De Jong (2004) PDF

NARRATORS, NARRATEES, AND NARRATIVES
IN ANCIENT GREEK LITERATURE

MNEMOSYNE
BIBLIOTHECA CLASSICA BATAVA
COLLEGERUNT
H. PINKSTER • H. S. VERSNEL
I.J.F. DE JONG • P. H. SCHRIJVERS
BIBLIOTHECAE FASCICULOS EDENDOS CURAVIT
H. PINKSTER, KLASSIEK SEMINARIUM, SPUISTRAAT 134, AMSTERDAM
SUPPLEMENTUM DUCENTESIMUM QUINQUAGESIMUM SEPTIMUM
IRENE DE JONG, RENÉ NÜNLIST, and ANGUS BOWIE

NARRATORS, NARRATEES, AND
NARRATIVES IN ANCIENT
GREEK LITERATURE
STUDIES IN ANCIENT GREEK NARRATIVE,
VOLUME ONE
EDITED BY
IRENE DE JONG, RENÉ NÜNLIST and ANGUS BOWIE
BRILL
LEIDEN • BOSTON
2004
This book is printed on acid-free paper.
Library of Congress Cataloging-in-Publication Data
A C.I.P. record for this book is available from the Library of Congress.
ISSN 0169-8958
ISBN 90 04 13927 3
© Copyright 2004 by Koninklijke Brill NV, Leiden, The Netherlands
Koninklijke Brill NV incorporates the imprints Brill Academic Publishers,
Martinus Nijhoff Publishers and VSP.
All rights reserved. No part of this publication may be reproduced, translated, stored in
a retrieval system, or transmitted in any form or by any means, electronic,
mechanical, photocopying, recording or otherwise, without prior written
permission from the publisher.
Authorization to photocopy items for internal or personal
use is granted by Brill provided that
the appropriate fees are paid directly to The Copyright
Clearance Center, 222 Rosewood Drive, Suite 910
Danvers, MA 01923, USA.
Fees are subject to change.
printed in the netherlands
TABLE OF CONTENTS
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
General introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi
Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv
Introduction. Narratological theory on narrators, narratees, and

narratives (I.J.F. de Jong) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
part one. epic and elegiac poetry
Chapter One. Homer (I.J.F. de Jong) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

Chapter Two. Hesiod (R. Nünlist) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Chapter Three. The Homeric Hymns (R. Nünlist) . . . . . . . . . . . . . . . . 35
Chapter Four. Apollonius of Rhodes (M.P. Cuypers) . . . . . . . . . . . . . . . 43
Chapter Five. Callimachus (M.A. Harder) . . . . . . . . . . . . . . . . . . . . . . . . . 63
Chapter Six. Theocritus and Moschus (R. Hunter) . . . . . . . . . . . . . . . . 83
part two. historiography
Chapter Seven. Herodotus (I.J.F. de Jong) . . . . . . . . . . . . . . . . . . . . . . . . . . 101

Chapter Eight. Thucydides (T. Rood) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Chapter Nine. Xenophon (V. Gray) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
Chapter Ten. Polybius (T. Rood) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Chapter Eleven. Arrian (T. Hidber) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
Chapter Twelve. Appian (T. Hidber) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
Chapter Thirteen. Cassius Dio (T. Hidber) . . . . . . . . . . . . . . . . . . . . . . . . . 187
Chapter Fourteen. Herodian (T. Hidber) . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
part three. choral lyric
Chapter Fifteen. Pindar and Bacchylides (I.L. Pfeijffer) . . . . . . . . . . . 213

vi table of contents
part four. drama
Chapter Sixteen. Aeschylus (J. Barrett) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235

Chapter Seventeen. Sophocles (I.J.F. de Jong) . . . . . . . . . . . . . . . . . . . . . . 255
Chapter Eighteen. Euripides (N.J. Lowe) . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
Chapter Nineteen. Aristophanes (A. Bowie) . . . . . . . . . . . . . . . . . . . . . . . . 281
Chapter Twenty. Menander (R. Nünlist) . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
Chapter Twenty-One. Lycophron (N.J. Lowe) . . . . . . . . . . . . . . . . . . . . . 307
part five. oratory
Chapter Twenty-Two. Antiphon (M. Edwards) . . . . . . . . . . . . . . . . . . . . 317

Chapter Twenty-Three. Andocides (M. Edwards) . . . . . . . . . . . . . . . . . 325
Chapter Twenty-Four. Lysias (M. Edwards) . . . . . . . . . . . . . . . . . . . . . . . . 333
Chapter Twenty-Five. Isocrates (M. Edwards) . . . . . . . . . . . . . . . . . . . . . 337
Chapter Twenty-Six. Demosthenes (M. Edwards) . . . . . . . . . . . . . . . . . 343
Chapter Twenty-Seven. Aeschines (M. Edwards) . . . . . . . . . . . . . . . . . . 349
part six. philosophy
Chapter Twenty-Eight. Plato (K.A. Morgan) . . . . . . . . . . . . . . . . . . . . . . . 357

Chapter Twenty-Nine. Xenophon (V. Gray) . . . . . . . . . . . . . . . . . . . . . . . 377
part seven. biography
Chapter Thirty. Xenophon (V. Gray) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391

Chapter Thirty-One. Plutarch (C.B.R. Pelling) . . . . . . . . . . . . . . . . . . . . 403
Chapter Thirty-Two. Philostratus (T.J.G. Whitmarsh) . . . . . . . . . . . . . 423
Chapter Thirty-Three. Aelius Aristides (T.J.G. Whitmarsh) . . . . . . . 441
part eight. between philosophy and rhetoric
Chapter Thirty-Four. Dio Chrysostom (T.J.G. Whitmarsh) . . . . . . . . 451

Chapter Thirty-Five. Lucian (T.J.G. Whitmarsh) . . . . . . . . . . . . . . . . . . . 465
table of contents vii
part nine. the novel
Chapter Thirty-Six. Chariton (J. Morgan) . . . . . . . . . . . . . . . . . . . . . . . . . . 479

Chapter Thirty-Seven. Xenophon of Ephesus (J. Morgan) . . . . . . . . 489
Chapter Thirty-Eight. Achilles Tatius (J. Morgan) . . . . . . . . . . . . . . . . . 493
Chapter Thirty-Nine. Longus (J. Morgan) . . . . . . . . . . . . . . . . . . . . . . . . . . 507
Chapter Fourty. Heliodorus (J. Morgan) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 523
Epilogue. Narrators, narratees, and narratives in ancient Greek

literature (I.J.F. de Jong and R. Nünlist) . . . . . . . . . . . . . . . . . . . . . . . . . . 545
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 555
Thematic index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 579
PREFACE
The germ of this book lies in a research project that I developed in

1988, after finishing my PhD. That project promised the narratological
investigation of narratives which are embedded in non-narrative genres,
notably the Euripidean messenger-speech, the myth of the Pindaric
victory ode, and the narratio in forensic speeches. As it turned out,
the Euripidean messenger-speeches offered such a wealth of material
that I devoted an entire book to them. After that I returned to my
‘first love’ Homer. I revived my plan in 1995, when I applied to the
Netherlands Organization for Scientific Research for a ‘Pioneer grant’,
i.e., a grant for a group of scholars working on an innovative subject. By
that time, the project had been re-christened Narrationis Ratio. A History
of Ancient Greek Narrative, and, as this title suggests, expanded to include
both narrative and non-narrative genres. In 1998 the opportunity came
to fulfil my old dream, when I was awarded a professorship of the
Royal Netherlands Academy of Arts and Sciences, which included a
modest budget. In September 1999 I organized a two-day workshop
in Amsterdam, which brought together prospective contributors, and
during which the principles and practical organization of the project
were discussed. One of the conclusions was that it would be impossible
to complete the task we had set ourselves in a single volume, and
that a series of volumes would be needed. On that occasion I asked
Angus Bowie to co-edit the first volume. But it was soon clear to
me that editing a multi-author and conceptually and methodologically
experimental work such as this was a huge task, and I asked René
Nünlist to act as—a second—co-editor of the first volume and to
become co-editor of the series as a whole.
I wish to thank Heleen Keizer and Linda Woodward for their assis-
tance in copy editing the texts. But my greatest thanks go to Michiel
Klein Swormink of Brill Publishers for the trust he has shown in this
truly pioneering project.
GENERAL INTRODUCTION
Someone interested in the history of ancient Greek literature has at

his disposal a large number of literary histories.1 Someone interested
more specifically in the development of Greek drama, historiography,
rhetoric, or literary criticism can also consult a series of handbooks.2
But someone interested in the forerunners of the most popular literary
genre of our own times, the novel, is left empty handed.3 The story of
ancient Greek narrative is as yet untold.
And yet the lack of a history of ancient Greek narrative is hardly
surprising, given the sheer quantity of material. A rough estimate tells
us that more than half of Greek literature is narrative, if we include
historiographical, biographical, and philosophical narrative and narra-
tives which are embedded in other genres. However, it appears that the
moment has come to undertake the task of writing such a history. The
reason is two-fold.
First, the twentieth century witnessed the rise of narratology, which
has provided literary scholars with a set of refined analytical and de-
scriptive terms. In the recent past, narratology has also taken root in
classical scholarship, which means that an important condition for the
writing of a history of ancient narrative has now been fulfilled: we have
the tools, as well as a growing number of people who know how to
wield them.
In the second place, literary history itself, after a difficult period
in the sixties and seventies of the previous century, when ‘the fall
1 E.g. Dihle 1967, 1989; Lesky 1971; Easterling and Knox 1985; Canfora 1986; and
Saïd, Trédé and Boulluec 1997.

2 Cf. e.g. (for drama) Jens 1971; (for historiography) Marincola 1997; (for rhetoric)
Leeman 1963 and Kennedy 1994; and (for literary criticism) Kennedy 1989.
3 Scholes and Kellogg 1966 offer interesting discussions on the ancient roots of
various formal aspects of the modern novel (plot, characterization, point of view),
but their corpus is limited and theoretical basis somewhat outdated. Doody 1996 also
explores the classical pedigree of the modern novel, but confines herself to the ancient
novel.
xii general introduction
of literary history’ was proclaimed,4 is making a come-back.5 One

of the new directions suggested is a move away from the traditional
biographical, ‘man and his work’ approach. That such a new direction
may also be fruitful for classics, has been demonstrated by the recent
literary history edited by Oliver Taplin, which focuses on the receivers of
the texts: readers, spectators, and audiences.6
The present history of ancient Greek narrative adopts yet another
approach, focusing on the formal devices within a text which authors
employ to enchant or persuade their audiences. Who introduces us to
the (fictional) world depicted, through whose eyes do we see it, how do
the events follow upon each other and combine to form a logical whole,
how are the people who live in this world presented to us, and how
are their words, prime indicators of their personalities, represented?
Traditional literary histories by and large present summaries of what is
known or generally thought about a work and author, but this one must
virtually start from scratch. For a number of texts some narratological
groundwork has been done, but the majority are as yet completely
unexplored. A series of Studies in Ancient Greek Narrative will investigate
the forms and functions of the main devices which narratology has
defined for us, such as the narrator and his narratees, time, focalization,
characterization, description, speech, and plot.
The aim of this enterprise is to combine the synchronic and the
diachronic, to offer not only analyses of the handling of a specific nar-
rative device by individual authors, but also a larger historical perspec-
tive on the manner in which techniques change over time, are put to
different uses and achieve different effects in the hands of different
authors, writing in different genres, and handling different material.
Each volume will contain an introductory chapter, which provides the
theoretical background to the topic under consideration, and an epi-
logue, which offers general conclusions about the forms and functions
of that topic through the ages and across the genres.
The first volume, which deals with such basic questions as what
actually constitutes a narrative and what kind of narrators present
those narratives, includes discussions of almost all genres and authors,
ranging from Homer to the novel, provided enough text has survived
to make possible a description of rules. In this way the corpus of the
4 Wellek 1973.
5 See, e.g. Ceserani 1990 and Perkins 1992.
6 Taplin 2000.
general introduction xiii
history of ancient Greek narrative is defined and the foundation laid

for subsequent discussion. Later volumes will be in a somewhat more
flexible format, and will not necessarily include all authors and all texts.
The series Studies in Ancient Greek Narrative is aimed primarily at classi-
cal scholars, but may also be of interest to a wider audience. As Wallace
Martin, one of the historiographers of narratology, writes:
… what appears to be new may simply be something that has been
forgotten, and scholarship on the Greek romances, medieval literature,
prose narratives of the seventeenth century, and non-western tales pro-
vides evidence useful for any general theory of narrative.7
In the same way, any general history of narrative should start at the
beginning and the Studies in Ancient Greek Narrative will explore that
beginning, from ‘the Greek romances’ all the way back to Homer.
IdJ.
7 Martin 1987: 28–29.

