in a single isolated dialogue (the Cratylus), and then either
explained away or ignored. But the theory does not behave itself. Plato suggests that words are
capable of falsity in dialogues as diverse as the Euthydemus (277d-278c), the Phaedrus (244c), and the
Philebus (12b-c). Even Plato’s masterpiece, the Republic, includes the objectionable theory. Socrates
continually mentions the importance of using true or correct names in the establishment of his ideal
πολιτεία. (See e.g. Rep. 359a, 387b, 405d, 443e, 560d-e.) Moreover, in a passage that seems almost
Orwellian, Socrates declares that Athenian demagogues make the mistake of calling things by the
wrong names, names which accord to the opinions of the “great beast” which personifies the demos.
Thus the demagogues call the opinions of the demos “wisdom,” and call those things which please
the demos “good” (493a-c).
Despite these passages, scholars have not generally assigned any significant role to Plato’s
philosophy of language in the Republic. When the Republic talks about people using the “wrong
names” for things, these scholars claim that these people express false beliefs about genuine realities.
And scholars are certainly right about that. And yet they are missing one critical point. Given a
fuller understanding of Plato’s philosophy of language – both in the Cratylus and elsewhere – the
9
distinction between an “error in naming” and an “error in belief” disappears. Plato holds not only
the odd view that names can be true or false, but also the unusual view that the truth of these names
depends upon the beliefs of those who have control of the names (the “standard setters” or
νομοθέται).1 Thus, errors in belief are made manifest in errors in naming; and errors in naming
perpetuate false beliefs.
In this chapter, I will argue that “false names” are central to the way that Plato thinks about
language throughout his middle period, and therefore that the truth and falsity of names ought to
play a more integral role in our understanding of both early education and dialectic. I will focus
most of my attention on material from the Cratylus; but the dialogue I want to illuminate is the
Republic. One key to understanding the role of false names in the Republic, I argue, must lie in an
understanding of Plato’s theory (in the Cratylus) that names are “images” or “imitations” – a theory
that has been misunderstood by modern readers of Plato.2 Once we understand precisely what Plato
means by saying that a name is an “imitation,” Plato’s theory that names can be false becomes – far
from plainly false – reasonably plausible, at least in the context of the theory of Forms. Moreover,
Plato’s views about truth and falsity in naming bear a striking and unappreciated similarity to
linguistic theories advanced by contemporary scientific realists like Hilary Putnam and Ted Sider.
Given his assumptions about the world, Plato’s proposals about language are not only coherent; they
are also sensible.
I owe the proposed translation “convention setters,” for νομοθέται, to Imogen Smith. Imogen Smith, “False Names,
Demonstratives and the Refutation of Linguistic Naturalism in Plato's ‘Cratylus’ 427d1-431c3,” Phronesis 53, no. 2
(2008): 125 - 151.
2
Throughout this chapter, I will use “image” and “imitation” as synonyms, following Plato’s seemingly indiscriminate
uses of the words in dialogues like the Sophist and the Cratylus.
1
10
The chapter will proceed in the following way. In the first section, I will investigate two
passages about language from the Sophist and the Euthydemus, and argue that these two passages
support the claim that Plato’s philosophy of language was developed as an antidote to sophistical
rhetoric. In section II, I will explain how the Euthydemus passage, in conjunction with certain
comments in the Cratylus, suggests that Plato held a view of language largely in keeping with the
views of certain modern philosophers of language – and defensible on the same grounds as these
contemporary views. In Section III, I will investigate the nature of words in the Cratylus in an attempt
to vindicate my interpretation in Section II, and I will discuss how Platonic names are content-rich.
In sections IV, V and VI, I will argue against several subtly mistaken ways of interpreting the Cratylus,
and I will argue that names can be false only insofar as they somehow embed false information.
Harmful rhetoric exploits this sort of false information, a point which Plato drives home in his
strong criticism of “education through names” at the end of the Cratylus. In section VII, I will argue
that education through names is a target of Plato’s ire only in a limited sense: an education through
names is incapable of generating knowledge, but it is capable (when guided by the dialectician) of
generating true belief. Finally, in section VIII, I will suggest that we might understand primary
education in the Republic as a form of education through names, precisely because it is capable of
generating true belief.
I. How Language Might Approach Reality
Early in the Sophist, the Eleatic Stranger makes what may represent Plato’s most
straightforward comment on the relationship between language and truth:
Now you must follow me in a joint inquiry [into things], beginning … with the
Sophist, and see if we can make his nature evident through discussion (ἐμφανίζοντι
λόγῳ τί ποτ᾽ ἔστι). After all, at this point, you and I only coincide in this single
respect: that we use the same name for him. But as to the actual thing (ἔργον) to
which each of us attaches this name, our conception of it may very well be different.
11
And so, it’s crucial that we always come to an agreement through our discussion about
the thing itself (συνωμολογῆσθαι…διὰ λόγων…πέρι τὸ πρᾶγμα αὐτὸ) rather than
proceeding without any preliminary discussion, agreeing only on a name. (218b-c)
I believe that this passage is the key to understanding Plato’s philosophy of language. In the passage,
the Eleatic Stranger proposes that intellectual inquiry (dialectic) will be ineffectual unless one first
engages in an investigation into the language of inquiry. This investigation has two requirements:
(1) When studying the nature of something named “F,” interlocutors must have some sort of verbal
agreement (συνωμολογῆσθαι … διὰ λόγων) about the thing being named. They must have roughly
identical referential intentions; that is, they must intend to indicate the same thing (ἔργον).3 (2) This
thing to which they refer must be a genuinely existent thing (it must have a τί ποτ᾽ ἔστι4) with
extensional boundaries that generally conform to the referential intentions of the speaker.
I will call the first requirement the Conventional Requirement; I will call the second the
Naturalistic Requirement.
The Conventional Requirement guarantees that language is used
consistently by both parties; the Naturalistic Requirement guarantees that there is a norm governing
this consistent usage5 – that is, it guarantees that the shared “meaning” of the interlocutors has an
appropriate connection with an actually existing object. If we do not conform to these requirements
concerning names, our dialectical investigations will not lead us to knowledge.
The use of the word ἔργον, in the Greek, introduces a traditional Greek contrast between “words” (λόγοι) and
“facts” or “deeds” (ἔργα). Some helpful insights about Plato’s engagement with this contrast, and its relationship to
the activity of the sophists, can be found in H.G. Gadamer, “Logos and Ergon in Plato’s Lysis,” in Dialogue and
Dialectic: Eight Hermeneutical Studies on Plato, trans. P. Smith (Yale U.P., 1980), 1-20.
4
The paradigmatic cases of things with a fixed nature are the Forms, but it may be that other non-Forms have fixed
natures as well. The is a ποτ᾽ ἐστίν ascribed to bees in the Meno (72b), for example, though it is surely controversial
whether there is a Form of the Bee.
5
The sort of norm I appeal to here need not be a value-laden norm. The norm acts as an anchor for the conversation,
a sort of target (or “quarry,” Stat. 290d) which is aimed at. When investigating the nature of a “sophist,” it must be
the case that there genuinely is a distinctive nature that sophists have, a nature which is subject to investigation. To
say there is no such nature is equivalent to saying (in terms familiar from Meno 72c-d) that there is no one thing that
different kinds of sophists have in common; i.e., that all sophists are sophists in different senses of the term.
3
12
These two requirements have been recognized by various interpreters as aspects of Plato’s
philosophy of language, but both requirements – the Naturalistic Requirement, in particular -- have
not received the attention they deserve. The requirements are not confined to the Sophist; in fact,
they are given a far more detailed treatment in the Cratylus, and they are either stated or implied by
the Euthydemus, the Phaedrus, the Seventh Letter, and the Republic.
Although satisfying these
requirements does not guarantee successful intellectual inquiry, failing to satisfy the requirements
leads to a sort of deep-seated miscommunication, exactly the sort of miscommunication that is
prominently on display by various Athenian sophists.
A dramatic account of how these requirements might be violated occurs in the Euthydemus.
Midway through the dialogue, Socrates indicts the sophists Dionysodorus and Euthydemus on the
charge of confusing the boy Clinias with their misleading language. Clinias has been asked whether
it is the knowledgeable person or the ignorant person who “learns” (μανθάνουσιν, 276c), and he
finds that the sophists refute both options. Socrates, however, takes exception:
My dear Clinias, do not be astonished if the arguments seem strange to you. … These
are the first steps in the initiations of the “wise.” For it is necessary, first of all, as
Prodicus says, to learn the right use of words (ὀνομάτων ὀρθότητος); indeed, this
is what the two visitors are showing you. You did not realize that people use the
word “learn” with two different purposes (ἐπὶ τῷ τοιῷδε). They call it learning
when one begins by knowing nothing about a subject, and then later understands
it, yes, but they also use the same name to describe when someone who already
possesses knowledge uses the knowledge to analyze the same subject with respect to
some action or discussion. The second is called understanding rather than learning,
but from time to time (ἔστι δ᾽ ὅτε) it is called learning. But this is something you
have not realized, as these fellows demonstrate. They hold the same word as
applying to people in opposite conditions, to one who knows and one who does not.
… All this is just a little game about learning, and that’s why I say they are playing
with you; I call it a game, because if one learned many such examples or even all of
them, one would be no nearer to knowing the condition of the things themselves
(τὰ πράγματα), but would be able to play with people because of fine distinctions
in words. (Euthyd. 277d-278c)
13
Socrates is not satisfied with the form of “communication” practiced by Euthydemus and
Dionysodorus.
He not only insists that the sophists’ question must be answered by a
disambiguation; he also hints that the word “μανθάνειν” has an independent standard of
correctness, since he claims that the phenomenon where a person who possesses knowledge uses
this knowledge to examine a particular case would ordinarily (that is, normatively) be said to involve
“understanding” (συνιέναι), not “learning.”6
The insistence on disambiguation invokes the
Conventional Requirement, which insists that words be used unambiguously in intellectual
discussions.7 The suggestion of an independent standard of correctness for words, on the other
hand, involves the Naturalistic Requirement, since it suggests that there is a norm governing
individual words.
With characteristic irony, Socrates puts a generous interpretation on the
intentions of the sophists, assuming that they (like Prodicus8) are merely trying to make sure we use
words with their proper meanings before engaging in philosophical discourse. Obviously, they have
no such thing in mind.
6
The particular terms here are not important to the normative point Socrates are making: that words should be
precise and unambiguous. If things were reversed, so that the word “συνιέναι” usually denoted learning and the word
μανθάνειν denoted understanding, it would make no difference. What matters is that there are two things instead of
one: learning and understanding. An argument that elides the difference between the two is bound to be misleading.
7
In the Gorgias, Callicles accuses Socrates of a similar charge: exploiting the meaning of a word “by nature” with its
meaning “by convention” (φύσει and νόμῳ, respectively; 482c-484a), in order to win arguments. Indeed, Callicles
indicates that this sort of equivocation is popularly associated with orators and sophists, since he compares Socrates’s
behavior to the behavior of a “demagogue” (482c). Socrates, in response, does not deny that this is a common tactic of
demagogues, but he does deny that he himself is engaged in it. He insists that it is more shameful to do than to suffer
wrong, both by nature and by convention (489a-b); in other words, he argues that our conventional judgments
correspond to the natural truth of the matter. I think this move in the argument is significant, even if we take
Socrates’s larger dialectical response to Callicles – that conventional judgments are noble because of the nobility
(=power) of the multitudes – as a not-too-subtle example of Socrates drawing out the absurdity of his opponent’s
assumptions. Socrates seems to hold himself to the dialectical standard Callicles proposes: that a refutation based on
the interlocutor’s agreement with conventional beliefs is only a valid refutation if these conventional beliefs
correspond to reality. This standard is either a corollary of the Naturalistic Requirement, or a close relative to it.
8
Again, this could be ironic, since some evidence suggests that the historical Prodicus considered the proper use of
words a worthy end in itself, not a means to further ends. Thus Plato refers slightingly to Prodicus’s pedantry in
several dialogues (Prot. 341a-c, Lach. 197d, Charm. 163d), and in the Theaetetus, Socrates recommends that people who
are not mentally pregnant go to Prodicus instead of him (151b).
14
This Euthydemus passage is best explained in context of the Cratylus. Socrates here says that
“people use the word ‘learn’ with two different purposes” (ἐπὶ τῷ τοιῷδε) 9, and the notion of using
a given word with a given purpose – a “referential intention” – plays a central role in the Cratylus.
In the course of questioning Hermogenes early in the dialogue, Socrates proposes an analogy: just
as we discover what the proper shuttle for a particular object is by adapting the general notion of a
shuttle to a particular “type of work” (e.g. weaving a tapestry, see Crat. 389b-c), we discover what the
proper name is for an object (or concept) by adapting the general notion of naming to the particular
purpose of the user of the name (the dialectician). This purpose is to indicate whatever the
dialectician needs to indicate. Guided by the dialectician, the lawgiver
must know how to put the natural and appropriate name for each thing into sounds
and syllables, and to create and assign names with a view to the genuine name of the
thing (βλέποντα πρὸς αὐτὸ ἐκεῖνο ὃ ἔστιν ὄνομα) … [Though] different
legislators will not assign the same syllables, … the legislator gives the appropriate
form of the name (τὸ τοῦ ὀνόματος εἶδος ἀποδιδῷ τὸ προσῆκον) in whatever
syllables. (389d-390a)
There is nothing simple about this passage, which is filled with language Plato elsewhere uses to
discuss immaterial Forms, but the passage does clearly indicate that there are standards for naming.
By these standards, there ought to be one, and only one, proper name for every purpose – every
“meaning” – that the dialectician may express in the course of a single discussion.10 When we
juxtapose this passage with the Euthydemus, we can see that the two sophists run afoul of proper
Blackson renders “ἐπὶ τῷ τοιῷδε” as “in two different circumstances,” a translation which is at the very least forced.
The LSJ specifies that ἐπὶ + dative only pertains to circumstances when it is accompanied by verbs of rest, and καλέω
is not a verb of rest. A more natural reading of the passage interprets the word ἐπὶ as an indication of purpose on the
part of the speaker. Thomas A. Blackson, “Plato and the Senses of Words,” Journal of the History of Philosophy 29, no. 2
(1991): 169-182. Henry George Liddell et al. A Greek-English Lexicon. (Oxford: Clarendon Press, [1843, 1940] 1996).
10
Socrates immediately qualifies this point, however, by saying that the one proper name of an object can be
embodied in various combinations of syllables. So far as I can tell, this means something like the following: the
“form” of a name can be instantiated in different phonemes, just as the same person might be composed of various
groupings of atoms at various times.
9
15
dialectical naming precisely insofar as they allow that, in the course of a rational investigation, two
separate purposes may be advanced by the use of the same name.
II. Description and Reference in Modern Philosophy
In this section, I will consider a number of resonances between Plato’s theory of language
and its 20th century counterparts. Some translators have – perhaps somewhat anachronistically –
translated “with two different purposes” (ἐπὶ τῷ τοιῷδε) by the phrase “in two different senses.”11
Although it is unclear whether these translators had in mind a Fregean understanding of “sense,”
applying the modern linguistic category of “sense” to the Euthydemus passage is instructive, all the
more so in the context of the Cratylus passage. Two central aspects of Frege’s notion of a “sense,” as
I understand it, are (1) that a name has content, or an “associated description,” and (2) that this
associated description determines the referent of the name. If there is no object answering to the
description, the name lacks a referent. Now, on Frege’s presentation of it, the description in #1
need not be analyzed for its correctness, since, if it were incorrect, it would fail to refer – and thus,
per impossibile, fail to be incorrect. Frege’s theory is a bad place to look for any notion of “true” or
“false” names, then. On any theory that allows names to be false, the name would have to have
content without this content determining the referent of the name.
Various contemporary philosophers have suggested views along these lines. Ted Sider argues
that certain joint-carving natural properties and relations are “intrinsically eligible meanings” that
can be referred to by terms whose intentional content does not necessarily align with the thing being
referred to.12 Loosely, we might understand this to be the view that a word’s “sense” is not reference-
Jowett, for instance, translates the passage with the word “sense,” and Gifford describes the passage as suggesting
that “the same word can be used in two different senses” (3). Plato & E.H. Gifford, Euthydemus, trans. Benjamin
Jowett (Oxford: Clarendon Press, 1905).
12
Theodore Sider, Writing the Book of the World (Oxford: Oxford University Press, 2011), 27.
11
16
fixing, and that the sense may thereby inaccurately describe its referent. A certain subset of natural
properties are “reference magnets,” which attract the reference of terms that have identical (or nearly
identical) extensions as the property being referred to. Supposing that 17th century English speakers
were to use the name “maggot” with the understanding that maggots are a type of larvae that
spontaneously generates, one might use the theory of reference magnetism to claim that these
speakers successfully refer to maggots despite having a false or inadequate notion of what a maggot
is. Instead of saying (as Frege might have to) that these people are correctly conceptualizing
something that doesn’t happen to exist, Sider would argue that they are incorrectly conceptualizing
something that does exist.
This is precisely the sort of linguistic relationship that exists between our words and the
Forms, on Plato’s view. The word “justice” somehow “latches onto” the Form of Justice, even though
nearly every human speaker will have a notion of justice that inadequately describes that Form. I
will be calling cases where someone uses a word with false embedded content to successfully refer to
a Form cases of “descriptively defective reference” (DDR).13
Without descriptively defective
reference, our language would be entirely irrelevant to any investigation of the nature of the Forms,
for reasons related to Meno’s Paradox. If you cannot learn anything about a Form without referring
to it, and if you cannot refer to it without already having a true description of the Form, then you
cannot learn anything about a Form. Any plausible notion of Platonic learning must be predicated
on the ability to talk about the Forms in some inferior, defective way. (Rep. 505e, 510d-e; Phaedrus
249b-c, 250a-b).
13
I owe the suggestion of the term “descriptively defective reference” to Eric Hiddleston.
17
The notion of descriptively defective reference is not only helpful to Plato, however. I would
argue that DDR is a necessary feature of any linguistic theory upon which people are noncoincidentally able to refer to natural kinds which they do not yet understand. The Fregean view
that a descriptive sense fixes reference – and thus that senses are necessarily accurate or vacuous
(since they apply to nothing) – cannot explain how scientific progress is possible. If Thales thought
that magnetism was “the action of an animate soul upon metals,” then descriptivists cannot say that
Thales had a false view of magnetism; instead they must say that Thales did not have a view about
magnetism, at all. Scientific understanding of concepts does not progress; instead, new concepts are
constantly being forged in order to meet the demand of changing beliefs. And the prior concepts
were not false, since they were not about anything.14 If we lack understanding of some natural kind,
we will not even be able to refer to the kind, unless we simply stumble upon the right description of
it.
Hilary Putnam, like Sider, is an “externalist” about reference; he believes that the
relationship between a word and its referent (when such a referent exists) is not exclusively
determined by the speaker’s referential intention. Putnam’s understanding of reference, however,
is causal. I will give the briefest sketch of this view. There are two stages necessary to establish that
a name N refers to an object O in a given language: (1) N is “baptized” to refer to O, usually
ostensibly, but sometimes using a description15, (2) the name N is passed from link to link, with each
new learner of the word intending to refer to O. This sketch surely describes something far closer
See Putnam’s discussion on 196-200. Hilary Putnam, “Explanation and Reference,” in Conceptual Change, eds.
Glenn Pearce & Patrick Maynard (Dordrecht: D. Reidel, 1973), 196-214.
15
According to Kripke, who also explicates this view, “The description used is not synonymous with the name it
introduces but rather fixes its reference. … Usually a baptizer is acquainted in some sense with the object he names
and is able to name is ostensively.” Saul Kripke, Naming and Necessity (Harvard University Press, 1980), note 42.
14
18
to a “naïve” or “pretheoretic” view of language than Frege’s view; and thus it is not surprising that,
as I will argue, Plato holds something like it. The naïve view would just hold that words refer to
objects because someone once gave the object the name, and the name “stuck.”
Putnam adds to this view (among other things) that the name which “stuck” to the object
has certain connotations attached to it: he calls these connotations the name’s “stereotype.” 16 A
stereotype is the descriptive component of a word: it is a set of properties conventionally associated
with a given term, whether or not these properties are properties of the referent of the term; given a
DDR framework, a stereotype will be a description of the referent that may or may not be accurate.17
As Putnam explains,
The properties mentioned in the stereotype … are not being analytically predicated
of each member of the extension, or, indeed, of any members of the extension. It
is not analytic that all tigers have stripes, nor that some tigers have stripes; it is not
analytic that all lemons are yellow, nor that some lemons are yellow; it is not even
analytic that I tigers are animals or that lemons are fruits. The stereotype is
associated I with the word; it is not a necessary and sufficient condition for
membership in the corresponding class, nor even for being a normal member of
the corresponding class.18
It may be that the objects stereotypically assigned a given property (or Form) do not in fact instantiate
that property. An action might fall under Meno’s stereotypical notion of “virtue” without actually
participating in the Form of Virtue. There is, then, a radical separation between our beliefs about
16
Stereotypes, for Putnam, only exist for names of types, not for proper names of individuals. Proper names do not
have associated stereotypes; their relationship to their referent is unmediated by any intervening information.
17
When our scientific understanding of something improves, this is a matter of our correcting false stereotypes.
Suppose that we live in 1900, and we believe that an “atom” is an indivisible building block of matter. Then, through
the experiments of physicists, we discover that the things we were calling “atoms” are divisible. Did we just discover
that these things are not “atoms,” or did we discover something new and surprising about atoms? Putnam prefers the
latter explanation, since our semantics must account for scientific progress in a way that accepts new descriptions of
the old things. Although competent speakers of a language might not share any single “sense” attached to the word
“atom,” it would appear much clearer that competent speakers in 1900 would share a common stereotype of atoms – a
stereotype which happened to be false. A term’s connection to its referent is not fixed by its sense, but instead
explained causally; when an old friend tells me about his wife Sally, I can begin to use the term “Sally” as a referring
term, despite the fact that I know almost nothing about Sally
18
Putnam, “Explanation and Reference,” 204-205.
19
properties and the characteristics of the properties themselves, but this separation does not make us
incapable of actually referring to and investigating the properties (i.e. the Forms).
My purpose in this section has been to explain how a view like Plato’s view on language
might be plausible – at least plausible enough, with certain adjustments, to merit the attention of
serious philosophers of language.19 I earlier wrote that “Plato holds not only (a) the odd view that
names can be true or false, but also (b) the unusual view that the truth of these names depends upon
the beliefs of those who have control of the names.” The first of these views is hardly outlandish, as
it turns out, since it is shared by any linguistic approach that allows for descriptively defective
reference. The second view is an aspect of causal theories of language, like those advanced by authors
like Putnam. These causal theories compare favorably to the Cratylus, since they describe how names
can be “baptized” – presumably by a namegiver – and how the notions of the namegiver can be
inscribed on the resultant name.
As it turns out, then, Plato is not merely an externalist about reference; he also appears to
be a causal theorist about reference. On his view, namegivers “baptize” names with their own
(possibly inaccurate) stereotype of the thing, but the name somehow sticks to the thing whether or
not the stereotype is accurate (Crat. 431d-e). Insofar as Plato gives us any notion of why this name
sticks, it must be a causal one: the namegiver has some acquaintance with the thing named – most
likely, he has an acquaintance with particular objects that participate in the Form being referred to.
Anyone who has seen beautiful things, then, is capable of giving a name for beauty; but only the
Contrast Robinson, who says that “the most fundamental falsehood” of Plato’s theory of names is that “it assumes
that the business of a name is to describe its nominate, whereas it is merely to refer to that nominate” (133).
Robinson’s criticism, if valid, would apply equally to the theories of Sider and Putnam. R. Robinson, Essays in Greek
Philosophy (Oxford, 1969).
19
20
person who has “seen” the Form of Beauty will be able to give an accurate name – that is, an accurate
stereotypical description.
When Socrates, in the Euthydemus, says, “People use the word ‘learn’ with two different
purposes,” then, he is saying that there are two different stereotypical descriptions that people assign
to the phoneme “learn,” and that the word is ambiguous between these two descriptions. Either of
these descriptions could be false of the referent, or could fail to refer to any referent. Plato gives us
a distinction – though we might like it to be sharper distinction – between name-as-phoneme and
name-as-sense. Thus, in the Cratylus, we are told that the name “Hermogenes” can be used to refer
to Hermogenes, despite the fact that the content of the name “Hermogenes” is false to Hermogenes
(429b-431d): Hermogenes is “not a son of Hermes” (Herme-genes), since Hermes is the ultimate
artisan, whereas Hermogenes is “not a skillful artisan of words” (408b). By distinguishing phonemes
from “purposes,” Plato places a particular emphasis on the beliefs and associations which speakers
of the language attach to the referent they intend to indicate.
Individual names20 have content, then, which is implied by the way they are used. A name
is false when the beliefs embedded in that name are false; that is, when the referent of the name
does not have the relevant property. Formalized, we might call this the
Correspondence Theory of Reference: The name “X” is true if and only if
the descriptive content associated with X is true of the Form of X.21
That is, the “purposes” of the names. From this point in the chapter onward, I will simply use the term “name” to
refer to the purpose of the name, as opposed to the mere phoneme which carries that purpose.
21
Cf. Partee: “The worth of [individual names] depends on their conformity to the eternal Forms” (120). The
connection between names and Forms should be qualified, however, by the mention of possible stable non-Formal
realities in note #4. The distinction between Formal and non-Formal realities in Plato’s thought – if there is one –
does not play any major role in this dissertation. As a matter of convenience, I will simply use the term “Forms”
throughout the dissertation, without continually making reference to stable non-Forms. Morriss Henry Partee, “Plato's
Theory of Language,” Foundations of Language 8 no. 1 (1972): 113-132.
20
21
According to this theory, names have standards of correctness despite the fact that the connection
between a name and its referent is predominantly conventional (Crat. 435b).22 The theory does not,
however, insist that every name admit of truth and falsity, since one can imagine names that do not
refer to any Form (or any fixed reality). Thus, in the Cratylus, Socrates expresses concern that words
like κάλος, ἀγαθός, or ἐπιστήμη might be false, but he does not appear concerned about errors in
the case of objects which are “never in the same state” (439e), e.g., ordinary material objects.23
The fact that we attach certain content to the names of unchanging objects, like virtues or
mathematical entities, paves the way for miscommunication, since this content is potentially false.
At the same time, we must communicate – indeed, if we are to do philosophy, we must sometimes
communicate about realities at an extremely precise level (cf. Theat. 184c). One central focus of the
Cratylus, then, is to clarify the nature of sophistical errors in language use, the kind of errors
manifested in the Euthydemus passage, and to explain exactly what role language can have in the
education of a citizen. I now turn to the question of how Plato approaches this task.
III. Plato’s View of Names in the Cratylus
In the Cratylus, Socrates draws an explicit distinction between two types of education: (a)
education through names and (b) education through things. One can either learn about things
Thus Gail Fine says that in the Cratylus, Plato is “assessing the adequacy of conceptual schemes. He is asking what
natural kinds there are and how divisions of words should match them. … A name is correct, then, not just so far as it
picks out a natural kind, but only if it also reveals the nature of that kind, or correctly describes it” (297). I think that
Fine is wrong, however, to say that Plato “rejects any causal theory of names. [He] will not allow ‘n’ to be a name of x
just in case there is an appropriate causal connection between ‘n’ and x. … For ‘n’ to name x, it must reveal the tupos
or outline of the essence of x; that is to say, it must correctly describe x’s essence. … The correctness of names consists,
then, in their descriptive adequacy” (298). In this passage, Fine blurs the distinction between a name referring, and a
name referring correctly. Plato’s insistence that names can be false, on the contrary, relies on the notion that a name
can refer without referring correctly, i.e. without correctly describing the nature of its referent. Gail Fine, “Plato on
Naming,” Philosophical Quarterly 27 (1977): 290–301.
23
Cf. “It would be especially likely to discover correct names pertaining to things that always naturally exist” (ἀεὶ ὄντα
καὶ πεφυκότα, Crat. 397b).
22
22
“through [their] names” (δι᾽ ὀνομάτων) or “through the things themselves” (τὰ πράγματα … δι᾽
αὐτῶν, 439a).24 These two possibilities are mentioned toward the end of the dialogue, in the
context of Cratylus’s insistence that the best – indeed, the only – way to educate a person in reality
(τά ὄντα) is through the use of accurate names. As Socrates expresses the view, “whenever someone
knows what kind of name something has … he also knows the thing, since the thing is like its name”
(Crat. 435d-e). Socrates comes to reject this view, as we shall see.
Cratylus’s view about the efficacy of education through names is motivated by the notion
that a name (ὄνομα) is an “imitation” (μίμημα, 423b) or an “image” (or “likeness,” εἰκόνας, 439a)
of its referent, a view that the dialogue appears to take quite seriously. Cratylus himself believes that
there are natural standards for names, but his conventionalist counterpart, Hermogenes, also agrees
with the theory that names are imitations. The theory that a name is an imitation, indeed, is
consistently a central focus of Plato’s comments about language. In the Timaeus, speech is seen as
an “imitation of the unwavering ways of God” (μιμούμενοι τὰς τοῦ θεοῦ πάντως ἀπλανεῖς, 47c)
and words are said to be “akin to the matter which they describe” (29b).25 In the Critias, all language
is seen as a matter of imagery and representation (Crit. 107b). And in the Republic, falsehood in
words is an “imitation” (μίμημά) of the soul’s unfortunate state (382b-c).
What does Plato mean by saying that a name is the εἰκών, or a μίμημα, of its referent? A
long digression in the middle of the Cratylus seems to take the imagistic account to refer almost
exclusively to the way that the sounds and structure of names can be said to imitate the real nature
of the things they describe. As in the Sophist, a name represents in speech (ἔν λόγῳ) what the object
For an illuminating discussion of this distinction, see Christine J. Thomas, “Inquiry Without Names in Plato's
Cratylus,” Journal of the History of Philosophy 46 no. 3 (2008): 341-364.
25
The entirety of this passage compares favorably to Rep. 400b-402a, which discusses music and speech as oriented
toward the imitation of good and bad moral character.
24
23
is in reality (ἔν ἔργῳ). Socrates and Hermogenes expound upon hundreds of etymologies, which
have been the subject of much critical consternation, centered around the question of why Plato
would write such a prolonged investigation on premises that he later seems to reject.
This rejection comes late in the Cratylus (435a), when Socrates, perhaps drawing on the
discussion of being reminded by similars or dissimilars in the Phaedo (74a), declares that we are
capable of communicating (or “giving indications,” δήλωμα γίγνεται) by making an utterance
(φθέγγομαι) that is unlike the object it indicates. He draws from this point the conclusion that,
although “words should as far as possible resemble things … this dragging in of similarity
(ὁμοιότητος) is a stingy26 thing, which has to be supplemented with an odious reliance on
convention” (435c). But even with this conventionalist turn, Socrates does not reject the imitative
model of names.27 In practice, we resort to convention (συνθήκη) in order to communicate, but
the use of likenesses would, if practiced, lead to “the most perfect state of language” (435c).
When he discusses the subject of the ability of names to educate, Cratylus contends that this
ability stems from their similarity to the objects they denote (435e), and Socrates does not disagree.
Quite the contrary: he seems to assume that it is an a priori truth that like can be used to teach about
like (435e; cf. Rep. 490b, 611e; Tim. 90d). Instead, Socrates’s attack on education through names
starts when he observes that, even on an imitative theory of names, “learning names” need not be
understood as the same thing as “learning things.” Whoever first gave things their names did so
The Greek word here is γλίσχρα, which literally means “sticky” or “gluey.” Shorey (note on Rep. 488) notes that the
word, though being extremely hard to translate, is often used in contexts of straining after something – in this case,
after resemblance. Spellman translates the word “shabby,” in this context (200). See Bernard Williams, “Cratylus’
Theory of Names,” in Language and Logos, eds. Schofield and Nussbaum. (Cambridge, 1982) 93. Lynne Spellman,
“Naming and Knowing: The ‘Cratylus’ on Images,” History of Philosophy Quarterly 10 no. 3 (1993): 197-210. Plato, and
Paul Shorey, The Republic. (London: W. Heinemann, 1930).
27
Socrates rejects the dragging in of “ὁμοιότητος” (similarity), which seems to imply that names are not the same sort
of things (ὁμός + οἷος) as things, but not that names are not images of things.
26
24
under the impression that the things named had certain qualities (οἷα ἡγεῖτο εἶναι τὰ πράγματα),
and so he assigned these same qualities (τοιαῦτα) to the names (436b). But this person may have
been mistaken about the object’s qualities, a mistake that would infect the qualities of the name, as
well. Thus, the person who learns by names alone is subject to being deceived. The implication is
that a language that accurately created likenesses of the objects named – that embodied a “most
perfect state of language” – could be used to teach accurately (whether or not it provided ἐπιστήμη),
but that Greek is not such a language.
This shortcoming on the part of the Greek language explains Socrates’s complaint that words
like βέβαιος or ἐπιστήμη embed false and ambiguous information, at 437a-c. For the sake of
argument, Socrates concedes to Cratylus that all things are in flux, but he points out that the names
themselves suggest something different. When the namegiver instituted the name βέβαιος, he
clearly must have thought that things were sometimes stable, since the word contains the notion of
fixedness (βάσεώς, 437a), and similarly with ἐπιστήμη, which suggests station (ἵστησιν).28 The
problem is not that the namegiver failed to understand what the syllables be-bai-on (or ep-hi-ste-me)
represented onomatopoeically, but rather that the namegiver had false beliefs about the relevant
objects: the namegiver falsely believed that “stable” earthly objects (or human knowledge) were not
subject to Heraclitean flux. Socrates, of course, need not endorse this judgment, which is merely a
consequence of Cratylus’s view. But the point about naming stands regardless.
It is hard to know what to make of such a position. Clearly, part of Plato’s problem with
sophists like Euthydemus is that they use words deceptively, which seems to suggest that their use of
language is “imperfect.” The imperfection of their language, however, does not always appear to be
The point about “ambiguity” in 437a is made because earlier in the text (412a) the interlocutors had judged
ἐπιστήμη to suggest motion, not rest. Apparently, it suggests both, and thus it is, as a name, doubly flawed.
28
25
etymological; it is at least sometimes semantic. Euthydemus and Dionysodorus do not proceed in
their sophistical routine by using deceptive primary names or errant etymologies; no, their problem
is either that (a) they exploit the fact that the word “learn” can be used with two different meanings,
or that (b) they exploit contradictory beliefs that Greeks ordinarily have about the concept of
learning. Either of these problems was, plausibly, very difficult to express, given the philosophical
terminology Plato had available to him: the one because it involves what we call “senses” of words;
the other because it involves the notion of a “concept.” Note, however, the resonance between these
two errors and the Conventional and Naturalistic Requirements. The first explanation violates the
Conventional Requirement, since it involves using a single word (i.e. phoneme) ambiguously. The
second explanation violates the Naturalistic Requirement, because one cannot consistently hold that
an object like learning, which genuinely has a nature (τί ποτ᾽ ἔστι), also has contradictory properties.
On either explanation, the sophist manipulates words (ὀνόματα), and this manipulation of
words acts as a form of obfuscation in the educational process. 29 But mistaken primary names or
etymologies do not make any appearance in either of these explanations, and it may be hard to see
how the Cratylus’s attack on education through names could possibly apply to ordinary nonphilological education that went on in Athens. Since I intend to argue that education through
names was the strategy of ordinary sophists – people like Protagoras and Prodicus – I will have to
now turn my attention to the question of whether education through names was necessarily a matter
of etymology. I do not believe that it is, and that is what I intend to demonstrate in section IV.
The example of this process in the Euthydemus is not unusual among Plato’s dialogues for its use of deceptive or
ambiguous language, but rather for the self-consciousness of this deception, and for Socrates’s exposure of the
ambiguities involved – hence the fact that Socrates innocently assumes that the sophists are trying to make a
methodological point. Sophists in other dialogues are no less likely to exploit deceptive assumptions about words.
The shifting positions of Thrasymachus (in Book I of the Republic) concerning the nature of justice are a good example
of such procedures. See Rachel Barney, Names and Nature in Plato’s Cratylus. (Routledge, 2001), 11.
29
26
IV. Scholarly Interpretations of the Cratylus
The Cratylus is filled with fanciful and unrealistic etymologies. The word areté, for instance,
is traced to the words ἀεὶ ῥέον, since these mean “always flow,” and according to Socrates, “the flow
of the good soul is always free” (415d). The central puzzle behind the Cratylus is this: if Plato means
to say something serious about language in the Cratylus, why does he focus so much on the
exploration of contrived etymologies? I have no innovative response to that puzzle here, but I will
object to several unsatisfactory ways of solving the puzzle. The responses I have in mind are various:
they may involve the claim that the bookends of the dialogue are merely a platform for a curious
game of etymologies, or the claim that the dialogue is wholly pessimistic about language, or the claim
that the dialogue is a reductio on the notion that names are imitations of their referents.30 The
common thread in these readings seems to be this: that the Cratylus does not mean to establish any
genuine connection between naming and knowledge, not even a connection rooted in the notion
of imitation. All of the readings, then, either (a) deny that Plato ever meant to imply that the certain
words can be correct or incorrect, insofar as they are guided by the dialectician (despite 390d), or (b)
argue that the later portions of the dialogue revise this view, problematizing the notion of names
being false.
The first sort of view is held by scholars who argue that the Cratylus contains few, if any,
views that Plato actually took seriously. Fowler, for example, peremptorily dismisses the Cratylus:
30
Shane Ewegen, for example, has recently written a book denying that any of the straightforward conclusions drawn
by Socrates in the Cratylus are meant seriously. On Ewegen’s view, the Cratylus is a sort of comedy of errors, in which
the views of Hermogenes and Cratylus are held up as instructive and amusing examples of how not to interpret names.
Ewegen denies that Socrates genuinely believes there is any such thing as “education through things.” Instead, he
argues that a central focus of the Cratylus is to in fact disprove the hypothesis that names are images of their referent.
This alleged disproof would establish that names cannot be images of their referents under conventionalist or
naturalist assumptions about language; thus, under the plausible assumption that conventionalism and naturalism are
the only two options, names cannot be images of their referents tout court. S. M. Ewegen, Plato's Cratylus: The Comedy
of Language. (Bloomington: Indiana University Press, 2013).
27
“The Cratylus cannot be said to be of great importance in the development of the Platonic system,
as it treats of a special subject somewhat apart from general philosophic theory.”31 Ewegen presents
an argument for a similar conclusion. According to him, the idea that the dialectician supervises
the creation of words (390c) has the seriously undesirable consequence that it implies some sort of
knowledge that exists independently from language.32 For Ewegen, however, “education through
things” is a ridiculous notion that Plato could not possibly hold. If knowledge is propositional, then
you cannot have knowledge without words. Since the namegiver did not have preexisting words,
the namegiver did not have preexisting knowledge (extralinguistic access to reality).
Ewegen’s argument fails because Plato does not have a straightforwardly propositional
understanding of knowledge.33 Plato does not assume that knowledge must be capable of being put
into words34, and he certainly does not assume that knowledge can be put into words without some
loss in fidelity. To the contrary, it is Plato who famously claimed that written words can do no more
31
Plato, and H. N. Fowler, Plato: Cratylus, Parmenides, Greater Hippias, Lesser Hippias, Loeb Classical Library (Cambridge
Mass.: Harvard University Press, 1926), 4.
32
Moreover, the comments about the limitations of language later in the dialogue, if taken seriously, would mean that
not even the dialectician could put knowledge into words or to express it. As Ewegen complains, “One would
condemn oneself to silence, unable to bring the look of a thing into speech: for the moment that one attempted to
speak of the thing one would submit it to the very hazards of human λόγος that one was seeking to evade” (173).
33
Gulley, for example, argues that Plato’s conception of δόξα is propositional, but his conception of episteme is not.
In a similar vein, Gonzalez defends a traditional interpretation of the objects of knowledge in the Republic against the
views of Gail Fine, arguing that episteme in the Republic functions on a sort of acquaintance model, whereby the objects
of knowledge are the Forms themselves, not propositions about the Forms. In this dissertation, I will not specifically
address Fine’s arguments against the acquaintance model of Platonic knowledge, but I will be largely operating in the
tradition that embraces the traditional interpretation against hers. Many of the things I say could be translated into
the language of a veridical reading of epistemology in the Republic; some of the things I say could not. For a good
overview of the controversy between Fine and her critics, see Allan Silverman’s entry on “Plato's Middle Period
Metaphysics and Epistemology” in the Stanford Encyclopedia. Norman Gulley, Plato's Theory of Knowledge
(Greenwood Press, 1962). Gail Fine, “Knowledge and Belief in Republic V,” Archiv für Geschichte der Philosophie 60 no.
2 (1978): 121-139. Francisco J. Gonzalez, “Nonpropositional Knowledge in Plato,” Apeiron 31 no.3 (1998): 235-284.
Allan Silverman, “Plato's Middle Period Metaphysics and Epistemology,” The Stanford Encyclopedia of Philosophy (Fall
2014 Edition), ed. Edward N. Zalta, http://plato.stanford.edu/archives/fall2014/entries/plato-metaphysics/.
34
Indeed, one plausible reading of the Theaetetus is that the dialogue is meant to prove that knowledge cannot be put
into words, by showing that every attempt to make knowledge propositional ultimately fails. Cf. Alexander Nehamas,
“Episteme and Logos in Plato's Later Thought,” Archiv für Geschichte der Philosophie 66 no. 1 (1984): 11-36.
28
than “remind one who knows” (Phaedr. 275d)35 – the same Plato who wrote in the Symposium that
διαλέγεσθαι (dialectic) points the way to “one single form of knowledge” (τινὰ ἐπιστήμην μίαν
τοιαύτην), a “wondrous vision,” which will be “neither some sentence (τις λόγος) nor some piece
of knowledge (τις ἐπιστήμη, Sym. 210d-211a).36 Clearly this wondrous vision cannot be put into
words. Thus the author of the Seventh Letter (Plato or one of his early followers) says that names are
not “stable” (βέβαιον, 343b), and that their instability makes any “reality that is expressed in words
or illustrated in objects liable to easy refutation by the evidence of the senses” (343c). The
comparison between words and illustrations is, if anything, profoundly suggestive of the
identification of words with images/imitations, as is the insistence that failures of words are
analogous to failures in illustration.37
A second group of authors acknowledge that the Cratylus aims at a serious result, but
nevertheless argue that the dialogue is ultimately pessimistic about the hope of finding correct
names. So, for example, Susan Levin contends that, although Plato had “serious intentions” in
composing the Cratylus, “these intentions do not result in the elaboration … of substantive positive
proposals. Instead, following a protracted investigation, Plato rejects approaches to the notions of
naturalness and appropriateness that are based on the constituency of onomata.”38 These readings,
Shortly after this passage in the Phaedrus, Socrates does indeed refer to a type of “living language” (λόγον … ζῶντα,
276a) that “accompanies” knowledge, but this living language is never said to be expressible into ordinary words,
whether spoken or written. Rather, the living language is “inscribed on the student’s soul” (276a); it is identified with
the ability (δυνατὸς) to defend one’s position, and to know when to speak and when to remain silent. The
implication here seems to be that a person who knows how to speak the living language necessarily possesses the τέχνη
of dialectic.
36
At 211a, Socrates discusses individual pieces of knowledge (τις ἐπιστήμη) as things clearly separable and inferior to
knowledge-as-such. Thus, the “single form of knowledge” referred to here cannot be identified with τις ἐπιστήμη. Cf.
Letter VII 342a-343e, where Plato implicitly makes the same distinction.
37
See Crat. 432a-c, where Socrates argues that no illustration (or statue) of Cratylus could be entirely accurate, since
the illustration would then have to have all the characteristics of Cratylus – in which case there would be two
Cratyluses, and no illustrations.
38
S.B. Levin, The Ancient Quarrel between Philosophy and Poetry Revisited. Plato and the Literary Tradition (Oxford: Oxford
University Press, 2001), 81.
35
29
in general, take the etymological portion of the dialogue very seriously, and think that the defeatism
about etymological accuracy at 435a-c indicates a thoroughgoing Platonic pessimism about the
correctness of names.39
V. Interpreters on Names and Education in the Cratylus
Rachel Barney is an outstanding example of this latter group. 40 An analysis of her reading
of the Cratylus will help to clarify the issues at stake in my own arguments, and – I hope –
demonstrate how one might both hold the view that education through names is generally deceptive
without also holding that names are merely conventional entities.
Barney says a number of useful things about the notion of “education through names,” a
process which she delightfully dubs “onomastikê.” Barney considers the scope of onomastikê to
correspond to the scope of rhetoric: Since “the application of primary names, complex names and
even logoi are all part of the craft of verbal ‘painting’, … the onomastic craft is thus also identical with
rhetoric properly understood” (99). She interprets Socrates’ naturalistic criticism of Hermogenes as
a hopeful stage of the Cratylus, during which Socrates entertains an idea he will later reject: that
names might reliably “[disclose] the natures of the things they name through a particular semantic
content” (97). The semantic content is possessed by a name “by virtue of its mimetic content,” and
39
Thus, Timothy Baxter describes the Cratylus as a criticism of sophistical ideas of naming, in which Socrates supports
no positive theses: “Greek thinkers and poets had consistently put too much trust in names, implicitly regarding them
as surrogates for the things they names. Such trust could only make sense when dealing with an ideal language, which
Greek was certainly not.” Baxter’s analysis makes a good deal of sense, but fails, I think, to explain why Socrates never
repudiates certain aspects within the sophistical repertoire: for instance, the claim that names are imitations, or the
claim that names admit of “correctness.” One can agree with Baxter that Plato engages in parody without believing
that the people he parodies (e.g. Prodicus) were wholly confused. Timothy Baxter, The Cratylus: Plato’s Critique of
Naming (Leiden: E.J. Brill, 1992), 184.
40
Barney, Names and Nature. David Sedley, like Barney, argues that the etymologies are of central importance to the
Cratylus, but Sedley thinks that Plato is much more optimistic about the capacity that etymologies have to successfully
describe the objects to which they refer. D Sedley, Plato's Cratylus (Cambridge: Cambridge University Press, 2003).
30
the semantic disclosure is accurate if “there is a natural relation of resemblance between the letters
and the nature of the object named” (98).
What Barney is pointing out here is a version of what I have called the Naturalistic
Requirement, the requirement that certain types of names (especially the names of Forms) link up
with the things they describe in a normatively adequate way. On Barney’s view, however, Plato
considers the mimetic content of a word to be present exclusively in its etymological structure.
Although there are multiple etymological structures that can all capture the essence of a single given
object (different words can correctly name a “tree”; 389d-390a, 394a-b), the nature of the object
named constrains the etymological structure of any name. Any kind of scissors will work to cut
paper, but an apple won’t work at all.41 The correctness of a name is wholly connected to its
etymology. If we apply this account to the final sections of the Cratylus, we will find that the inability
of onomastikê to educate is rooted in its etymological deficiencies. On Barney’s view, then, naturalism
in language – even though Socrates praises it as a desideratum (435c) – ultimately plays no
meaningful role, certainly not a role in education. Although earlier, in her analysis of the hopeful
section of the text, Barney had suggested that onomastikê could be a “subsidiary of philosophical
dialectic” and that the giver of names, guided by the dialectician, might create “a language strictly
governed by a philosophical understanding of the natures of things” (101-102), later in the dialogue
Barney takes Socrates to reject all such optimistic fervor. “Since names as such turn out to be images,
no … strong distinction between correct and incorrect [names], between the real thing and the
41
Thus, Barney connects the etymological section very closely with the Tool Analogy at 388a-390e. For an interesting
survey of reasons why the Tool Analogy is meant quite earnestly in the Cratylus, see Imogen Smith, “Taking the Tool
Analogy Seriously: Forms and Naming in the Cratylus,” The Cambridge Classical Journal 60 (2014), 75-99.
31
imitation, can properly be drawn” (139). In the Cratylus, “a perfectly correct language is impossible,
and a partially correct language, therefore, of no particular use” (141).
This formidable interpretation of the Cratylus relies on the premise that the sole criterion of
correctness of names is etymological accuracy. There is evidence, however, that Socrates’s position
on this premise changes in the latter portion of the dialogue.42 Consider, for example, the nature
of the namegiver’s error at 436b-c.
Soc. Why, clearly he who first gave names gave them according to his conception of
the things which they signified – did he not?
Crat. True.
Soc. And if his conception was erroneous (μὴ ὀρθῶς ἡγεῖτο43), and he gave names
according to his conception, what do you think will happen to us, when we follow
him? Shall we not be deceived by him?
The error manifests itself as an error in primary names and etymology, but the actual error is an error
in belief: the namegiver has false beliefs about the object named. The namegiver must “know” the
things he names (438a), and a namegiver who fails in this respect is a bad “craftsman of names”
(431e). The problem is not that the bad namegiver does not know how to give names correctly, nor
is it that etymologies are useless; rather, it is that the namegiver does not properly conceptualize the
object he is naming. As Putnam would have it, the namegiver does not have an accurate stereotype,
and so he crafts a name that enshrines his misunderstanding of the object named. He gives the
wrong name to an object because he does not recognize what type of object it really is.
42
Spellman has a sensible argument for this point. Observing that words are portrayed as images throughout the
dialogue, she nevertheless argues that Socrates rejects the notion that this notion of a word-as-image can be cashed out
in terms of etymology. Therefore, “if the dialogue is not to be dismissed as just silliness, one must find another way
for language to be an image.” Spellman, “Naming and Knowing,” 200.
43
I translate the verb ἡγεῖτο with the noun “conception.” I do this, not out of any desire to overload the Greek with
the notion of a “concept,” but merely to indicate the set of beliefs that one has about an object. I will use the word
“conception” in that way throughout the dissertation.
32
It is plausible, then, as Jeffrey Gold has argued, that the word ὀνόματα is used ambiguously
in the Cratylus.44 On the one hand, an ὄνομα could be a mere sound: a sensory object subject to
the ordinary limitations of sensory objects, though (unlike most other objects) used symbolically.
On the other hand, however, an ὄνομα is a semantic unit that plays a certain role in communication;
Gold suggests that this meaning of ὄνομα roughly corresponds to the modern notion of a “word
type.” By this latter understanding of names, the English word “hat” would be literally the same
ὄνομα as the French word “chapeau.”45
The important thing to notice about the move Gold is making, in introducing types, is that
types are content-rich.46 Whereas a mere sound can only refer to a thing, a type plausibly carves out a
semantic space, such that tokens are tokens of that particular type only if they occupy that space.
Not every curved thing is a ball, but any competent speaker of the language would immediately know
that a non-curved object could not be a ball. Thus, in the Cratylus, the namegiver individuates names
on the basis of their “force” (δύναμις, 394a-b), where the force of a word is clearly something
content-rich, but at the same time something distinct from the referent of the word. 47
44
Gold’s
I shall leave it to the reader to determine whether Plato egregiously and ironically violates the Conventional
Requirement, in using the word “name” ambiguously. For more on Gold’s view, see Jeffery B. Gold, “The Ambiguity
of ‘Name’ in Plato’s Cratylus,” Philosophical Studies 34 no. 3 (1978): 223-251.
45
Cf. Wittgenstein: “One could say that the real name of an object was what all symbols that signified it had in
common. Thus, one by one, all kinds of composition would prove to be unessential to a name.” Ludwig Wittgenstein,
Tractatus Logico-Philosophicus: German and English Edition, trans. C.K. Ogden (Routledge, 1981), 3.3411.
46
As will become clear from my analysis, I certainly do not believe that scholars should understand the Forms
themselves to be types or concepts. Plato’s theory of Forms posits that the joints of reality are not subject to human
conventions (Stat. 262b, Crat. 438d). Since types are plausibly formed along at least partially conventional lines, types
cannot plausibly be identified with Forms.
47
The introduction of the term “δύναμις” is a response to the question of how names could have correctness
(ὀρθότης, 391b), a question which Socrates previously answered (at least in part) by an appeal to the notion of a
“form of the name” (τὸ τοῦ ὀνόματος εἶδος) at 390a (compare “αὐτὸ ἐκεῖνο ὃ ἔστιν ὄνομα” at 389d) and to the
“true natural names of each thing” (τὸ ἑκάστῳ φύσει πεφυκὸς ὄνομα) referred to at 389d. For the purposes of this
chapter, I will not enter into the controversy about what Socrates means by these phrases. Whatever these phrases
mean, however, it is clear that they are not meant to indicate mere sounds, and it is clear that they are evaluative in
nature. According to 389-390, the namegiver must be able to somehow embody these “forms” or “natures” in
ordinary words. This task of embodiment is clearly the same task Socrates discusses with Cratylus at the end of the
dialogue, though the end of the dialogue emphasizes that the task is prone to error.
33
approach is further exemplified by J.L. Ackrill, in a paper called “Language and Reality in Plato’s
Cratylus.”48 As Ackrill puts it:
We might call the name identified by the idea that it expresses the ideal name (as
opposed to the ordinary name).49 An ordinary name is made of particular sounds or
letters; but the ideal name is a semantically defined unit not made of sounds or
letters. It is in effect the meaning of all the ordinary names in a group of synonyms:
and one might well call it the name-as-concept, in that what synonyms all express is
the same concept.50
Ackrill emphasizes the roles that names with false conceptual content play in Socrates’s refutation
of Cratylus, but both Ackrill and Gold use Socrates’s refutation of Hermogenes to ground their
notion of a “name-as-concept.” When the later section of the Cratylus discusses names as being
semantically deceptive in the hands of sophists, then, the earlier section of the Cratylus has already
set in place a framework from which to evaluate a word’s correctness on the basis of semantics, not
merely on the basis of etymology.51 Indeed, we see this framework playing at least one very concrete
role in Socrates’s discussion with Cratylus: the name “man” is said to be incorrect in cases where it
assigns a “likeness of a woman” to a “man,” despite the fact that we are not told that any etymological
error is being made; the error is a purely semantic one (430b-d). “Man” is a concept-rich name and,
as such, it is not the proper tool with which to indicate a woman.52
Ackrill’s interpretation is certainly plausible, since it does not assign to the character Cratylus the perfectly bizarre
view that the only way to learn anything is by doing etymology. J.L. Ackrill, “Language and Reality in Plato’s Cratylus,”
reprinted in J. L. Ackrill, Essays on Plato and Aristotle (New York: Oxford University Press, 1997): 33-52.
49
Compare Crat. 390a.
50
Ackrill, “Language and Reality,” 44.
51
It is not clear, of course, that Plato realizes the extent to which he is equivocating on the word “ὄνομα.” Thus Evans
comments that Plato “does not recognize any fundamental distinction between the linguistic act of applying a name to
a thing and the mental act of applying a concept to a thing” (323). Matthew Evans, “Plato on the Norms of Speech
and Thought,” Phronesis 56 no.4 (2011): 323.
52
Barney points out that Hermogenes’s claim, early in the Cratylus, that the names “horse” and “man” are
interchangeable “is hardly plausible if [these] ‘names’ are taken to include the senses or meanings which, we normally
think, should guide their use.” She calls these senses “semantic content.” Now the fact that “man” and “woman” are
not interchangeable, according to Socrates’s later statement, indicates that he believes genuine names in a language
have semantic content. Barney thinks this semantic content is undermined by the weakened role etymology plays later
in the dialogue, but nothing in the dialogue indicates any such thing. The belief that a name has semantic content is
completely independent from the belief that this semantic content is exclusively onomatopoeic and compositional. I
48
34
Although the syllables assigned to various conceptual ὀνόματα may be merely conventional,
the content of the name can be correct or incorrect based on very precise standards: the Forms. This
is why Kretzmann, in his highly influential paper on the Cratylus, takes Plato to be presenting an
“objective taxonomic correctness” of names53 – what I am calling the Naturalistic Requirement.
Although Barney is quite right to emphasize that names are sensory imitations of the things they
refer to, she is mistaken to suggest that Socrates primarily measures their correctness
compositionally, on the basis of their etymological structure.
A fuller treatment of Barney’s book is not possible here, but it is worth mentioning one
commitment my interpretation has, in contrast to Barney’s interpretation. Early in the Cratylus,
Socrates suggests to Hermogenes, in the course of the “tool argument,” that conceptual names are
in principle distinguishable on the basis of their force (δύναμις). Barney argues that the later
portion of the Cratylus rejects this standard for linguistic individuation of names, and replaces it
wholesale with the notion of names as images which mimic their referents compositionally.
According to my interpretation, however, Plato maintains the same basic (conceptual) model of
names as images throughout the dialogue, despite an abortive attempt to add a further level of
naturalism through the use of onomatopoeic imitation.
VI. Semantics in the Cratylus
The introduction of conceptual ὀνόματα introduces a new framework through which we
might understand the action of the namegiver at 436b-d. Prior to naming, the namegiver has certain
beliefs about a particular object, beliefs that could be true or false. These beliefs guide the namegiver
think Socrates maintains the first view (consistent with his responses to Hermogenes), but eventually rejects the second
view. Barney, Names and Nature, 7.
53
Norman Kretzmann, “Plato on the Correctness of Names,” American Philosophical Quarterly 8 no. 2 (1971): 128.
35
in crafting a (conceptual) ὄνομα for the object54; false beliefs about an object lead to the creation of
a deceptive ὄνομα. This ὄνομα is deceptive both etymologically and conceptually. The result is
that sophists need not use misleading etymologies to deceive; they can also use equivocation, or they
can exploit culturally false beliefs about words.
Once we combine the notion that a name could be a “type” or a “concept” with a
pretheoretic, causal view of linguistic reference, we end up with a view almost indistinguishable from
Putnam’s view, which I described above. Each name (i.e., name type) is packed with information,
containing a certain stereotype about its referent. The relationship between the name and its
referent is historically fixed by an initial baptism. The referent of the name does not change
depending on what sorts of things the stereotype actually describes: “Pythagoras” will still refer to
Pythagoras, even if the name “Pythagoras” embeds information that is false of Pythagoras and true
of Zeno. The information contained in the stereotype will be true if it accurately describes the
referent, and false otherwise.
The picture that emerges from my analysis of the Cratylus is this, then: though it is desirable
that words display the nature of their referents through etymology, they may also capture this nature
by “packing in” semantic stereotypes which correspond to it. Names become content-rich. The
namegiver should be faulted not only when he uses syllables that do not resemble their referent, but
also when his names contain false semantic content.55 To use an example: the word δικαιοσύνη
(justice), as it is used in the Greek language, carries with it certain information: that the thing it
Thus Spellman says that “what is put into the matter which will become words are correct beliefs about Forms.”
Spellman, “Naming and Knowing,” 203.
55
Thus, the Eleatic Stranger in the Sophist identifies the sophist as an imitator, and says that the sophist is able to
“create with his pencil representations bearing the same name as real things, [and thus] deceive the innocent minds of
children” (234b).
54
36
indicates is a virtue, for instance.
Without knowing what justice essentially is, one might
nevertheless know that – provided one accepts the Greek stereotype of justice – one is committed to
the belief anything that is not a virtue is not justice. Indeed, this is precisely what Callicles accuses
Socrates of doing in the course of an elenchus: harnessing stereotypical meanings of terms to his
advantage in producing a contradiction, whether or not he genuinely endorses those stereotypical
meanings. (The stereotypical meaning is “by convention,” νόμῳ; the genuine meaning is “by
nature,” φύσει, Gorg. 482c-483d).
What I have written here is, admittedly, only a sketch, but it provides us with the outlines of
a Platonic analysis of how language can be false or misleading at the level of words. Barney believes
that the link between a name and its referent is a matter of etymology: a matter of the syllables of a
word (sensory images) imitating the object named. But the notion that names have attached
stereotypes opens the way to a much broader understanding of the Naturalistic Requirement. On
this latter understanding, there ought to be an intimate semantic correspondence between a concept
(an abstract image, cf. “ideal name” at 389d) and the thing conceptualized; and indeed, this is a
connection that actually exists between content-rich concepts like “man” and individual men.
Moreover, a person could make false judgments about this connection, as when a student mistakenly
claims that hydrogen is inflammable. Once we recognize this, we are better positioned to understand
how Plato’s critique of language in the Cratylus opens the door to understanding the role of the
sophists in what we might call “miseducation.” And it is to education and its lack to which I now
turn.
VII. Education Through Names and Education Through Things
37
At this point in the chapter, we should be well situated to take a fresh look at the concepts
of “education through names” and “education through things,” as they are contrasted in the last few
pages of the Cratylus. Education through names relies on the accuracy of imitative names – that is,
the accuracy of linguistic stereotypes – to transmit to the student accurate beliefs of the objects which
are imitated by the names (Crat. 435d-e). Education through things, by contrast, is a mode of
investigation that involves looking at the “existing kinships” between objects, as well as directly
examining the “objects themselves” (αὐτὰ δι᾽ αὑτῶν, 438e). Any successful engagement with
education through names presupposes a prior successful engagement with education through things,
since names will be accurate when bestowed by someone knowledgeable, and otherwise inaccurate.
The individual who knows things in themselves should clearly be identified with the person who
“employs” (χρήσεται) names, described earlier in the dialogue at 390d, the “dialectician”
(διαλεκτικόν).
Education through names relies upon content-rich images, and this is its problem, since
images can be inaccurate. The same very point is made about the educational aspirations of the
lovers of sights and sounds in Book V of the Republic: they consider beauty (τὸ κᾶλον) to be that
which it resembles (αὐτὸ ἡγῆται εἶναι ᾧ ἔοικεν, 476c), and thus their opinions are fallible (μὴ
ἀναμαρτήτῳ, 477e).56 The culturally influenced stereotype that I have of “electron,” for example,
would not be a perfect guide to scientific investigation: I should study the thing itself, and not assume
that my idea of the thing is (prior to investigation) fully accurate. In Book X, the poet is compared
56
I am certainly not saying that Plato is, in Book V, making a point about language and naming. What I am saying is
that the fact that images (including names) represent things imperfectly is a motif in Plato’s thought that connects a
number of different types of people, including whoever Plato intends to indicate by the “lovers of sights and sounds.”
Their notion that beauty could be known wholly through its instances is similar to Cratylus’s notion that objects can
be known through their names.