GLOSSARY
actorial motivation: the ‘why’ of the development of the story anal-

ysed in terms of the aims and intentions of a character. An actorial
motivation usually is explicit. Compare narratorial motivation.
analepsis (flashback, ‘Rückwendung’): the narration of an event which

took place earlier than the point in the story where we are. A distinc-
tion can be made between internal analepses (narrating events which fall
within the time limits of the main story) and external analepses (nar-
rating events which fall outside those time limits), in the case of internal
analepses, between repeating ones (narrating events also presented else-
where) and completing ones (narrating events which are not presented
elsewhere), and between narratorial and actorial analepses (those pre-
sented by the narrator and those presented by one of the characters).
argument function: the function or significance which an embedded

narrative has for the characters. Compare key function.
characterization: characterization may be explicit or implicit (when

personality traits have to be inferred by the narratees), narratorial or acto-
rial, synoptic (when a detailed introduction is given, often at a character’s
first appearance) or gradual (when pieces of information are released
only at intervals and have to be collected and turned into a composite
whole by the narratees themselves).
description: descriptions can be static (the story comes to a halt and

the outward appearance or nature of a person or thing is described)
or dynamic (when the history of an object, place, or person is told
or when an object etc. is described while in the process of being
made), narratorial or actorial (given by the narrator or by one of the
characters).
embedded narrative: a narrative which is embedded in the main

story; it is either told by the primary narrator or by a character
xvi glossary
acting as secondary narrator. It usually takes the form of an analepsis

or prolepsis. See also argument and key function.
embedded or secondary focalization: when the narrator repre-

sents in the narrator-text a character’s focalization, i.e., his percep-
tions, thoughts, emotions, or words (indirect speech). Embedded focal-
ization can be explicit (when there is a shifter in the form of a verb of
seeing, thinking, or a subordinator followed by subjunctive or optative)
or implicit.
fabula: all events which are recounted in the story, abstracted from
their disposition in the text and reconstructed in their chronological
order.
focalizer: the person (the narrator or a character) through whose ‘eyes’

the events and persons of a narrative are ‘seen’.
frequency: events may be told singulatively (telling once what happened

once), repetitively (telling more than once what happened once), or itera-
tively (telling once what happened more than once).
interlace technique: the technique of interweaving different story-

lines through regular switches between them.
key function: the significance which an embedded narrative has for

the narratees. Compare argument function.
main story: the events which are told by the primary narrator
(minus external analepses and prolepses).
narratees: the addressees of the narrator. We may distinguish between

external and internal, primary and secondary (tertiary etc.), and overt and covert
narratees. Compare narrator.
narration: we may distinguish between subsequent narration (following

after the events have taken place), simultaneous narration (at the same
time when the events are taking place), and prior narration (when the
events still have to take place).
narrator: the person who recounts the events of the story and thus
glossary xvii
turns them into a text. We may distinguish between external narrators

(who are not a character in the story they tell) and internal narrators
(who are), primary narrators (who tell the main story) and secondary
(tertiary etc.) narrators (who tell embedded narratives), overt narra-
tors (who refer to themselves and their narrating activity, tell us about
themselves, and openly comment upon their story) and covert narrators.
A special type of overt narrators are self-conscious narrators, who are
aware that they are narrating and reflecting on their role as narrator.
All narrators are also focalizers.
narrator-text: those parts of the text which are presented by the

primary narrator, i.e., the parts between the speeches. We may fur-
ther distinguish between simple narrator-text (narrator presents his own
focalization) and embedded focalization (narrator presents focaliza-
tion of a character).
narratorial motivation: the ‘why’ of the development of the story

analysed in terms of the aims and intentions of the narrator. The nar-
ratorial motivation often remains implicit. See also actorial motiva-
tion.
order: the chronological order of the fabula may be changed in the

story, for instance to create prolepses and analepses or any other
anachrony.
paralepsis: a speaker provides more information than, strictly speak-

ing, he could, e.g., when the narrator intrudes with his superior knowl-
edge into the embedded focalization of a character or when a character
knows more than is logically possible. Contrast paralipsis.
paralipsis: a speaker provides less information than he could; details

or events are left out, to be told at a later, more effective place. Contrast
paralepsis.
periphrastic denomination: a reference to a character not by prop-

er name but by a form of indirect description.
prolepsis (foreshadowing, ‘Vorauswendung’): the narration of an

event which will take place later than the point of the story where we
are. We may distinguish between internal prolepses (referring to events
xviii glossary
which fall within the time limits of the main story) and external pro-
lepses (which refer to events which fall outside those time limits), and
between narratorial and actorial prolepses. See also seed.
rhythm: the relation between story-time and fabula-time. An event

may be told as a scene (story-time=fabula-time), summary (story-time <
fabula-time), retardation (story-time > fabula-time), or ellipsis, i.e. not told
at all (no story-time matches fabula-time). Finally there may be a pause,
when the action is suspended to make room for an extended description
(no fabula-time matches story-time).
seed (hint, advance mention): the insertion of a piece of information,

the relevance of which will become clear only later. The later event thus
prepared for becomes more natural, logical, or plausible.
story: the events as dispositioned and ordered in the text (contrast:

fabula). The story consists of the main story+embedded narra-
tives. In comparison to the fabula, the events in the story may differ
in frequency (they may be told more than once), rhythm (they may
be told at great length or quickly), and order (the chronological order
may be changed).
text: the verbal representation of the story (and hence fabula) by a

narrator.
introduction
NARRATOLOGICAL THEORY ON NARRATORS,

NARRATEES, AND NARRATIVE1
I.J.F. de Jong
The narrator
Perhaps the most central concept in narratology is that of the narrator.

For most narratologists, his presence is the main criterion for calling a
text a narrative (as opposed to drama, where we are also dealing with
the representation of characters and events): ‘A narrative text is a text
in which a narrative agent tells a story.’2 It is an important principle of
narratology that this narrator cannot automatically be equated with the
author, even when he bears the same name; rather, he is a creation of
that author.3
Given that every narrative text has a narrator, the next step is to
describe and analyse that narrator. There are, in fact, many types of
narrators.4 The first thing to ask oneself is whether or not the narrator
is a character in his own story: if he is, we speak of an internal narrator,
if not, we speak of an external narrator.5 Internal narrators used to
be called first-person narrators, but this is a less fortunate term, since
external narrators can also refer to themselves as ‘I’, as witness, e.g.
1 For a historical overview of narratology see Martin 1987; for a systematic overview
of key terms see Prince 1987.

2 Bal [1985] 1997: 16. Cf. e.g. Friedemann 1910: 26; Genette 1979, [1983] 1988: 14;
Stanzel 1982: 15, and from the point of view of drama Pfister [1977] 1988: 2–4. For
narratologists who adopt a broader definition of narrative, which does not require the
presence of a narrator, see the section on narrative below.
3 Friedemann 1910: 21–22; Kayser 1958: 91; Genette [1972] 1980: 213–214; Stanzel
1982: 25–28.
4 The following analysis is based on Genette [1972] 1980: 212–260 and Bal [1985]
1997: 19–77.
5 Genette actually uses the terms homodiegetic (internal) and heterodiegetic (exter-
nal). Bal replaced these terms by character-bound and external, and in De Jong 1987:
33 I have systematized the terminology into internal and external.
2 introduction
the ‘tell me, Muse, of the man’ of the Odyssey’s external narrator. In
fact, all narratives are in principle recounted by a narrating subject—
even if this narrating ‘I’ nowhere refers to himself—so that this is not a
watertight criterion for distinguishing narrators.
Next, we must determine the level of narration at which the narrator
finds himself: the narrator who recounts the main story and whose
voice is usually the first we hear when the story begins, is the primary
narrator. This primary narrator may hand over the presentation of
events to a character who recounts a story in direct speech, in which
case we speak of a secondary narrator. When this character in turn
embeds another narrative in his own narrative, we are dealing with a
tertiary narrator, and so on.6
Together the criteria ‘internal’-‘external’ and ‘primary’-‘secondary’
suffice to describe most narrators in world literature. A good example
of an external primary narrator is the anonymous narrator in Jane
Austen’s Pride and Prejudice, while we find an internal primary narrator
in the person of Pip in Charles Dickens’ Great Expectations. We have an
external secondary narrator in Scheherazade in Tales from a Thousand
and one Nights (a primary narrator introduces the frame-narrative of
Scheherazade, who must tell the sultan stories in order to save her life;
she herself plays no role in the stories she recounts). An example of
an internal secondary narrator is Odysseus in the Odyssey (an external
primary narrator recounts the last phase of Odysseus’ return from Troy,
in the course of which the hero himself tells the Phaeacians about his
earlier experiences on the way home).
Having established the ‘identity’ of the narrator, we can go on to
investigate his role and his attitude.7 It is convenient to start with an
overt narrator, i.e., a narrator who clearly manifests himself as narrator
throughout the text. His presence can take various forms: he may be
dramatized (given a life and personality of his own), or comment on
the events he relates, or may be self-conscious (showing awareness that
he is telling a story and reflecting on his activity as narrator). When
the narrator displays none of these characteristics, we speak of a covert
narrator.8 Another aspect worth analysing is the narrator’s privileges:
6 Again, I follow Bal, who introduced these terms to replace Genette’s extradiegetic,
intradiegetic, and metadiegetic.

7 The following discussion is based on Booth [1961] 1983: 149–165; Genette [1972]
1980: 255–259; and Chatman 1978: 196–262.

8 This distinction between a covert and an overt narrative style is often referred to
i.j.f. de jong – narratological theory 3
does he know and reveal the outcome of his story, does he have access
to the inner thoughts of his characters, and can he move freely and
rapidly to and from all the locations in his story? An external narrator
will by definition have more privileges than an internal one, but the
latter may use his hindsight to supplement his knowledge.
In the case of an internal narrator,9 it is also relevant to examine
the role he plays in his own story, which may range from protagonist
(Odysseus or Pip) to mere witness (Zeitblom in Thomas Mann’s Doktor
Faustus or Carraway in Scott Fitzgerald’s The Great Gatsby), as well as
everything that lies in between.
A special—and fairly rare—phenomenon is second-person narra-
tion,10 which means that a narrator recounts the acts of a character
in the ‘you’ form. This form of narration may be confined to a brief
section, often triggered by an apostrophe, but it may also determine the
shape of an entire novel, as happens e.g. in Michel Butor’s La modifica-
tion, which has been called an extended apostrophe.
One aspect that is relevant to all narrators and all forms of narration
is the temporal relationship between narrator and narrated events.
Narration can be subsequent (when the narrative follows the action),
as is most often the case; simultaneous (when the narrative runs parallel
to the action, when someone reports to another what he is seeing);
or prior (when the narrative precedes the action, as in prophecies or
dreams).11
Some narratologists posit one more agent to account for the signify-
ing process of a narrative text: the implied author, the ‘ideal, literary,
created version’ of the author, who is responsible for the moral evalua-
as ‘showing’ versus ‘telling’ (the distinction, though not the precise terminology, derives
from Lubbock [1921] 1926: 62, 67).
9 Romberg 1962; Stanzel 1982: 71–72, 109–148; Sturrock 1993.
10 Fludernik 1994a and Kacandes 1994. Strictly speaking, this is just as unfortunate
a term as first-person narration: we are still dealing with a subject of narration, who
may be an internal, external, primary, or secondary narrator. For a detailed discussion
of how second-person narration might be integrated into Stanzel’s and Genette’s
typologies, see Fludernik 1994b. However, since it would be somewhat cumbersome
to speak of ‘the situation of a primary (etc.) narrator using second-person verb forms
and pronouns to refer to the actions of a character’, the shorthand term ‘second-person
narration’ will be used. I note that Genette [1983] 1988: 133 speaks of ‘second-person
narrating’ in inverted commas.
11 Genette [1972] 1980: 215–223.
4 introduction
tion of the story.12 Following Bal and Genette,13 I do not think that—for
a narratological analysis—such an extra agent is necessary:14
[A] narrative of fiction is produced fictively by its narrator and actually
by its (real) author. No one is toiling away between them, and every
type of textual performance can be attributed only to one or the other,
depending on the level chosen. For example, the style of Joseph and His
Brothers can be attributed only (fictively) to the celestial narrator who is
supposed naturally to speak in that pseudo-biblical language or to Mr.
Thomas Mann, a writer in the German language, winner of a Nobel
prize for literature, etc., who makes him speak that way … No place
here for the activity of a third person, no reason to release the real
author from his actual responsibilities (ideologic, stylistic, technical, and
other) … (Genette [1983] 1988: 139–140)
This quotation briefly reintroduces the author into the realm of narra-
tology, if only to distinguish the responsibilities of author and narrator,
and to show that the notion of an implied author can be dispensed
with. But it also has another relevance. Intertextuality is a factor of
importance in many ancient Greek narratives. As a narratologist, one
is immediately confronted with the question of who is responsible for
this intertextuality, the narrator or the author? In the former case, we
must posit the persona of a learned narrator, while in the latter case the
intertext does not enter the narrative universe and, strictly speaking,
falls outside the scope of a narratological analysis. There is no ready-
made solution to this problem, and each case must be decided on its
own merits.
Narratees
Storytelling is an act of communication and every narrator presupposes

an addressee, or narratee.15 More specifically, for every primary nar-
rator there is a corresponding primary narratee, for every secondary
narrator, a secondary narratee, and so on. When we turn to the ques-
12Booth [1961] 1983: 74–75.