38
to a painter, and is said to use words and phrases to craft colorful images of arts like shoemaking or
warfare so that “others of a similar mind, who see through words (ἐκ τῶν λόγων θεωροῦσι), will
praise him exceedingly” (601a). The following warning at Crat. 440c, then, may very well apply not
only to the itinerant etymologist like Cratylus, but also to the lover of sights and sounds, the painter,
the poet, and (no doubt) the sophist: “No intelligent person (νοῦν ἔχοντος ἀνθρώπου) will put
himself or the nourishment of his soul under the guardianship (ἐπιτρέψαντα) of names.”57
Education through things, in contrast, approaches the objects of knowledge themselves, in
some sense unmediated by words. Socrates picks out two qualities of education through things: (1)
Things are known through themselves (αὐτὰ δι᾽ αὑτῶν, 438e), and (2) Things are known through
any similarities they have with each other (δι᾽ ἀλλήλων γε, εἴ πῃ συγγενῆ ἐστιν, 438e). For the
purposes of this chapter, it will not be necessary to understand with any precision what Socrates
means by these phrases, except to say that these two activities are clearly the types of things that
require the capacity to measure and engage in mathematical reasoning – two capacities that are never
mentioned in the context of the study of names. Both in Socrates’s insistence that the objects of
knowledge transcend imagery and in his insistence that they be encountered “in themselves,” his
discussion of education through names strikingly recalls the culmination of Plato’s description of
the Divided Line, in the Republic, where Socrates says that the highest section of the Line refers to
“that which reason itself grasps by the power of dialectic … entirely spurning the use of the senses,
Notice that it is the person with νοῦς that refuses to nourish himself with names. This is not an indication that
education through names is useless (if properly administered with guidance from the dialectician), but rather that
education through names is not the type of thing the philosophical nature will pursue.
57
39
but only using Forms in themselves, through themselves, unto themselves, and resulting in Forms”
(εἴδεσιν αὐτοῖς δι᾽ αὐτῶν εἰς αὐτά, καὶ τελευτᾷ εἰς εἴδη, 511b-c).58
I suspect that the nature of education through things might come into a sharper contrast,
however, if we merely consider what it is not. It is not an inquiry into words, neither in the sense
of being an etymological investigation nor in the sense of being an analysis of our stereotypes (or
perhaps “folk concepts”) of various words. When a person inquires into the nature of things, she
does not take the semantic qualities of words as her starting point. In this way, her actions are
fundamentally at odds with the common practice of linguistic/conceptual analysis, as it has been
practiced by various philosophers throughout the ages, including many sophists and perhaps even
the historical Socrates.59
Peter Ludlow offers a remarkably clear thinking statement of this practice, in his discussion
of contextual theories of epistemology: he says that “any investigation into the nature of knowledge
which did not conform to some significant degree with the semantics of the term ‘knows’ would
simply be missing the point.”60 This claim, which strikes Ludlow as “obvious,” is expressly denied
in the Cratylus. The ordinary semantics of “knowledge” could, according to Plato, be dramatically
disconnected with the nature of the genuine referent of the word. The person who pursues
The passage at Phaedrus 249b, which refers to “speaking according to Forms” (κατ᾽ εἶδος λεγόμενον), is especially
evocative here, though its meaning is the subject of much dispute, a dispute which I need not enter into, for the
purposes of this chapter. Hackforth translates 249b-c in the following way: “Only the soul that has beheld truth may
enter into this our human form – seeing that man must needs understand the language of forms, passing from a
plurality of perceptions to a unity recognized by reasoning.” On such a translation, it would appear that Plato
explicitly acknowledges the sort of hypothetical perfect language suggested by the conclusion of the Cratylus: a language
in which there would be a 1-to-1 function between the Forms and the names in the language. Plato, and R.
Hackforth, Plato: Phaedrus (Cambridge: Cambidge University Press, 1952).
59
My understanding of education through names, here, is largely in keeping with that of Christine Thomas. When
attempts at education through names are not focused on etymologies, Thomas describes them as “inquiries focused
eventually on natures or essences but which take linguistic items as their starting points.” That sort of characterization
makes them sound strikingly like linguistic or conceptual analyses. Thomas, “Inquiry Without Names,” 346.
60
Peter Ludlow, “Contextualism and the new linguistic turn in epistemology,” in Contextualism in Philosophy:
Knowledge, Meaning, and Truth, eds. Gerhard Preyer and Georg Peter (Oxford University Press, 2005), 13.
58
40
education through names, in the context of epistemology, runs the risk of being duped by the
vagaries of human belief, and thus never developing acquiring a genuine understanding of
knowledge at all.61 And this sort of error would seem to plague any investigation into some ultimate
truth that started from our conception of (or our intuitions about) the proper uses of words.
VIII. Toward an Understanding of Language in Plato’s Republic
It would appear, then, that Plato denies that education through names plays a central or
defining role in a proper education. But then, why write a dialogue about it? Does my interpretation
involve just as much of a pessimistic reading of the Cratylus as other interpretations, despite my
insistence that Plato maintains the dialogue to advance a number of positive theses? I don’t think
so: for I believe that I can explain why these positive theses are important in the context of Plato’s
middle period as a whole, in particular why they have important ramifications for the Republic.
The Cratylus is at pains to demonstrate that our names could be false. But when guided by
the “user” (the “person who knows,” the “dialectician”), these names are true. It is possible for one
person (say, Aristotle) to investigate things, and subsequently bestow accurate names upon them,
whereupon another person (say, Alexander) might come along and become informed about the
world merely by learning the nature of Aristotle’s names, without ever personally engaging with
objects in their own right. Aristotle’s way of learning is proclaimed by Socrates to be “finer and
61
Cf. Thomas: The problem with relying on names in inquiry is not that one is necessarily misled. Presumably those
engaged in inquiry with names can both acquire new concepts and revise old ones in such a way as to make epistemic
progress. But there is risk involved. If one takes names as a guide one might be deceived into embracing a thoroughly
false picture of the ultimate nature of reality, a picture that may be difficult to relinquish or to overcome.” Thomas,
“Inquiry Without Names,” 349.
41
more distinct” (καλλίων καὶ σαφεστέρα, 439a) than Alexander’s way, but Alexander has
nevertheless learned something.62
In the Cratylus’s repudiation of sophistry, then, there remains logical space for a positive use
of education through names. The problem with sophistry is not simply a method of education: it is
that sophists have false beliefs, both about the nature of particular objects and the nature of the
world. Sophists and rhetoricians, by and large, do not admit that there is such a thing as education
through things (Soph. 246a-b, 254a). As such, their opinions are blind. They might stumble upon
a true belief, but they do not admit the existence of the standard that makes this true belief true. 63
A namegiver who was guided by a dialectician would not have this problem. He would have true
beliefs about the objects of knowledge, and he would believe that his names could only be reliable
if they accorded with these true beliefs.
We have seen how this sort of model for true names is suggested in dialogues throughout
Plato’s middle period: not only the Cratylus, but also the Euthydemus and the Sophist. (The Phaedrus
and the Gorgias, with their discussions of noble and ignoble rhetoric, could be added to this list). It
is also very much present in the Republic. The Republic roundly chastises those who pursue
knowledge through sophistical devices, but it also presents a vision for how accurate nomenclature
could play a role in the creation of an ideal state. Those who “see through words” (ἐκ τῶν λόγων
θεωροῦσι, 601a) are criticized, and yet it is precisely the human tendency to see through words
upon which the primary educational system of the Republic is built. The sort of totalitarian control
To deny that Alexander has learned is to deny the soundness of Socrates’s argument against Cratylus. For Socrates
argues that education by names fails because the names themselves are unreliable, which implies that at least some
level of education could be achieved through reliable names.
63
Contrast this with the ordinary citizens of Kallipolis, whose true beliefs are not blind, since their beliefs have been
molded by the education of the guardians. See Rep. 590d-591a.
62
42
of language advocated for in Books II, III, and X is aimed at the establishment (τίθεσθαί, 484d) of
true beliefs in the soul, and yet it clearly does not involve the student in anything resembling an
education through things. This latter type of education is reserved for those who practice dialectic.
The Republic, then, should be read in the context of the theory of language developed in the
Cratylus. Since false names and false beliefs are inseparable, the constraints that Plato puts on
naming in the Cratylus ought to play a significant role in informing the constraints that Plato puts
on primary education and rhetoric in the Republic. In the next chapter, I will embark upon a rather
unusual journey, setting out to discover traces of Plato’s philosophy of language in the Republic, and
– in particular – attempting to demonstrate that Platonic censorship was aimed at precisely the sorts
of sophistical misrepresentations that the Cratylus identifies.
43
CHAPTER 2: ONOMASTIC EDUCATION IN PLATO’S REPUBLIC
Although the Republic is not known as a dialogue about language, the workings of language
are never far from its political and moral vision. A large portion of the dialogue is concerned with
censoring bad and harmful poetry, which is spoken of as a sort of toxin that inevitably poisons the
healthy human soul. Both the demos and the demagogue are chastised for their use of language, and
– perhaps most intriguingly – one of the sharpest criticisms Plato levels at irrational people is that
they “see through words” (ἐκ τῶν λόγων θεωροῦσι, 601a) – that their opinions are shaped by
words and arguments, not by facts or realities. Nevertheless, aside from certain perennial questions
pertaining to questions of poetry and censorship, scholars of the Republic rarely focus their attention
on Plato’s comments about language. This ought to change.
I believe that the Republic has a great deal to say about language; indeed, I hold the strong
view that the Republic cannot be adequately understood without making reference to that curious
dialogue about language, the Cratylus. For the purposes of this chapter, however, I restrict my
attention to one narrow (but far-reaching) claim: that an important goal of primary education in the
Republic is to train all citizens in the objectively correct usage of names. The incorrect usage of names
inevitably results in false beliefs, and false beliefs are both destructive of virtue and inimical to
dialectic. The role of accurate and unambiguous nomenclature may be more explicitly described in
other dialogues, but it is certainly implicit in discussions of the positive uses of rhetoric in the
Republic.
The structure of my argument is reasonably straightforward: I will begin by laying out
forcefully, and in some detail, my case for the position that the Republic is concerned with the correct
usage of names, and that it is correctness of naming and belief that explains the effectiveness of
44
primary education.
Second, I will explore the relationship between naming in the Republic and
naming in other dialogues, most centrally the Cratylus. Third, I will respond to an important
objection, the objection that we should not identify the normative language surrounding names in
the Cratylus with the normative language surrounding persuasive rhetoric in the Republic. Finally, I
will explain how correct naming plays a central role not only in the virtue of the citizens of Kallipolis,
but also in the purely scholarly pursuits of the philosopher kings, as they prepare for dialectic.
Under the Spell of the Demos
Scholars have sometimes suggested that the sophists, as Plato saw them, were those who
manipulated the demos in order to accomplish their own individual goals: power, pleasure, and the
like.1 This suggestion attempts to combine two clear accusations against the sophists into one neat
package: (1) the accusation that sophists practiced manipulation, and (2) the accusation that the
sophists were motivated by things like power, reputation, and pleasure.
The truth is much more complicated. Sophists do practice manipulation, and they do pursue
inferior aims, but it is not quite accurate to say that they manipulate the demos. In fact, as James
Wilberding has demonstrated, it would be far more accurate to say that the demos manipulates them.2
Thus, in the Gorgias, Socrates accuses Callicles of being “under the spell” of the demos: “If any
statement of yours is contradicted by the Athenian demos, you change your position and say what it
wishes” (481e). Sophists are panderers to the masses; indeed, Socrates identifies the politicallyminded sophist with the demagogue (δημολογικόν, Soph. 268b).3 Plato’s dialogues depict the
For example, see Helen North, “Inutilis sibi, perniciosus patriae: A Platonic Argument Against Sophistic Rhetoric,”
Illinois Classical Studies 06 no. 2 (1981).
2
James Wilberding, “Prisoners and Puppeteers in the Cave,” Oxford Studies in Ancient Philosophy 27 (2004): 117-39.
3
In what way, then, do sophists practice an “art of deception” (ἀπατητικήν, Soph. 264d)? The demos, presumably,
does not need to be deceived, no more than a drunkard needs to be convinced to take a drink. No, the objects of this
deception cannot be the demos at large; instead, they must be either citizens in their private capacity or children who
1
45
masses as embroiled in error, a type of error that the sophist imbibes, exploits, and encourages. The
ordinary Athenian politician is the pawn of the people precisely because he benefits from responding
to their whims.
In the Republic, the picture of the politically-minded sophist is strikingly similar. The
democratic masses are compared to a “great beast” which bullies its handlers into catering to its every
whim. Socrates turns the common prejudice against sophists on its head, counterintuitively
portraying the sophists as victims of the populace:
Do you really think, like the masses do, that the youth are corrupted by sophists in
their private practice of sophistry, to any degree worth mentioning? Or don’t you
agree that it the people who say these things [i.e. the masses] who have a greater
claim on the title “sophist,” since they give the most complete education, molding
young and old, man and woman, into whatever sort of shape they please? [When
do they do that?] Whenever the masses gather all together, seated at assemblies or
law-courts or theatres or battlefields … vehemently condemning some words and
actions and vehemently praising others, striving to outdo one another in shouts or
applause, and even the rocks and the walls of their gathering place resound with the
double tumult of censure and praise. Don’t you think that the young man’s heart,
in such a situation, is moved within him? (492a-c)
Sophists in professional practice are so far from being the impetus of corruption that Socrates
emphasizes how they themselves are corrupted. The public exploits the sophist’s desire for pleasure
and gain, and gets the sophist to use language according to its own whim:
Each one of these paid professionals, whom ordinary people treat as rivals, calling
them “sophists,” inculcate no other doctrine (δογμάτα) except the doctrines which
the masses pronounce when they gather en masse, which they call “wisdom” (σοφία).
It is just like if a man were acquainting himself with the passions and desires of a
great strong beast whom he was taking care of, learning how to approach it and
touch it, learning when and why it becomes manageable or unmanageable, learning
the reasons for each of the sounds of it is accustomed to make, and what sorts of
sounds will soothe or aggravate it. … Now, though [the trainer of the beast] doesn’t
genuinely know a thing about which of these beliefs (δογμάτων) and desires is
beautiful or ugly, good or bad, just or unjust, he applies all these names (ὀνομάζοι) in
have not yet come to “think with” the demos. The sophist “imposes [beliefs] upon the young who are still far from
things as they really are” (Soph. 234c). On Plato’s view, then, sophists justify the views of the demos to those who have
not yet accepted them.
46
accordance with the opinions of the great beast, calling the things that delight it
good, and the things that irritate it bad. (493a-c)
The passage is explicitly about sophists in private practice, but it is easily applicable to all those who
curry public favor – demagogues, politicians, poets, and the like. Let us call this type of person a
“panderer.”
The problem with panderers, according to this passage, is that they apply names
incorrectly. The panderer applies names according to the beliefs of the demos – beliefs which may
well be false, since they are based on the desire for pleasure, not the desire for truth. The name
“good,” according to the great beast, is understood to mean “whatever is pleasing to me”; whereas
the name “good,” in reality, refers to a fixed reality with objective standards of correctness. Since
“being good” does not genuinely mean “being pleasing to the demos,” something is amiss.
In this chapter, I draw upon my Chapter I analysis of Plato’s philosophy of language, and I
argue that this analysis can be applied to his comments about language in the Republic. The way this
plays out in the Republic can be seen in the above quotation: the demos attaches a certain content to
the word “good,” a type of content which may or may not be etymologically based. This content is
similar to Frege’s notion of “sense,” or Putnam’s notion of a “stereotype,” in that it involves the
ordinary understandings that speakers of a language have about the referent of a word. The content
the demos attaches to “good,” however, is false of the Form of Goodness; as such, teaching the
content to the professional panderer is a form of miseducation.4
Cherished Beliefs in the Republic
At this point, I will draw out and analyze a number of passages about education and
miseducation (cf. 514a) in the Republic, in the hopes of discovering the role that naming might play
4
It would remain a form of miseducation even if, by some dumb luck, the demos chanced upon an extensionally
adequate set of beliefs about the good (See Rep. 506c). The problem is that the demos has an unreliable method; it is
not pursuing truth, only pleasure.
47
in the dialogue. In its discussion of musical education, the Republic identifies two sorts of poetry:
poetry that harms its listener, and poetry that encourages its listener in virtue. The former sort of
poetry inclines a person toward base desires and diverts a person from the path of Truth. Bad poetry
both fails to establish true beliefs in the soul, and fails to insulate the true beliefs the soul already
possesses from the seduction of pleasure and the forcible influence of pain (see 413a-d). Such
failures are failures in musical education, the kind of education that prepares the soul for dialectic
(402a).
Socrates describes the beliefs that the soul acquires through musical education as “dogmas”
(Rep. 413c; 414b) – the same sorts of things (δογμάτων, 493b) possessed by the masses and
pandered to by inferior lawmakers at 493b.5 The most striking account of the origin of these sorts
of beliefs comes late in Book VII, in the context of the story of the rich and suppostitious son. This
son is brought up subscribing to the beliefs of his parents, which are compared to the ordinary beliefs
Athenian children acquire as they grow:
We have, at any rate, certain beliefs (δόγματα) from childhood about the just and
the fine (δικαίων καὶ καλῶν), under whose influence we have been raised as if they
were our parents, and we obey and revere them. (538c)
When he discovers that these are not his real parents, however, his allegiance to the δόγματα wanes.
Flatterers (i.e. sophists) swoop in and he adopts their way of life and, no doubt, their beliefs about
matters of value.
The story complicates matters somewhat, since it envisions the sophists as active, not (as in
the great beast image) passive. But there can be no doubt that the δόγματα mentioned at 538c are
not the same beliefs as those of the demos; for they do not sway sophists, but instead are swayed by
5
For the ancient Greek, the word “dogma” did not indicate that a belief was irrational.
48
them. And indeed, Socrates is defensive on behalf of these beliefs.6 Even if false, these beliefs do
not seem to be primarily influenced by the prospects of pleasure and pain; indeed, if we read 538c
closely with 413a-d, these δόγματα are precisely the sorts of things that would ideally resist the lures
of pleasure and pain. The “great beast” of the demos does not delight in such beliefs. The person
who is taught dialectic too soon is dissuaded from these beliefs about justice and goodness and made
to believe that a certain thing “is no more fine than base” (538d-e). He no longer obeys or reveres
the δόγματα, and thus becomes, in essence, an immoralist. Moreover, he becomes a perfect target
for the demos.
If we cast these two passages at 492-493 and 538-539 narrowly in terms of language, they
present the reader with two points about conventional beliefs about words. First, they express that
certain young men’s hearts respond with heartfelt emotion to the way the masses use language,
causing them to fall in line with popular errant language in an effort to maintain influential positions
in the city. The influence of the lower soul parts – the desire for base pleasures, for honor, for
popularity – distorts the content of the young man’s language at the same time as it distorts the
constitution of his soul. Second, the two passages insist that other young men (possibly, though not
necessarily, distinct from the first group) are taught by dialectic – or perhaps, the sophistical
doppleganger of dialectic – how to destroy the foundations of the conventional beliefs of the masses.
These conventional beliefs are incorrect, and likely becoming more incorrect with each succeeding
As Adam has pointed out in his commentary, Socrates’s endorsement of conventional morality here is almost
certainly not a wholescale endorsement. The guardians who have seen the Forms are said to “take care to insert into
human beings both in public and in private the habits which he sees [in the world of Forms]” (ἃ ἐκεῖ ὁρᾷ μελετῆσαι
εἰς ἀνθρώπων ἤθη καὶ ἰδίᾳ καὶ δημοσίᾳ τιθέναι, 500d), which certainly suggests that these habits are not identical
to the habits valued by ordinary Athenians. The passage at 538-539, however, is unambiguous that conventional
Athenian values are far superior to conventional immoralism. Plato, and James Adam, “The Republic of Plato, edited
with Critical Notes, Commentary and Appendixes” (Cambridge University Press: Cambridge, 1965).
6
49
generation of democracy,7 but they are not radically incorrect enough to merit their complete
destruction.8 They are relics of an earlier time, a time when cultural conventions (and names) were
more amenable to truth and virtue. Moreover, in contrast to the beliefs of many practitioners of
sophistry (e.g. Protagoras), conventional beliefs clearly presuppose that there are objective facts
answering questions like, “What is justice?” Nevertheless, these beliefs are not firmly entrenched
enough in the souls of young men to withstand the deleterious effects of premature dialectical
inquiry.
The combination of these two factors yields an interesting result. Conventional Athenian
beliefs are mediocre guides to virtue, but they are extraordinarily better than the results of an eristic
that makes argumentation the slave of the lower impulses. The conventionalist – a person like
Cephalus or Charmides – is incorrect, but is nevertheless more correct than his sophistical critic. 9
The error of the practitioner of eristic necessarily follows from her method, since the person willing
to destroy any belief will, as a consequence, be left without true beliefs.
The error of the
conventionalist is, by contrast, contingent, since the conventionalist would not use words falsely if
the convention setters did not first possess false beliefs about the referents of these words.
Education Through Names
7
The ruling powers of a democracy, according to Plato, are focused on sating their appetites, by effectively using
money and legal intervention to obtain what they want. They “care nothing for virtue” (556c), they “trample virtuous
habits underfoot” (558b), and they avoid education (560b). Their engagement in philosophy is just done for
appearances (ὡς ἐν φιλοσοφίᾳ διατρίβων, 561d).
8
As Hall observes, Plato “suspected the arbitrary character of most men’s judgments when they think for themselves,
and suggests that they are better off accepting conventional standards.” Dale Hall, “Interpreting Plato's Cave as an
Allegory of the Human Condition,” Apeiron 14 no. 2 (1980): 76-77.
9
Throughout the dialogues, Socrates consistently displays this sort of carefully circumscribed but sincere appreciation
of conventional values and beliefs. Consider the Euthyphro, where Euthyphro’s rebellion against conventional values is
roundly ridiculed.
50
Plato’s view of naming in the Cratylus is contingent in precisely this sort of way, as we have
seen in Chapter One. The names given by legislators in the Cratylus happen to be false, but – if the
namegiver had known the truth about the world – they could have been true. This is the case
because the objects of knowledge “have their own characteristic and independent essences, which
are not in relation to us or up to us (οὐ πρὸς ἡμᾶς οὐδὲ ὑφ᾽ ἡμῶν), pulled up and down on the
basis of how they appear to us, but independent” (386d-e). A name is a “tool for teaching and
distinguishing essences” (388b-c), which is given by the νομοθέτης – the “rule setter” or the
“lawgiver.” The name can be given well or given badly, and it will be given badly just in case the
νομοθέτης has a mistaken conception of the qualities of the object named (436b). The only truly
reliable form of education is that which does not rely on the beliefs of the namegiver: education
through things, not education through names.
In the Republic, then, we should be able to see what has gone wrong in the education of the
panderers in Book VI. The problem is not that the panderers lack advanced education in the
mathematical or dialectical arts. The problem is that the panderers encounter naming practices that
are, to whatever degree, false and unreliable. It is not the names themselves (as phonemes) that
mislead the sophists; rather, it is the fact that those who laid down the meanings of these names
have imbued them with a set of false beliefs or stereotypes. The panderers are not the creators of
false names, but rather the victims of such names: they are people who are being educated by names
laid down by an ignorant lawgiver. The panderers “put [themselves] and the nourishment of [their
souls] in the guardianship” (Crat. 440c) of inferior lawgivers. They commit the sin of Cratylus, by
assuming not only that names “are given for instruction” (428e), but also that “he who knows the
names also knows the things [they refer to]” (ὃς ἂν τὰ ὀνόματα ἐπίστηται, ἐπίστασθαι καὶ τὰ
51
πράγματα, 435d).10 Then, once they are “educated,” they pass themselves off as experts in the
nature of reality, not just in the nature of names.
What is the source of their false opinions? The answer may be a bit complex. According to
Rep. 538-539, children are brought up imbibing the beliefs of their elders; but, according to Rep.
492-493, the beliefs of the demos, in a democracy, cause the names we use to become corrupted. In
the Cratylus, names are given by the lawgiver (νομοθέτης). There is no indication that new lawgivers
cannot give new names, names that conflict with the beliefs of their ancient counterparts. I should
like to hypothesize, then, that the lawgivers of a democracy – the demos – are just as capable as any
other of laying down names, and that the systematic misnaming of virtue terms at Rep. 493 is an
instance of democratic namegiving. This hypothesis may seem ambitious, but I believe that it unifies
the concerns of these two dialogues in a striking and plausible way.
I would argue, then, that the mistaken usage of “good” or “just” on the part of the panderers,
derives from the opinions (δογμάτων) of the great beast. The demos is the νομοθέτης in a
democracy, and as such, it is the source of various rules, norms, and customs – including the norms
of linguistic usage. These norms are circumscribed by a set of conventional beliefs that the demos
attaches to the names we use: making “wisdom” connote “the things that the people pronounce
when they gather” and “evil” connote “the things which irritate the demos.” We might understand
these beliefs in various ways: as the results one would arrive at through a linguistic analysis of the
word, as the Wittgensteinian rules of a “language game,” as Putnam’s notion of a “stereotype.” All
I haven’t the space here to counter various interpretations of the Cratylus that would call into question my
assumption that Socrates’s views at Crat. 440 are sincere views that Plato himself subscribed to. For readings of the
Cratylus which are consistent with my own, see Kretzmann, “Plato on the Correctness of Names”; Ackrill, “Language
and Reality”; and Gold, “Ambiguity of ‘Name’.”
10
52
these understandings have in common two things: (1) that they involve packing a word with various
embedded beliefs, and (2) that these beliefs are contingent in the way I described above.
When we put this in the context of Platonic ontology, the resulting conclusion is this:
“beautiful,” “ugly,” “good” and “bad” genuinely refer to stable realities. Nevertheless, within the
Greek language, these names are content-rich, packed full of conventional Greek beliefs about the
given concepts. In a proper πολιτεία, these beliefs would be accurate, dictated by the lawgiver in
accordance with the knowledge of the dialectician. But, since Athens is a democracy, the views of
the masses become enshrined in linguistic customs; instead of imbibing true belief and demotic
virtue from their youth, the children of democracy “have lawlessness implanted in them” (τὴν
ἀναρχίαν ἐμφυομένην, 562e). The beliefs attached to the names they use are both unstable and
inaccurate. This instability and falsehood infects the education of those who learn these names, in
much the way that Socrates bemoans popular education through names at the end of the Cratylus.
In this section, then, we have seen how false beliefs attached to names are manipulated by
sophists and panderers, both in the Cratylus and in the Republic. In the next section, I will discuss
the power that these false beliefs have upon individual souls, and I will suggest that there might be
a positive counterpart to this sort of corruption: a way that virtue could be nurtured through the
correct usage of names.
True and False Words in the Soul
Jakub Jirsa11 has convincingly argued that the anarchic (“leaderless”) nature of a democracy
has its analogue in democratic language. Virtues are given the names of vices, and these vicious
names create a pretext through which the drones of a democracy can drive the virtues out of the city
11
Jakub Jirsa, “Sophists, Names and Democracy,” Croatian Journal of Philosophy 12 no. 2 (2012): 125-138.
53
(560d). The sophists are capable of manipulating names with such alacrity only because the polis
has become distanced from its ancient lawgivers12; we might say that the democratization of Athens
has left a semantic void, a void which becomes filled by sophistry and demagoguery – a dynamic
through which the leaders please the people, and the people flatter the leaders (see 494c).
This process is effected almost exclusively through the use of words. First, the son of the
oligarch “samples the honey of the drones and falls in with fiery and cunning beasts who have the
resources to assemble a limitless panoply of richly textured pleasures” (559d), inspiring a “brood of
passions” which “take over the acropolis of the young man’s soul, since they gather that it is bereft
of education and noble habits and λόγων ἀληθῶν (560b).13 Then “false and unreliable words and
opinions dash in and take hold of [him]” (ψευδεῖς δὴ καὶ ἀλαζόνες οἶμαι λόγοι τε καὶ δόξαι ἀντ᾽
ἐκείνων ἀναδραμόντες κατέσχον, 560c), becoming leaders for the young man’s soul. (Notice
again the intimate connection between words and beliefs). Finally, he returns to the drones – here
called “lotus-eaters” – and “lives openly with them” (ibid.). The toxic discourses of the drones “shut
the gates of his kingdom and … will not even admit the wise words of ambassadors of a bygone age”
(560c-d). Now that true words and beliefs have been rejected, the son follows the drones in the
systematic misnamings described in 560d-561a. (I will explore these misnamings in more detail, a
little later.)
12
The Athenians considered Solon to be their lawgiver, and a great many of their laws did derive from his written
code, even into the 4th century BC. Nevertheless, in Plato’s depiction of democratic Athens, every person “has the
authority to do as he pleases” (557b) and may defy or evade any law that claims authority over him (557e-558a).
Democracy “tramples underfoot” all solemn ideals and pronouncements (558a-b), and recasts the profound reverence
ordinary people feel for the law as “foolishness” (560d). Clearly, Plato would deny that Solon is responsible for the
genuine customs (νομοί) that are exemplified by his contemporary Athenians.
13
I intentionally leave the phrase “λόγων ἀληθῶν” untranslated. It could be translated “true discourses,” “true
arguments,” or “true language”; since these three translations are all compatible, I am inclined to believe that all three
of these meanings of λόγος are relevant here. The third translation, “true language,” most straightforwardly supports
my thesis.
54
This passage of the Republic has clear and unambiguous resonances in the Cratylus, where
Socrates warns that names in the hands of inferior legislators are deceptive, and that “no sensible
person will put himself or the nourishment of his soul under the guardianship of names” (440c).
Both passages are a manifestation of Plato’s characteristic antipathy toward politicians: the giver of
names (ὀνομάτων θέτης) and the giver of laws (νομοθέτης) – the establisher of conventions
(νομοι) – are said to be one and the same.14 We would hardly expect Socrates to recommend that
Cratylus and Hermogenes trust in the linguistic fabrications of orators, nor would we expect
someone as suspicious of politics as Plato to be optimistic about the trustworthiness of future
lawgivers. And yet elsewhere in this very dialogue, the Republic, Plato is unambiguously optimistic
about the educational potential of political rule. Would Plato’s warning not to trust the lawgiver
apply equally to the citizens of the perfect city, the Kallipolis?
Certainly not. The philosopher in Kallipolis “looks intently on whatever things are fixed
and eternal” and imitates them, thus becoming himself “divine and graceful,” and “if it should be
required (ἀνάγκη) of him in his engagement with society to insert into human beings both in public
and in private the habits which he sees there… he will [not] prove a poor craftsman of temperance
and justice and every kind of demotic virtue” (Rep. 500c-d). This passage parallels the passage at the
conclusion of the cave image, where we are told that the philosopher who has seen the Good must
then be compelled (προσαναγκάζοντες) back into the cave, where they will discern the nature of
14
My own sense of this identification of the lawgiver with the name-giver is that Plato is here anticipating something
like the postmodern understanding of language as a tool that the powerful use to manipulate the masses, or that
various powerful in-groups use to create disempowered out-groups. In Plato’s dialogues, however, there is always a
question of who is manipulating whom, however, since – as we have seen – the politicians themselves are often
portrayed as mere puppets in the hands of the opinionated people (the demos).
55
the shadows projected onto the cave wall “infinitely better than the people there” (520c, see 519c520d). Or consider 484c-d:
Do you think, then, that there is real difference between blind people and those
who are genuinely deprived of the knowledge of each thing that exists, since they
lack any manifest pattern in their souls and so they cannot even like a painter
consult the genuine article, considering the spectacle of it as the most precise
example and standard for imitation, and accordingly establishing the norms (τὰ
νόμιμα … τίθεσθαί) of beauty and justice and goodness where they are currently
lacking, while taking responsibility for the maintenance of already existing norms
(τὰ κείμενα)?
The message is abundantly clear: those who do have a pattern of the Forms in their souls are capable
of constructing demotic norms, norms that are naturalistically adequate for the generation of
accurate naming conventions and true beliefs among the populace.
There is some complexity in these passages, however, which is certainly worth
consideration.15 In the Republic, the term νομοθέτης is almost exclusively used to describe the
founders of the city (Socrates, Glaucon, and Adeimantus), not its guardians.16 Indeed, the very term
“guardian” gets at this notion, since the guardians are not selected for their ability to establish laws
or customs, but rather their ability to be a “guard of himself and the musical education he has
received” (φύλαξ αὑτοῦ ὢν ἀγαθὸς καὶ μουσικῆς ἧς ἐμάνθανεν, 413e). It would be a mistake,
then, to say that the stamping or insertion (τίθεσθαί) of the virtues referred to in 484c-d and 500cd indicates the original imposition of customs; rather, it refers to the particular grafting of already
existent normative customs onto a soul that does not yet possess these customs.
Now, if we were to import from the Cratylus the intimate connection between the giving of
customs and the giving of names, we would find that a Platonic critique of names necessarily informs
Both passages anticipate Stat. 309c-d, which certainly supports the view that the νομοθέτης is involved in the
insertion of accurate terms and true beliefs into the citizen’s soul.
16
458c, 497d, 530c.
15
56
the naming conventions of the founders of the city (the “lawgivers”), and that the guardians are
charged with the job of applying these names accurately in the process of paideia.17 To apply the
names accurately means (a) to have the right beliefs about the Forms corresponding to these names,
and (b) to use these beliefs to determine which objects “fall under” the name (e.g. which people are
courageous).18 Primary education in Kallipolis, then, is not simply aimed at making the soul
amenable to the rule of reason. It is also centrally concerned with training the novice in the proper
use of names, which is to say the proper beliefs about the Forms. This education has at least two
purposes: (1) to improve the moral fiber of each citizen, since correct beliefs about the virtues are
essential to – and, in the case of courage, perhaps even constitutive of – right action, and (2) to
prepare the student for the practice of dialectic, which will both reaffirm and refine the true beliefs
about the Forms that a proper education through names consists in. The student who engages in
dialectic at the proper time, with the proper natural dispositions and the proper entrenched true
beliefs, is capable of becoming a philosopher king. But without the generally accurate names, the
task of higher education is impossible.
Correct and Incorrect Naming in the Republic
There is a great deal of evidence for this view in the Republic.19 According to Book VII, the
conventional dogmata of individuals raised in a given culture are not primarily – as we might expect
– views they have heard from their peers or their parents or the culture at large, but rather they are
17
There is no need, of course, for the founders to create a completely new language. It will be sufficient to keep
current words, but simply to change our beliefs about those words: that is, to change the content that is culturally
inscribed on these words.
18
For practical purposes, then, true belief is just as useful as knowledge, even in Kallipolis. (Cf. Meno 97b-c).
However, true belief is not as stable as knowledge, since it is blind and untethered to a λόγος (Rep. 401e-402a).
Moreover, one cannot have reliable true beliefs unless these beliefs have been implanted by a person who possesses
knowledge.
19
Pace Jirsa, who writes that “within the Republic there are not many passages dealing with the correct approach to
naming.” Jirsa, “Sophists, Names and Democracy,” 134, note 20.
57
the views they have heard from the νομοθέτης (538d). On the surface, of course, this is completely
implausible, since most children would have never spoken to a legislator, much less an august
νομοθέτης like Solon. But the Cratylean view that the entire fabric of a language proceeds from
the lawgiver may clear things up, here: the meanings that are inscribed onto Athenian names
manifest the moral beliefs of the demos, moral beliefs of the culture as a whole, and these are precisely
the beliefs that proceed from the names imposed by the lawgiver (in Athens, the demos). It is of vital
importance that these names be correct, or the child will be ill-equipped to proceed on the ladder
from ignorance to knowledge.20
Time after time, Plato emphasizes this point, as his characters discuss the use of language
and names (ὀνόματα) among the citizens of his cherished republic.21 First of all, in discussing the
censorship of the poets, Socrates somewhat curiously focuses his critique on terminology: “All the
names (ὀνόματα) of [certain] terrors and fears must be banished” (387b). The passage makes a
connection between truth and normative values: the terrors and fears imagined are in fact false
A comparison with Aristotle’s notion of ἀρετή is instructive here. In the Nicomachean Ethics, Aristotle insists that
(demotic) virtue proceeds almost entirely from proper training, which is the province of the lawgiver: “Legislators make
the citizens good by forming habits in them, and this is the wish of every legislator, and those who do not effect it miss
their mark, and it is in this that a good constitution differs from a bad one. Again, it is from the same causes and by
the same means that every virtue is both produced and destroyed, and similarly every art; for it is from playing the lyre
that both good and bad lyre-players are produced. … For if this were not so, there would have been no need of a
teacher, but all men would have been born good or bad at their craft. This, then, is the case with the virtues also; by
doing the acts that we do in our transactions with other men we become just or unjust. … It makes no small
difference, then, whether we form habits of one kind or of another from our very youth; it makes a very great
difference, or rather all the difference” (2.1). Aristotle, and W.D. Ross, The Nicomachean Ethics of Aristotle, Translated
and Introduced by Sir David Ross (Oxford University Press, 1961).
21
Note that it is the use of names is regulated here. The assumption is that the names themselves instituted by the
founders are properly informed by dialectic, but that the populace must now be made to use these names correctly. As
Reeve says, with respect to the Cratylus, “Through his association with the dialectician, the name-maker has completed
the first of his tasks: he has made naturally correct names for all the forms. It remains for him to carry out the other
part of his task and assign those names correctly to things. He must ensure that the name of a form is assigned only to
things which have a nature that imitates it. This he does by setting rules or conventions governing the use of names”
Plato, and C.D.C. Reeve, Cratylus, Translated, with Introduction and Notes, by C.D.C. Reeve (Indianapolis: Hackett, 1998),
lii.
20
58
(386c), because in a perfect society they do not exist.22 That which fails to edify the future guardian
is considered to be false: thus, the harrowing stories of the poets are “neither true nor beneficial to
future warriors” (386c). There is a similar passage later in Book III, where Socrates mentions how
the indulgent diets and lifestyles of certain Athenians wreaked such havoc on the body as to force
the sons of Asclepius to coin names (ὀνόματα τίθεσθαι) for the resultant diseases (405d), names
which Glaucon calls “unnatural” (ἄτοπα) – since presumably they would not be needed in Kallipolis.
Of course, Plato’s central preoccupation, in terms of naming things, pertains to moral
terminology. The demagogue, as we have seen, is criticized for applying names like “beautiful” and
“just” according to the δόξα of the multitudes (493b-c). Moreover, whereas the uneducated men
emerging from Glaucon’s primitive state of nature mistakenly “name the things enjoined by the law
‘lawful’ and ‘just’” (ὀνομάσαι τὸ ὑπὸ τοῦ νόμου ἐπίταγμα νόμιμόν τε καὶ δίκαιον, 359a), those
subject to primary education in Kallipolis “name an action ‘just’ and ‘fine’ when it preserves and
contributes to [their harmonized] condition [of the soul]” (ὀνομάζοντα δικαίαν μὲν καὶ καλὴν
πρᾶξιν ἣ ἂν ταύτην τὴν ἕξιν σῴζῃ τε καὶ συναπεργάζηται, 443e). Correct standards of naming
are present in the virtuous state; indeed, incorrect standards for naming moral attributes seem to be
a touchstone of corrupt constitutions.
22
Although I shall not argue the case fully here, my sense is that this particular passage anticipates modern
poststructuralist thought in an interesting way, since it tends toward the view of the “Sapir-Whorf hypothesis,” which
contends that language influences thought. Chris Swoyer recently characterized this line of thought as “linguistic
relativism,” the view that “different languages carve the world up in different ways, and that as a result their speakers
think about it differently” (http://plato.stanford.edu/entries/relativism/supplement2.html). One could not simply
classify Plato as a linguistic relativist, however, because although he agrees (in the Cratylus, Republic and Statesman, at
any rate) with the descriptive claim that different languages carve up the world in different ways, he would insist that
there are objective standards against which to judge the worth or accuracy of these different divisions of reality. Chris
Swoyer, Maria Baghramian, Adam Carter, "Relativism,” The Stanford Encyclopedia of Philosophy (Summer 2015 Edition),
ed. Edward N. Zalta, http://plato.stanford.edu/archives/sum2015/entries/relativism/.
59
This process of renaming exists not only in the state, but in the soul. The story of the son
of oligarchy in Book VIII, which I have already alluded to, describes a man seized by a sort of
psychological sophistry, making his soul democratic, and therein giving ugly names to the virtues
and sweet names (or “pet names,” ὑποκοριζόμενοι, 560e) to the vices. Plato personifies the young
man’s epithumetic desires, and alleges that these desires call attributes of the soul by the wrong
names, calling a sense of shame “folly,” temperance “effeminacy,” but calling impudence
“refinement,” license “liberty,” wastefulness “magnificence,” and shamelessness “manliness” (560de; cf. Laws 821d-822c). Thucydides makes a strikingly similar complaint, while discussing the
manipulation of language during a time of war:
Words had to change their ordinary meaning and to take that which was now given
them. Reckless audacity came to be considered the courage of a loyal ally; prudent
hesitation, specious cowardice; moderation was held to be a cloak for unmanliness;
ability to see all sides of a question, inaptness to act on any. Frantic violence became
the attribute of manliness; cautious plotting, a justifiable means of self-defence. The
advocate of extreme measures was always trustworthy; his opponent a man to be
suspected. To succeed in a plot was to have a shrewd head, to divine a plot a still
shrewder; but to try to provide against having to do either was to break up your party
and to be afraid of your adversaries.23
The problem here is, again, not with the particular syllables that are used, but with the semantic
content housed behind these syllables. Calling license “liberty” entails a number of false beliefs
about a genuine standard for action: the virtue of liberty. If a young man believes that drinking
exorbitant amounts of wine is a manifestation of liberty, then his soul (not to mention his body)
becomes subject to serious harms. The beliefs embedded in words almost inevitably manifest
themselves in actions.
23
Thucydides, History of the Peloponnesian War, trans. Richard Crawley (New York: Dover, 2004), III.84.4.
60
What is the relationship between the Cratylus and the Republic, then? In the Cratylus, Plato
constructs a notion of linguistic correctness, and specifies that education through names can only
be of any use to a person who can be assured of the names’ corresponding to that notion of
correctness. In the Republic, Plato constructs a system of lawgiving whereby the conceptual content
of names is dialectically constrained (cf. Crat. 390c) so as to be morally and conceptually accurate.
By doing this, Plato discovers a way in which education through names can consist in learning what
is objectively true, since the conceptual names of Kallipolis embody true beliefs about the nature of
genuine objects (the Forms).24
Philosophy and Rhetoric
Up until this point in the essay, I have defended a view about Platonic naming that centrally
hinges on the idea that primary education in the Republic – at least insofar as such education deals
with words – should be identified with the kind of onomastic education Plato writes about in the
Cratylus.25 The viewpoint I have advanced is, I believe, compelling, but it can only be as strong as
this premise about the close relationship between the two types of education. What reason, then, is
there to believe that Plato’s concerns in the Cratylus were similar to his concerns in the Republic,
much less to believe that musical education in the Republic is an application of his educational
musings in the Cratylus? That is the question to which I will now turn.
To begin to answer this question, we must first remember why it is possible to compare the
two types of education, in the first place. Both onomastic and musical education are concerned with
These beliefs are externally reliable, though their internal justification is lacking. They lack an internal λόγος.
Thus, they are “true belief,” but not knowledge.
25
Non-verbal aspects of primary education can hardly be called “onomastic,” though an argument could be made that
even instructions on the modes of music and the physical movements of the body aim at true belief. They likely aim at
such belief indirectly, however: by removing the obstacles to true belief that would exist in a disorganized and
disharmonious soul.
24
61
persuasion. Both of them have the aim that the student might use language according to certain
norms. In good onomastic/musical education, words are used accurately; in bad onomastic/musical
education, words are used inaccurately. In both cases, good educators are identified with noble
lawgivers or philosophers, and bad educators are identified with demagogues or sophists.
Since both pedagogies revolve around persuasion, I intend to do three things to defend my
interpretation in the remainder of the essay: (1) provide a basic survey of Plato’s understanding of
persuasion, (2) argue that all verbal, but non-dialectical, persuasion should be identified as education
through names, and (3) argue that, since primary education is non-dialectical verbal persuasion, it is
a variety of education through names. A surprising consequence will fall out from this analysis: it
will become clear that, to Plato, the hallmark of non-dialectical inquiry is that it disregards the nature
of things, in preference for discussions about the words themselves.26 This sort of persuasion
through mere language has its purposes, though it bypasses any strict notion of rationality. It would
appear, then, that irrational (or perhaps arational) persuasion is a driving force behind the
constitution of Plato’s Republic; and this is an interesting result indeed.
Platonic Persuasion
Persuasion, as an activity, consistently appears as a central theme of various Platonic
dialogues. In the Philebus, Protarchus quotes Gorgias as saying that “the art of persuasion is far
superior to the other arts, for it makes all things willingly subject to it, without making recourse to
force” (58a-b). There are a multitude of types of persuasion (e.g., persuasion in mathematics, Gorg.
453d-454a), but Plato tends to focus on the kind of persuasion that is “concerned with right and
wrong” (Gorg. 454b, cf. Soph. 223a). Value-laden words like “right” and “wrong” are the subject of
26
Surely the interlocutors do not always know that their subject matter is mere words, however.
62
much public disagreement, and the use of these words is “bound to fluctuate” (Phaedrus 263a) in
common parlance. The instability of ordinary beliefs about these words make them subject to
manipulation in the hands of the rhetorician. Those who practice persuasion “contend with words
(ἀντιλέγουσιν) … about justice and injustice,” and they are capable of making “the same things
appear to the same people at one moment just, and then, when they choose, unjust” (Phaedrus 261cd). Perhaps the most illuminating comment Plato makes on this sort of persuasive activity comes in
the Phaedrus, where Socrates describes rhetoric as “a distracting of the soul through language”
(ψυχαγωγία τις διὰ λόγων, lit. “soul-bringing,” 261a).
There are two kinds of people who seek to persuade, according to the Phaedrus. First, there
is the individual who, though he “does not understand good and evil,” nevertheless “uses persuasion
to subdue a society just as clueless as he is, … and makes evil out to be good” (260c). This person
has “gone hunting for beliefs”; instead of saying that he practices an “art of words” (λόγων …
τέχνην), it is more appropriate to say that “his art is laughable, and turns out to be completely
artless” (262c). Doubtless, this individual could be justly compared to the tragic poet or the mob
orator in the Gorgias, who desires only to “please his countrymen,” who “despises the common good
in pursuit of his own interests … and doesn’t care in the slightest whether his actions make others
better or worse” (502e-503a), and who “caters to our desires without recognizing any goodness or
nobility concerning them” (518c).
Second, the Phaedrus describes the individual who (a) understands the nature of his audience,
and (b) has the capacity to “divide things into kinds and embrace each individual thing under a
single form” (273d-e) – in other words, the dialectician. This individual is the “true rhetorician”
(269d), who is capable of persuading someone with truth, or deceiving someone with falsehood,
63
precisely because he has “knowledge” (259e). These orators consistently speak “with an eye to what
is best, their aim being to discover any way their words might make citizens as good as possible”
(Gorg. 502e, cf. 503d-e). “The person who would become a proper rhetorician must be just himself
and be fully conversant in just things” (ἐπιστήμονα τῶν δικαίων, ibid. 508c). The rhetorician
treats the soul as the doctor treats the body, with the goal that their patient become “well ordered”
(κεκοσμημένον, 504a) according to “a certain form” (εἶδός τι, 503e). The goal of proper rhetoric
is clearly arête.
Now, since it is clear that one of the goals of dialectic is also virtue, the question arises: is
philosophy synonymous with proper rhetoric? In the Phaedrus, the dialectician and the “true
rhetorician” might seem to be the same person. The passage from 501d-504e, in the Gorgias, could
be read along the same lines. Thus Griswold identifies the philosopher with the true rhetorician,
and claims that this identification is based on the fact that – whereas false rhetoricians instill belief
– true rhetoricians instill knowledge.27 Griswold is wrong to identify the philosopher and the true
rhetorician, however, because even good rhetoric (in keeping with the later section of the Cratylus)
is incapable of producing knowledge. It may be the case that a certain philosopher is also a true
rhetorician, or vice versa, but the two categories are conceptually separate – just as the same woman
might be both a wife and a mother, despite those two identities being conceptually distinct.
The philosopher is distinguished from the true rhetorician in dialogues like the Cratylus, the
Statesman, and the Republic. In these dialogues, Plato presents a picture by which the practitioner of
dialectic instructs the lawgiver in either the use of names (Cratylus) or right opinion (Statesman,
Republic), as a tool for proper governance. The dialectician and the rhetorician are conceptually
Charles L. Griswold, “Plato on Rhetoric and Poetry,” The Stanford Encyclopedia of Philosophy (Fall 2014 Edition), ed.
Edward N. Zalta, http://plato.stanford.edu/archives/fall2014/entries/plato-rhetoric/.
27
64
separate (Crat. 390d; cf. Rep. 601e-602a), even if they may coexist in the same person. In the
Statesman, rhetoric is portrayed as being “auxiliary” (ὑπηρετοῦν) to statesmanship; statesmanship is
the “art that is in charge of rhetoric” (τῇ τῆς πειστικῆς ἀρχούσῃ, 304d-e). Proper rhetoric does
not provide formal instruction. Instead, it involves telling the masses the appropriate “myths”
(304d), myths which inculcate “firmly entrenched true beliefs about beauty, justice, and goodness”
(309c).28
In the early books of Plato’s Laws, the Athenian repeatedly emphasizes the role of persuasion
in the art of ruling the state. He says that the lawmaker has two available instruments of rule:
persuasion and compulsion (cf. Gorg. 517b). The problem with contemporary legislators is that they
“proceed by compulsion unalloyed” (722c). They do not make people agree with the law; they only
make people follow the law. Elsewhere, Plato appeals to the close connection between desire and
action, and the idea that a person might be made captive by her own desires (Crat. 403, Phaedo 82e83a; cf. Rep. 517a) – and moreover, that such a captivity might be used to positive effect, if what the
citizen desires is virtue (Rep. 590e-591a). The appeal to persuasion, in the Laws, is an application of
this notion of the constraining power of desire. The Athenian recommends that the lawgiver work
in “conjunction” (κοινούμενος, 720d) with the citizens, “using persuasion to somehow win them
to his side” (ἐπέταξεν … πῃ συμπείσῃ) – a task which is done through the introduction of a prelude
(προοίμιον) to each law. These preludes are highly suggestive of the latter portion of the Phaedrus,
Plato’s generalized attitude toward persuasion and rhetoric is obscured by the fact that he sometimes (e.g. Gorg.
454e) seems to consider genuine teaching a form of persuasion. Even in such instances, however, he distinguishes
persuasion that aims at knowledge from persuasion that aims at belief. For my purposes here, I shall consider
persuasion proper to be the second sort of persuasion; when I refer to the first sort of persuasion, I will call it
“dialectic.” The relevant point is that the statesman does not force the ordinary citizen to engage in dialectic, although
he does persuade the ordinary citizen.
28
65
as is the Athenian’s suggestion that the lawgiver study the natures of individual citizens, so as to
more effectively implant convictions into their minds (Laws 720d; cf. Phaedrus 271b-e).
Non-dialectical persuasion is always focused at this goal of implanting convictions. We are
told in the Timaeus that knowledge arises in the soul through instruction (διδαχῆς), true belief
through persuasion (πειθοῦς, Tim. 51d-e); this is the central distinction between knowledge and
true belief. Knowledge occurs μετ᾽ ἀληθοῦς λόγου, with true “language” or “reasoning,” whereas
true belief is senseless (ἄλογον).29 Although some scholars30 have argued that persuasion tending
toward true belief may include rational elements, it is abundantly clear that these rational elements
stop short of a λόγος; or rather, even if the educator possesses the λόγος, this λόγος is not one of
the things communicated through non-dialectical persuasion. So long as an instance of nondialectical persuasion ends in true belief, it is successful.
Onomastic Persuasion
I will now argue that the sort of “education through names” described in the Cratylus can
be identified with the practice of persuasion described above – in other words, that the category of
“education through names” subsumes both proper and improper rhetoric.
Commentators have often assumed that “education through names” is merely a matter of
etymology. Such a proposal, however, seems to neglect a number of places where the Cratylus
suggests that the art of naming and the arts of rhetoric/sophistry are intertwined – centrally 425a,
where Socrates explicitly compares the onomastic art with rhetoric: “We shall make speech by the
29
Looked at from this angle, the Theaetetus is concerned with distinguishing true beliefs that arise from instruction
(knowledge) from true beliefs that arise from persuasion (mere opinion). The word “λόγος” captures this difference.
The distinction between instruction and persuasion is very prominent in the Theaetetus. At 201a-c, for example, true
belief and knowledge are distinguished on the basis of their origins, completely in keeping with Tim. 51.
30
Christopher Bobonich, “Persuasion, Compulsion and Freedom in Plato's Laws,” Classical Quarterly 41 no. 02 (1991):
365-.
66
art of naming, or rhetoric” (τῇ ὀνομαστικῇ ἢ ῥητορικῇ). Naming, in the Cratylus, is continually
connected with figures that Plato associates with sophistry: Prodicus (383b), Protagoras (386a),
Euthydemus (386d), Homer (391d; cf. Prot. 316d), and Euthyphro (396d-e). The suggestion seems
to be that sophists take advantage of the meanings of words in order to establish their arguments:
thus Socrates waxes poetic about what the arch-sophist Protagoras has learned about the “correctness
of names” (391b-c); elsewhere, in the Euthydemus, Socrates complains that Euthydemus and
Dionysodorus are exploiting various customary meanings of words in order to confuse the young
Clinias (277d-278c). The accusation here is clearly not that sophists take advantage of certain
etymologies; no, it is that sophists take advantage of the associations that words have, whether or
not these associations are enshrined etymologically. This makes sense of the fact that Socrates is still
concerned with a student’s being deceived by the meaning of names even after he has established
that names are largely conventional.31
If such considerations do not make fully clear that education through names is not
exclusively etymological, other considerations from the text will. Consider, for example, the nature
of the namegiver’s error at Crat. 436b-c.
Soc. Why clearly he who first gave names gave them according to his conception of
the things which they signified – did he not?
Crat. True.
Soc. And if his conception was erroneous, and he gave names according to his
conception, in what position shall we who are his followers find ourselves? Shall
we not be deceived by him?
31
Barney would appear to be a proponent of the thesis that onomastikê is essentially etymological, but even she
acknowledges that “[since] the application of primary names, complex names and even logoi are all part of the craft of
verbal ‘painting’, … the onomastic craft is … identical with rhetoric properly understood.” Unless Barney wants to
limit “rhetoric properly understood” to etymologies, however – which is wildly revisionist – her position fits best in the
context of an understanding of onomastikê which embraces more than just etymological theorizing. Barney, Names and
Nature, 99.
67
The error in this case manifests itself as an error in primary names and etymology, but the actual error
is an error in belief: the namegiver has false beliefs about the object named. The problem, in modern
terminology, is not that the namegiver does not know how to give names correctly, but rather that
he does not properly conceptualize the object he is naming.32 The namegiver gives the wrong name
to an object because he does not recognize what type of object it really is.33
The conclusion of the Cratylus makes a broad claim about education through names:
education through names is unreliable, since the names used are inadequate. These names are
inadequate precisely because they embody false beliefs. Once we juxtapose this conclusion with the
survey of persuasion above, a fascinating result emerges. Education through names is flawed for
precisely the same reason that persuasive rhetoric is flawed: because it involves the learner in an
engagement with false (or at least unreliable) beliefs. The best explanation of this correspondence
is, obviously, that “education through names” is just another name for rhetoric – exactly the sort of
claim we see at 425a.
This interpretation explains and illuminates other Platonic texts, as well. The Sophist and
the Statesman both contain numerous references to a distinction between two types of education
that occur in the city: the sort of education practiced by sophists and demagogues, and the sort of
education practiced by philosophers and statesmen (Soph. 268a-d). The former sort of education
involves the use of images and phantasms, and exploits the possibility that words can be false. Both
these qualifications are the hallmarks of the discussion about onomastikê from the Cratylus, since
Cratylean names are “images” (439a) or “imitations” (423b) which can be either true or false. The
That is, he does not have the correct “set of beliefs” about the object. See Chapter 1, note 42.
For a helpful analysis of the relationship between naming and conceptualizing, in the Cratylus, see Gold, “The
Ambiguity of ‘Name’.”
32
33
68
latter sort of education can be characterized by dialectic, but need not be: since the educational
actions of a statesman aim at true belief, not knowledge, they must be a type of non-dialectical
persuasion.
The Statesman juxtaposes two types of rhetoric: the rhetoric of demagogues (303a-c) and the
rhetoric of statesmen (or those directed by statesmen, 303e-304e). The rhetoric of statesmen is
impressively similar to the uses of language described in the hopeful portions of the Cratylus, when
names are described as tools of education (Crat. 428e).34 This latter form of rhetoric is uniquely
qualified to educate because of it “implants” a “true and certain opinion” (ἀληθῆ δόξαν μετὰ
βεβαιώσεως, Stat. 309d) in the minds of the talented young people of the polis. It generates “right
opinion” about topics like goodness, justice, and benefit (309c). The difficulty inherent in the wrong
opinions is discussed earlier in the Statesman, though not in relation to value terms35. The Stranger
complains that the nature of the statesman has “eluded them in the course of naming” (κατὰ τὴν
ὀνομασίαν ἐκφυγών, 275d), through a series of argumentative steps prompted by false beliefs.
Several examples are given; for instance, the belief that animals all have something in common (263c)
and the belief that statesmanship involves “art of nourishing herds” (ἀγελαιοτροφικήν) instead of
“art of looking after herds” (ἀγελαιοκομικὴν, 275d-275e). With such false “elements” (στοιχεῖα,
34
At 429b, Socrates suggests to Cratylus that some legislators do their job of naming better than other legislators.
Cratylus disagrees, thereby introducing a digression where Socrates criticizes the sophistical commonplace that all
names are rightly imposed, and thus that all conventions are equally valid. This digression has interesting resonances
with the exploration of how falsehood is possible, in the latter portion of the Sophist. More importantly, however, I
would note that the digression obscures the point Socrates surely believes: that some legislators are better at naming
(i.e. assigning conventions or laws) than others. And this is the point from the Cratylus that is picked up in the
Statesman.
35
I follow one reasonably standard interpretation of the Statesman, according to which the dialogue investigates
something relatively unimportant (what is a “statesman”) in an effort to demonstrate adequate methods for
investigating more difficult topics.
69
278d), the dialectical formation of a “single true belief” (μίαν ἀληθῆ δόξαν, 278c) about the object
of inquiry is said to be impossible.36
The Sophist, likewise, contains too many resonances with the Cratylus to ignore. In the
Sophist, a name ought to “express the nature” (Soph. 221d) of the thing to which it refers. The
problem that sophists have, as the interlocutors describe it, is that they somehow manage to
participate in falsehood, despite the fact that (by their standards) falsehood is impossible. Since the
sophist is not guided by the dialectician, his claims about the world are haphazard, like a blind man
groping in darkness. The sophist “creates with his pencil representations bearing the same name as
real things” (Soph 234b). As in the Cratylus, the sophist is likely enough to get the name right, but
the meaning wrong. Indeed, the sophist could hardly manage to deceive unless he had a knack of
getting the name right. False beliefs are said to be “imitations” (μιμήματα) of real things, and false
beliefs account for the practice of deception (Soph 264d). False attempts at virtue are based on false
beliefs about the nature of virtue (cf. Soph 267b-c).
It is abundantly clear, then, that Crat. 425a, which identifies rhetoric with onomastic
education, should be taken extremely seriously. Every type of verbal persuasion that does not involve
dialectic is onomastic – it is persuasion through mere language. Having established this, it is a small
step to say that musical education in the Republic is onomastic. Musical education is clearly not a
form of dialectic, though it also clearly involves persuasion. I make no claim, of course, that musical
education involves only persuasion; it involves various other things, like an engagement in
harmonies, rhythms, and forms of speech. But the persuasive power of musical education is
An illuminating discussion of 277a-278e – where the Stranger discusses the relationship between letters, words, and
true belief – can be found in Kenneth M. Sayre, Metaphysics and Method in Plato's Statesman (Cambridge University
Press, 2006).
36
70
intimately related to its moral power to form citizens into demotically virtuous individuals, and this
persuasive power is a form of onomastikê. As at Crat. 428e, names are indeed given to instruct; and
if these names derive from the discernment of the dialectician, they are quite well suited to instruct
a person in virtue.
Bridging the Gulf Between Knowledge and Opinion
How do the true beliefs and true names inculcated by onomastic education relate to the
more mature knowledge and wisdom of the philosopher-kings of the Republic? This is the question
I should now like to turn to. My contention is that the sort of stable true beliefs produced by
education through names are needed in order for a student to successfully engage in dialectic. I will
proceed, first, by setting out why true beliefs about the Forms must play a key role in the Republic;
second, by appealing to evidence in the Republic to support my thesis; finally, by appealing to similar
evidence in other dialogues.
In a recent paper on Republic V, Katja Vogt argues against what she calls the “traditional”
view of Socrates’s conversation with the lover of sights and sounds. 37 This view holds that, since
knowledge and opinion are separate faculties set over (ἐπί, Rep. 477b) different kinds of objects, it
is impossible to possess an opinion about a Form. Vogt’s concerns are quite broad, but I should
like to emphasize a single point that she makes. According to Vogt, the traditional view makes the
very notion of Platonic education through the dialectical process impossible, since such education
is consistently portrayed as involving the movement from a belief about X to knowledge about X.
The classic case of this process is the depiction of the slave boy in the Meno (82-85), whose “true
opinions” have been awakened by questioning, and who – through further questioning – will attain
Katja Maria Vogt, “Belief and Investigation in Plato’s Republic” in Belief and Truth: A Skeptic Reading of Plato
(Oxford: Oxford University Press, 2012), 53.