13Bal [1985] 1997: 18 and Genette [1983] 1988: 136–150.
14 As Genette observes, in literary theory in general it may make sense to use the
concept of the inferred author, ‘everything the text lets us know about the author’
(1988: 148).
15 See Prince [1973] 1980 and Genette [1972] 1980: 259–262, [1983] 1988: 130–135.
tion of the narratees’ involvement in the story being told (whether they
are external or internal), we see that many combinations are possible.16
Let us start with primary narratees. An external narrator usually
addresses external narratees. This fairly common narrative situation is
found, for example, in Jane Austen’s Pride and Prejudice, where neither
narrator nor narratees play a role in the events recounted. As in the
case of the narrator and the author, it is tempting simply to equate
these external narratees with the listeners or readers, not least because
they are often referred to as ‘dear reader’. But we only have to think
of the ‘Madam’ and ‘Sir’ in Laurence Sterne’s Tristram Shandy to realize
that here again we are dealing with a product of the author’s imagina-
tion. Sometimes we find the combination of an external narrator and
an internal narratee, for example in ancient hymns, when a narrator
recounts the deeds of a god while addressing that god (the so-called Du-
Stil, which is a form of second-person narration). An internal narrator
may have a corresponding internal narratee, i.e. a person who was also
involved in the events recounted, as in Hesiod’s Works and Days, where
‘Hesiod’ recounts the story of their quarrel to his brother Perses. Again
we may find second-person narration, which in a novel like Oriana Fal-
laci’s A Man is even continued throughout the entire book. But more
often an internal narrator tells his story to persons who have not wit-
nessed the events themselves, i.e. to external narratees, as Pip does in
Charles Dickens’ Great Expectations (cf. his sudden address to ‘those who
read this’ at the end of chapter 9).
Primary narratees are either overt or covert, like primary narrators.
Overt narratees may be found in the narrator’s ‘reader’ or ‘you’, or
represented in the text by anonymous witnesses (‘the pensive character
which the curtained hood lent to their heads would have reminded the
observer of some early Italian conception of the two Marys’: Thomas
Hardy Tess of the D’Urbervilles) or anonymous interlocutors (‘there some-
one could object’). When narratees are covert, we may still sense their
presence, for example, in explanations that the narrator inserts on their
behalf, or negated passages, where their—implied—expectations are
contradicted or their curiosity piqued.
I turn to secondary narratees. The most common situation is where
character A informs character B about something A has experienced
16 The following analysis of narratees is my own; curiously enough, Genette dis-
cusses only primary (extradiegetic) and secondary (intradiegetic) narratees, not internal
(homodiegetic) and external (heterodiegetic) ones.
6 introduction
(internal secondary narrator–external secondary narratee). An example

is the couple Nelly Dean and Mr. Lockwood in Emily Bronte’s Wuthering
Heights. Next, we have character A telling character B about events in
which neither has participated (external secondary narrator–external
secondary narratee); this, too, is a fairly common situation and may be
illustrated by Scheherazade’s telling her thousand-and-one tales to the
sultan. Then there is character A recalling in the presence of character
B something they have both experienced (internal secondary narrator-
internal secondary narratee; here we may find ‘we’ forms or ‘you’
forms); an example of this situation is Zeus reminding Hera how he
once punished her (Il. 15.18–33). Finally, there is character A telling
character B about something B participated in (external secondary
narrator–internal secondary narratee; again we may find ‘you’ forms);
this is not a common situation, but one example is Achilles reminding
Thetis of her support of Zeus in the past (Il. 1.396–406).
It will be clear that narratees, both primary and secondary, are a
powerful instrument for influencing the reception of a text, in that they
provide the readers with figures to identify with or distance themselves
from.
Narrative
I have already referred to the fact that most narratologists see the
presence of a narrator as one of the conditions for calling a text a
narrative. The other main condition is a sequence of at least two real or
fictional events (as in Forster’s celebrated example: ‘The king died and
then the queen died’).17
Some scholars adopt a broader definition of narrative, which also
includes drama. Ricoeur, for example, says ‘I am not characterizing
narrative by its “mode”, that is, by the author’s attitude, but by its
“object”, since I am calling narrative exactly what Aristotle calls mu-
thos, the organization of the events.’18 This broader definition is fol-
lowed by the classical scholars Gould, Goward, and Markantonatos.19
17Forster [1927] 1979: 87.

18Ricoeur [1983] 1984: 36 and Chatman 1990: 109–118. Cf. also Barthes [1966]
1977: 79, who mentions drama as a form of narrative.
19 Gould 2001; Goward 1999: 10–13; and Markantonatos 2002. Cf. also Lowe 2000:
163–164.
Their main arguments are as follows. (1) The dichotomy between nar-
rative and drama is not clear-cut, since narrative may contain speeches
and drama may contain narrative. While this is true, no one would
dream of calling the Homeric epics a play and analysing them in terms
of drama theory; so why call drama a narrative and analyse it in terms
of narratology? (2) Drama displays the same devices found in narrative,
such as analepses and prolepses, choice of setting, and differences in
pace. These, they claim, can only be explained by assuming a central
controlling and selecting mind, a ‘narrator’ and can only be analysed
in narratological terms. Let me start with the latter point. Drama the-
ory, e.g. the highly systematic one devised by Pfister, offers a panoply
of critical terms by which to analyse prolepses and the like.20 It is
therefore simply not necessary to turn to narratology when discussing
drama texts. Indeed, applying narratology to drama dilutes the speci-
ficity of narratology and stretches its concepts to such a degree that
they become meaningless; every character on stage becomes a narrator
and a focalizer. Given that prolepses and the like are equally at home in
drama, there is no need to postulate a ‘narrator’ (even in inverted com-
mas). We may safely ascribe them to the author, and this is in fact what
Gould and Markantonatos ultimately do, when they equate the control-
ling and selecting mind or ‘narrator’ with the playwright-director!21 In
actual practice, both Gould and Goward concentrate largely on narra-
tives that are embedded within drama, such as messenger-speeches or the
recollections and prophecies of characters. This is perfectly acceptable,
and it is also the line followed in the Studies in Ancient Greek Narrative.
But drama is not the only literary genre which, while not narrative
itself, does include narratives. Since this is a recurrent phenomenon in
ancient Greek literature, a brief extension of my earlier discussion of
narrators and narratees is in order here.22 In the case of the victory
ode and oratory, it is most logical to say that the poet or orator, at the
moment he turns to his myth or narratio, becomes a primary narrator,
and to see his audience (primarily the victor and jury, but in the end
all listeners) as the primary narratees. In the case of the dramatic dia-
logues of Plato and Lucian, that is to say, those dialogues which lack a
narrative frame and consist solely of speeches, it seems most sensible—
20 I note, e.g. sections on ‘the perspective structure’, ‘characterisation’, ‘story and
plot’, ‘structures of time and space’.

21 Gould 2001: 333 and Markantonatos 2002: 6–7.
22 There appears to exist no discussion of this question in narratological theory.
8 introduction
in view of their close similarity to dialogues with a narrative frame—to

see them as narratives with a suppressed primary narrator and sup-
pressed primary narratees.23 The same reasoning can be applied to the
so-called ‘mimetic’ poems of Callimachus and Theocritus. In the case
of drama, the intra-dramatic narrators are best seen as secondary nar-
rators, even though there is no narrative frame and hence no sugges-
tion of a suppressed primary narrator. The chorus or characters they
address are intra-dramatic secondary narratees, which means that the
role of primary narratees is reserved for the spectators, who as such
can, e.g. savour instances of dramatic irony which elude the secondary
narratees on stage.
Just as non-narrative texts may contain narratives, narrative texts
may contain non-narrative elements. The first category is the dramatic
element of speech. We have already seen that speeches can be carriers
of embedded narratives by characters. In general, from Plato onwards,
narrative has been described as the genus mixtum, a combination of
narrative and speech, and speeches have traditionally been included
in narratological discussion. A second category is description, which
usually involves the present tense and temporarily brings the story to
a standstill. Again, narratologists such as Genette, Bal, and Hamon24
have argued for the relevance of descriptions within a narrative, and
for their inclusion in narratological theory. Both speech and description
will be the subject of later volumes of the series of Studies in Ancient Greek
Narrative.
A final question to be considered here is the status of historiograph-
ical texts. Whereas Barthes, one of the founding fathers of narratology,
without further discussion considered historiography a form of narra-
tive and hence as belonging to the domain of narratology,25 two other
narratologists, Genette and Cohn, have recently opened the discus-
sion on whether factual texts, specifically historiographical texts, have
a special status and whether, in the words of Cohn, we do not need a
‘historiographical narratology’.26 Their arguments include (1) analepses
and prolepses are always functional in historiography, whereas in fiction
they may be included purely for aesthetic reasons; (2) detailed scenes,
including speeches, in historiography are a sign of fictionalization; (3)
23 Genette [1972] 1980: 236–237 speaks of the ‘pseudo-diegetic’ form of narration.

24 Genette [1972] 1980: 99–106; Bal [1985] 1997: 36–43; Hamon 1993.
25 Barthes [1966] 1977: 79 and 1970.
26 Genette 1991: 65–93 and Cohn 1999: 109–131.
in the case of historiographical texts, we should postulate not only the

triad fabula–story–text, but also an extra, referential level: ‘the more or
less reliably documented evidence of past events, out of which the histo-
rian fashions his story’ and which in the text may manifest itself in the
form of a ‘testimonial stratum’; (4) embedded focalization in historiog-
raphy, according to Genette, can only occur in a modified or qualified
manner (‘X thought … as we know from his diaries’ or ‘X feared, as I
suppose’), while in Cohn’s view it is unacceptable; (5) in historiographi-
cal texts the narrator is to be equated with the author.
Without going into the question of whether these arguments are
valid for modern historiographical texts, I think they are certainly not
all pertinent in the case of ancient historiography, where—presumably
under the strong influence of Homer—we do find speeches (2) and
embedded focalization (4), without this being considered a sign of fic-
tionality. Equating the figure which, say, Herodotus projects of himself
in his Histories with the real Herodotus (5) seems a dangerously naïve
thing to do, not to mention Thucydides and Xenophon, who expressly
forestall any identification of author and narrator by employing an
external rather than an internal narrator, even though they themselves
play a role in the historical events. The strict functionality of analepses
and prolepses (1) is a more relevant point, and one that requires fur-
ther exploration. Finally, positing an extra, referential level (3) makes
sense, but is not unproblematic. There may be very little material from
which to reconstruct this level, as is often the case in ancient history,
with its relative lack of documents and archives. Also, this level itself is
no narrative (yet): it is the historiographer who turns historical events
into a narrative, for one thing by deciding what is ‘the beginning, mid-
dle, and end’. In other cases, this level may consist of the narratives of
oral sources or written predecessors.27 All in all, I am inclined to follow
Quintilian, who considers ancient historiography as ‘close to poetry’
and ‘in a sense a kind of prose poem, which is told to narrate, not
to win a case’.28 Historiographical and biographical texts will therefore
take up an important place in the Studies on Ancient Greek Narrative.
27 Seeing that most ancient Greek literature is traditional in the sense that the same
stories are recounted time and again, it may be relevant to posit an extra, referential
level in those cases as well, which would then consist of the narratives of predecessors.
28 Quintilian Institutio Oratoria 10.1.31: est enim [historia] proxima poetis, et quodam modo
carmen solutum est …