37
71
knowledge (85c-d). Vogt argues that this same sort of process takes place in the Republic, when
Socrates reluctantly offers his opinion about the offspring of the Good, in the Sun, the Line, and
the Cave metaphors. Even though Socrates insists to Glaucon that “true beliefs without knowledge
… are at best blind” (506c), he nevertheless goes on to share his own opinion (μοι φαίνεται, 506e)
about the Good in analogical form. Vogt’s conclusion is that, although belief is properly directed
at the believable (δοξαστόν, 478b), “this does not mean that [belief] cannot – deficiently – be
directed at what [it] is not adequate for.”38
Notice that Vogt puts this point in the passive voice: belief may be directed. This, I would
argue, is precisely the educational programme of the Republic. If a person’s beliefs are blind – if they
“direct themselves” – then, although they may be practically useful and even true (Meno 97b-c), they
are “ugly” (αἰσχραί, 506c). But true beliefs directed either by a wise teacher (as in the Meno) or by
a virtuous constitution (as in the Republic) are essential steps in obtaining knowledge.
The distinction between mere true beliefs and what we might call “directed” true beliefs is
explicit in the text, too. In Book IV, courage is understood as the “preservation of the customary
belief (δόξης) produced by education … about what sort of things should be feared” (429c). Then,
a moment later, Socrates opines that true beliefs about what should be feared, on their own, do not
constitute courage: “accurate opinion about the same matters, if it has not come about through
education … you would apparently call something other than ‘courage’ ” (430b). Educational
direction must be the means by which a true belief may be, in Vogt’s sense, “directed” at the Forms.
I would like to suggest, then, that Vogt is actually wrong to fault the traditional
interpretation, despite the fact that she is right to suggest that beliefs may be (in some sense) directed
38
Ibid., 64.
72
at the Forms. The information-rich names created by the lawgiver and supervised by the dialectician
are images of the Forms, and it is these that contain the raw materials of true belief. As such, Vogt
is right to say that the opinions of the students of Kallipolis are directed at the Forms in an inferior
way, but only if she qualifies this by saying that these beliefs are directed at the inferior verbal
imitations of the Forms.
The lawgivers of Plato’s perfect state provide educational direction through musical
education: that is, an accurate education through names presided over by the dialectician. Although
the content contained within these educational names is not exact, and is certainly subject to revision
(see 504b), the names are nevertheless the indispensable stepping stones to virtue and knowledge.
Returning to Socrates’s discussion of courage, he there says that
We were contriving [in the case of musical education and physical training] precisely
this: the noblest and most excellent way for [the future guardians] to be convinced
by the laws and receive them like a dye, so that their belief about what should feared
and all the rest might be steadfast (lit. dye-fast), held in check by their nature and their
virtuous upbringing, lest their dyes be washed out by the most cutting detergents:
pleasure, … pain, fear, and desire. (430a-b, italics mine)
The phrase “and all the rest” indicates here that it is precisely the stability of their true beliefs about
virtues in general that classify the guardians as worthy to advance in their education. As Cross and
Woozley put it, at this point in the Republic, “the mark of the ruler was true belief”39 – and, as I
discussed in the earlier section, true belief cannot be obtained without accurate nomenclature.
Book IV, then, provides a strong basis for the claim that true opinions and true names are
necessary for advancement, and thus necessary for knowledge. There is narrative evidence for this
R.C. Cross and A.D. Woozley, Plato’s Republic: A Philosophical Commentary (New York: St. Martin’s Press, 1964), 200.
There are many other suggestions that a central goal of musical education is to instill true beliefs about virtue. Take,
for example, Socrates’s insistence that children not hear “any chance stories told by random people,” lest they “receive
opinions in their souls that are mostly contrary to those which we want them to possess in adulthood” (377b).
Moreover, we are told that these stories will “mold [the] children’s souls … much more than [the nurses and mothers]
mold their bodies by handling them” (377c).
39
73
claim, as well. The characters of Glaucon and Adeimantus enter into the conversation of Books IIX with the firm and unshakeable belief that justice is beneficial to its possessor, despite the fact that
they are able to produce a very strong argument that it is not. Socrates praises them, saying that
there must be something godlike in them if they “are not persuaded that injustice is better than
justice though [they] can argue so capably on its behalf” (368a). Now, although it almost certainly
going too far to say that Glaucon and Adeimantus hold all the true beliefs taught in Kallipolis
(despite their notable education and military experience40), it is nevertheless reasonably clear that
Socrates’s praise for the irrational/arational true beliefs of Plato’s brothers foreshadows the critical
role that such dogmatic beliefs play in his constitution.
The notion that true opinion is necessary for knowledge appears elsewhere in Plato’s
dialogues, as well. The Eleatic Stranger in the Statesman balks at the possibility of a person acquiring
knowledge without first believing correctly (278d-e). In her extended study on the Symposium, Frisbee
Sheffield convincingly argues that the philosophical activity Eros instigates in the rational soul results
in two things: accounts (logoi) and true beliefs.41 True belief is midway between ignorance and
wisdom (Symp. 202a), and the philosopher’s journey from ignorance to wisdom involves a sort of
ratcheting back and forth between aporia (resourcelessness, “no way”) and euporia (resourcefulness,
“good way”), whose offspring Eros is portrayed as (203c-e).42 When the soul realizes its own aporia,
it schemes after some piece of wisdom and creates a λόγος, but this λόγος turns out to be unstable,
since it is said to “ebb away” (203e). Through such progressions from aporia to euporia, engaged in
40
See Sandra Peterson, Socrates and Philosophy in the Dialogues of Plato (Cambridge University Press, 2011), 151-159.
Frisbee C. C. Sheffield, Plato’s Symposium: The Ethics of Desire (New York: Oxford University Press, 2006): 64 and
note 31.
42
Sheffield, ibid., 65.
41
74
repeatedly over a long period of time, the philosopher moves from ignorance to true belief, and then
from true belief to knowledge.
An almost identical progression is envisioned in the Seventh Letter, the work of either Plato
or an ancient author steeped in Platonic tradition.43 The Letter presents a rather complicated picture
of education, and I will not attempt to work out the details of this picture here.44 The author of the
Letter divides the progress of education into five classes: the name, the definition45, the likeness, the
piece of knowledge, and the thing itself. (In Greek, ὄνομα, λόγος, εἴδωλον, ἐπιστήμη, and ὃ δὴ
ἐστιν, 342b.) A course of instruction that continually leads a person up and down through these
lower four classes “just barely manages to breed knowledge (ἐπιστήμην ἐνέτεκεν) of a naturally
good object in a naturally good individual” (343e).
Now if genuine knowledge is only obtained at the fifth step, clearly the label ἐπιστήμη for
the fourth step is somewhat misleading. The author of the Letter helps us out, however, by clarifying
that the fourth step is a step that involves “both thought and true opinion” (καὶ νοῦς ἀληθής τε
δόξα, 342c), and that even this step depends upon the evidence of the senses (343c). It is no stretch,
then, to propose that all four classes exist at the level of Doxa. The author emphasizes how unstable
and unreliable the four lower classes are (343b), and this instability recalls multiple passages in the
43
For authorship, see Plato and Glenn R. Morrow. Plato's Epistles (Indianapolis: The Bobbs-Merrill Company, Inc.,
1962). For an excellent discussion of the problems raised by Plato’s epistemology in the Letter, see Kurt Von Fritz,
“The Philosophical Passage in the Seventh Platonic Letter and the Problem of Plato's Esoteric Philosophy,” in Essays in
Ancient Greek Philosophy, eds. John Peter Anton, George L. Kustas and Anthony Preus (New York: State University of
New York Press, 1971).
44
Rachel Barney’s book on the Cratylus contains a chapter entitled “The Cratylus and After: Names and Logoi,” which
explores the connections between the Cratylus and the Letter, and has influenced my comments here. Barney, Names
and Nature.
45
The meaning of λόγος here is, per usual, somewhat unclear. In context, it is clear that the λόγοι of this passage
have more epistemic dignity than mere words (ὄνοματα), but less dignity than likenesses (εἴδωλα). Plato gives an
example of a λόγος: the λόγος of a circle is “the thing whose edges are removed by the same distance from the center
in every direction” (342b). I do not think we will go amiss by saying that such a λόγος is a “description” or a
“definition.”
75
Republic about the difficulties with Doxa, including 479a-c, where “things” are said to “equivocate”
(ἐπαμφοτερίζειν), and the three fingers passage at 523a-526b, which describes the senses as
“deficient” (ἐνδεῶς, 523e).46
Once again, then, true belief is a prerequisite for knowledge.47 In the Letter, moreover,
increasingly correct ὄνοματα are explicitly necessary for the progression from ignorance to belief to
knowledge. The Seventh Letter explicitly places the blame for the instability of Doxa on the
inadequacy of language.
These four [classes] make an effort to disclose the particular instance of a thing (τὸ
ποῖόν τι) quite as much as the genuine nature of the thing itself (τὸ ὂν ἑκάστου),
because of the inadequacy of language (διὰ τὸ τῶν λόγων ἀσθενές). … Now we
affirm that none of our names are at all stable (βέβαιον); nothing prevents the
things we now call “round” from being called “straight,” and vice versa, nor would
the names have any less stability (οὐδὲν ἧττον βεβαίως) if they were changed from
their current meanings. (342e-343b)
The conventionality of language echoes the latter part of the Cratylus. And yet, clearly the goal being
sought is some level of objective correctness, a correctness which is only possible in the stability that
links concepts with real objects, not in the stability that links phonemes to concepts. It is the
objective correctness of names that allows one to advance forward through the four stages; this
objective correctness, therefore, is a necessary precondition for knowledge.
Conclusion
My analysis above has illuminated at least this much: The Cratylus is not an isolated dialogue,
and certain theories Plato advances there play important roles in other dialogues. The idea that
46
The mention of the senses here is relevant, because the Seventh Letter passage squarely places the blame for the
instability of the four lower classes on the fact that they rely on the senses.
47
It is worth noting, at least in passing, how striking the resemblance between this passage and the passage I discussed
above, from the Symposium, is. Both passages describe a ladder toward knowledge that involves both downward and
upward movement, a realization of insight and a realization of ignorance. Both passages make explicit mention of logoi
and true belief as stepping stones to knowledge.
76
names are imbued with meaning by the lawmaker is a characteristic assertion Plato makes
throughout his middle period, and it is particularly important in the Republic. The negative attitude
Plato has toward education through names in the Cratylus is a result of his antipathy toward sophists,
not an indication that no sort of education through names could be efficacious. For, as we have
seen, the proper method of educating the guardians, in the Republic, begins with an education
through names.
My explorations in this chapter and the previous one go some way to explaining the role of
philosophy of language in Plato’s system of primary education, but they do not situate language in
the broader metaphysical vision of the Republic. In the following two chapters, I proceed under the
hypothesis that language can be situated in that context, that we can understand language in the
context of various other mainstays in Platonic ontology, like appearances and Forms. If I can succeed
in demonstrating that, it will make an even stronger case for what I have been saying here: that the
regulations on permissible speech in Kallipolis are manifestations of the standards of naming
expressed in the Cratylus.
77
CHAPTER 3: MIDDLE VIRTUE
ACCURATE “POETRY” IN REPUBLIC X
In this chapter, I should like to propose a number of novel suggestions about the nature of
virtue in the final book of the Republic. In all the voluminous debates about Rep. X’s apparent
reversal on the issue of poetry, Plato’s real concern – which is virtue, not poetry – has been somewhat
lost. My argument here is that the sort of understanding of Platonic language I have proposed in
Chapters I and II illuminate the discussions about virtue in Book X. In Book X, Plato envisions a
notion of virtue which is neither the virtue manifested in imitative poetry nor the virtue possessed
by a philosopher. This “middle” virtue, I argue, is a familiar object in disguise: it is a word. Both
this chapter and the next will attempt to do something few people have attempted to do: they will
attempt to situate words as metaphysical entities in the metaphysical vision of Plato’s Republic. In
my work on these chapters, I hope to shed light on what sort of existence words have, in Plato’s view.
The chapter will proceed by moving from a discussion about couches to a discussion about
virtue, and then from a discussion about virtue to a discussion about words. The first section will
focus on the role of the craftsman in Book X’s depiction of the three kinds of couch, and will draw
a number of connections between this discussion and the Cratylus. The next section will consider
how we can use two emphases of Book X – couches and poetry – to infer the truth about something
Plato did not emphasize: what I will call “intermediate virtue,” the type of virtue that corresponds
to the craftsman’s couch. I will argue that intermediate virtue should be identified with demotic
virtue, and that Plato envisions it as a kind of sensory object that can, in some sense, genuinely
educate a person about right action. I will then argue that the virtue words are objects that fit that
precise description, and I will identify four common characteristics between these words and
78
intermediates. Finally, I will draw these four commonalities into an argument that virtue words and
intermediate virtues are, for Plato, the same thing.
That is the plan for the chapter. I will begin, however, by discussing upholstery.
Section 1: On Couches
At the beginning of Book X, Socrates describes three couches and three people associated
with couches. The first type of couch has its being “in nature” (ἐν τῇ φύσει οὖσα, 597b) and is
created by God. God made “only one of these” and it has “real existence” (597c). The second couch
is created by the craftsman (δημιουργός), who looks toward the Form of a couch in order to fashion
it (πρὸς τὴν ἰδέαν βλέπων οὕτω ποιεῖ, 596b).1
This “intermediate” couch has two aspects: a
“reality” (τὸ ὄν, 598b) to the way it is, and an “appearance” (τὸ φαινόμενον, 598b) which changes
depending on the perspective of the viewer (598a). The imitative painter does not produce his couch
by looking toward the Form, nor by looking to the “reality” of the intermediate couch, but rather by
looking to the appearance of the intermediate (πρὸς τὸ φαινόμενον, 598b). The painted couch is
not genuinely the “sort of thing” (τοιούτου, 597d) God and the craftsman produce, since it does
not serve the purpose of a couch (e.g. you cannot sit on it). It is merely an imitation of one particular
appearance of the intermediate. 2
The three people associated with couches are these: the user, the craftsman, and the painter.
(See 601d-602c). The action of the painter, we have seen, is to imitate appearances. The action of
the “user” – and, indeed, what Plato means by this term – is somewhat obscure; I shall not say much
1
Similar language is used in the Cratylus, in reference to the artificer of names: the legislator puts names into syllables
by “looking toward the genuine name of the thing itself” (βλέποντα πρὸς αὐτὸ ἐκεῖνο ὃ ἔστιν ὄνομα, 389d); and
the true namegiver is “the one who sets his gaze on the natural name of each thing” (ἐκεῖνον τὸν ἀποβλέποντα εἰς
τὸ τῇ φύσειὄνομα ὂν ἑκάστῳ, 390e), and puts this name into appropriate syllables.
2
See Jessica Moss, “What is Imitative Poetry and Why is it Bad?” in The Cambridge Companion to Plato's Republic, ed. G.
R. F. Ferrari (Cambridge: Cambridge University Press, 2007), 415–44.
79
about it here, except that the user appears to be in direct contact with the Form of the Couch.3 My
concern is with the craftsman. The craftsman takes information acquired from the user, and uses
this information as a template for his “making” (ποίησις). Socrates’s mention of the user at 601c,
then, must be a slight revision of what is said at 596b. The craftsman, strictly speaking, does not
“look toward” the Form of the Couch, but rather receives information about the couch from someone
who has looked toward the Form. It is this information that provides the craftsman with the true
beliefs that allow him to produce a “good instance” of a couch (where a “good instance” is a material
couch that somehow conforms to or instantiates the Form of Couch).
Although it has seldom been mentioned in the scholarship of Book X, the commonalities
between the passages I just described and the opening of the Cratylus are extraordinary. Both
dialogues distinguish explicitly between the “maker” (or “craftsman”) and the “user” (Crat. 388e390d; Rep. 601d-602b). Both dialogues argue that the maker employs the knowledge of the user as
a guide in making (Crat. 390d; Rep. 601e-602a). Both dialogues suggest that there are standards of
use by which the product of the craftsman may be normatively evaluated (Crat. 389d-390a, 431d;
Rep. 601e, 598e). The subject matter of the two passages, however, would appear to be quite
different: the Cratylus is discussing words, whereas the Republic is discussing couches. Nevertheless,
in both cases, it is heavily implied that both words and couches have standards for judgment, and at
least some of the time these standards for judgment seem to be the Forms (Crat. 438d-e, 439d; Rep.
597d).
It is obvious that the “user” is a person who, say, sits on the couch, and Plato clearly puts the user in an exalted
cognitive position in respect to the couch. But it is hard to see why the user would have any specialized knowledge
about the couch, much less the sort of knowledge that would be absolutely essential to the craftsman. Perhaps the
claims about the user make the most sense when we consider a context where the craftsman is not capable of using the
relevant object: fighter jets and skateboards might be good examples.
3
80
In Republic X, of course, Socrates is not really concerned with interior decorating; he is
concerned with the nature of virtue, and the vicious effects of imitative poetry. The entire
conversation about imitating material objects aims to reveal how deficient poetic imitations of moral
objects (the virtues) are. When the discussion turns to virtue, we can understand how the
connections between the Cratylus and Republic X can illuminate the nature of virtue in the Republic.
I believe that the moral analogue of the craftsman’s couch is a linguistic object: the particular word
we use to describe each individual virtue. The craftsman of virtue uses refined and accurate poetry
(“making”) to teach true beliefs about this word, I will argue, and the ordinary citizen acquires
demotic virtue as a result of this interaction.
Section 2: On Virtue
As we have seen, Book X presents an “intermediate” couch: a couch that is neither the
“Form” of the Couch, nor any particular sensory appearance of a couch. 4 This intermediate couch
is the creation of the craftsman. The couch only matters, however, insofar as it is analogous with
the genuine topic of interest in Book X: virtue. Socrates aims to demonstrate that the imitator of
virtue (the poet) makes products quite as deficient as the imitator of couches (the painter). For this
analogy to work, however, there must be both three-part divisions that I discussed above, in the
context of virtue as well as couches.5 And this is what we find in Book X, although Plato’s depiction
By “appearance,” I mean either (a) the type of sensory experience caused by an intermediate (e.g. a couch as seen
from a particular angle), or (b) a reflection, shadow, or image of an intermediate object. Perhaps unexpectedly, Plato
does not claim there is an important distinction between seeing a couch from one perspective and seeing a painting or
a reflection of the same couch. All these are phantasmata. For an illuminating discussion of the nature of
appearances, see again Moss, “What is Imitative Poetry?”.
5
A number of commentators have recognized this, including Burnyeat, in his third lecture on Culture and Society in
Plato’s Republic. Burnyeat talks about intermediate virtue as an “ordinary, second-level” (314) version of understanding
virtue, and says that “the nonphilosopher’s second-level knowledge of values is ‘knowledge’ only by courtesy, better
called opinion” (315). M.F. Burnyeat, “Culture and Society in Plato’s Republic,” The Tanner Lectures on Human
Values 20 (Salt Lake City, 1999), 314-315.
4
81
of certain aspects of the analogy is not always quite clear. The poet is the analogue of the painter;
the lawmaker is the analogue of the craftsman; and the philosopher is the analogue of the “user.”6
Where the craftsman makes an intermediate couch, the lawmaker makes “intermediate virtue.”
USER
COUCHES
VIRTUES
Form of the Couch
Form of the Virtue
Intermediate couch
Intermediate virtue
(Product of craftsman)
(Product of craftsman of virtue?)
Painting of a couch
Virtue portrayed by an imitative poet
MAKER
IMITATOR
If this were all the information the Republic gave us about intermediate virtue, an analysis of
its nature would be speculative indeed. But we are given one additional clue earlier in the Republic,
and this clue is a decisive one. In Book VI, Socrates explicitly calls the guardians of his imagined
polis “craftsmen of temperance and justice and the whole of demotic virtue” (500d, δημιουργὸν ...
σωφροσύνης τε καὶ δικαιοσύνης καὶ συμπάσης τῆς δημοτικῆς ἀρετῆς). This passage clearly
connects the Book II-IX themes of “molding” or “shaping” (πλάττειν, 377b-c, repeated at 500d)
citizens into virtuous activity with the Book X theme of “making/crafting virtue.” In Kallipolis, the
craftsman of virtue is the guardian, and the product he produces is demotic virtue.
6
Imitators of virtue (poets) are contrasted with lawgivers like Lycurgus, Charondas and Solon; innovators like Thales
and Anacharsis; practitioners of a “way of life” (600b) like Pythagoras. The common thread here is that these latter
people claim to “understand what sorts of habits make people better or worse” (599d), and that the beneficiaries of
their craft “take part in” a certain type of “training” (παιδείας, 600d) which they believe is indispensable to them
(599e; 600c-e). Intermediate virtue is the object of this training.
82
What is demotic virtue, then? In Platonic tradition, it is understood as the type of civic
virtue that “comes from habituation without philosophy and reason.”7
My analysis reveals
something more. Demotic virtue manifests speech and action that is somehow in accordance with
the Forms, without involving knowledge (Phaedo 82a-b; cf. Rep. 619c, 518d). As Veltman and Barney
(among others) have argued, demotic virtue is available to nonphilosophers who possess the correct
beliefs about virtue – and this arguably includes every single citizen of Kallipolis.8 Athenian
lawgivers, like Solon, cannot effectively craft demotic virtue because they do not know the truth
about the Forms, but (as I argued in the first two chapters) Kallipolan lawmakers have no such
barrier. The guardian crafts demotic virtue in a way that is fully adequate for the moral education
of a citizen; as we see in Book III, this involves a thorough indoctrination in certain true beliefs
about virtue that are judged essential to the city.9 A craftsman of virtue, then, is someone who
devotes his life to the education of others in a particular way, by instituting norms in the form of
true beliefs, or by instilling habits. This craftsman has “correct belief” (601e) about matters of value.
As he bestows these true beliefs upon others, they are empowered to act virtuously; this is
intermediate virtue.
To summarize, then, an intermediate virtue is a product which the craftsman of virtue
provides the broader public by means of (largely verbal) moral education. Ideally, this virtue will
somehow conform to the Form of the relevant virtue, though this connection will only occur when
David Keyt, “Plato and the Ship of State,” in The Blackwell Guide to Plato’s Republic, ed. Gerasimos Santas (Oxford:
Blackwell): 202.
8
Andrea Veltman, “The Justice of the Ordinary Citizen in Plato's Republic,” Polis: The Journal of the Society for the Study
of Greek Political Thought 22 no. 1 (2005): 45-59. Rachel Barney, “Comments on Sarah Broadie, Virtue and Beyond in
Plato and Aristotle,” Southern Journal of Philosophy 43 (2005): 115-25.
9
Ordinary citizens of Kallipolis receive certain lies from their rulers, but none of these lies directly pertain to virtue.
For more on the relationship between demotic education and deception in Plato, see Nicholas D. Smith and Thomas
Brickhouse, “Justice and Dishonesty in Plato's Republic,” Southern Journal of Philosophy 21 no. 1 (1983): 79-95.
7
83
the process is guided by a dialectician. The intermediate virtue manifests itself in the beliefs of the
individual, and these beliefs explain the moral decisions of the individual. 10 The ordinary person
possessed of an intermediate virtue will believe in certain norms, and possess certain habits, and
these norms and habits will tend to make her a better person (599d). Education may be inferior in
either (or both) of two ways: (1) by inculcating intermediate virtues that are not themselves in
complete accordance with the Forms (i.e. the laws of Solon or the teachings of Protagoras), or (2) by
operating at one further remove from reality, as the imitative poets do. 11 Once we understand that
(on Plato’s view) poets like Homer are both using defective models (defective intermediates) and
operating at a further remove, we may be able to see more clearly why Plato has such a negative
attitude toward poetry. We will certainly see why some sort of reformed musical education is a
critical ingredient of his republic.
Section 3: On Words
My position in this chapter is that intermediate virtue plays a more central role in Plato’s
vision of philosophical investigation than it is ordinarily given credit for. Intermediate virtues are
This point that action manifests belief may be simply thought of as a corollary to Plato’s intellectualism about moral
behavior. On the traditional interpretation, Plato is an intellectualist about virtue: he believes that virtue is
knowledge, and he therefore believes that one cannot learn that Xing is wrong without thereby ceasing to X. When we
extend this view to true and false beliefs, we would have the view that “demotic virtue is true belief,” that “one cannot
acquire a belief that X is wrong without thereby ceasing to X.” This interpretation is consistent, so far as I can tell,
with every mention of belief in Plato’s dialogues, and it is positively confirmed by one of them. In the Meno, Socrates
makes the argument that true belief has all the same effects as knowledge (97a-d). Since the action-compelling nature
of knowledge is one of knowledge’s effects, it stands to reason that true belief would be action-compelling, as well.
11
One might think that primary education in Kallipolis operates at this more distant level, as well, but I think that is a
misunderstanding. The “poetry” told to the children there is not imitative (in the sense the word is used in Book X).
This, I would speculate, is because it consists exclusively of true beliefs. As I will discuss in Section 3, and Chapter IV,
the distinction between merely apparent objects and “trustworthy” objects (objects of pistis) is determined by their
objective accuracy.
A number of articles have dealt with the difficult question of why the additional level of imitation manifested in
poetry would lead people astray. Pictures provide only partial information about their subject matter, but this partial
information is not inaccurate merely because it is partial. For ways of addressing this problem, see Moss “What Is
Imitative?” (above).
10
84
examined through the “serious” (σπουδαῖόν, 603c) or intellectual portion of the soul12, they are
measureable (603e; cf. 602d-603a), and they have standards of correctness that are determined by
the Forms (598d-e). Intermediate objects cause appearances to appear, and yet they themselves are
“caused,” in some sense of the word, by the Forms (601e-602a; cf. 509b). They occupy a place
between what is fully rational (Forms) and what is irrational (appearances). Intermediate virtues are
similar to the Forms in that they can be understood through the intellect, but dissimilar to the Forms
in that they are objects of sensory experience. Their status as sensory objects that play an educational
role has not been emphasized in the literature.
Once we realize that intermediate virtues are sensory objects that play an educational role, it
is a small yet ambitious step to say that they are words. I shall take that step by appealing to the
Cratylus, and specifically the claims I have made about Plato’s philosophy of language in Chapters I
and II. In those chapters, I believe I have established that (1) the Cratylus was a serious attempt at
philosophy of language, even if the etymologies were not always meant seriously, (2) the dialogue’s
consistent assertion that words are “images” or “imitations” is a genuine view that Plato held, and
(3) that, when the Cratylus addresses the connection between “things” (πράγματα) and their names,
the “things” being talked about are the Forms.13
There is some dispute among scholars about how the twofold distinction between the “serious” (603c) and the
“irrational” (605d) parts of the soul in Book X should be reconciled with the threefold distinction between reason,
spirit, and appetite in Book III. Competing proposals can be found in Jessica Moss, “Appearances and Calculations:
Plato's Division of the Soul,” Oxford Studies in Ancient Philosophy 34 (2008): 35-68. Todd Ganson, “The Rational/NonRational Distinction in Plato's Republic,” Oxford Studies in Ancient Philosophy 36 (2009): 179-197. For a proposal that
the distinction in Book X is a distinction within the rational soul, see Alexander Nehamas, “Plato on Imitation and
Poetry in Republic 10,” in Plato on Beauty, Wisdom, and the Arts, eds. J. M. E. Moravcsik and Philip Temko (Rowman
and Littlefield, 1982).
13
This assumption is supported by the discussion of “true beauty” at 439c-d, which is strikingly similar to Plato’s
discussion of the “form of beauty” in the Symposium.
12
85
I have argued that Socrates portrays intermediate virtues as sensory objects that somehow
“link up” with Forms, in something like the way that the couch made by a carpenter “links up” with
the Form of the Couch. Plato describes this link by using the term mimesis. The intermediate object
corresponding to the Form of Courage is an imitation of Courage that is nevertheless (unlike the
Form) a sensory object – in particular, an auditory object. Words are, on a straightforward
understanding of Plato’s philosophy of language, precisely this sort of object. They are “imitations”
(Crat. 423b) or “images” (Crat. 431d) of the Forms; they are audible objects; they have the power to
refer to real things (πράγματα, Crat. 436a) that exist in a way that they themselves do not. The
contrast between the Forms of virtues and intermediate virtues, then, is the contrast between purely
cognitive objects, on the one hand, and sensory objects that refer to cognitive objects, on the other.
In the schema given by Book V, intermediates are objects of opinion, not knowledge. They may be,
at best, the objects of true belief; they can never be known. Nevertheless, they are “imitations” of
things that can be known. In terms of qualities, they are remarkably similar to words.
I will put forward four significant claims about how intermediates are similar to words: (1)
Intermediates and words “are” in a sense that mere appearances are not. (2) Both intermediates and
words are objects of sensory experiences, in the sense that they “show themselves” to the senses. (3)
Both intermediates and words are the products of the craftsman’s true opinions. (4) Intermediates
and words have a level of stability that is manifested in the fact that they can be measured, and that
they can be incorrect. For the remainder of this chapter, I will explicate those four claims. The
combination of the four will, I hope, go a long way toward establishing my thesis.
1. Intermediates and words “are” in a sense that mere appearances are not.
86
My first comparison between intermediates and words turns on an often overlooked passage
in Book X. In this passage, which I have already touched on, Socrates makes the controversial
suggestion that objects at the intermediate level are real in a way that appearances are not14:
Please tell me this about the painter: Does it seem to you that he attempts to copy
each thing itself as it genuinely (ἐν τῇ φύσει) exists, or the works of the craftsman?
The works of the craftsman. As they are, or as they appear to be? [As they appear].
(597e-598a)
Three categories are indicated: (1) realities that do not appear (Forms), (2) intermediates that possess
both an appearance and a reality, and (3) appearances detached from any reality.15 What can
Socrates mean by saying that the couch made by the carpenter has a “reality” (τὸ ὄν, 598b), a word
that we might have expected him to reserve for immaterial Forms? What is this “reality”? It is
obviously not the sort of robust reality that the Forms have, since Plato expressly specifies that the
Form of the Couch has a sort of “genuine existence” (ὄντως οὔσης) that a particular couch (κλίνης
τινὸς, 597d) doesn’t have. Though it is difficult to discern Plato’s purpose in using the term τὸ ὄν
in this context, I will suggest and defend the following understanding: in relation to the Forms, the
carpenter’s couch is “coming to be” (γίγνεσθαι), but in comparison with appearances, it “is” (εἶναι).16
I follow Vlastos, among others, in considering the language of “being” in Book X to be a matter of genuineness or
“degrees of reality,” not a matter of literal existence and nonexistence. A painting of a couch clearly exists, but its
existence (insofar as it is a “couch”) is derivative in a way that the couch’s existence is not. Gregory Vlastos, “A
Metaphysical Paradox,” Proceedings and Addresses of the American Philosophical Association 39 (1965): 5-19.
15
Tate claims that a fourth category is implied by the question “as they are, or as they appear?” He thinks that such a
question suggests a type of imitation (especially in the realm of poetry) that imitates the reality of things. But I would
argue that this proposal is incoherent in the realm of painting, since a painter who copied things as they really are
would be a craftsman, not a painter. (See the argument about two Cratyluses at Crat. 432a-b.) So it is unclear why
there could be poetic imitations of the reality of things, when there can be no pictographic imitation, and it is far
more plausible that Plato considers all poetry to be an imitation of appearances. Belfiore provides a helpful responses
to Tate. J. Tate, “‘Imitation’ in Plato's Republic,” CQ 22 (1928): 16-23. Elizabeth Belfiore, “A Theory of Imitation in
Plato's Republic,” in Ancient Literary Criticism, ed. Andrew Laird (Oxford: Oxford University Press, 2006): 133-135.
16
In my use of “is” and “comes to be,” I intend to preserve any ambiguities that might be found in the Greek variants
of these terms. For an interesting argument that Greek philosophy systematically benefited from the ambiguity
contained in the verb εἶναι, see Charles Kahn, “The Greek Verb ‘To Be’ and the Concept of Being,” Foundations of
Language 2 (1966).
14
87
The same sort of contrast is found in the Timaeus, where Timaeus describes a “third and intermediate
kind of existing thing” (οὐσίας εἶδος), which is created out of the combination of that “which exists
always in the same way” (ἀεὶ κατὰ ταὐτὰ ἐχούσης οὐσίας) and that “which is divided into physical
entities” (τῆς … περὶ τὰ σώματα γιγνομένης μεριστῆς, Tim. 35a). The use of οὐσίας to describe
the intermediate resonates with this usage of τὸ ὄν in Book X.
The sense in which the carpenter’s couch is coming to be, I think, should be reasonably clear:
the carpenter makes it, and, once it is made, it remains subject to change (e.g., it could be
reupholstered). The sense in which the carpenter’s couch “is,” perhaps, is less clear. In order to
understand this, we shall have to consider a consistent (and perhaps annoying) trait in Plato’s
discussions of ontology. Plato often seems to insert epistemological arguments where we might
expect metaphysical ones. Book X is no exception. In his discussion of poetry in Book X, he
distinguishes the ontological status of objects on the basis of our cognitive access to them. We have
access to objects in two primary ways: (1) Through measurement (603a) and/or an understanding of
the purpose of things (601e-602a) and (2) Through an appeal to mere appearances detached from
any sort of reality (602a-b).17 The first type of access only allows for an engagement with objects that
“are” (in some relevant sense): the Form of Couch and the couch made by the carpenter. Things
that have content independent from their appearances can be measured and understood; other things
cannot. Appearances can only be accessed through other appearances.
Plato seems to be suggesting that, since we do not cognitively engage with couch-appearances
in the same way that we do with couches, there must be two types of objects with which we engage.18
17
A third form of access is mentioned at 601e: we might access cognitive objects through correct testimony about a
thing’s nature or purpose.
18
This may be begging the question against the bundle theorist, who says that couches just are a bundle of
appearances, but that is an objection Plato would plausibly have missed. If one does give that part of the argument a
88
This point applies equally well to words, as well as couches. Insofar as a word is an appearance, it
consists of various sounds with different qualities; but what makes the phoneme a word and not a
succession of sounds is its capacity to refer. Modern philosophers might explain this property of
language by saying that there is something interesting about our intentions when we use particular
words. Plato appears to have thought, to the contrary, that there was something interesting about
the words themselves. In the Cratylus and elsewhere, he expressed this point in terms of imitation.
Words are imitations of the things they refer to. They are distinguished from other sensory objects,
then, not only insofar as they are referential, but also insofar as they are representational. Just as a
painting is an image of what it represents, so a word is an image of the Form to which it refers (if
there is such a Form).19
The language of representation is a major motif in Book X. With respect to couches, we are
told that the craftsman of the couch makes “something that is in some way like reality (τὸ ὄν), but
is not (ὄν δὲ οὐ)” (597a). The craftsman’s couch has its level of “reality,” then, precisely because it
resembles real being. A couch will be a good couch if it resembles the Form closely, a bad couch if
the resemblance is strained. A word will be a good word if it resembles its referent closely – that is,
if the beliefs which are somehow “bound up” with that word are true of its referent. Both
intermediates and words, then, have the capacity to “borrow” some level of real existence, insofar as
they attain to the standards which govern them.
pass, the conclusion that couches are ontologically superior to couch appearances would appear considerably more
sensible. There is something airy and ephemeral, even “unreal,” about an object that available to experience, but
obstinately eludes analysis.
19
This claim may seem objectionable, in that I speak as if the theory of Forms was already functional in the Cratylus.
That is not my intention, however. Although it seems likely that the Cratylus is engaged with Plato’s specialized
doctrine of Forms – see, for instance, 439d-e, where Socrates’s discussion of “the beautiful itself” echoes Symp. 211a-b
– the theory of naming discussed in the Cratylus need not make any specific mention of Forms, only “things” (cf. the
discussion of investigating things “through themselves” at 438e-439b). If we apply that theory to an ontology that does
include Forms, we find that Forms take the place of the “things” which clearly do play a central role in the Cratylus.
89
2. Both intermediates and words are objects of sensory experiences, in the sense that they
“show themselves” to the senses.
Despite the fact that Socrates says that the intermediate couch has a certain sort of being
(598a-b, 599a), he is also at pains to clarify that it exists in a way that is inferior to the Forms. The
couch “does not differ from itself in the slightest” (598a) depending on the angle from which it is
viewed, but it is nevertheless an object that is subject to the senses: it “presents different appearances”
(ibid.) in different situations. The couch emits “phantasms” or “appearances” (see 598b) which are
the immediate object of our sensory experience, just as shadows are the immediate object of the
prisoner’s experience in the cave.
The nature of the intermediate couch might be understood using the language of
dimensionality. We often say that we can “see” three-dimensional objects, but it is nevertheless false
that we can ever see all three dimensions at once. What we see, at any given moment, is a twodimensional visual field which we know – through testimony and inference, presumably – contains
three-dimensional objects. The third dimension is never simply “given” to us in experience; and
thus, the existence of the three-dimensional object itself is not given to us in experience, either.20 In
just the same way, the craftsman’s couch is not directly accessible to the senses, but only indirectly
accessible. When we talk about “the couch,” we are talking about the three-dimensional object, but
none of our experiences contain the object in its full dimensionality. “The couch” is the thing that
causes couch appearances, just as the three-dimensional (or “real”) apple is what causes my twodimensional experience of an apple from various perspectives.
20
Two-dimensional images are
The point applies much more straightforwardly to the sense of sight than the sense of touch, however. And it is
much more difficult to present an argument for something like the “compresence of opposites” when we consider the
sense of touch, as opposed to the sense of sight, since touch does not seem to admit of perspectives in the same way. I
take it for granted that Plato was not particularly concerned with the cognitive implications of touch.
90
straightforwardly an object of the senses; three-dimensional objects are only objects of the senses
insofar as they cause such two-dimensional images to exist.
I would like to propose that this notion of dimensionality can be fruitfully compared to the
sort of distinction I discussed in Chapter I, the distinction between a “name-as-phoneme” and a
“name-as-concept.” Socrates, in the Cratylus, mentions the sharp distinction we must make between
an “ideal name” and particular phonetic structures in human languages – more than one of which
may accurately embody (or instantiate) this name. We might call this ideal name a “conceptual
name,” the name that carries with it information about the stereotypes speakers apply to the thing
being named. This information is opaque to a speaker of the language unfamiliar with the word,
until they draw conclusions from contextual clues about the beliefs embedded in the concept, clues
which emerge from a close observation of how the word is used. Just like the intermediate couch
presents different appearances at different times, a conceptual name is appealed to (by the use of a
phoneme) in a number of different scenarios, which appear quite different. A “red” couch does not
look much like a “red” stop light. Nevertheless, just as the observer acquires beliefs about the couch
by accumulating and analyzing couch appearances, the linguistic learner acquires the culturally
endorsed concept of redness by observing and analyzing the circumstances in which the phoneme
“red” is used.
Moreover, there seems to be a sense in which the usages of the phoneme “red” are caused by
the concept of redness, similar to the way that couch-appearances are caused by the intermediate
couch. The reason that people use “red” instead of “potato” to apply to certain objects is that they
have certain stereotypical understandings of what redness is, and their concept of redness applies to
those objects. Note that their action of using the phoneme is not caused by the Form of Redness (at
91
least not directly), because they could correctly apply their understanding of the concept redness
without describing something that actually participates in the Form. Thus a medieval doctor might
make a speech act blaming an illness on “phlegm,” where the use of this word (rather than another)
is very much caused by the doctor’s concept of phlegm, not based on the actual properties of phlegm.
The word “cause” here, of course, does not refer to any notion of an efficient cause, but rather to an
explanatory cause.21 There is truly a sense in which our concepts – even our flawed concepts –
explain our actions.
The interpretation I propose here explains what might otherwise be a puzzling aspect of Book
X’s ontology: what can Plato mean by suggesting that intermediates are “more real” than appearances
– in particular, that intermediate virtues are more real than appearances of virtue? If we follow Plato
in saying that “real” things are precisely those things that cannot be experienced through the senses,
then it should follow that an object will be “more real” insofar as understanding it involves abstracting
away from the senses – as we do when we come to understand a concept, as an inference from the way
a phoneme is used.22 Plato believes that the knowhow that goes into making (or conceptualizing) a
couch is not purely sensory, and he certainly believes that our understanding of words involves a
type of cognition that transcends a merely sensory encounter. He calls this type of cognition pistis
(511e), and he regularly suggests that this is the type of cognition associated with true opinion. And
so I will turn now to true opinion.
3. Both intermediates and words are the products of the craftsman’s true opinions.
Obviously, calling it a “formal” cause would be unhelpful, since I am precisely insisting that the relevant speech acts
are not (directly) explained by the Forms.
22
Although I feel that the subject is too large and controversial to broach here, the discussion of “moving forms” in
Book III, in the Timaeus, and elsewhere, surely has some relevance to the topics I discuss here. A moving form, like a
concept, has a substantial resemblance to a Form without being a Form. Moreover, since it is moving, it would appear
to be a sensible object of some sort. See Chapter IV for more on moving forms.
21
92
What does it mean, in Plato’s ontology, to say that something is an object of sensory
apprehension that nevertheless has some important relationship to the Forms? It almost certainly
means, for one, that we should identify the object as an object of “pistis”: it is the sort of object we
might “trust” (πιστεύειν), even if it is not the sort of object that might be known. Both in the
Republic and elsewhere, this sort of object is constantly being referred to as an object of “true
opinion.” Objects of true opinion have a level of reliability that exceeds that of other sensory objects,
as evidenced by the word pistis itself, which often denotes a “pledge of good faith” or an
“assurance.”23 Nevertheless, Plato never gives any suggestion that true opinions “transcend the realm
of the senses” or manage to be directed at the Forms. Indeed, Plato’s conception of belief is, to
modern ears, quite curious for its constant appeal to the senses. In the Sophist and the Theaetetus,
for instance, belief is envisioned as the outcome of a sort of internal dialogue within the soul (Soph.
263e-264a, Theaet. 189e-190a); so far from being a “propositional attitude,” belief is envisioned as a
spoken word.
True opinion is contrasted with knowledge, but it is also contrasted with the type of
ignorance that is not informed by any sort of authority whatsoever. In Book X, Plato dramatizes this
distinction by appealing to the “user,” the “maker” (or craftsman), and the “imitator.” The “user”
of a couch has knowledge of the Couch itself (602a); the craftsman of the (intermediate) couch has
true belief about couches (πίστιν ὀρθὴν, 601e; δόξαν ὀρθὴν, 602a); and the imitator of the couch
only has familiarity with “what appears good to the many, who know nothing” (602b; cf. 493a-c).
The user’s knowledge is (among other things) knowledge of “the sorts of good and bad things that
arise from [an object’s] use” (οἷα ἀγαθὰ ἢ κακὰ … ἐν τῇ χρείᾳ ᾧ χρῆται, 601d-e) and of the type
23
Henry George Liddell et al. A Greek-English Lexicon. (Oxford: Clarendon Press, [1843, 1940] 1996).
93
of condition that could make it fine or defective (κάλλους τε καὶ πονηρίας, 602a). We should
assume, then, that the craftsman of a couch does not know either (a) the Form of the Couch, nor (b)
the properties that would make an individually crafted couch a “good couch” or a “bad couch.”24
Rather, we should say that the craftsman has true beliefs about how to make or how to identify a physical
object that manages to be an imperfect instance of the Form.25 If a couch is made well, it will embody
these true beliefs. The painter, by contrast, merely knows how to create sensory representations that
imitate her sensory experience of the couch.
The craftsman’s true opinions distinguish him from the imitator, and they also, no doubt,
distinguish the craftsman’s product from the imitator’s product. 26 Whatever it is that makes the
intermediate couch more real than the painting of a couch, this quality adheres to the intermediate
couch because the craftsman used his true beliefs to put it there. The “reality” (τὸ ὄν) of an intermediate
object is caused to be there by the action of the craftsman of that object. The craftsman’s true beliefs explain
why the object is not merely a multitude of various appearances, but rather a relatively stable object
of cognition. Moreover, it is the craftsman’s beliefs – not just technical expertise – that are being
evaluated whenever we use standards of correctness to evaluate whether an individual couch is a
24
It may be that knowing the Form of the Couch precisely is knowing the conditions that make individual couches
good or bad couches.
25
This analysis of true belief conforms to other discussions of true belief in the dialogues. So, for instance, the Meno
says that true belief (so long as it remains true) is functionally adequate for making judgments about action (97c) and
the Symposium describes true beliefs as intermediate between knowledge and ignorance (202a), and (on a plausible
reading) suggests that true opinions about the beauty of individual sensible objects are stepping stones to knowledge of
the insensible Form of Beauty (210a-d).
26
Are these true opinions about the Forms? I don’t think so. Socrates uses the example of flute playing, and says that,
whereas the user knows what sorts of flutes lend themselves to skillful use, the maker of the flute merely has true belief
about the excellence or deficiency of the individual flute (602a). The craftsman is not even interested in the Forms for
their own sake, but must be “compelled” (601e) to learn the properties that make flutes good or bad. Once he is
familiar with these properties, however, he has a facility with various materials that might be shaped into an instance
of a flute.
94
“good” or a “bad” couch, since the couch is an embodiment of these beliefs. The couch will be good
insofar as it reflects the Form accurately, and bad insofar as it reflects the Form inaccurately.
This vision of the nature of a craftsman will be familiar to any careful reader of the Cratylus.
The craftsman of words in the Cratylus (the legislator) lays down words in accordance with her own
beliefs: if these beliefs are true, the resultant words will also be “true.” Words are meaning-rich
entities that are only accurate or “true” insofar as the beliefs of the namegiver are true. The
determiner of the truth of a word is, as in the case of couches, the Form that corresponds to it. We
might therefore divide words into two classes: correct and incorrect words. Correct words, like
intermediates, have a rugged connection to the things themselves: they are first-order imitations of
those things. If we start from the obvious assumption that a craftsman necessarily makes something,
and we likewise assume that the product made by the craftsman of virtue manifests his true opinions,
it should begin to appear more and more likely that the product made by the craftsman of virtue is
the (meaning-rich) name of a virtue.
Moreover – and this is perhaps the most striking connection between the Cratylus and Book
X – the same exact distinction between a “user” and a “craftsman” is made in both works. We have
seen, above, that the user in Book X informs the craftsman about the standards by which an
intermediate ought to be made. This is exactly the picture we get in the Cratylus, as well, in relation
to words: “Who would best guide the lawgiver’s activity and judge when it is complete? … Would
this not be the user?” (390c). In both cases, the “user” is identified with the dialectician (ibid.), and
in both cases, the craftsman appears to have an inferior grasp of the object, as compared with the
user. Given that the focus of the later portion of the Cratylus is the opinions of the craftsman of
95
words, it is no stretch to say that the craftsman of words holds true opinions, though not necessarily
knowledge.
4. Intermediates and words have a level of stability that is manifested in the fact that they can be
“measured,” and that they can be incorrect.
The intermediate couch is not simply a sensory object (an “appearance”), but it certainly is
not a purely intellectual object. It is a sort of hybrid, just as a word is a hybrid between a sensory
object and a concept.
The intermediate couch consists of both a reality and a plurality of
appearances, both of which are in some way accessible to the human mind. How do we cognitively
approach the reality of the intermediate couch? Plato seems to leave it open that we might approach
it through its appearances, but it cannot be approached merely through appearances. This point is
not made explicit by Plato, but it is implicit in the image of the couch. “Does a couch differ from
itself in the slightest, if you should view it from the side or the back or any direction at all, or does
it remain identical, while appearing different? … It remains identical” (598a). There are a set of facts
about the couch that are stable, no matter which particular appearances the spectator is engaging
with at the moment. Although these facts may be subject to change in one sense (e.g. the couch
could be destroyed), they are not subject to perspectival change. One cannot say of the couch: “I
moved to the other side of the room; now the couch is no longer rectangular.” The couch has some
sort of stable existence, though the nature of this existence is not given by any particular sensory
experience.27 This stable nature can, at best, be inferred to by certain appearances – an inference
that is not done through the senses (see the discussion of “provocatives” at 523a-535a).
27
Nor even, arguably, through the sum of the appearances. One might not understand the nature of a watch merely
by perceiving it in all possible sensory contexts. Indeed, Plato seems to think that knowing an object’s nature is
synonymous with being able to make that object; and, on that view, sensory perception would certainly not teach a
person to understand watches.
96
The relatively stable existence of the intermediate couch makes it possible for one to make
judgments about the couch as a whole that are true or false. As we have seen, the Form of Couch
acts as a standard of correctness from which to judge the individual couch.28 Epistemologically, the
stable nature of the intermediate couch makes it, unlike appearances, subject to measurement.
Measurement helps us to ascertain the genuine qualities of the couch as a whole, just as
measurement is of assistance to the craftsman who “looks to the Form” in order to produce some
(complete) product.29
How does this notion of measurement apply to words? Here the answer is by no means
obvious. The notion of “measuring” words is nearly incomprehensible, and the idea that a word
can be incorrect or “false” seems hardly more explicable. The latter claim, however, is clearly a thesis
that Plato takes quite seriously30, as we have seen in Chapter One. Just as a couch is a bad couch if
it diverges from the nature of the Form of Couch, a word is a bad (or “false,” Crat. 430c-d) word if
it fails to be an accurate image of the thing it refers to. “Justice” is a false word if it conveys to the
listener information that is false of the Form of Justice.31 Names can be incorrect without failing to
refer, on Plato’s view.
28
In the case of appearances, by contrast, it is puzzling how an appearance could be false, qua appearance, since
appearances lack the sort of stability involved in grounding such a judgment. (See Phil. 59b; Crat. 440a; Theaet. 152154.) One cannot sensibly say that “the stick’s appearing bent when submerged” is false. But it is perfectly coherent
to ask the question of whether the stick is in fact straight, despite its bent appearance (602c-d). Moreover, even if an
appearance might be true (as perhaps suggested by Theaet. 166e-167a), it is necessarily incomplete, since it does not
capture the whole of the object that generates it.
29
Such a claim is puzzling, since human measurements rely on the assumption that at least some of our sensory
experiences are accurate. If no experience is veridical, then no scientific claim can be advanced about the things which
are subject to the senses. The field is ripe for skepticism.
30
Euthydemus (277d-278c), Phaedrus (244c), Philebus (12b-c).
31
Through much of the Cratylus – though perhaps not the whole dialogue – this information is envisioned as being
encoded onomatopoeically in the syllables of the word. See Spellman, “Naming and Knowing.”
97
The claim about measurement remains a puzzle, however. If intermediate virtue is linguistic
in nature, then how does it admit of measurement? We can begin to address this question by looking
at what Socrates says about the behavior of the virtuous individual, in Rep. X:
When it falls out by chance that a reasonable man should, say, lose his son or
anything else he values most highly, we said that he would bear it more easily than
anyone else. … This is not to say that he would not grieve at all, for that is
impossible, but rather that he would act moderately in response to his pain
(μετριάσει δέ πως πρὸς λύπην, 603e).
Those who indulge in tragedy are taken to task because they do not “moderate” their grief. The
word for “moderate” here is μετριάζω, and it is clearly meant to be the moral equivalent to μετρέω
at 602d, a word used to describe the way that intermediate objects like couches are subject to
measurement.32 Socrates says that this sort of moderation of grief proceeds from “reason and law”
(λόγος καὶ νόμος), whereas lack of restraint would flow from “the misfortune itself” (αὐτὸ τὸ
πάθος, 604a-b). Moderation and “measured action” proceeds from the intellectual part of the soul;
dissolute and diffuse action from the “vexatious part” (τὸ ἀγανακτητικόν, 604e).
What does all this have to do with words? For the person learning demotic virtue, a lot. In
Book VIII of the Republic (560c-561a), Plato makes it clear that errors in moral language underlie
errors in moral education. The man who believes that temperance is effeminate (560d) fails to act
in a temperate fashion, precisely because he has the wrong beliefs about temperance. As we have
seen above, a training in words is a training in actions. Plato’s view, in this respect, is strikingly
similar to comments Protagoras makes about education, in the dialogue that takes his name:
As soon as a child can understand what is said to him, nurse, mother, tutor, and the
father vie with each other to make him as good as possible, instructing him through
everything he does or says, pointing out, “This is right and that is wrong, this
honorable and that disgraceful, this holy, that impious; do this, don’t do that.” …
The teachers take good care of this, and when boys have learned their letters and are
ready to understand the written word as formerly the spoken, they set the works of
32
S. Halliwell, Plato: Republic X, Translation and Commentary. (Warminster: Aris and Phillips, 1989): 137.
98
good poets before them on their desks to read and make them learn them by heart,
poems containing much admonition and many stories, eulogies, and panegyrics of
the good men of old, so that the child might be inspired to imitate them and long
to be like them. (Prot. 325c-326a)
On Protagoras’s view, the child learns the application of the words “right,” “holy,” and “honorable”
by being given one example after another, until she is able to apply the words accurately of her own
accord; this education is begun by her instructors and completed by exposure to exemplary poetry.33
The child learns the scope of these words as precisely as possible. In Wittgensteinian terms, she
learns the proper conditions of the word’s use; but, packed into these lessons, she learns a thorough
course in the art of demotic virtue, which is the practice of following these rules: doing things that
fall under some virtue term, avoiding things that fall under some vice term.34 In the terminology of
measurement, she learns to “measure” her actions in accordance with these words.
In this way, we can understand how someone who “sees through words” – that is, a person
lacking in philosophical acumen or potential – might still be able to act in a morally acceptable way.35
The waywardness of such a vulnerable soul finds its guidance in the guardian, who is a “craftsman
of temperance and justice and the whole of demotic virtue” (500d).36 When such a person is taught
the right words, she is able to see whether her actions “measure up” to these words. If she finds that
Plato would certainly not endorse the particular poems Protagoras has in mind, but he would – I think –
nevertheless endorse the method Protagoras puts forward.
34
There must be a “meta” level of teaching in musical education, then, as indicated by the passage from Protagoras.
Not only must the student be taught “this is holy”; she must also be taught “do this.” She must be taught that “the
holy actions” are actions that are commanded, and “the unholy actions” are the actions that should be avoided. One
might say that a character like Thrasymachus fails on this meta level, since he (at one point, at least) seems to agree
with Socrates on the scope of just actions, but disagree with him as to whether just actions should be done.
35
I do not mean to imply that only the non-philosophical need education in demotic virtue. Philosophers also need
this education, though not precisely for the same reason. The beginning guardians need such training primarily
because, as Nehemas puts it, “all of us have an analogue to the [unruly] crowd within our own soul … the appetitive
part” (217). The education in true beliefs about virtue envisioned in Book III puts the appetitive soul in its place.
Alexander Nehamas, “Plato and the Mass Media,” The Monist 71 no. 2 (1988): 214-234.
36
Belfiore fails to notice this passage when she claims that the craftsman of virtue, in Book X, should be identified
with the “reasonable man” (140) – by which she seems to indicate individuals who act in appropriate ways. Such an
interpretation fails to realize the importance that Plato puts upon the lawgiver, as the one who implants virtue in the
citizenry. Belfiore, “Theory of Imitation.”
33
99
a certain action falls under the heading “loyal,” she will engage in it; if she finds that the action falls
under “traitorous,” she will eschew it.37 There is, then, a way of acting morally which engages
centrally with language, but proceeds without anything beyond a linguistic understanding.
Individuals at this level are at the mercy of their educators, as Plato certainly believed.
Conclusion
In the above sections, I believe that I have made a plausible case that accurate language plays
a central role in the inculcation of virtue, and that accurate language concerning virtue is a significant
(though embedded) theme of Book X. The guardians do not forcibly control the actions of the
demos, nor do they acquaint the demos with the Forms; rather, they “instill” (τιθέναι, 500d) virtue
by teaching the populace the method of “moderating” their actions and “calculating” whether an
action falls under a certain virtue. Their students have been “dyed” with the correct beliefs about
whether or not any given action falls under the heading “virtuous,” even though the students cannot
give an account of virtue itself (402a). They know the scope of virtue, as it pertains to ordinary
conduct.38 All they need to be good people, according to Plato, is a good moral vocabulary. Though
I imagine they would like a good couch, too.
37
There is more than a hint, here, of another doctrine that clearly unifies the notions of morality and measurement:
Aristotle’s “mean between two vices.” For a discussion of the mean in the context of the ordinary Greek
understanding of measurement, see Richard Bosley, “Aristotle’s Use of the Theory of the Mean: How Adaptable and
Flexible is the Theory?” Apeiron 25 no. 4 (1992): 35-66.
38
Similarly, in a recent article, Moss claims that “we reach pistis as the conclusion of calculation that aims to determine
whether x is in fact F” (227). Moss makes a helpful point for my purposes here: When Plato calls our cognitive
relationship to intermediates “pistis,” he is using a variant of peithesthai, “to be convinced” which implies that pistis,
unlike eikasia, is the endpoint of a process. J. Moss, “Plato's Appearance-Assent Account of Belief,” Proceedings of the
Aristotelian Society 114 no. 2 (2014): 213-238.
100
CHAPTER 4: LANGUAGE IN THE CAVE
The cave image was explicitly constructed to address education and miseducation (Rep. 514a).
It is surprising, then, that Plato never makes it clear where the bad teachers are in his image; he
constructs an image, but leaves it to us to seek out the sophist inside. This chapter is an attempt to
find that elusive character, and I would propose that, as we find him, we shall become aware that
the cave image has a great deal more to say about his specialty, language, than most commentators
have realized. The cave is, I will argue, the region of education that is both permeated by, and
circumscribed by, language. Beyond the cave, in the outside world, philosophers are free to learn
things other than language, things that an “education through names” (Crat. 439a) could not even
begin to approach.
I will be providing a number of (I hope) compelling reasons to believe that language is a key
to understanding the cave image, reasons why such an interpretation either supplements or
supersedes various other interpretations of the cave. First, however, I wish to simply present a puzzle.
In Book VI of the Republic, Socrates comments that the sun gives life and visibility to the natural
world, and poses an analogy to the Form of the Good in the following way: “This thing that provides
truth to the things known and gives the person who knows his ability to learn, call this the ‘idea of
the good’” (508e). This characterization, presumably, is meant to apply to the “upper portion” of
the divided line, where the Form of the Good makes everything that is intelligible intelligible.
But what of the lower portion of the divided line? The objects in this region are not
intelligible or knowable; it is not even clear whether they are “true” or “genuine” (ἀλήθεια). They
do have some cognitive status, however: they are “opinable.” But if Plato thinks that there must be
some thing that makes intelligible things intelligible, shouldn’t we expect that there is also some thing
101
that makes opinable things opinable? What is this thing? Whatever it is, it should be the sort of
thing that (adapting the above) “provides plausibility to the things opined and gives the person who
opines his ability to opine.” The best candidate for such a thing, I would propose, is language.
Language is the light that illuminates the darksome cave; goodness is the light that illuminates the
real world. The whole of my dissertation is an attempt to validate the proposal that language plays
such a central role in the cave. I shall make that case in the space below.
Part I: Why the Puppets are Virtue Words
This final chapter will draw a great deal on the three previous chapters, so it will be prudent
to review those chapters here, and briefly locate them in the context of the cave image:
(1) In the first chapter, I discussed the contrast between “education through names” and
“education through things.” Education through things results in a knowledge of the things
which names refer to; education through names results, at best, in true belief about those
things, and it will only result in true belief if the namegiver was originally guided by the
dialectician.
Although names link up with their referents by convention, there may
nevertheless be a correctness of names, since the names themselves somehow have
information embedded in them, and this information may be false of the referent.
If I am right about those theses, then I do not think there is any difficulty
understanding them in the context of the cave image. On nearly every interpretation of the
cave, the objects that exist inside the cave are ordinary sensory objects. Since names are
sensory objects, they exist inside the cave, whereas things (i.e. Forms) exist outside the cave.
Education through names is, at the very least, one type of education that goes on inside the
cave. The resemblance relation between a name and its referent must be something like the
102
relationship between puppets/shadows in the cave and the real things outside the cave.
When the resemblance is a poor one, bad education ensues. Even a good education inside
the cave, however, will not result in knowledge, only true belief.
(2) In the second chapter, I argued that Plato applies his theory of naming to various aspects of
primary education in the Republic, although he does not explicitly tell the reader he is doing
this. Sophists and demagogues are criticized for their use of false names, as when they call
license “liberty” or shamelessness “courage” (560e-561a). The notion of the guardian as a
legislator (νομοθέτης) who “puts norms” into citizens is made much more cogent when we
understand these norms to be embedded in the language of Kallipolis itself, in the sorts of
terms provided for us in the Cratylus.
It should be quite obvious that both true and false beliefs exist within the cave, and
therefore (on the view I propounded in Chapter I) the cave contains both true and false
names. The contrast between sophists and demagogues, on the one side, and philosophers
and statesmen, on the other, is a contrast that manifests itself in two very different kinds of
“onomastic” education: one which is based on falsehoods, the other which is based on truth.
The philosopher’s expertise in Forms extends to a correspondent expertise in names: “Down
you must go [into the cave] … for once your eyes adjust you will discern [the things there]
infinitely better than its denizens, and you will know the nature of each image (εἴδωλα) and
its original” (520c). The contrast between sophist and philosopher, then, may be played out
in the distinction between the type of guidance about shadows that the freed prisoner gives
when he returns to the cave and the type of guidance prisoners give to one another.
103
(3) In the third chapter, I argued for a reading of Republic X which placed an emphasis on what
I called “intermediate virtue.” Intermediate virtue occupies a place parallel to the craftsman’s
couch, and it should be understood to be the creation of another kind of craftsman: the
craftsman of virtue, whom the Republic identifies with the guardian. For any individual
virtue, I argued, the corresponding intermediate virtue was precisely the word in the language
used to refer to that virtue: thus the word (or word-as-concept) “courage” is an intermediate
virtue corresponding to the Form of Courage. Drawing on my work in Chapter I, I argued
that these virtue names have meanings embedded in them by the lawgiver, and that these
names can be correct or incorrect, depending on the beliefs of the lawgiver.