10 introduction
Embedded narrative
The primary narrator may decide to embed another narrative into his
narrative, either doing the narrating himself (Marcel recalling the story
of Un amour de Swann in Marcel Proust’s A la recherche du temps perdu),
or turning one of the characters into a secondary narrator (Odysseus’
Apologue).
These embedded narratives can fulfil various functions in relation to
the main narrative:29 they may be (1) explanatory (when they take the
form of an analepsis which recounts how the present has come to be);
(2) predictive (when they take the form of a prolepsis, which announces
what will happen); (3) thematic (when there is a resemblance between
embedded and primary narrative); (4) persuasive (when the embedded
narrative is intended to influence the further course of events in the
main narrative); or (5) distractive (when there is no relationship at all,
but the embedded narrative is told to entertain, as is often the case
in frame-narratives like the Canterbury Tales or Decamerone). Needless to
say, an embedded narrative can fulfil more than one function at the
same time. In the case of an embedded narrative told by a character, it
may also be relevant to distinguish between the function it has for the
secondary narratee(s), the character(s) who are listening (the ‘argument’
function), and for the primary narratees (the ‘key’ function).
Having introduced the theory of narrators and narratees, it is now
time to turn to the practice of ancient Greek narratives and investigate
the manifold guises that their narrators take.
29 The following discussion is based on Genette [1972] 1980: 231–234, [1983] 1988:
92–94; Bal [1985] 1997: 52–60; and Andersen 1987a.

part one
EPIC AND ELEGIAC POETRY

chapter one
HOMER
I.J.F. de Jong
The primary narrator
Our first acquaintance with the Homeric narrator1 is deceptive: in the

proems of both epics (Il. 1.1–7 and Od. 1.1–10) he steps forward openly.
This suggests an overt narrator, who will make clear his presence as a
narrating and focalizing subject throughout. However, after the proems
the narrator largely withdraws into the background.
Even the proems themselves, upon closer inspection, yield little infor-
mation about the persona of the narrator: neither name nor place nor
date. The only thing which we can deduce, in view of the fact that
he addresses the Muses, is that the narrator is a professional singer
(comparable to the characters Phemius and Demodocus in the Odyssey):
these alone invoke the Muses, the ‘patron’ goddesses of their art, where-
as ‘amateurs’, such as Achilles in Iliad 9.186–191 or Odysseus in Odyssey
9–12, do not.
The remainder of the poems offers a few scattered pieces of infor-
mation. The ‘such as men are now’ passages (e.g. Il. 5.302–304),2 an
instance of the ‘reference to the narrator’s own time’ motif, make it
clear that the narrator, and by implication his narratees, belong to a
later period than the characters, in other words, that his narration is
subsequent. This difference in time is also suggested by his occasional
use of absolute tote, ‘then, at that time in the past’ (Il. 6.314; 14.287;
Od. 8.74), the expression ēmati keinōi, ‘on that remote day’ (Il. 2.482 and
4.543), and his—single—reference to the heroes of his tale as hēmitheōn
genos andrōn, ‘a race of semi-divine men’ (Il. 12.23).
1 De Jong 1987: 41–53 and Richardson 1990.

2 Cf. Il. 12.378–385, 445–449; 20.285–287. This type of passage does not occur in
the Odyssey.
14 part one – chapter one
The Homeric similes have traditionally been seen as windows on the

narrator’s own world. However, many of them describe events of all
times, such as rain, storm, or snow, in short are omnitemporal (as is
also suggested by the presence of gnomic aorists and epic te), and as
such link the heroes of the past, the narrator and his narratees, and us,
the later readers.
The fact that the narrator belongs to a much later time than the
events he is recounting immediately implies that he is external, does
not himself play a role in those events. He is also omniscient, in that
he knows—and reveals—the outcome of events beforehand in numer-
ous prolepses (e.g. ‘this was the beginning of Patroclus’ downfall’: Il.
11.604),3 and in that he has access to the inner thoughts and emotions
of his characters, primarily in the numerous instances of embedded
focalization (e.g. ‘and Odysseus pondered within him, whether to go
for Melanthius with his cudgel and kill him or pick him up like a
jug and break his head on the ground. But in the end he endured
and restrained himself ’: Od. 17.235–238).4 In some places the narrator
suppresses his omniscience, introducing instead anonymous spokesmen:
e.g. ‘Athena went to Olympus, where they say (fasi) is the abode of the
gods, unmoving and eternal’ (Od. 6.42).5
The Homeric narrator is also omnipresent: he recounts what hap-
pens among the gods on Mount Olympus and among the heroes on
earth, in the Greek camp and in Troy, on Ithaca and such remote
places as the island of Calypso, regularly switching back and forth
between the different locations.
An external, omniscient, and omnipresent narrator is in fact the
archetypal narrator of early storytelling. Interestingly enough, the Ho-
meric narrator accounts for his omniscience: it is the result of his
collaboration with the Muses, who are eyewitnesses of everything that
happens in the world (Il. 2.485), and whom he calls on not only in
his proems but in various other places throughout his narrative (see
below).6 This explicit motivation of his omniscience is related to the
3 Narratorial prolepses are found in equal measure in Iliad and Odyssey; discussions
and inventories in Duckworth 1933; Hellwig 1964: 54–58; de Jong 1987: 86–89; and
Richardson 1990: 132–139.
4 Embedded focalization takes up about 5 per cent of the texts of Iliad and Odyssey;
de Jong 1987: 101–148, 1994; Richardson 1990: 126–132.

5 Cf. Il. 2.783 and 17.674.
6 For the Homeric Muses, see Svenbro 1976: 11–45; Murray 1981; de Jong 1987:
45–53; and Ford 1992: 57–89.

i.j.f. de jong – homer 15
proto-historiographical function of the epics.7 Although brought as en-

tertainment (cf. e.g. Od. 8.62–99), they are much more than that. They
tell of the deeds of heroes of the past, for whom their status, the respect
paid them by their peers, is a central concern. Next to land and riches,
one of the most important means of gaining status is kleos, one’s glorious
reputation as ‘doer of deeds and speaker of words’ (cf. Il. 9.443). In
the oral society in which these heroes live, there are two things that
preserve this kleos for generations to come: grave-mounds (cf., e.g. Il.
7.86–91) and tales. Most of these tales are no more than hearsay (cf.
Il. 2.486), but professional singers like the narrator can claim to offer
the historical truth, precisely because of their collaboration with the
Muses.
The Homeric concept of historical truth is not the same as the mod-
ern one. What is aspired to is a vivid evocation of the past, rather than
an accurate and painstaking reconstruction of that past.8 The narrator
brings the past alive (e.g. by frequently allowing the heroes themselves
to speak) and transports his narratees back to that past (by narrat-
ing events so graphically that they almost become eyewitnesses). This
means that he allows himself a great deal of amplification and inven-
tion, while leaving intact the core of the story (which was presumably
quite small). Here the Homeric narrator sets the tone for all traditional9
and historical literature to come: lyric poets, tragedians, and historians
alike will take the liberty of amplifying and adapting the events they
relate.
After the proems the Homeric narrator withdraws into the back-
ground and narratorial interventions are rare. On occasion we find:
– Muse-invocations: e.g. ‘tell me now, Muses, who have your homes
on the Olympus, whoever came first against Agamemnon, either
of the Trojans or of the illustrious allies?’ (Il. 11.218–220),10 which
not only explain and authorize the narrator’s omniscience, but
also serve to enhance the significance of a scene or hero. Closely
related to these are utterances such as ‘there whom first and whom
7 Cf. Strasburger 1972 and Ford 1992.

8 Andersen 1987b; Finkelberg 1990; Ford 1992: 49–56; Bowie 1993: 8–20; Pratt 1993:
1–53; and Bakker 1997: 156–183.
9 I use ‘traditional’ here to refer to all literature that makes use of a body of
given stories. Only comedy and Hellenistic literature will introduce non-traditional,
i.e. purely invented stories.
10 Cf. Il. 2.484–493, 761–762; 11.218–220; 14.508–510; 16.112–113; never in the Odys-
sey.
last did he kill?’ (Il. 5.703–704),11 which though ostensibly questions

(addressed to the Muses) are in fact expressive statements.
– apostrophes, when the narrator addresses one of his characters
with a vocative and then switches (for one or two lines) to second-
person narration: e.g. ‘Then who killed you first, who last, Patro-
clus, when the gods called you to your death?’ (Il. 16.692–693).
Though not all scholars agree, this marked way of narration seems
a sign of compassion or sympathy on the part of the narrator.12
– gnomic utterances: e.g. ‘Zeus’ will is always the stronger; he terri-
fies even a brave man, and takes away his victory, easily, after he
has himself urged him to fight’ (Il. 16.688–670).13
– evaluative comments: e.g. ‘the dream went away and left Aga-
memnon believing things in his heart that were not to be accom-
plished. For he thought that on that very day he would take Priam’s
city; fool, who knew nothing of all the things Zeus planned to accom-
plish, who was yet minded to visit tears and sufferings on Tro-
jans and Greeks alike in strong battles.’ (Il. 2.35–40), or ‘this man
spoke among the overbearing suitors’ (Od. 20.291).14 While these—
sparse—comments in the Iliad tend to stress either the glorious or
the tragic nature of the heroes (who march towards their death
without knowing it), in the Odyssey they are negative or criti-
cal (mainly concerning the suitors, who persist in their criminal
behaviour despite numerous warnings). Sometimes we are dealing
with metanarrative comments: ‘it is difficult for me to relate all
these things like a god’ (Il. 12.176), which can be considered an
early instance of the ‘aporia’ motif, or ‘But the mass I could not
describe nor mention by name, not even if I had ten tongues, ten
mouths, an unbreakable voice and a bronze heart … The leaders
of the ships, however, I will name and the total sum of their ships’
(Il. 2.488–493), an instance of the ‘recusatio’ motif.
– ‘if not’-situations: e.g. ‘Now wretched Odysseus would have per-
ished, beyond his destiny, had not the gray-eyed goddess Athena
11 Cf. Il. 11.299–300 and 16.692–693.

12 Yamagata 1989, which also discusses older literature.
13 Cf. Il. 20.264–266. In fact, these are the only two instances, since gnomic utter-
ances are almost exclusively voiced by characters; Lardinois 1997: 229–233.

14 Cf. Il. 2.38, 873; 4.194; 6.234, 262; 12.113, 127; 13.569; 16.46; 16.685–691; 17.293;
18.4, 311–313; 20.411, 445; 22.5; 22.158–161, 402–404; Od. 1.8; 2.156, 324 = 4.769 =
17.482 = 20.375 = 21.361; 4.627 = 17.169, 772; 13.170; 17.233; 19.62; 22.31–33; 23.152;
24.469. De Jong 1987: 18–20; Griffin 1986; and Richardson 1990: 158–166.
given him forethought’ (Od. 5.436–437).15 They create tension (is

the ‘Odysseus’ story coming to an end here, with the death of the
hero?) or pathos (once again our hero is facing death).
In the eighteenth century the covert nature of the Homeric nar-
rator led to the communis opinio that the narrative style of the Iliad
and Odyssey was distanced and impassive, and that the events told
themselves. It is only in the last decades that studies have shown
this view to be questionable: though ‘invisible’, the Homeric nar-
rator is very active, ‘rigorously controlling our beliefs, our inter-
ests, and our sympathies’.16 Some of the more important means of
implicit evaluation or emotionalization by the narrator are:
– descriptions: e.g. Eumaeus’ yard and home, which he built himself
in order to take proper care of his master’s pigs (Od. 14.5–28),
immediately characterize him as a loyal servant; Andromache’s
headdress, ‘which golden Aphrodite had given her on the day
when Hector of the shining helmet led her from the house of
Eetion, after he had given countless gifts’ (Il. 22.468–472), evokes
the pathos of her situation following Hector’s death.17
– comparisons and similes: e.g. the comparison of two young war-
riors to tall pine trees at the moment of their death (Il. 5.539–540)
evokes pathos.18
– motifs: e.g. the death of the Trojan Hippothous in Il. 17.300–303
(‘he collapsed … far away from generous Larisa, and he could
not render again the care of his dear parents; he was short-lived,
beaten down beneath the spear of high-hearted Ajax’) combines
the three pathetic motifs of ‘far from home’, ‘short life’, and
‘bereaved parents’.
– juxtaposition; e.g. by narrating the suitors’ gleeful anticipation of
Telemachus’ trip abroad and Euryclea’s concern one after the
15 Cf. further Il. 2.155–156; 3.373–375; 5.22–24, 311–313, 679–680; 6.73–76; 7.104–
108, 273–276; 8.90–91, 130–132, 217–219; 11.310–312, 504–507; 12.290–293; 13.723–725;