In this chapter, I will be arguing for a number of theses which situate the vision of virtue
presented in Chapter III in the context of the cave, so I will only say a few words here. First of all,
the notion that there is a certain “elevated” form of opinion is a notion confined to two parts of the
Republic: (1) the line and the cave, and (2) Book X’s discussion of poetry.1 The divided line envisions
a level of belief called “pistis,” in which one’s beliefs are, presumably, more trustworthy (πιστός); this
level surely locates the intermediates. As I will argue, there is too much evidence here to argue that
Book X and Books VI-VII are quite independent. Secondly, we can see in the cave image that the
intellectual movement from shadows to puppets is an upward movement, which proceeds from a
region of instability to a region of greater stability. This upward movement would parallel the
movement from imitative virtue toward demotic virtue in Book X. If I am right, then the upward
1
Both of these passages may seem to be in tension with the stark division between opinion and knowledge presented
in Book V. But, as I have argued in Chapter III, intermediates are precisely the sorts of things that objects of opinion
are: they are sensory objects. The fact that they have properties that other sensory objects don’t have – greater stability,
standards of correctness – does not disqualify them from being objects of opinion.
104
movement in both Book VII and Book X frees a person from an education mediated by sophists,
poets, and the like, and allows the person to contemplate the outlines of “real” things, even if these
real things are (from another perspective) nothing but what we might call “toy models” of the Forms.
These suggestions are tentative, but I would like to put them on more sure footing. In the
course of this chapter, I will do that and more, focusing on the cave image itself and the conclusions
about language that might be drawn from it. In the first half of the chapter, I will survey how the
literature about the cave addresses the issue of “second-level” objects, and I will argue that both the
Sophist and the Cratylus introduce a distinction within Doxa that will help illuminate the cave. In
the second half of the chapter, I will explain how my view has significant explanatory power, and I
will respond to two interesting objections to my view.
Surveying the Cave
Although the literature on Plato’s cave image is extensive, there has been comparatively little
scholarship on the nature of the puppets on the second level of the cave. Scholars have focused,
instead, on the nature of objects at the purely sensory level of the cave: the shadows cast on the wall.
Their understanding of objects at this first level (which I will call C1) tends to dictate their
understanding of objects on the second level (C2).2 I will proceed in exactly the opposite way. As
far as I can tell, the key to understanding Plato’s cave image begins at the second level, not the first.
An analysis of the second level will illuminate the darker regions of the cave beneath it.
2
For examples of such cursory and derivative analyses of C2, see Raven 172, Nettleship 261, and Cross/Woozley 226227. In contrast, Hall rightly observes that the nature of the cave “can become clear only if we recognize the status of
the upper level of the objects [the objects at C2]” (82). J. E. Raven, Plato's Thought in the Making: A Study of the
Development of His Metaphysics (Greenwood Press, 1965). Richard Lewis Nettleship, Lectures on the Republic of Plato (R.
West, 1937). Cross/Woozley, Plato’s Republic. Hall, “Interpreting Plato's Cave.”
105
All analogies are to some degree imperfect. Nevertheless, I take it that good analogies will
tend to be those which either have the fewest irrelevant details, or those which clearly indicate which
details are irrelevant. Since Plato does not, in the cave image, flag any aspects of the image as
irrelevant, it is perhaps not unreasonable to hope that the image includes few, if any, irrelevant
details. Interpretations of the cave image, then, which account for as many aspects of the image as
possible, have an interpretive advantage. The text includes the following passages about second-level
objects:
Between the fire and the prisoners there is an upper path, and I want you to imagine
a wall buttressing the path, arranged just the way that circus performers place
barriers in front of themselves, above which they display their curiosities. (514b)
These people are holding up (φέροντας) a wide assortment of equipment along the
wall: men and animals of all sorts, fashioned out of metal and wood and every
material, no doubt some of the (puppet-masters) making noise and others quiet.
(514c-515a)
Do you imagine men of this sort would have seen any characteristic of … [the
puppets] … except what is directly in front of them: the shadows that strike the cave
wall? (515a-b)3
[Consider] someone who is released and abruptly forced to stand up and turn his
head around and walk, and turn his eyes toward the light. Doing all this would
cause him pain and he would be unable because of the dazzling light to fasten his
gaze on the objects whose shadows he had seen before. And what do you think [the
prisoner] would say, if someone … held up various puppets and forced the prisoner
to answer what the objects were? Don’t you think he would get flustered and have
the opinion that the things he is used to seeing are more genuine than the things
now pointed out to him? (515c-d)
Based on these passages, C2 objects have the following characteristics within the image:
1. They are manmade representations, and they can be used to produce misleading images that
do not resemble either themselves or the objects at C3/C4.
3
The passage is interpolated from two questions Socrates asks successively, but the second question is impossible to
render without the first.
106
2. They are subject to change (indicated by “φερόμενον”; for usage, see Crat. 404d, 411c).
3. One cannot become (painfully) accommodated to seeing real objects unless one first becomes
(painfully) accommodated to seeing the objects at C2.
4. They are more real (or more “true”) than the objects at C1, but less real (true) than the
objects at C3/C4.
Existing interpretations of the objects at C2 do not explain all of these features of the analogy. When
Cross and Woozley say that C2 is the level where liberated captives are free to draw their own
conclusions from the facts, their interpretation does not explain #4; for why should conclusions one
reaches independently be more real than conclusions that are thrust upon one?4 Ferguson’s view
that both C1 and C2 are but dual aspects of a corrupt political system runs afoul of #3, since there
is no reason to believe that any aspect of such a corrupt system would help one proceed out of the
cave.5 Reeve’s position that the objects at C2 instantiate “modes” which represent folk wisdom is
not designed to handle #1, since Reeve seems to think that C1- and C2-objects differ merely in
degree of resemblance to the Form, not in kind.6 Morrison proposes that the objects are “moving
forms,”7 but this claims seems difficult, if not impossible, to square with #1. Morrison’s moving
forms are abstract objects, not manmade objects.8 Wilberding offers the unique interpretation that
the objects at C2 are the opinions of the multitude, which sophists at C1 cater to, but this
interpretation gets #4 wrong. Pace Wilberding, sophists are not more ignorant of reality than
Cross/Woozley, Plato’s Republic, 220-221.
A. S. Ferguson, “Plato's Simile of Light. Part II. The Allegory of the Cave,” The Classical Quarterly 16, no. 1 (Jan
1922): 15-28.
6
C. D. C. Reeve, Philosopher-Kings: The Argument of Plato's Republic (Hackett Pub. Co., 1988): 70.
7
J. S. Morrison, “Two Unresolved Difficulties in the Line and Cave,” Phronesis 22 no. 3 (1977): 227.
8
See, for a variant of this criticism, J. Malcolm, “The Cave Revisited,” Classical Quarterly 31 no. 01 (1981):60-.
4
5
107
ordinary people; they are at the same level (Phaedrus 260c, Rep. 601a).9 Moreover, none of the views
mentioned above adequately explain the pain that the prisoner experiences in the ascent from C1
to C2. This pain, however, is one of the most striking and consistent themes in the whole of Plato’s
own rendering of the image.
Dale Hall’s interpretation comes closest to explaining all four attributes.10 Hall does not see
the second level of the cave as deceptive, in any way whatsoever, but he also does not believe that
the people at the second level are thinking critically. C2 depicts “the natural state achieved when
men are freed from apaideusia by the right paideia.”11 It “represent[s] the condition of men when
ruled philosophically.”12 Hall’s argument for this point is detailed and subtle, but it centrally turns
on a relatively basic point that some others seem to have missed. Since the Republic itself contains
two different types of education (primary/musical and secondary/dialectical), Hall contends that the
cave image ought to envision two types of education. The prisoner’s approach to the objects at C2
is clearly a form of paideia, but it does not involve a genuine encounter with real things (Forms). It
stands to reason, then, that the approach to C2 is meant to symbolize the primary education of the
citizens of Kallipolis – an education received by all citizens, not merely guardians. 13 All people in
Kallipolis are educated to the level of the puppets, though only guardians engage in the second,
dialectical stage of education, advancing to the level of real things.
9
Wilberding says that sophists are more ignorant than the demos, in that sophists are subjectivists about truth, whereas
ordinary people are objectivists. I appreciate this point, but it still does nothing to explain #4. For in the cave image,
it is not merely the case that people at C2 are closer to reality than those at C1; it is also true that the objects they
experience are more real, in themselves, than the objects at C1. Wilberding does not explain how this could be true in
the context of his interpretation. Wilberding, “Prisoners and Puppeteers.”
10
Hall, “Interpreting Plato's Cave.”
11
Ibid., 82.
12
Ibid.
13
Ibid. 82-83. Most scholars of the cave image would agree with this somewhat minimal claim. For dissenting views,
see Wilberding “Prisoners and Puppeteers” (above) and J. R. S. Wilson, “The Contents of the Cave,” Canadian Journal
of Philosophy, supp. vol. 2 (1976): 117-27.
108
By my reckoning, Hall’s analysis, though otherwise stellar, falls short in two ways. First, Hall
fails to account for #1, since he does not explain, nor even attempt to explain, how the objects at
C2 might be used to mislead. He argues – quite in accordance with the text – that C2 is not
fundamentally a deceptive region, since it constitutes a necessary step in genuine education. But he
fails to recognize that objects that are not intrinsically deceptive (puppets modeled after real things)
might be made deceptive through their use (as when one uses two fingers to project a shadow image
of a rabbit).14 Hall also hastily and indefensibly rejects the suggestion of Ferguson and others that
the puppets at C2 might be manipulated by sophists, demagogues, and the like. Second, although
Hall gives an account of #3 that persuasively explains the cognitive superiority of C2, his account
does not tell us why the move from C1 to C2 would be painful. Is there any reason to believe that
philosophical paideia would be unpleasant? If not, then why does Plato include pain in his account
of the cave?
In this chapter, I advance a modification and expansion of Hall’s position that will answer
these questions. I will present a novel and plausible explanation for why Plato included all four of
the features mentioned above in the cave image. The thesis, put simply and somewhat inaccurately,
is that the objects at C2 are words. Words are changeable, manmade artifacts which are imitations of
real objects (the Forms), and yet – as I have demonstrated in the prior three chapters – these
imitations can be used to deceive. Words approach reality more successfully than other sensory
objects, and yet words themselves are always deficient – and sometimes dangerous – ways of
approaching reality (see Chapter III). Words (especially virtue words) are subject to change in a way
Malcolm has something like this in mind, when he uses passages in the Sophist (234, 264) to argue that, on Plato’s
view, false things may be images of true things.
14
109
that their referents may not be. Finally, words clearly contain representational content, content that
is a necessary prerequisite to the acquisition of knowledge.
I believe, then, that my thesis would unify a whole set of concerns that are obviously central
to Plato in Book VII. There is nevertheless a problem with my interpretation in toto, and a very
obvious one at that: Plato could have explicitly said that he meant the objects at C2 to represent
words, and yet he never did. This omission gives me a significant burden of proof. It will not suffice
to say that words have a number of things in common with the objects at C2, for different objects
can nevertheless possess a number of the same properties.
In addition to listing these
commonalities, then, it is also incumbent upon me to demonstrate that words play precisely the
educational role that Plato wants the objects at C2 to play, in the context of Platonic education.
Socrates’ primary concern in proposing the cave image was surely not to make a point about
language, and I should never claim that it was. My claim, rather, is that the metaphysical status that
Plato gives to words in other dialogues is completely compatible with the status of C2 objects in the
cave. Plato did not pause in the middle of Book VII to explain precisely how language relates to
knowledge. Nevertheless, it would be hard to imagine how the type of vision of Platonic education
that we see in Book VII could fail to include the proper use of language, at the very least as a
precondition for dialectic – which is to say, as a precondition for one’s escape from the cave. Indeed,
a large portion of the Republic is explicitly concerned with the proper uses of language in education,
including (but not limited to) Books II, III, and X.15 My interpretation explains how these books,
and Plato’s comments on reformed poetry, may be directly incorporated into the cave image.
Why Should We Think that the Puppets Are Words?
15
See Chapter II.
110
At this point, I will advance the thesis for the chapter directly, by pointing out the advantages
to my view. In prior chapters, I have developed the idea that words themselves (for Plato) embody
beliefs, and that primary education is an “education through names” that communicates to students
the beliefs of the namegiver. This idea can be applied, basically in its entirety, to the cave image.
On this view, the movement from the first to the second level of the cave would be the movement
from one (inferior) type of education through names toward another (superior) type of education
through names. So long as we stay in the cave, however, all we learn about is words and how to apply
them. Individuals who advance to the second level have a sort of acumen that allows them to discern
which things “fall under” which words, and they also have a moral commitment to act according to
those words which are taught to be virtues. They have no ἐπιστήμη with regard to moral objects,
and yet they are capable of behaving as a virtuous person would. They have been “dyed” with the
proper true beliefs, beliefs which are embedded in the very language that they speak.
This is the view. I shall now begin to list a number of advantages to this view.
The first major advantage to the view is that it explains how one might be in some sense
learning about justice in the cave, while in another sense learning about a mere sensory object. The
word “justice,” as I have explained, is a sensory object that is patterned after the Form of Justice – at
least, in the context of any adequate educational system, it is. I am not familiar with any other
interpretation of the cave on which the students of Kallipolis are genuinely learning about some
sensory object at the second level, and yet not learning about the Form of Justice. At best, scholars
have argued that the people (or objects) that somehow embody or instantiate Forms are sensory
objects; but this interpretation simply doesn’t make sense in the context of the cave image, which
separates the actor (the puppet-master) from the act (the puppet). Words are used by someone. “The
111
just man,” in contrast, is not “used” in any way. It is, therefore, hard indeed to imagine how Plato
might envision the just man as a puppet.
Another significant advantage is that my view makes sense of the notion that the objects at
the second level of the cave are manmade. One reasonably straightforward way of reading the cave
image connects it closely with the Timaeus, and argues that C2 objects are the sort of things that the
Demiurge creates when shaping the recalcitrant receptacle into material things. That explanation
may seem to work in the case of animals and plants, but it does not explain how the objects are
manmade, and it does not seem to make any sense in the context of virtues like justice or
moderation. The Demiurge does not create these things “after the pattern of the Forms,” since he
does not create them at all. Insofar as they can be said to be created, they must be the creations of
the namegivers, who (ideally) look to the Forms when they give names to things. The second-level
objects that Plato cares about are not the handiwork of a god or daimon; they are produced by a
human being, and this is a reason to believe that they are words.
The third advantage of my view is that it unifies and illuminates the comments Plato makes
about language in the Republic, the Cratylus, the Sophist, and a variety of other dialogues. I have made
the case for this sort of consilience throughout this dissertation, but I would like to crystallize that
case in the space below. What we will find, I think, is that Plato found the broad notion of Doxa
insufficiently precise, and so he consistently used the dialogues to subdivide that notion into what
we might call “higher opinion” and “lower opinion.” The cave dramatizes that division and
encapsulates various comments about imitation in the dialogues mentioned above. Higher opinion
is the realm of accurate language, the place where imitation can genuinely be a path to truth – even
if this truth does not consist of ultimate realities.
112
I shall make the case for consilience in the space below.
Part II: Unifying the Evidence
I will first demonstrate that my reading illuminates certain issues found elsewhere in the
Republic. In Rep. VI, Socrates draws an analogy between the sun and the Good, whereby, just as the
sun is the medium by which visible things become visible, the Good is the medium by which
intelligible things (Forms) become intelligible. The Good makes education possible. The sun
analogy is, of course, central to understanding the Republic. But one aspect of the analogy often goes
overlooked. In the Cave image, Socrates describes the underground region of the cave as illuminated
by a light which presumably makes things in the cave visible, at least insofar as they are visible.
Scholars tend to see this light inside the cave as an analogue to the sun, and thus divide the cave
world and the exterior world in the following way:
Inside the Cave
Sun
Visible World
Visible Objects
Outside the Cave
Good
Intelligible World
Intelligible Objects (Forms)
This division of things is simple and attractive. But it won’t do. Although it is true that the cave
ostensibly represents the region revealed “by sight” (δι᾽ ὄψεως, 517b), Socrates makes it clear that
the opinions that people in the cave have are not exclusively about visible objects. In point of fact,
they hold opinions about things like justice, beauty and goodness (517e). Is justice a visible thing?
Surely not. And yet a shadow of justice appears at the first level of the cave.
Now what kind of object is justice? On Plato’s view, no doubt, it is an intelligible object. To
say that “S has an opinion about justice” should be equivalent to saying that S has an opinion about
an intelligible object – that is, a Form. But this won’t do either. For according to Book V, opinions
113
are set over opinables, and knowledge is set over knowables. We cannot say that the people at C1
hold an opinion about a knowable. But what can we say? Logic compels us to say that the opinions
about justice held at C1 are not, strictly, opinions about the Form of Justice, but rather opinions
about some inferior object that is somehow related to the Form of Justice. 16 What is this inferior
object?
My analysis in Chapter III clears the way for a potential answer. An opinion about justice is
an opinion about a word17: a sensory object that has a degree of stability superior to the mere
appearances, but inferior to Forms. Words belong in the cave, because they are aspects of sensory
experience. But they also “bridge the gap” between the cave and the real world, since they reliably
indicate (and, ideally, reliably describe) real things. As we have seen in previous chapters, Plato cashes
out these notions of reference and description through the concept of an “imitation.” Despite being
sensory, words point toward what is exclusively intelligible.
These intermediate objects, it might be objected, are completely unlike the other objects at
C2: people, animals, and artifacts. But this objection misses the point. Words are sensory artifacts,
16
Harte argues that the words of the cave-dwellers genuinely refer to Forms, but her argument stems on a questionable
assumption. She plausibly argues that Rep. 515b should be understood to imply that the cave-dwellers words do not
refer to shadows, but she then assumes without argument that these words must genuinely refer to Forms. There is, of
course, a third option: that these words not to shadows, but to puppets. I myself remain agnostic on the question of
what the prisoners’ words refer to. All I insist upon is that the cave-dwellers do not have opinions about Forms, whether
or not they unwittingly refer to Forms. Harte’s interpretation is very much on the right track, in that she sees a
normative line stretching from Forms down to ordinary words in the language. As I argued in Chapter I, Plato holds
the curious view that words can be true or false, since they are somehow linked to Forms which provide standards for
their truth or falsity. Verity Harte, “Language in the Cave” in Maieusis: Essays in Ancient Philosophy in Honour of Myles
Burnyeat, eds. Myles Burnyeat and Dominic Scott (Oxford University Press, 2007): 195--215.
17
It would not be entirely inaccurate to say that it is “an opinion about an Opinion,” where an “Opinion” with a
capital “O” denotes something with linguistic (as opposed to ontological) stability. This may be roughly the same type
of stability that philosophers often ascribe to terms like “begging the question,” when people use the term incorrectly.
The philosopher is not saying that the phrase “begging the question” has a natural meaning discoverable by
investigation, but is rather insisting that its conventional meaning is not subject to a Humpty-Dumpty type of linguistic
subjectivism.
114
even if they are not three-dimensional objects.18 As we have seen, they are the works of a craftsman
who excels in the ability to put reality “into appropriate syllables” (Crat. 389d). The tremendous
virtuosity of the wordsmith is his ability to put all kinds of things into words (389d), just as the
implied craftsman of the objects at C2 has made miniature imitations of all kinds of things (Rep.
515a; cf. Rep. 596c). What is needed at C2, however, is not only virtuosity at imitation; it is – as in
the Cratylus – some sort of understanding of the thing being imitated. If there were no such
understanding, then characteristics #3 and #4 above would not make sense. How can movement
from C1 to C2 represent genuine intellectual progress if the maker of puppets is no more
knowledgeable than the maker of shadows?
By reading Book VII alongside Book X, and by applying a Cratylean understanding of
imitation, we can reveal what sorts of artifacts exist at C2. They are the handiwork of a statesman
like Solon, who sets the terms of discourse. This is not to say that the objects at the second level of the
cave will always resemble the Forms. They will not – and they will be especially unlikely to resemble
the Forms if the lawmakers (as in democratic Athens) have been infected with the corrupted beliefs
of the poets and the sophists. Thus, as we saw in Chapter II, a corrupt populace may, over time,
acquire corrupt beliefs about the virtues, by the working of an inferior education through names.
But C2 objects have imitative standards that, when met, genuinely enable the populace to acquire
demotic virtue. In an effort to make all these claims more plausible, I will now turn our attention
to a series of passages in the Sophist and the Cratylus, passages which confirm and develop my
discussion of the Republic.
18
This fact may be indicated straightforwardly by Plato, when he describes the use of language by both prisoners and
puppeteers in the Cave. (See Rep. 515b). The passage is a reasonably vexed one, however, since it does not make clear
who is doing the speaking.
115
Eikastikê Versus Phantastikê
There is a fascinating passage about imitation in Plato’s Sophist, a passage which is often read
alongside Republic X, but rarely compared with Republic VII. In the passage, the Stranger claims that
the sophist possesses “a kind of appearance-based (δοξαστικὴν) knowledge about everything, but
not the truth” (233c), which enables him to produce plausible appearances of everything, though he
does this in a fashion similar to the graphic artist, who “devises imitations which bear the same name
as real things” (μιμήματα καὶ ὁμώνυμα τῶν ὄντων, 234b). The sophist does the same thing in
the context of language (περὶ τοὺς λόγους, 234c), “beguiling his auditors by displaying spoken
facsimiles of all things” (234c), in an effort to produce certain advantageous beliefs in the young,
without concern for their truth.
The passage is almost a direct parallel to the passage at Republic 601a-b, with its emphasis on
the imposition of appealing falsehoods upon the undiscerning, although clearly the Sophist passage
is targeted at sophists instead of poets. More to the point, however, the passage once again repeats
the cave image’s emphasis on the virtuosity of the imitator, producing “all things”; but here, the sort
of production involved is not nearly as multifaceted as the production of the craftsman. The sophist
uses a single medium, whereas the craftsman of Book VII builds out of “metal and wood and every
material” (515a). This is why Plato calls the sophist’s work “shadow play” and “illusion.” Plato gives
the name “phantastikê” to the monochromatic imitation practiced by sophists (and presumably
poets), and he contrasts this with a type of imitation called “eikastikê,” which “designed according
to the pattern of its model in all three dimensions, giving the appropriate color to each part” (235de).
116
It is prima facie plausible, from what I have said above, that the phantastikê/eikastikê
distinction in the Sophist can be – at least roughly – imported into the context of the cave and divided
line. (I am assuming, without argument, that the cave and line images relate to each other in a
straightforward manner, as implied by Socrates’ comments at 517b). The realm of shadows is
phantastikê; the realm of puppets eikastikê. If anything, this scheme resonates even more with the
divided line. The lowest portion of the line, after all, denotes “picture thinking” (eikasia), which is
contrasted with what we might call “stable thinking” (pistis).19 Pistis is set over things ordinary
material things, which are (presumably) three dimensional and polychromatic; an imitation that
renders in three colorful dimensions would, presumably, also be a subject of pistis.
Above I have established a reasonably small point: that Plato consistently thought of
imitation in binary terms, and that the same basic distinction among types of imitation is reiterated
in the Republic and the Sophist.20 A more consequential point comes next: I am going to argue that
the phantastikê/eikastikê distinction plays a critical role in the Cratylus. The argument is more
speculative than the above, but it also, if plausible, it will provide a powerful example of the sort of
unifying work that my thesis accomplishes.
Rhetoric and Poetry in the Cratylus
In earlier chapters, I have argued that the Cratylus depends upon two fundamental theses:
(1) that words are images, and (2) that words embed information that can be true or false. I have
19
I think this connection is the most plausible one, despite the obvious etymological connection between eikastikê
(which I place on the second level) and eikasia (which I place on the first). In Book X of the Republic, Plato continually
uses the word phantasmata to describe artistic objects, which clearly belong at the lowest level of the divided line. The
production of phantasmata would obviously be phantastikê; this leaves eikastikê to describe the production of something
else.
20
I find it likely that this distinction is ubiquitous in Plato’s dialogues, though elsewhere it may be somewhat harder to
discern. The Phaedrus’s distinction between superior and inferior rhetoric, for example, might be understood along
the same lines.
117
not, however, discussed the greater portion of the dialogue, a portion where Socrates offers a set of
etymologies that range from thoughtful to ridiculous. A central question of the dialogue, however,
revolves around the following question: If Plato meant the linguistic theories espoused in the Cratylus
to be taken seriously, why did he illustrate them with apparently facetious examples?
I do not have any complete or satisfying answer to that important question, but I can make
an observation: the Cratylus is far more concerned with poetry, and rhetoric in general, than critics
usually allow. This will become clear when we juxtapose Socrates’s concluding diatribe about
education with the actual content of various etymologies in the dialogue. As various commentators
have observed, there is something of a drama being played out through the etymologies, a
competition between the notion that all things are in Heraclitean flux and the notion that there is
an “abiding nature” (440c) in some things. This drama is played out in the etymologies: competing
etymologies of the same words yield different results, some Heraclitean, some Platonic. 21 At the
conclusion of the dialogue, Socrates becomes fed up by the unreliability of any educational scheme
that relies on investigating such words: “No intelligent person will put himself or the nourishment
of his soul under the guardianship of names” (Crat. 440c). In typical fashion (cf. Socrates’s
discussion with Hippocrates in the Protagoras), Socrates is concerned lest the talented Cratylus be
overly persuaded by sophistical techniques predicated on linguistic analysis, and thus neglect an
“education through things.”
In Chapter I, I argued that Socrates does not merely have teachers of etymologies in mind,
at 440c; he has in mind other sophists too, since sophists acquire reputations for wisdom by
exploiting the meanings of words. Socrates does not merely target sophistry with these words,
21
There is no indication, however, that our existent words must be correct. Quite the contrary. It may be that our
existent word “epistêmê” indicates movement, despite the fact that its referent (“knowledge itself”) is stable.
118
however; he also targets the teachers of the sophists. From the dialogue, we can have no doubt who
these teachers are: they are the poets. As Socrates goes through the etymologies, he continually –
and almost exclusively – appeals to Homer and Hesiod as authorities on the usage of Greek words;
and from their usage, he develops his etymologies.22 If these etymologies were fully endorsed by the
text, this habit might make us think that Plato has a newfound respect for the intellectual standing
of the poets. But, given the general aura of jesting in the etymologies – and given the fact that most
of the given etymologies endorse Heracliteanism (see Crat. 411d-412c, e.g.) – it is no great leap to
imagine that here, as elsewhere, Plato is once again being critical of the poets. They are surely among
the targets of the sardonic criticism that
the primeval givers of names … [are] like our modern philosophers, who, in their
search after the nature of things, are always getting dizzy from constantly going
round and round. … And this appearance, which arises from their own internal
condition, they suppose to be a reality of nature; they think there is nothing
permanent, but only flux and motion. (411b-c, trans. Jowett)
We find a more detailed catalog of the mistakes of the inferior namegiver toward the end of
the dialogue. There, we are told that no namegiver can capture the complete essence of any object
in a name, since images that are complete cease to be images (432a-c), but that the good namegiver
will capture the “basic character” (τύπος, 432e) of the thing named, whereas the bad namegiver gave
names according to a false conception of their referents (436b). The errors of the inferior namegivers,
then, are not errors of being bad “renderers,” but of having bad models from which to render. Surely
this is also the case for the poets, who do not – in Book X of the Republic – portray virtues falsely out
of spite, but rather because their own impression of that virtue is shallow and two-dimensional.
22
The examples are almost too numerous to count. Socrates appeals to Hesiod as an authority on the uses of words at
397c-398c, 396c, 402b, and 406d. He appeals to Homer at 391d-393b, 402a, and 407a-c. He appeals to the poetic
tradition, in general, at 395c and 404b, to the “poets” at 412b, and to the Orphic poets at 400c. The entire discussion
is concentrated on poetic sources.
119
As I see it, then, there is a straight line running through the practice of naming in the
Cratylus, the production of images in the Sophist, the action of “makers” in Republic X, and the
fabrication of objects of Doxa in Republic VII. In every case, a two-dimensional inferior imitation is
contrasted with a three-dimensional adequate imitation which is more intellectually respectable –
and more educationally sufficient – than its poetic/sophistical counterpart. The discussion of
dimensionality necessarily becomes more metaphorical and obscure when applied to normative
terms instead of physical objects, but it is (as I argued in Chapter III) no less applicable. A twodimensional virtue is a merely the appearance of virtue on a given occasion, which may, indeed, be
identical to the appearance of vice on another occasion. A three-dimensional virtue is a cognitive
attitude that embodies stable true beliefs about the (Formal) virtues. When sophists and poets are
faulted for being bad educators, they are faulted precisely because their understanding of virtue tends
toward the two-dimensional, because there is something fundamentally “perspectival” (or perhaps
“relativistic”) in their notions of virtue. Sophists and poets practice phantastikê, not eikastikê. As I
have argued earlier in my dissertation, the type of verbal education that tends toward “threedimensional” virtue (what the Sophist calls eikastikê) is the prerogative of legislators, or philosophers
insofar as they play the role of legislators.
More from the Sophist and the Cratylus
The connections between these two dialogues and the cave image run deeper. English
translations of the cave image blithely state that the puppets are being “carried” above the wall at
C2. This translation, though adequate, misses a great deal about the meaning of the word φέρειν
in Greek, and, in particular, the way Plato uses the word. Two particular uses of the word matter
for my purposes here. First, the word φέρειν is commonly used to mean “moving.” Thus Plato
120
often, in discussions of Heraclitean flux, uses this word to mean not that “all things are being
carried” but that “all things are in motion” (Theaet. 182c). Second, the word φέρειν is occasionally
used – particularly by Plato – to denote the application of names. Thus at Sophist 237c, the Stranger
talks about “correctly applying” (φέρων ὀρθῶς) names, a phrase that might more conventionally
take the word τιθέναι (Crat. 389a, 406b). If I might speculate, however, the word φέρειν might be
used when the connotations of τιθέναι are odious – when the author wishes to explicitly describe
using a word, as opposed to establishing it.23
Both these usages of φέρειν support my interpretation here. The first usage suggests, in the
context of the cave, that the entire cave (including the second level) is a place of constant movement,
not unlike the universe imagined by self-styled Hericlitean philosophers like Cratylus himself.
Socrates says that the things named “are moving (φερομένοις) and flowing and coming into being”
(Crat. 411c) and that “names indicate their nature to us, at least if we assume that all things are in
motion (φερομένου) and flux” (436e). Now suppose that these names were correct, and all things
were in motion. In that case, there would be nothing other than the cave – nothing other than
“coming to be” – then we need nothing more than names to know all the truth there is. But Socrates,
both early (386d-e) and late (438e-440a) in the Cratylus, calls this line of thinking into question: Not
only are many words suggestive of stability instead of motion, but there are clearly standards of things
like beauty and goodness, and these things are standards precisely because they do not move. If we
understand the puppets at C2 to be words, we can understand these standards to be standards of
linguistic usage. Motion belongs in the cave, but the “things carried” can be modeled after the things
23
See Chapter One, note 50.
121
beyond. When a person “applies” (φέρειν) a word correctly, the word “carries” (φέρειν) the right
denotation; this is musical education, properly understood.