15.121–127,459–464; 16.698–701; 17.70–73, 319–325, 530–532,613–614; 18.151–152+165–
168; 20.288–291; 21.211–213, 544–546; 22.202–204; 23.154–155, 382–384, 490–491, 540–
542; 733–734; 24.713–715; Od. 5.426–427; 14.32–34; 16.220–221; 21.125–128, 226–227;
23.241–242. De Jong 1987: 68–77; Richardson 1990: 187–191; Nesselrath 1992; and
Louden 1993.
16 Booth [1961] 1983: 4–5 and de Jong 1987.
17 Griffin 1980: 1–49.
18 Griffin 1980: 103–143.
other (Od. 2.301–381), the narrator underlines the depravity of the

former, the affection of the latter.19
The main story
There is a remarkable difference between the emphatic and explicit

openings of the Iliad and Odyssey (the proems) and their unobtrusive
and unmarked endings. The proems, which as we have seen, take the
form of Muse-invocations, serve to mark the transition from the real
world to the narrated world by introducing the mediator between the
two (the narrator), to give an idea of what the story which follows is
about, and to indicate its starting point.
The lack of formalized endings has been linked to the oral back-
ground of the Homeric epics:20 as is clear from Od. 8.536–543, where
the singing of Demodocus is abruptly interrupted by Alcinous, singers
always had to reckon with the possibility of suddenly having to end
their song; this circumstance was not conducive to the development
of elaborate endings. This does not mean, however, that the Iliad and
Odyssey have no closure. This is effected by ring-composition (books 1
and 24 of the Iliad both feature a divine assembly, a father coming to
the enemy camp to release his child, and summarized periods of nine
and eleven days; books 1–2 and 24 of the Odyssey both feature divine
councils and Ithacan assemblies), by a curtain call of the major charac-
ters in Iliad 23–24 (notably the Games) and Odyssey 23–24, and by the
presence of natural closural motifs such as burial (of Hector, the suitors,
and Achilles) and reconciliation (between Priam and Achilles and—
temporarily—the Greeks and the Trojans; and between Odysseus and
the families of the dead suitors).21
Secondary narrators and embedded narratives
In addition to the primary narrator, the Homeric epics feature a host

of secondary narrators. In the first place there are heroes recounting
events from their own past: e.g. Nestor (Il. 1.267–273; 7.132–156; 11.671–
19 Goldhill 1988a.
20 Van Groningen 1958: 70–77.
21 Whitman 1958: 249–284; Macleod 1982: 28–34; Taplin 1992: 251–284; de Jong
2001: Introduction to Book 24.

761; 23.629–643), Phoenix (Il. 9.447–484), and Eumaeus (Od. 15.403–

486). The Odyssey in particular contains many long embedded narra-
tives, which made ‘Longinus’ remark that ‘most of the Odyssey is nar-
rative’ (On the Sublime 9.13). An important group are the nostoi-stories
of Nestor (Od. 3.130–185), Menelaus (Od. 3.276–302 +311–312; 4.351–
586), Agamemnon (3.193–198; 4.512–537; 11.405–504), and of course
Odysseus himself (Od. 3.153–164; 4.555–560; 9–12), which are told in
several instalments and foreshadowed in Phemius’ song in 1.326–327:
‘he sang of the Achaeans’ bitter homecoming from Troy which Athena
had inflicted upon them’. A special case is that of Odysseus’ many lying
tales (the first in 13.253–286),22 which serve to back up his disguise as
a beggar and which display a shrewd and ever-changing mix of true
and invented elements (not seldom allomorphs of his real adventures).
Except for Athena, all the listeners totally believe these tales, which
led Aristotle to comment that ‘Homer most of all has taught the other
poets how to tell lies properly’ (Poetics 1460a18–20).
These internal narrators are more restricted in their knowledge than
the omniscient and omnipresent external narrator: they do not know
the future, cannot read the minds of other characters, and cannot
be present everywhere. Occasionally these restrictions are effectively
exploited: the fact that on his way home from Troy Nestor becomes
separated from Odysseus makes it impossible for him to tell Telema-
chus about the whereabouts of his father (Od. 3.160–185), and therefore
necessitates the youth’s visit to Menelaus. But more often they are
circumvented by allowing the internal narrators to use their ex eventu
knowledge. Thus Nestor knows what lies ahead of him in Od. 3.160–
161, Odysseus can read the minds of his companions in Od. 10.415–417,
and knows what took place on Mt. Olympus in Od. 12.374–390. Only
their references to divine interventions are, almost without exception,
non-specific (theos, theoi, daimōn, Zeus).23
An important difference between the external primary narrator and
the internal secondary narrators is their use of emotional and evalua-
tive words. As we have seen above, the narrator seldom makes explicit
comments; characters, however, quite regularly employ emotional lan-
22 Cf. further 14.192–359; 17.415–444; 18.138–140 and 19.75–80; 19.165–202+221–
248+268–299; and 24.244–279+303–314. For secondary literature, see de Jong 2001: ad

13.253–286, to which Grossardt 1998 should be added.
23 Jörgensen 1904.
guage in their narratives (as in their speeches in general).24 Thus Achil-

les refers to Apollo’s arrow, which caused the plague, as a ‘bad mis-
sile’ (Il. 1.382); Odysseus calls the stormy sea on which his raft floats
‘unspeakable’ (Od. 7.273), Circe’s potion which changed his men into
pigs ‘destructive’ (10.394), the meal which the cannibalistic Laestrygo-
nians made of his men ‘unpleasant’ (10.124), and himself and his men,
under attack by the Ciconians, ‘unfortunate’ (9.53).
Occasionally we find the rare form of second-person narration: e.g.
when Zeus reminds Hera how he once punished her (Il. 15.18–30) or
when Menelaus recalls how Helen nearly betrayed the Greeks inside
the Wooden Horse (Od. 4.266–288).25 This form of narration, which
flows from the fact that the narrator’s addressee plays a role in the
events recounted, also increases the urgency or impact of what is told:
Zeus’ story functions as a threat, Menelaus’ story as an implicit accusa-
tion.
But characters also tell stories about events in which they themselves
have not been involved, in which case they are external secondary nar-
rators. An example is the ‘Meleager’ story, as told by Phoenix in Il.
9.527–599, which he introduces as ‘an event of old times, not recent’.26
Here we find the kind of naïve, or rather unmotivated omniscience and
omnipresence familiar from fairytale and myth, of which I talked ear-
lier. Another characteristic of this type of embedded narrative is that
it is usually told allusively and elliptically: because the stories are well
known, their narrators can suppress details, motives, prehistory, etc.,
relying on the narratees to fill them in, while at the same time stressing
or even expanding those points which are most relevant to the situa-
tion at hand. An example is the story of the battle between the Cen-
taurs and Lapiths, as recounted by Antinous in Od. 21.295–304: all we
are told is that the Centaur Eurytion became drunk in the palace of
Peirithous and that a fight ensued; the larger context (the Centaur’s
attempt to rape Peirithous’ bride Hippodamea) is omitted. Antinous
does dwell on the punishment which the Centaur suffered at the hands
of the Lapiths (they cut off his ears and nose), obviously because this
is the punishment he has in mind for ‘the beggar’/Odysseus, to whom
24 Griffin 1986; de Jong 1988, 1992.

25 Cf. Il. 10.285–290; 20.188–194; 21.441–457; Od. 24.37–92.
26 Cf. Il. 5.385–404; 19.95–133; 24.602–617; Od. 21.295–304.
he is telling this story. This allusive style will become typical of choral
narration (→ Pindar, → Aeschylus, → Sophocles) and Hellenistic poetry
(→ Theocritus).
Embedded narratives are told either in answer to a question from
the narrator’s interlocutor (e.g. Od. 3.247–252, where Telemachus asks
Nestor to tell him the ‘Oresteia’ story) or spontaneously. In the latter
case they often have an ‘argument’ function and serve either as a horta-
tory paradigm (e.g. when Nestor recalls how the Lapiths, when fighting
the Centaurs listened carefully to his advice, in order to persuade the
Greeks to do the same now: Il. 1.259–274), or as a dissuasive paradigm
(e.g. when Phoenix holds up to Achilles the negative example of Melea-
ger). For the primary narratees these embedded narratives may have
an additional ‘key’ function. Thus, the ‘Meleager’ story prepares the
narratees for the development of the story: even more than Meleager,
Achilles will come to regret his refusal to join the fighting in time.
Embedded narratives are clearly marked off as independent units
through the use of ring-composition (e.g. ‘Now you and I must remem-
ber our meal. For even Niobe remembered to eat … But she remem-
bered to eat when she was worn out with weeping. … Come then,
we also must remember to eat’: Il. 24.601–602, 613, 618–619), an emo-
tional preamble, which often takes the form of the ‘recusatio’ motif (e.g.
‘I could not tell you all the exploits of enduring Odysseus, so many as
there are. But here is something he did and endured …’: Od. 4.240–
243)27 or the ‘aporia’ motif (‘what shall I tell you first, what last; for the
gods have given me many sorrows’: Od. 9.14–15), or some form of con-
clusion (‘all this I told you in truth, sorrowful though I am’: Od. 7.297).28
In Od. 11.328–332 Odysseus uses the same ‘recusatio’ motif to conclude—
somewhat abruptly—his narrative; this anticipates the later Abbruchs-
formel (→ Pindar).
A special type of embedded narrative is formed by the songs of
Demodocus (Od. 8.73–92, 266–366, 499–520). Strictly speaking, he is
not a secondary narrator, since his songs are quoted in indirect rather
than direct speech (Demodocus ‘began singing the story about the love
of Ares and Aphrodite, how they first lay together in the house of
Hephaestus, secretly …’), which after a few lines becomes an indepen-
dent construction (‘And Ares gave many presents and defiled the bed
of Hephaestus’). In this way the voices of primary and secondary nar-
27 Cf. Od. 3.113–114; 7.241–243; 11.517–520.

28 Cf. Il. 7.155; Od. 8.516–520; 11.328–330; 12.450–453; 14.359.
rator merge. The first and third song, dealing with episodes from the
beginning and the end of the Trojan War respectively, also provide us
with the rare situation of an external narrator—albeit unwittingly—
addressing an internal narratee (Odysseus). The figures of the singers
Demodocus and Phemius can be considered narratorial alter egos, in
that they are a mirror of the primary narrator, who as we saw earlier
himself is a professional singer.
Primary and secondary narratees
Like the primary narrator, the primary narratees are largely covert.
Only the Iliad contains explicit traces of them, in the form of the
‘indefinite second person’ device (‘there you could have seen’, e.g. Il.
4.223–225)29 and the ‘anonymous witness’ device (‘and there no more
could a man have disparaged the fighting’, Il. 4.539–542).30 But their
implicit presence is unmistakable and essential: they are the active
recipients of the narrative devices of the narrator, the ones who pick
up the pathos or feel the suspense he creates. Their presence can also
be sensed in:
– gar-clauses, which anticipate their questions (e.g. the explanation
that gods do not eat bread or wine anticipates the question of why
the gods have no blood but ichor: Il. 5.539–542).31
– ‘presentation through negation’ passages, which contradict exist-
ing expectations or create new ones (e.g. the fact that Patroclus did
not take Achilles’ spear with him in Il. 16.140–144 both contradicts
an expectation based on other arming scenes and creates tension:
what will be the role of this spear, which he has so conspicuously
left behind?).
– rhetorical questions (e.g., ‘of the others, who could mention their
names, so many of the Greeks as aroused battle behind them?’: Il.
17.260–261).32
– a very special way in which the existence of the narratees is
evoked, is when characters foresee that they themselves or people
29 Cf. 4.429–431; 5.85–86; 15.697–698; 17.366–367. Discussion in de Jong 1987: 54–57
and Richardson 1990: 174–178.

30 Cf. Il. 13.343–344; 4.421; de Jong 1987: 58–60.
31 De Jong 1987: 91–93 and Richardson 1990: 141–148.
32 Cf. Il. 22.202–204 and Od. 22.12–14.
around them one day will become the subject of song ‘for people
to come’ (essomenoisi): Il. 6.357–358.33
More tangible are the numerous secondary narratees, characters who
listen to stories told by other characters. Their function is more than
just that of passive listener; in many cases we see how secondary narra-
tors adapt their story to the recipient. This is most apparent in the case
of repeated stories, since here we are able to make comparisons. Thus
when the ‘Oresteia’ story is recounted to Telemachus, the emphasis is
on the role of Orestes, whose example he should follow (Od. 1.298–302
and 3.193–200); when recounted to Odysseus, the role of Agamemnon
is stressed, whose fate at the hands of Aegisthus and Clytemnestra he
must avoid (Od. 11.409–456).34 But even a story which is told only once
may be clearly tailored to its addressee: the song of Ares and Aphrodite
(Od. 8.266–366), which recounts how after a faux pas one man makes
amends to another, is intended to appease ‘the stranger’/Odysseus,
who has been insulted by Euryalus, but will soon receive conciliatory
words and gifts.
The embedded narratives, especially those of the Odyssey, trigger a
wide range of reactions from the secondary narratees, from aesthetic
admiration (e.g. Alcinous in 11.363–369) and emotional involvement
(e.g. Eumaeus in 14.361–362), to enchantment (e.g. Phaeacians in 13.1–2
or Eumaeus in 17.515–521). These reactions may help us to determine
the intended reaction of the primary narratees to the Iliad and Odyssey
themselves, which I see as a combination of these three.
One last aspect of the Homeric narrative style that deserves our
attention here is the brilliant handling of the hierarchy of narrators
and narratees, especially in the Odyssey. Firstly, the primary narrator
carefully distributes the presentation of one and the same story over
different secondary narrators (notably in the case of the ‘nostos’ sto-
ries of Agamemnon, Menelaus, and Odysseus, which are recounted in
several instalments by different speakers).35 Secondly, he arranges for
the interests of primary and secondary narratees to coincide (e.g. in
Od. 8.572–576, where Alcinous’ request to ‘the stranger’/Odysseus to
at last tell him about his wanderings reflects the curiosity of the pri-
mary narratees, who since the proem have also been waiting to hear
33 Cf. Od. 3.203–204; 8.579–580; 24.196–202.

34 The same phenomenon can be observed in the case of the ‘Tydeus’ stories (Il.
4.372–399; 5.801–808; and 10.285–290) and Odysseus’ lying tales.
35 De Jong 2001: Appendix C.
about those wanderings). Thirdly, he quells the curiosity of a tertiary

narratee and a secondary narratee at the same time (e.g. in Od. 4.535–
560, where Proteus’ information about Odysseus’ whereabouts is as rel-
evant to Menelaus, Proteus’ addressee in the past, as to Telemachus,
Menelaus’ addressee in the present). Fourthly, he superbly adapts the
interests of a secondary narrator to those of himself (e.g. when in his
Apologue Odysseus comes to the point of his stay with Calypso, and
stops narrating; for Odysseus it would have been the second time he
recounted this episode, for the primary narrator the third time …).
In conclusion, I would like to draw attention to what is perhaps the
greatest masterstroke of that covert but almighty Homeric narrator. He
repeatedly compares the hero of the Odyssey to a singer, notably in
a simile at the climax of the story, when Odysseus, who is stringing
the bow with which he will strike the first—decisive—blow against
his enemies, is compared to a singer stringing his lyre (Od. 21.406–
409).36 The message implied by this comparison is relevant to the entire
history of ancient Greek narrative: heroes may perform great deeds,
but their eternal fame depends on the narrators who turn those deeds
into stories.
36 Cf. 11.363–369 and 17.518–521. Moulton 1977: 145–153; Thalmann 1984: 170–173;
and Murnaghan 1987: 148–154.

chapter two
HESIOD
R. Nünlist
Theogony
Of all the surviving Greek texts, the Theogony is the first to give its
narrator a name: Hesiod.1
And once they [sc. the Muses] taught Hesiod fine singing, as he tended
his lambs below holy Helicon. This is what the goddesses said to me first,
the Olympian Muses, daughters of Zeus the aegis-bearer.
(Th. 22–25, transl. West)
Despite the transition from third to first person (‘Hesiod … me’) there
can be little doubt that they designate the same person.2 The transition
can be paralleled from the opening sections of Hecataeus, Herodotus,
and Thucydides.3 Hesiod’s case is admittedly somewhat different be-
cause he becomes a character in his own story (i.e. an internal narra-
tor).
Hesiod’s Dichterweihe (on which more below) forms part of the so-
called ‘Hymn to the Muses’ (Th. 1–103), which is best separated from
the rest of the Theogony for a narratological analysis. Whereas the main
part of the Theogony (104–end) is much closer to the Homeric epics
(see below), the ‘Hymn to the Muses’ shares many features with the
‘Homeric hymns’.4
In the first line of the ‘Hymn’, a self-conscious narrator proclaims
in the first-person plural (‘let us begin our singing’) that the Muses
1 This is not to say that Hesiod actually was the first Greek poet to mention his
name (so e.g. Jaeger [1933] 1954: 111). This question is better left open because the texts
of Homer’s and Hesiod’s predecessors have not been transmitted to posterity.
2 Differently and, to my mind, unconvincingly Ballabriga 1996.
3 Hecataeus 1 F 1a FGH, Herodotus proem vs. 1.5.3. etc., Thucydides 1.1.1. vs. 1.1.3.
etc.
4 For a detailed comparison see Friedländer 1914.
26 part one – chapter two
are to be his subject matter. The narrator then ‘disappears’, and the
subject matter is immediately expanded after an introductory relative
pronoun (Th. 2). The same structure (narrator’s ‘I’ as subject of the first
sentence, subject matter+relative pronoun) is to be found in a number
of ‘Homeric hymns’ (→).5 The relative pronoun starts off the narrative,
which in the present case is simultaneous and iterative.
The term ‘simultaneous iterative narration’ and the claim that Th.
2–21 is actually narrative need a brief explanation, because many schol-
ars work on the basis of the following equations: ‘past tense = nar-
rative’ vs. ‘present tense (and “timeless” aorists) = non-narrative’.6 It
is, however, preferable to apply Genette’s distinction between ‘singu-
lative’ and ‘iterative’ narration. Singulative narration means ‘narrating
once what happened once’, iterative narration means ‘narrating once
what happened n times’.7 This distinction, then, is to be combined with
a second, namely between ‘simultaneous’ and ‘subsequent’ narration
(→ Introduction). Admittedly, simultaneous iterative narration is rather
uncommon, but in the present case Hesiod’s choice was dictated by the
subject matter. Unlike singular events (birth, encounters, etc.), recurring
activities require a simultaneous iterative narration because gods are
immortal.8 The equation ‘present tense = non-narrative = descriptive’
is also misleading, because it does not do justice to a dynamic activity
like the dance and song of the Muses (Th. 2–21, 36–52) or Apollo’s jour-
ney to the assembly of the gods with subsequent dancing and singing
(h.Ap. 182–206). Scenes like these should not a priori be equated with the
gods’ more static ‘appearance, possessions, haunts and spheres of activ-
ity’.9 For these I suggest expanding Genette’s model with the notion
of durative (or omnitemporal) narration (‘narrating once what happens
permanently’; e.g. Th. 60b–1, 63–64, etc.).
5 Alternatively, the narrator may begin with a request to the Muses (→ Homeric
Hymns).
6 E.g. Janko 1981: 11; Miller 1986; West 1989. However, Janko and West disagree on
the status of Th. 5–21 (see below n. 9).

7 Genette 1980: 113–116 = 1972: 145–148; an example of iterative narrative is Il.
1.488–492.
8 Cf. e.g. Il. 5.746–747 and 750–751.
9 Janko 1981: 11. As for the status of Th. 2–21 specifically, West’s (1989) explanation
of steikhon (10) as a present-stem injunctive (similarly the augmented aorists in 7–8) is

more attractive than a slightly awkward transition from simultaneous iterative (2–4) to
subsequent singulative narrative (5–21) which is ‘required by the fact that they (sc. the
Muses) cannot always be singing this particular song’ (Janko 1981: 20). West 1989: 135
rightly points to the very similar song in 36–52.
r. nünlist – hesiod 27
After ‘disappearing’ in Th. 2, the narrator marks his presence again

emphatically in the ‘autobiographical’ Dichterweihe (22–34).10 There, the
Muses instruct him to sing a Theogony, which is framed by hymns to
themselves, ‘And they told me to sing of the family of blessed ones who
are for ever, and first and last always to sing of themselves’ (Th. 33–
34). The Abbruchsformel in 35 (‘But what is my business round tree or
rock?’) then leads to a fresh start in 36 (‘from the Muses let us begin’),
comparable to the one at the very beginning. After that the narrator
‘disappears’ again, and simultaneous iterative and subsequent singula-
tive narration alternate until the end of the ‘Hymn’.11 It is noteworthy
that the ‘Hymn’ does not contain an apostrophe (except for the self-
apostrophe in 36) and therefore does not show explicit signs of a narra-
tee.12
The main body of the Theogony is preceded by the narrator’s invoca-
tion of the Muses (104–115) which in principle makes use of the same
rhetorical devices and fulfils the same function as in the Homeric epics:
the external narrator addresses the Muses, asks them to give (him) a
song, and thereby gives an outline of the content and subject matter of
this song. Hesiod’s proem is, therefore, in principle similar to Homer’s.
There are, however, a few differences: most important are the more
frequent addresses of the Muses (Th. 104, 105, 108, 114, 115) and the
specific instruction to relate the story ‘from the beginning’ (Th. 115),
whereas the Homeric epics (→) do not start from the beginning.
When the story begins in Th. 116, the obvious and prominent mark-
ers of the narrator’s presence, his own ‘I’ and the ‘you’ of the Muses,
disappear from the text much as they do in Homer. The exceptions are
also comparable: two further invocations of the Muses in Hesiod mark
the transition to a new section with new subject matter.13 Apart from
10 Th. 27–28 seem to discuss the question of fiction, but scholars widely disagree on
the exact meaning of the lines, see most recently Katz and Volk 2000 and the literature
cited there.
11 Simultaneous iterative (and durative): 37–52, 60b–7, 71–72, 79–103; subsequent
singulative: 53–60a, 68–70, 73–79. (→ ‘Homeric hymns’ for further details about the
narrative style of the ‘Hymn’.)
12 A similar observation can be made with respect to a number of Homeric hymns
(→), which do not address the deity until the epilogue (E), which again is comparable
in function to Th. 104–115 (Friedländer 1914).
13 Th. 963–968 (unions of goddesses with mortal men), 1019ff. (unions of gods with
mortal women) which probably sets off the ‘Catalogue of Women’ (→ Homer for the
invocations). A majority of scholars doubt the authenticity of Th. 901–1022, but cf. West
(1966: 399): ‘The most likely explanation is … that the later poet received a complete
the three invocations, the Hesiodic narrator—like most narrators in

antiquity—presents himself as omniscient.14 This includes coverage of
aspects that strictly speaking he cannot know, e.g. an elaborate descrip-
tion of the underworld.15
In some respects, the narrator’s presence is even less discernible than
in the Homeric epics: the narrator of the Theogony never addresses
his narratee(s), except for five occurrences of the particle toi.16 Other
traces of communication between narrator and narratee(s) do not occur
frequently either:
(1) Presentation through negation (by which the narrator contradicts
the expectation of the narratee)17 does occur (e.g. Th. 488, 529,
687), but is more frequent in Homer.
(2) There is a single instance of ‘if not’-situation (again alluding to a
different direction the story could have taken):18
A thing past help would have come to pass that day, and he [sc. Typhon]
would have become king of mortals and immortals, had the father of
gods and men not taken sharp notice. (Th. 836–838)
(3) Frequent analepses and prolepses are not to be expected because
the Theogony’s structuring principle is primarily genealogical and
not chronological.
(4) Explanations for the benefit of the narratee (in particular expla-
nations and ‘etymologies’ of names) are given, but not very often:
e.g. Th. 195–200, 209–210, 234–236.19
If this seems to point towards a generally covert narrator, the impres-
sion is substantially contradicted by other characteristics of the Theogo-
Theogony … and that he remodelled the end in his own style, but following the outlines
of the original’, which, one could add, may well have contained the invocations to the
Muses.
14 A rare qualification of his omniscience is a reference to anonymous spokesmen
(‘they say’: 306), which is, however, problematic because the authenticity of the entire
passage is doubtful.
15 Th. 720–819. Contrast the Homeric narrator, who subtly sidesteps the question of
how he knows what the underworld looks like by making divine characters describe the
underworld (Il. 8.13–16 [Zeus], Od. 10.508–515 [Circe]).
16 Th. 126, 448, 873, 986, 1015. The particle toi is virtually absent from the Homeric
narrator-text (of seven Iliadic occurrences in the narrator-text, six concern a character
and only one the narratee: 10.316).
17 De Jong 1987: 61–68.
18 De Jong 1987: 68–81.
19 Examples only refer to explicit explanations. Obviously, the entire Theogony is
pervaded by a didactic intent.