This may have been the sort of thing Plato talked about at Rep. 402, where he likens children
who learn to discern their letters “in all the contexts that carry them around (περιφερόμενα)” to
the guardians, who learn to recognize the “forms” (εἰδή) of the virtues when they are “carried around
everywhere” (πανταχοῦ περιφερόμενα). There is no need to understand “forms,” here, as
anything more than sensible manifestations of virtues. Indeed, to claim that Plato is speaking about
“the Forms” presents insuperable problems, since Plato is here discussing knowledge that proceeds
from musical education (402c), not from dialectic. No, his point here is not that the guardians must
know the Forms; but rather that the guardians must learn how to apply a sort of “language,” not
reflected through distorted sophistical φαντάσματα (see Rep. 402b, Soph. 234c), but in its original
and uncorrupted form. This would work roughly in the way I described in Chapter III: there are a
set of Forms of virtues, and a set of intermediate virtues which are linguistic imitations of these
Forms. The linguistic imitations do not have a “nature” in the full sense of the term, but they are
“measureable,” in the sense that they have a restricted range of things they apply to. There are a set
of actions that “fall under” courage, and a set of actions that fall outside it. In primary education,
the student acquires the ability to measure which are which. This is the sort of “knowledge” – or at
least awareness – that people acquire when they have ascended to the second level of the cave.
Language and Immanent Forms?
There is one additional way that my interpretation unifies our existing understanding of
various dialogues. For decades, Plato scholars (e.g. Vasiliou, Keyt, Fahrnkopf, Wilson, and Grube)
have discussed the notion of “moving” or “immanent” forms, as a category in Platonic thought
122
inferior to Forms but somehow cognitively significant in a way that appearances are not. So, for
example, there has been much discussion of the “equals themselves” (Phaedo 74c) – as distinguished
from the Equal Itself – and there has been speculation that this passage in the Phaedo is referring to
the same thing the Timaeus seems to describe as a moving form (50c).
Although I cannot defend the thesis here, there is an interesting case to be made that the
Republic engages with some or all of these passages in a passage I have already discussed. At 402b,
Socrates insists that we must recognize the “forms” of the virtues wherever they may be found moving
(περιφερόμενα). I argued above that this passage can be fruitfully applied to the assumption that
these “moving forms” are words.24
But then, the notion immediately occurs: what if
immanent/moving forms are words throughout the dialogues? I have no real argument to make about
the question; it goes too far afield from my topic (the Republic), and it would take a great deal of
analysis to untangle. Nevertheless, my thesis suggests a way to explain something quite obscure
(“immanent forms”) in terms of something quite ordinary and explicable to us (words). It is, after
all, rather plausible that Plato considers words – at least certain words – to be “the likenesses of
eternal realities modeled after their patterns in a wonderful and mysterious manner” (Tim. 50c,
trans. Jowett). If my view suggests such a unifying and overarching solution to a Platonic problem,
this is certainly something in its favor.
Objection #1: Are the Sophists Puppeteers?
I will now discuss two objections to my thesis, and my response to them. Here is the first
objection.
This claim has the added advantage of incorporating many of the arguments Morrison gives for situating “moving
forms” at C2. Unlike Morrison, however, I claim that moving forms are sensory objects with the capacity to refer, not
purely abstract objects.
24
123
In the course of this chapter, I have suggested that the sophist can be found at the second
level of the cave, along with his counterpart, the statesman. Both statesman and sophist work with
the same raw materials – words – but the content of their educational programs are quite different.
The sophist (whether intentionally or unintentionally) deceives the public by presenting false words
that look like true words, just as the shadow-puppeteer exhibits shadows that look nothing like the
things that cast the shadows. The statesman, on the contrary, tells young people of the demos
“suitable stories” (Stat. 304d) in order to instill a “bond of true conviction” (Stat. 309d, trans. Skemp)
among them about “the fine, the just, and the beneficial” (Stat. 309c).
There would seem to be a problem with this interpretation, however. It places the sophist
(along with the statesman) at the second level of the cave, and presumably at a higher cognitive state
than the people who see everything through the medium of shadows. And yet elsewhere, Plato goes
to great lengths to emphasize that ordinary sophists, rhetoricians and poets know nothing more than
their audiences (Phaedrus 260c, Rep. 601a). Moreover, the nature of words, on my rendering of the
cave image, is puzzling. Are words stable entities which, when we encounter them, are a step in
leading us “out of the cave”? Or are words the unstable, shifty sort of entities that might be freely
exploited by sophists and tricksters? My interpretation must address these issues.
The best response, I believe, begins by observing something I discussed at length in the first
two chapters. When a teacher “teaches the student a word,” she is teaching him a set of beliefs about
a concept – and, as in Chapter III, when a virtue word is “taught,” a pattern of behavior is being
taught as well. Now, it should be obvious that there are at least two ways of acting: virtuously and
viciously. Neither virtuous nor vicious action is identical to the appearance of virtuous or vicious
action, just as an apple is not identical to the appearance of an apple. The vicious person’s vicious
124
actions manifest his false beliefs about virtue (see Chapter III), false beliefs that, again, are not
identical with the appearances that proceed from these false beliefs. Thus the shameless actions of
a grave robber do not in appearance resemble Euthyphro’s nonchalant prosecution of his father,
despite the fact that both actions manifest the same vice: impiety.
An interesting result emerges. A vice is not a model upon which people might base their
lives (akin to virtue), but rather it is the way a person acts when she has false beliefs about virtue – or, in
terms of Platonic psychology, when her beliefs about virtue are determined by her appetitive or
spirited soul.25 These false beliefs could, perhaps, be native to the person who holds them (as a child
might naturally think the earth is flat), but they could also be imposed upon a person in childhood,
in the form of corrupt education. This is where poets and sophists come in. Like guardians, poets
and sophists take it upon themselves to mold children into an image, but the image they mold them
into is not patterned after the Forms. One might call it a sort of “demotic virtue,” but only with the
proviso that this kind of demotic virtue is not wholly good – indeed, it could be quite bad.
We see, then, that there must be at least two possible “versions” of the cave. The first version
is the “Kallipolan cave,” which consists of accurate intermediate virtues which reliably guide the
student toward (genuine) demotic virtue. The second version we might call the “sophist’s cave.” In
this cave, the blind lead the blind, and education is a perverse sort of parody of genuine education.26
Like always, the content of the words at C2 is determined by the νομοθέτης, but this νομοθέτης
I do not say that “this is Plato’s theory of vice,” but only that I am drawing out one clear aspect of that theory,
whatever it is. Vice necessarily proceeds from false belief – though it may not necessarily proceed from false beliefs in
the rational soul, since the lower soul parts are capable of determining action in an unharmonized individual. The
connection between vice and false belief is guaranteed by Plato’s intellectualism.
26
This should not be surprising, since (at least among students) sophistry often presents the appearance of genuine
philosophy, and Homeric poems often present the appearance of philosophically educational myths. Ordinary people
are completely incapable of knowing “which cave they are in,” which is precisely why such care must be taken to make
sure that the constitution of the state is guided by philosophers who do know the difference.
25
125
has false beliefs about the things the words describe. Education, in this cave, can at best access the
flawed opinions of one lawgiver; and it will be unlikely to do even this, if sophists or poets keep
ordinary people “in the dark.”
In some sense, then, the sophist is at the second level of the cave. But in the sophist’s cave,
there is no way “up” and no way “out.” As the sophists themselves taught (at least, as Plato
understood them), there is simply no “Truth” that human beings have access to, only the world of
opinion wherein the opinions of the powerful prevail. One does not win court cases or elections by
acting virtuously or speaking truly, but only by manipulating the beliefs of the public to one’s
advantage. The language the sophist uses will not tend to lead a person toward Truth, since it is not
designed by people who have access to the truth.
These observations become even more compelling when we recall the name Plato gives to
“second-level” cognitive states: trust (πίστις). In the Cratylus, Socrates warns of the immense dangers
one incurs by “entrusting” (ἐπιτρέψαντα, 440c) oneself to an education through names. This
warning is echoed manifold times in Platonic discussions of the sophists (e.g. Prot. 312c), and was
even heralded in a comment Heraclitus makes about the masses: “What understanding or
intelligence have they? They put their trust in popular bards and take the mob for their teacher,
unaware that most people are bad and few are good” (B104). The best that ordinary people can
hope for is to have good teachers, teachers that will instruct them in the language of right and wrong.
They have no way of confirming the veracity of these teachers; instead, they must simply “have faith”
(πίστις).
Objection #2: Why Does the Prisoner Experience Pain?
126
The second objection is a much simpler one. In this chapter, I set out to give a reading that
explains all the characteristics of the cave image, with respect to C2. But, thus far, I have not done
so. I have not explained why ascending to the second level of the cave is described as painful. And
it does not seem that it would be painful. The children of Kallipolis, after all, are simply hearing
purified stories, listening to specific rhythms, and imitating good characters; this is their primary
education. A critic might find those activities otiose, but hardly painful.
The best response, I believe, is based on the close affinity between the description of
ascending to C2 and descriptions of experiencing Socratic elenchus, elsewhere in the dialogues (e.g.
Meno 80a-b). It is painful to endure elenchus. On the other hand, elenchus (at least in its fledgling
non-dialectical stage) does not seem to be necessary in Kallipolis. Elenchus purges the mind of
harmful moral beliefs (Apol. 36b-d; Apol. 41e-42a), but the very purpose of Platonic censorship is to
prevent children from encountering such beliefs. We might expect, then, that Athenian individuals
would benefit from elenchus a great deal, whereas Kallipolan individuals would not. Might we
suppose that the cave image, as we have it, is a description of Athens, not Kallipolis?
I think so. Let me appeal to the “two caves” I discussed above, the sophist’s cave and the
Kallipolan cave. Individuals in the sophist’s cave can only ascend to C2 through an arduous process
of elenchus. Individuals in Kallipolis, however, casually move to C2 in the course of ordinary
education. This interpretation would be open to various interpreters (Malcolm holds a version of
it), but it resonates particularly well with my own. For on my view, elenchus in Athens would be the
purification of one’s language – in essence, the process of overcoming one’s ordinary understanding
of words. Since elenchus does not end in a doctrine, one’s ordinary understanding is not replaced
with anything, though one may (after elenchus) stand in a better position to see the harmful exploits
127
of the poets and sophists. The puppets in the sophist’s cave do not accurately represent anything.
It is obvious why such a process would be startling; after all, it might seem to land one in a sort of
moral skepticism. Kallipolan primary education, on the contrary, would result in a variety of largely
comforting and stable beliefs that are building blocks to knowledge. There is no reason for that
process to be painful.
Conclusion: What is the Cave?
I hope my analysis has left the cave roughly as it was before, though maybe a bit brighter
inside. Perhaps it was already obvious enough that the cave was the region of language, and that the
outside was the region of reality-beyond-language; but one can be forgiven for stating an obvious
claim that others seem to have overlooked. I believe that I have found where the sophist must be
situated in the cave, and, as a consequence, precisely why the ideal city can have absolutely no place
for sophists (or imitative poets, for that matter!). All such individuals not only have a disregard for
the truth (since they are motivated by their appetites and spirit), but they also have false and
unreliable beliefs about the truth, beliefs which inevitably infect their practice. The cave will only
be well governed when these beliefs are rooted out and replaced with intermediate objects (words)
which point to and describe the inspiring and wholly authentic world outside the cave. As long as
citizens are kept from any access to that world, demagoguery and democracy will ensue.
128
CONCLUSION: PROPAGANDA IN PLATO’S CAVE
There is reason to think that, in the time surrounding when he wrote the Republic, Plato’s
thoughts about the purposes of communication were experiencing some amount of upheaval.
Whereas the Protagoras, the Gorgias, the Apology, and the Euthydemus all seem to call into question
the validity of non-rational forms of persuasion, the Phaedrus radically reverses course, carving out a
place for a positive rhetoric. This rhetoric “plants λόγους scientifically, … which give seed to other
λόγους which nourish habits [ἤθεσι] in the listener and … make him as happy as a person can
possibly be” (Phaedrus 276e-277a). As we have seen, the Republic envisions a similar role for primary
education, a kind of education which persuades a person to believe in what is good and true, yes,
but persuades using a method (poetry, storytelling) that is capable of presenting either truth or falsity.
It is, in that sense, non-rational. Likewise, the persuasive capacities of the statesman, in both the
Statesman and the Laws, are (at least partially) employed not in an attempt to make the public learn
– in the strict sense of “come to ἐπιστήμη” – but rather in an attempt to nourish and sustain certain
true beliefs that redound to the stability and virtue of the polis.
The 20th/21st century poststructuralist movement argued (among other things) that language
is a political tool which shapes the thought of its auditors. This claim is controversial and often
misunderstood, but from what I have argued here, it would seem that Plato would wholeheartedly
endorse it. Plato, like the poststructuralist, tends toward the view that all rhetoric is propaganda.
Nevertheless – again, like many poststructuralists – Plato is of the opinion that some propaganda is
better than other propaganda, and moreover that indoctrinating young people into the right moral
viewpoints is of critical importance. To some, this might appear to be an artificial imposition on
the minds of the young, who (as we say) ought to be permitted to “form their own opinions.” And
129
yet, in practice, we find that many moral reformers, even in the modern day, do not leave individuals
to form their own opinions. They strategize ways to appeal to both the rational and the irrational
aspects of a person’s psychology, shaping their rhetoric to suit the psychic structure of the audience
in much the way Plato describes in the Phaedrus. The good goal of producing a morally enlightened
individual is used to justify the questionable method of indoctrination. And indeed, this may seem
justified, especially in the context of raising children. It is hardly possible not to raise a child to
acquire certain moral biases, and it would certainly be surprising if all such indoctrination (e.g.
“don’t spit on people”) was wrong.
The staggering audacity of Plato’s vision in the Republic, however, is that everyone is
indoctrinated in the same way, by the same “parents” (see Rep. 538c). On this scale, indoctrination
rightly takes on the name “propaganda.” According to one modern handbook, propaganda is the
“deliberate and systematic attempt to shape perceptions, manipulate cognitions, and direct behavior
to achieve a response that furthers the desired intent of the propagandist.”1 Kallipolis is a polis
utterly circumscribed by propagandistic language, and only those who advance beyond primary
education are freed from the spell of such language. French philosopher Jacques Ellul said that
propaganda
tries to surround man by all possible routes, in the realm of feelings as well as
ideas, by playing on his will or his needs, through his conscious and his
unconscious, assailing him in both his private and his public life. It furnishes him
with a complete system for explaining the world, and provides immediate
incentives to action. We are here in the presence of an organized myth that tries
to take hold of the entire person.2
1
Garth Jowett and Victoria O'Donnell, Propaganda and Persuasion (Thousand Oaks, Calif.: Sage, 2006): 6.
Jacques Ellul, “Characteristics of Propaganda,” in Readings In Propaganda And Persuasion, eds. Garth Jowett and
Victoria O'Donnell. (Thousand Oaks, Calif. Sage, 2006): 7.
2
130
One could hardly describe a more thorough picture of Plato’s cave image, which wholly surrounds
its occupant through its use of media, and only allows occupants to access reality insofar as it can be
mediated through words. Those inside the cave “see through words,” and are only capable of as
much good behavior as the words allow them. Nevertheless they have a “complete system of
explaining the world” and they have no doubt about why they should do what is required of them.
When the details of this indoctrinary scheme become quite clear, Plato’s thought is
remarkable both for its arrogance and for its insight. Does anyone doubt that advertising, media
programming, and censorship work in almost exactly the ways that Plato anticipated – especially
among children? We no longer need a philosopher to advance that claim; it is plain enough from
experience. And it is true in more subtle contexts as well, the very contexts that poststructuralists
would place emphasis on. How many of us would object to propaganda, and yet we are firmly
attached to our own chosen terminology to describe certain morally controversial concepts? We
agree about reference, but bicker over sense. What one person calls “affirmative action” another
calls “reverse discrimination.” What one calls a “freedom fighter” another calls a “terrorist.” What
one calls a “panderer” another calls a “responsive politician.” Even the term “propaganda” itself is
malleable in this way: where does rhetoric end and propaganda begin?
These are important questions about the malleability of human language, and the dangers
of relying on an education that relies overmuch on the names which we use. And yet, the arrogance
of Plato’s solution to these dangers is quite astonishing. The beliefs of the philosopher are imposed
upon the populace, thereby stripping the populace of its moral autonomy in the interests of the good
of the state. Plato does not envision a world without the cave, but a world with a better cave. It is
worth noticing that, if Plato is right that there are Forms, and if a person who has seen the Forms
131
comes to rule the city, and if this person desires the good of the city, then Plato may be quite right
in claiming that this sort of indoctrination is justified – at least insofar as this person could “craft”
intermediate virtues that tend toward the best interests of each citizen. But these three conditions
require a great deal, and Plato himself arguably came to believe that they would never be satisfied in
the real world.
Nevertheless, Plato’s full-bodied support for propaganda might give us pause. Do we
honestly reject moral propaganda in modern society, or are our airwaves full of it? How should the
government respond when its own values are being undermined by the diametrically opposed values
of a smaller social community?
How do the linguistic choices of modern media sources
surreptitiously affect the beliefs of their audience, and how important is it that the audience become
aware of these effects? Is it morally permissible to air public service announcements that appeal to
irrational impulses, not rational aims?
Questions like these come into sharper focus when we put them in the context of the cave
image. The discussions we have about them in the cave, however, can only take us so far. If any
truth is to be found to guide our propaganda, it will not be found in words and shadows, but in the
things themselves.
132
BIBLIOGRAPHY
Ackrill, J.L. “Language and Reality in Plato’s Cratylus.” In Essays on Plato and Aristotle, edited by J.
L. Ackrill, 33-52. New York: Oxford University Press, 1997.
Aristotle, and W.D. Ross. The Nicomachean Ethics of Aristotle, Translated and Introduced by Sir David
Ross. Oxford University Press, 1961.
Barney, Rachel. “Comments on Sarah Broadie, Virtue and Beyond in Plato and
Aristotle.” Southern Journal of Philosophy 43 (2005): 115-25.
Barney, Rachel. Names and Nature in Plato’s Cratylus. Routledge, 2001.
Baxter, Timothy. The Cratylus: Plato’s Critique of Naming. Leiden: E.J. Brill, 1992.
Belfiore, Elizabeth. “A Theory of Imitation in Plato's Republic.” In Ancient Literary Criticism, edited
by Andrew Laird, 133-135. Oxford: Oxford University Press, 2006.
Blackson, Thomas A. “Plato and the Senses of Words.” Journal of the History of Philosophy 29, no. 2
(1991): 169-182.
Bobonich, Christopher. “Persuasion, Compulsion and Freedom in Plato's Laws.” Classical
Quarterly 41 no. 02 (1991): 365-.
Bosley, Richard. “Aristotle’s Use of the Theory of the Mean: How Adaptable and Flexible is the
Theory?” Apeiron 25 no. 4 (1992): 35-66.
Burnyeat, M.F. “Culture and Society in Plato’s Republic.” The Tanner Lectures on Human Values 20
(1999), 215-324.
Cross, R.C. and A.D. Woozley. Plato’s Republic: A Philosophical Commentary. New York: St. Martin’s
Press, 1964.
133
Ellul, Jacques. “The Characteristics of Propaganda.” In Readings in Propaganda and Persuasion, edited
by Garth Jowett and Victoria O'Donnell, 1-49. Thousand Oaks, CA: Sage, 2006.
Evans, Matthew. “Plato on the Norms of Speech and Thought.” Phronesis 56 no. 4 (2011): 322-49.
Ewegen, S. M. Plato's Cratylus: The Comedy of Language. Bloomington: Indiana University Press,
2013.
Ferguson, A. S. “Plato's Simile of Light. Part II. The Allegory of the Cave.” The Classical Quarterly
16, no. 1 (Jan 1922): 15-28.
Fine, Gail. “Knowledge and Belief in Republic V.” Archiv für Geschichte der Philosophie 60 no. 2
(1978): 121-39.
Fine, Gail. “Plato on Naming.” Philosophical Quarterly 27 (1977): 290–301.
Gadamer, H.G. “Logos and Ergon in Plato’s Lysis.” in Dialogue and Dialectic: Eight Hermeneutical
Studies on Plato, translated by P. Christopher Smith, 1-20. Yale U.P., 1980.
Ganson, Todd. “The Rational/Non-Rational Distinction in Plato's Republic.” Oxford Studies in
Ancient Philosophy 36 (2009): 179-197.
Gold, Jeffery B. “The Ambiguity of ‘Name’ in Plato’s Cratylus.” Philosophical Studies 34 no. 3 (1978):
223-251.
Gonzalez, Francisco J. “Nonpropositional Knowledge in Plato.” Apeiron 31 no.3 (1998): 235-284.
Griswold, Charles L. “Plato on Rhetoric and Poetry.” The Stanford Encyclopedia of Philosophy (Fall
2014 Edition), edited by Edward N. Zalta,
http://plato.stanford.edu/archives/fall2014/entries/plato-rhetoric/.
Gulley, Norman. Plato's Theory of Knowledge. Greenwood Press, 1962.
134
Hall, Dale. “Interpreting Plato's Cave as an Allegory of the Human Condition.” Apeiron 14 no. 2
(1980): 74-86.
Halliwell, S. Plato: Republic X, Translation and Commentary. Warminster: Aris and Phillips, 1989.
Harte, Verity. “Language in the Cave” in Maieusis: Essays in Ancient Philosophy in Honour of Myles
Burnyeat, edited by Myles Burnyeat and Dominic Scott, 195--215. Oxford University Press,
2007.
Jirsa, Jakub. “Sophists, Names and Democracy.” Croatian Journal of Philosophy 12 no. 2 (2012): 125138.
Jowett, Garth, and Victoria O'Donnell. Propaganda and Persuasion. Thousand Oaks, CA: Sage,
2006.
Kahn, Charles. “The Greek Verb ‘To Be’ and the Concept of Being.” Foundations of Language 2
(1966): 245-265.
Keyt, David. “Plato and the Ship of State.” In The Blackwell Guide to Plato’s Republic, edited by
Gerasimos Santas, 189-213. Oxford: Blackwell, 2008.
Kretzmann, Norman. “Plato on the Correctness of Names.” American Philosophical Quarterly 8 (2)
(1971): 126-138.
Levin, S.B. The Ancient Quarrel between Philosophy and Poetry Revisited. Plato and the Literary Tradition.
Oxford: Oxford University Press, 2001.
Liddell, Henry George, Robert Scott, and Henry Stuart Jones. A Greek-English Lexicon. Oxford:
Clarendon Press, 1996.
135
Ludlow, Peter. “Contextualism and the new linguistic turn in epistemology.” In Contextualism in
Philosophy: Knowledge, Meaning, and Truth, edited by Gerhard Preyer and Georg Peter, 11-50.
Oxford University Press, 2005.
Malcolm, J. “The Cave Revisited.” Classical Quarterly 31 no. 01 (1981):60-.
Morrison, J. S. “Two Unresolved Difficulties in the Line and Cave.” Phronesis 22 no. 3 (1977): 212231.
Moss, Jessica. “Appearances and Calculations: Plato's Division of the Soul.” Oxford Studies in
Ancient Philosophy 34 (2008): 35-68.
Moss, J. “Plato's Appearance-Assent Account of Belief,” Proceedings of the Aristotelian Society 114 no.
2 (2014): 213-238.
Moss, Jessica. “What is Imitative Poetry and Why is it Bad?” in The Cambridge Companion to Plato's
Republic, edited by G. R. F. Ferrari, 415–44. Cambridge: Cambridge University Press, 2007.
Nehamas, Alexander. “Episteme and Logos in Plato's Later Thought.” Archiv für Geschichte der
Philosophie 66 no. 1 (1984): 11-36.
Nehamas, Alexander. “Plato and the Mass Media.” The Monist 71 no. 2 (1988): 214-34.
Nehamas, Alexander. “Plato on Imitation and Poetry in Republic 10.” in Plato on Beauty, Wisdom,
and the Arts, edited by J. M. E. Moravcsik and Philip Temko, 47-78. Rowman and Littlefield,
1982.
Nettleship, Richard Lewis. Lectures on the Republic of Plato. R. West, 1937.
North, Helen. “Inutilis sibi, perniciosus patriae: A Platonic Argument Against Sophistic Rhetoric,”
Illinois Classical Studies 06 no. 2 (1981). 242-271.
136
Partee, Morriss Henry. “Plato's Theory of Language.” Foundations of Language 8 no. 1 (1972): 113132.
Peterson, Sandra. Socrates and Philosophy in the Dialogues of Plato. Cambridge University Press, 2011.
Plato, and James Adam, The Republic of Plato, edited with Critical Notes, Commentary and Appendixes.
Cambridge University Press: Cambridge, 1965.
Plato, and H. N. Fowler. Plato: Cratylus, Parmenides, Greater Hippias, Lesser Hippias. Loeb Classical
Library Cambridge Mass.: Harvard University Press, 1926.
Plato & E.H. Gifford. Euthydemus. Translated by Benjamin Jowett. Oxford: Clarendon Press,
1905.
Plato, and R. Hackforth. Plato: Phaedrus. Cambridge: Cambidge University Press, 1952.
Plato and Glenn R. Morrow. Plato's Epistles. Indianapolis: The Bobbs-Merrill Company, Inc., 1962.
Plato, and C.D.C. Reeve. Cratylus, Translated, with Introduction and Notes, by C.D.C. Reeve.
Indianapolis: Hackett, 1998.
Plato, and Paul Shorey. The Republic. London: W. Heinemann, 1930.
Putnam, Hilary. “Explanation and Reference.” In Conceptual Change, edited by Glenn Pearce and
Patrick Maynard, 196-214. Dordrecht: D. Reidel, 1973.
Raven, J. E. Plato's Thought in the Making: A Study of the Development of His Metaphysics. Greenwood
Press, 1965.
Reeve, C. D. C. Philosopher-Kings: The Argument of Plato's Republic. Hackett Pub. Co., 1988.
Robinson, R. Essays in Greek Philosophy. Oxford, 1969.
Sayre, Kenneth M. Metaphysics and Method in Plato's Statesman. Cambridge University Press, 2006.
Sedley, D. Plato's Cratylus. Cambridge: Cambridge University Press, 2003.
137
Silverman, Allan. “Plato's Middle Period Metaphysics and Epistemology.” In The Stanford
Encyclopedia of Philosophy (Fall 2014 Edition), edited by Edward N. Zalta,
http://plato.stanford.edu/archives/fall2014/entries/plato-metaphysics/.
Sheffield, Frisbee C. C. Plato’s Symposium: The Ethics of Desire. New York: Oxford University Press,
2006.
Sider, Theodore. Writing the Book of the World. Oxford: Oxford University Press, 2011.
Smith, Imogen. “False Names, Demonstratives and the Refutation of Linguistic Naturalism in
Plato's ‘Cratylus’ 427d1-431c3.” Phronesis 53, no. 2 (2008): 125 - 151.
Smith, Imogen. “Taking the Tool Analogy Seriously: Forms and Naming in the Cratylus.” The
Cambridge Classical Journal 60 (2014), 75-99.
Smith, Nicholas D. and Thomas Brickhouse. “Justice and Dishonesty in Plato's Republic.” Southern
Journal of Philosophy 21 no. 1 (1983): 79-95.
Spellman, Lynne. “Naming and Knowing: The ‘Cratylus’ on Images.” History of Philosophy Quarterly
10 no. 3 (1993): 197-210.
Swoyer, Chris, Maria Baghramian, and Adam Carter. "Relativism." The Stanford Encyclopedia of
Philosophy (Summer 2015 Edition), edited by Edward N. Zalta,
http://plato.stanford.edu/archives/sum2015/entries/relativism/.
Tate, J. “‘Imitation’ in Plato's Republic.” CQ 22 (1928): 16-23.
Thomas, Christine J. “Inquiry Without Names in Plato's Cratylus.” Journal of the History of
Philosophy 46 no. 3 (2008): 341-364.
Thucydides. History of the Peloponnesian War. Translated by Richard Crawley. New York: Dover,
2004.
138
Veltman, Andrea. “The Justice of the Ordinary Citizen in Plato's Republic.” Polis: The Journal of the
Society for the Study of Greek Political Thought 22 no. 1 (2005): 45-59.
Vlastos, Gregory. “A Metaphysical Paradox.” Proceedings and Addresses of the American Philosophical
Association 39 (1965): 5-19.
Vogt, Katja Maria. “Belief and Investigation in Plato’s Republic.” In Belief and Truth: A Skeptic
Reading of Plato. Oxford: Oxford University Press, 2012.
Von Fritz, Kurt. “The Philosophical Passage in the Seventh Platonic Letter and the Problem of
Plato's Esoteric Philosophy.” In Essays in Ancient Greek Philosophy, edited by John Peter
Anton, George L. Kustas and Anthony Preus, 408-447. New York: State University of New
York Press, 1971.
Wilberding, James. “Prisoners and Puppeteers in the Cave.” Oxford Studies in Ancient Philosophy 27
(2004): 117-39.
Williams, Bernard. “Cratylus’ Theory of Names.” in Language and Logos, edited by Schofield and
Nussbaum, 83-93. Cambridge, 1982.
Wilson, J. R. S. “The Contents of the Cave.” Canadian Journal of Philosophy, supp. vol. 2 (1976):
117-27.
Wittgenstein, Ludwig. Tractatus Logico-Philosophicus: German and English Edition. Translated by C.K.
Ogden. Routledge, 1981.
139
ABSTRACT
THE SOPHIST IN THE CAVE
EDUCATION THROUGH NAMES IN PLATO’S REPUBLIC
by
DANIEL PROPSON
May 2016
Advisor: Dr. Joshua Wilburn
Major: Philosophy
Degree: Doctor of Philosophy
The Cratylus is often considered an isolated dialogue in Plato’s corpus, and the major theses
of the Cratylus are often seen as disposable and problematic elements in Platonic thought. When
one carefully compares this dialogue, however, to Plato’s comments elsewhere about rhetoric and
dialectic, a set of fascinating connections emerge. In this dissertation, I argue that the Republic ought
to be read in light of the Cratylus. In the former dialogue, Plato is vitally concerned with the use of
accurate language in his republic, a fact most clearly brought out by his accusation against
demagogues: that they “give names” to things on the basis of the beliefs of the populace, not on the
basis of reality. I argue that this sort of popular false nomenclature should be identified with Plato’s
discussion of deceptive names in the Cratylus. Moreover, I explain how Plato’s discussions of
sophistical manipulation of names in the Cratylus, the Euthydemus, and the Sophist can be used to
illuminate the epistemological landscape of the Republic. In particular, I contend that the cave image
can be best understood only when we understand that the cave itself is the realm of human language,
140
a realm that includes two very separate regions: the shadow language which is presented by the
sophist, and the truer language represented by the puppets which cast the shadows. This second
region of language, I argue, plays a central role in the education of the guardians.
141
AUTOBIOGRAPHICAL STATEMENT
Upon the receipt of his doctorate, Daniel Propson will be an Adjunct Assistant Professor
of Philosophy at Oakland University in Rochester Hills, Michigan. He grew up in Rochester Hills,
and attended Bishop Foley High School. He earned his Bachelor of Education degree at the
University of Michigan in Ann Arbor, and taught as a high school English teacher for seven years
in Ypsilanti. During that period, he met and married Priscilla Greathouse; they have five energetic
children. Daniel entered graduate school at Wayne State in the Fall of 2009. In his time there, he
published papers on Hume and bioethics, though he has since fallen in love with ancient
philosophy.