ny’s narrator. These make clear that he actually ‘hides’ much less than
his notoriously covert Homeric counterpart. These features are:
(1) Considerable reduction of secondary focalization: virtually the
entire Theogony is presented from the narrator’s point of view. This
goes together with the remarkable absence of secondary narratives
from Th. 116–end.
(2) Unrestricted use of evaluative terms in the narrator-text: the Ho-
meric (→) distinction between ‘character language’ and ‘narra-
tor language’ virtually confines evaluative terms to the speeches.
The same cannot be said about Hesiod. The narrator-text of the
Theogony contains many words which belong to the Homeric ‘char-
acter language’, e.g.
They [sc. the Titans and the Olympian gods] had been fighting each
other continually now for ten full years, and the fight gave them pain at
heart (thumalgēs); and to neither side came solution of the bitter (khalepos)
strife … (Th. 635–637, cf. also 590–593 quoted below, which is part of the
notorious misogynistic judgment about women and marriage: 590–612)20
(3) Direct comments by the narrator, notably their frequency, position
and type. As with his regular use of evaluative terms, the narrator
of the Theogony does not refrain from commenting on his own
narrative. The possible functions of these comments are
(3a) to make clear the structure of the text, e.g. by summarizing
the preceding section, e.g. ‘That is the descendance of Ceto
and Phorcys’ (Th. 336, cf. 263–264, 362–363, 448–449, 613,
at the end of the Prometheus episode).
(3b) to make statements of ‘eternal truth’ (gnomes, aetiological
explanations), e.g. ‘For from her [sc. Pandora] is descended
the female sex, a great affliction to mortals as they dwell
with their husbands—no fit partners for accursed Poverty,
but only for Plenty’ (Th. 590–593, cf. 556–557 quoted below).
(3c) to evaluate the act of narrating itself, e.g. ‘It is hard for a
mortal man to tell the names of them all, but each of those
peoples knows them that live near them’ (Th. 369–370).
20 For lack of statistically reliable material (only 8 speeches with a total of 34 lines),
it cannot be ruled out with certainty that Hesiod distinguished between ‘character
language’ and ‘narrator language’, but in the case of evaluative terms this seems highly
unlikely.
(3d) to refer to the continuity of, say, cult practice down to the
narrator’s own time, e.g. ‘Even now, when an earthly man
sacrificing fine offerings makes ritual propitiation, he invokes
Hecate’ (Th. 416–418).
Not that comments like these are totally absent from the Homeric
texts,21 but Hesiod uses them in a rather un-Homeric way, e.g. in the
following passage from Prometheus’ deception of Zeus.
And he [sc. Zeus] grew angry about the lungs, and wrath reached him
to the spirit, when he saw the white oxbones set for a cunning trick.
Ever since that, the peoples on earth have burned white bones for the immortals on
aromatic altars. In great ire Zeus the cloud-gatherer said to him, ‘Son of
Iapetus, …’ (Th. 554–559)
The aetiological explanation of the Greek sacrifice in 556–557 may not
be inconceivable in Homer, but the narratorial interruption at such a
dramatic point in the scene is unparalleled in Homer.
In terms of narrative style, the Theogony is an extended catalogue
of characters (mostly gods), who may or may not become ‘heroes’
of a narrative section. In the latter case, the narrator simply states
the ‘facts’ (mostly the birth or genealogy of the particular divinity,
regularly expanded by his or her domain or particular achievement).
In the former case, one in a group of divine characters (often the last
mentioned) triggers off a narrative section: Cronus and the castration
of Uranus which leads to new births (Th. 137–206), Zeus overthrows
Cronus (457–505), Prometheus tricks Zeus (521–589), the Hundred-
Handers support Zeus in the Titanomachy (624–720), Typhon fights
against Zeus (836–868).22
Unlike the narrator, the narratee is completely covert. It is, therefore,
impossible to deduce a directly applicable model of how the Theogony’s
narrator envisages an ideal reception of his poem. But it is a fair
assumption that the general description of the singer (aoidos) in the
‘Hymn to the Muses’ is an implicit self-portrait which includes the
desired effect of his poetry:
Though a man’s heart be withered with the grief of recent bereavement,
if then a singer, the servant of the Muses, sings of the famous deeds of
21 Compare e.g. Th. 369–370 with Il. 12.176.

22 The material does not lend itself to a defensible description of how the narrator
introduces and ends these narrative sections. A major stumbling block is the often
disputed authenticity of (groups of) lines in the transmitted text.
men of old, and of the blessed gods who dwell in Olympus, he soon
forgets his sorrows and thinks no more of his family troubles, quickly
diverted by the goddesses’ gifts. (Th. 98–103)23
Works and Days
As with the Theogony, it is best to separate the introductory hymn to

Hesiod’s second poem for the purposes of a narratological analysis,
although the hymn covers ten lines only. Its narrative form and func-
tion are comparable to that of the Theogony (see above). Important dif-
ferences (apart from the length) are that in Works and Days the narrator
starts with an invocation of the Muses24 and that this hymn has two
addressees, first the Muses (1–2) and then Zeus (9–10).
In many respects the Works and Days is rather different from the
Theogony. From a narratological point of view, the most remarkable
difference concerns the status of the narrator, who in the Works and
Days is an internal narrator. That is to say, the narrator relates ‘in
his own name’ events in which he himself takes part. The numerous
signs of the narrator’s ‘I’ and the narratee’s ‘you’, which are spread
over the entire poem, sometimes refer to the characters, in particular
the notorious dealings of the two brothers (35–37, 396–397). More
often, however, these signs refer to the ‘I’ as narrator and to the ‘you’
as narratee.25 This led to the impression that the poem resembles a
speech and therefore ‘ist eine einzige verselbständigte und zum Epos
erweiterte “Rede” ermahnenden Charakters’.26 Although Jaeger does
not say so explicitly, this impression is due not only to the ubiquitous
signs of the ‘I’ and the ‘you’, but also to the preponderance of the
23 Similarly, the birth of the Muses is described as ‘oblivion of ills and respite from
cares’ (Th. 55).

24 ‘Muses from Pieria, who glorify by songs, come to me, tell of Zeus your father …’
(W&D 1–2) instead of ‘From the Muses of Helicon let us begin our singing’ (Th. 1), cf.
‘Homeric hymns’ (→).
25 Signs of the narrator are pronouns and verb forms in the first person (W&D 10,
35, 37, 106, 174, etc.), signs of the narratee apostrophes (27, 213, 274, 286, 299, 397, 611,
633, 641; 248, 263), pronouns and verb forms in the second person (33, 34, 35, 37, 38,
43, 44, 45, etc.). Perhaps the most important signs are the numerous imperatives (and
infinitives used as imperatives). Through them the presence is felt both of the narrator
and of the narratee because they can be paraphrased as: ‘I tell you to do x and not to
do y.’
26 Jaeger [1933] 1954: 101 with n. 1.
present tense (= simultaneous narration),27 whereas the narrator-text

of most narrative texts is dominated by the past tense (= subsequent
narration). In short, the primary story of the Works and Days is one
of the relatively few examples in Western literature of a simultaneous
narration with a dominant internal narrator and a well-represented
internal narratee.
The resemblance to speech is not so surprising if one takes into
account parallel texts from the Near East,28 which to a large extent
consist of speech (sometimes in dialogue). An important difference from
Hesiod is, however, that most of these parallel texts contain a framing
narrative, in which a (usually external) narrator introduces the speaker
and his addressee.29 However, in Works and Days this framing narrative
is lacking,30 but the last line of the proem (‘I should like to tell Perses
words of truth’: W&D 10) seems to indirectly fulfill this introductory
function, as if the text read ‘I (will) tell Perses the (following) words of
truth: “I see there is not only one strife …”’31 In other words, Works and
Days comes close to what Gérard Genette calls ‘reduced metadiegetic’
or ‘pseudo-diegetic’, i.e. the elimination of the framing narrative level.32
In view of this narrative trick, one should speak of the ‘primary’ and
‘secondary’ stories of the Works and Days in inverted commas only. In
terms of the narrative situation, it is legitimate to compare the Works and
Days to elegies by Tyrtaeus or Solon, which also ‘suppress’ the framing
27 E.g. ‘for Hunger goes always with a work-shy man. Gods and men disapprove of
that man who lives without working. […] It is from work that men are rich in flocks
and wealthy, and a working man is much dearer to the immortals. Work is no reproach,
but not working is a reproach’ (W&D 302–311).
28 West 1978: 3ff.
29 Cf. e.g. the Sumerian Instructions of Suruppak, part of which is quoted in translation
by West (1978: 4): ‘The intelligent one, who knew the (proper) words, and was living
in Sumer, […] Suruppak gave instructions to his son: “My son, let me give you
instructions, may you pay attention to them!”’ Here and in many other cases, the
addressee is the speaker’s son.
30 Schmidt 1986: 18.
31 The speech-like quality of W&D 11ff. is underlined by the particle ara in 11, which
is operational at the level of interaction between speaker and addressee (Duhoux 1997;
Wakker 1997: 212–213). On W&D 11 see also Most 1993: 77–80 and Scodel 1996: 72–79.
32 ‘These forms of narrating where the metadiegetic way station, mentioned or not,
is immediately ousted in favor of the first narrator, which to some extent economizes on
one (or sometimes several) narrative level(s)—these forms we will call reduced metadiegetic
(implying: reduced to the diegetic), or pseudo-diegetic’ (Genette 1980: 236–237 = 1972:
247). The archetype of this form is found in Plato’s (→) Theaetetus: Eucleides represents
Socrates speaking directly to Theodorus and Theaetetus, thereby omitting the interca-
lated speech formulae like ‘I said’ or ‘he said’.
narrative and immediately start with the ‘speech’. However, subject

matter and the frequent apostrophes of the narratee Perses point more
in the direction of Theognis and his narratee Cyrnus.33
As for the narratee of the Works and Days, it should be emphasized
that Perses is the dominant but not the only explicit narratee. W&D
248 and 262–263 (and implicitly also 202) are addressed to the ‘kings’
(basilēs). And although the explicit narratee mostly is his brother Perses,
the narrator presupposes a wider audience, especially in the passages
that contain instructions of a general type.34
The subject matter and rhetoric of the Works and Days are didac-
tic and instructive in nature. The narrator, therefore, is almost liter-
ally omniscient. In order to back his argumentation, he incorporates a
number of ‘secondary’ narratives (the myth of Pandora, the five gen-
erations, the fable of the hawk and the nightingale), which are set off
explicitly from the ‘primary’ narrative. E.g. ‘If you will, I will summa-
rize another tale for you, well and skilfully—mind you take it in’ (W&D
106–107).
The passage caps the preceding myth of Pandora and at the same
time introduces that of the five generations. Both ‘secondary’ narratives
are called a logos. Similarly, the fable of the hawk and the nightingale is
introduced as an ainos (202). Given the simultaneous nature of the ‘pri-
mary’ narrative, these ‘secondary’ narratives by necessity are instances
of external analepsis. The only exception is the fifth and last genera-
tion to which the narrator himself belongs. Its dreadful fate is described
in a long pessimistic prolepsis (‘they will not do x, they will not do y’:
176–201). The generally analeptic character of the ‘secondary’ narra-
tives has as a consequence that—unlike the examples given above—
they can also be introduced by a verb in the past tense only, that is,
by means of a temporal shift, e.g. the myth of Pandora, ‘But Zeus con-
cealed it [sc. men’s food], angry because Prometheus’ crooked cunning
had tricked him’ (W&D 47–48). The ubiquitous ‘I’ of the narrator (see
33 Tyrtaeus and Solon are compared e.g. by Jaeger ([1933] 1954: 100–101), Theognis
e.g. by West (1978: 23). Phocylides is a different case because the recurring half-
line ‘And the following is by Phocylides too’ (fr. 1.1 etc. Gentili-Prato) in function
resembles the introductory narrator-text. Conversely, the Precepts of Chiron appear to
start immediately with Chiron’s ‘speech’ (‘Hes.’ fr. 283 M-W). Further parallels to
W&D may be found in Parmenides and Empedocles and their narratees (anonymous
and Pausanias, respectively). In both cases, the fragmentary status precludes a decision
about the presence or absence of a framing narrative.
34 Schmidt 1986.
above) inevitably eclipses the other indications of his presence (evalua-

tive language etc.), which are treated in a similar way as in the Theogony
(see there).
To summarize: after a short introductory hymn, the Works and Days
consists of a pseudo-diegetic and simultaneous narrative, in which a
highly discernible internal narrator is almost constantly at the elbow of
his narratee Perses and, through him, of his narratees in general.
chapter three
THE HOMERIC HYMNS
R. Nünlist
The so-called Homeric hymns were written by different poets and

their dates of composition seem to vary greatly.1 It is, nevertheless,
justifiable to treat them as a group here because of their structural
similarities, especially in terms of narrative technique. For virtually all
the Homeric hymns have the following narrative structure: (A) The
narrator introduces (B) his subject matter (usually the god to whom
the hymn is dedicated), (C) followed by a relative pronoun, which (D)
sets off the primary story. At the very end, (E) the narrator addresses
the god in question in an epilogue. In fact, the invocation (E) of the
god provides formal closure to the preceding primary narrative and in
function mirrors the relative pronoun (C) at the beginning.2
The introduction (A) by the external narrator (which in form and
function resembles the proems of Homer or Hesiod) can take two
forms:3
(A’) The narrator self-consciously announces (B) his subject matter in
the first person, e.g.
‘Of Demeter the lovely-haired, the august goddess first I sing …’
(h.Dem. 1, transl. West)4
(A”) The narrator invokes the Muse(s) to sing a song about (B) the
subject matter, e.g.
1 About the authors of the Homeric hymns one knows next to nothing (cf. however
West 1975 for an attempt to identify the author of h.Ap.). Their dates of composition are
vague and/or controversial. The most promising approach seems to be a combination
of linguistic and historical arguments (Janko 1982, with a table on p. 200).
2 An address to the narratee is a typical means with which to end a narrative.
3 Unlike the narrator of Hesiod’s ‘Hymn to the Muses’, who once turns into an
internal narrator (Th. 22–34; → Hesiod), the narrators of the Homeric hymns without
exception are external: they are not themselves a character in their stories.
4 This type is found in h.Dem., h.Ap., h. 6, 7, 10, 11, 12, 13, 15, 16, 18, 22, 23, 25, 26,
27, 28, 30; cf. Hes. Th. 1, Ap.Rh. 1.1–2.

36 part one – chapter three
‘Sing of Hermes, Muse, the son of Zeus and Maia …’ (h.Herm. 1)5
The ‘I’ of the narrator and the ‘you’ of the narratee are absent from
parts (B) through (D) of the hymns, which for that reason contain little
more information about the narrator than that his story is located
in a distant past which precedes him by an unspecified number of
years. In other words, the bulk of the hymns consists of subsequent
narration.6 Only the short introduction (A) and the similarly short
epilogue (E) show traces of ‘I’ and ‘you’, the most obvious signs of
the presence of the narrator and the narratee. This observation holds
for the majority of the hymns, the most important exception being
the Hymn to Apollo with its remarkable mixture of ‘second person’ and
‘third person’ narrations.7 Two explanations of the general rule are
conceivable, which are not mutually exclusive: (i) The Homeric hymns’
general similarity to the Homeric epics, which includes diction etc., also
extends to their primary narrators, who are similarly covert and equally
reluctant to address the narratee directly.8 (ii) As the epilogue (E) in
each case shows, the narratee of the hymn is the god to whom the
hymn is dedicated, but who is also the ‘hero’ of (D) the primary story.
(Needless to say, the ultimate narratees of the hymns are, of course, the
human audience, who, however, cannot be addressed, lest the phthonos
theōn [envy of the gods] be roused.) More frequent references to the
god-narratee would, therefore, make the narrator switch back and
forth between ‘he/she’ and ‘you’. This, in fact, is what happens in the
exceptional Hymn to Apollo (and probably also in the fragmentary Hymn
to Dionysus 1). But its effect is rather odd, which may be the reason why
the other poets avoid it altogether.
5 The second type is found in h.Herm., h.Aphr., h. 9, 14, 17, 19, 20, 31, 32, 33; cf.
Hom. Il. 1.1, Od. 1.1, Hes. W&D. 1. A third type begins with an invocation of the god
to whom the hymn is dedicated: h. 8 (which is exceptional in several other respects and
probably does not belong to the collection: e.g. Richardson 1974: 3), 21, 24 and 29.
6 For temporal markers (in addition to the ubiquitous past tense) see h.Dem. 10, 97,
451, h.Ap. 101, h.Herm. 73, 233, 513, h.Aphr. 54 (in all eight cases the adverb tote ‘then’).
Conversely, markers which point to the narrator’s own time are rare: h.Herm. 125–126,
508. Cf. the notorious hoioi nun brotoi passages in Homer (→).
7 On the Hymn to Apollo, which is exceptional in several respects, see below; the
other exceptions are h. 8, 21, 22, 24, 29 and 30 (cf. the exceptions in n. 5).
8 This reluctance to address the narratee extends to an avoidance of the particle
toi in the narrator-text (for Homer and Hesiod see chapter on Hesiod). Apart from the
exceptional Hymn to Apollo, there are only three instances of toi in the narrator-text,
all from the Hymn to Hermes (25, 111, 138); cf. Denniston 1954: 537, although his list is
incomplete.
r. nünlist – the homeric hymns 37
As it is, little can be gathered from the texts about the narrators
themselves. Both introduction (A) and epilogue (E) are short and rather
standardized. The epilogue, which does not, of course, occur in either
Iliad or Odyssey, often contains a wish to the deity to grant a special
favour, none of which is particularly informative, except perhaps for
‘food in exchange for the song’9 and for ‘victory in the present contest’
(h. 6.19–20). The important exception here is again the Hymn to Apollo
(see below).
The Homeric narrator (→) avoids using evaluative terms in the
narrator-text to such an extent that one can differentiate between a
‘character language’ and a ‘narrator language’. The Homeric hymns
are too short to provide statistically reliable results in such a compari-
son. It can, however, be said that explicit comments or evaluations by
the narrator are not much more frequent than in Homer and hardly
ever obtrusive. The most extensive (and for that reason not repre-
sentative) example is the instruction not to divulge the secrets of the
Eleusinian rites.10
… the solemn mysteries which one cannot depart from or enquire about
or broadcast, for great awe restrains us from speaking. Blessed is he of
men on earth who has beheld them, whereas he that is uninitiated in the
rites, or he that has had no part in them, never enjoys a similar lot down
in the musty dark when he is dead. (h.Dem. 478–482)
Except for direct comments, the narrator’s controlling function can
also be gathered from (1) presentation through negation, (2) ‘if not’-
situations, (3) ana- and prolepses, all of which do not occur frequently
either. As for (1) presentation through negation, the most remarkable
instance comes from the Hymn to Aphrodite, where the narrator first
mentions three goddesses (Athena, Artemis, Hestia) whom Aphrodite
could not seduce (7–32). This priamel, which is summarized in 33–35,
functions as a foil for Zeus, who was and is a victim of her machi-
nations (36–40). However, the thoroughly developed expectation of
the narratee is eventually disappointed because the narrator does not
expand on Zeus’ love affairs as expected, but instead on his revenge on
Aphrodite by making her fall in love with a mortal (45–291). As for (2)
9 h.Dem. 494; the wish is, of course, particularly apt after the famine described in
the hymn.
10 For other narratorial comments cf. h.Dem. 111, 243, 246, 291, 451, 486–489, h.Ap.
227–228, 237, h.Herm. 76, 125–126, 316, 396, 576–578, h.Aphr. 26, 167.
‘if not’-situations, there are only two in the entire corpus, both in the
Hymn to Demeter and both marking an important crisis in the plot:
Indeed she [sc. Demeter in the guise of an old woman] would have
made him [sc. Demophon] ageless and deathless, if in her folly fair-girt
Metaneira [sc. Demophon’s mother] had not waited for the nighttime
and spied from her fragrant chamber. (h.Dem. 242–245)
Indeed, she [sc. Demeter] would have destroyed humankind altogether
by grievous famine, and deprived the Olympians of their honorific privi-
leges and their sacrifices, had Zeus not taken notice, and counselled with
his heart. (h.Dem. 310–313)
Even the longest among the Homeric hymns are short (max. 580
lines) compared to other narrative texts, and their primary narratives
are straightforward in structure and narrated in roughly chronological
order. Frequent ana- and prolepses (3) would therefore be more difficult
to explain than their rarity. The straightforwardness of the plots is
also indicated by the fact that they do not contain more than two
principal elements. The Hymn to Demeter contains two: the main element
(Persephone abducted, Demeter in search of her, famine, Demeter
reconciled) frames a second element (Demeter in Eleusis, foundation of
the Eleusinian rites). The Hymn to Hermes also contains two elements (cf.
the programmatic summary in 17–18): the main element (Hermes steals
Apollo’s cattle) is framed by the story of the lyre (its invention, means
of consoling Apollo). The Hymn to Aphrodite contains one element only
(Zeus makes Aphrodite fall in love with a mortal). The Hymn to Apollo
is again exceptional. Although two principal elements can be discerned
(birth; foundation of Delphic oracle), the narrator manages to fit in en
passant a number of other things, notably the Telphousa episode (244–
276) and the long analepsis about Typhon (305–355).
Like their Homeric predecessor, the narrators of the Homeric hymns
regularly leave the stage to their characters by quoting their speeches.
It is noteworthy, however, that indirect speech is more frequent than
in Homer.11 A greater prominence of reported speech instead of direct
speech inevitably leads to greater salience of the narrator’s controlling
function.12 The most remarkable passage is probably Demeter’s threat-
ening declaration:
11 Cf. h.Dem. 171–173, 207–209, 297–298, 331–333, 443–447, h.Herm. 57–61, 391–394,
427–433, 521–523, 525–526, h.Aphr. 28.

12 Genette 1980: 171–172.
She said she would never set foot on fragrant Olympus, or allow the
earth’s fruit to come up, until she set eyes on her fair-faced daughter [sc.
Persephone]. (h.Dem. 331–333)
Given its importance, the Homeric narrator would, no doubt, have

quoted her speech verbatim.
Despite the frequency of direct speech, secondary narrators do not
feature prominently in the Homeric hymns. When characters do tell
‘stories’, they usually cover an aspect of the primary narrative itself,
e.g. repeating internal analepsis (h.Dem. 407–432, h.Herm. 340–364, 370–
374). The most remarkable counter-example is the story which Aphro-
dite tells Anchises about two other human lovers of a divinity. Gany-
medes was abducted by Zeus, who compensated the mourning father
with swift horses (h.Aphr. 202–217). And Eos, who fell in love with
Tithonus, asked Zeus for eternal life but forgot to ask for eternal youth
too (218–238). Both exempla have an obvious persuasive function and are
meant to soothe Anchises’ anxiety.13
The narrators of the hymns have confidence in the cooperation of
their narratees. For they show a tendency to leave rather substantial
‘gaps’ (Leerstellen) in the narrative, which the narratee is to fill in for
himself or herself. An instructive, because ‘un-Homeric’, example is
a passage from the Hymn to Hermes: Apollo’s actual discovery that his
cattle have been abducted from Pieria, a corner-stone of the story, is left
out of the narrative (between 183 and 184), but can be ‘reconstructed’
from Apollo’s speech to the old man in Onchestus (190–200).
All in all, one can conclude that, despite a number of minor varia-
tions, most of the Homeric hymns do not depart from Homer’s narra-
tive technique in a substantial way. Their narratives form a steady flow,
one event ‘automatically’ motivates the next, and the narrator’s dis-
cretion leads to that well-known impression that ‘the story appears to
tell itself ’.14 The title ‘Homeric hymns’, though misleading in terms of
authorship, is not inappropriate in terms of narrative technique—with
one important exception.
13 One could, however, argue that the argumentation of the second exemplum is to
some extent undercut, because Tithonus’ fate is after all not very appealing. This may,
nevertheless, be deliberate on the part of Aphrodite (‘argument function’). For she later
ends her speech with an unmistakable warning not to disclose their lover’s hour (286–
288).
14 The notorious phrase by Lubbock ([1921] 1926: 113), which describes a narrator
who ‘shows’ (unlike the narrator who ‘tells’).

Hymn to Apollo
It has been stated more than once in the previous part of this chap-
ter that the Hymn to Apollo differs remarkably from the other hymns.15
Its most striking difference is the departure from the structural scheme
as explained above. References to the narrator’s ‘I’ and above all to
the addressee’s ‘you’, which are restricted to sections (A) and (E) in
the other hymns, frequently recur in the Hymn to Apollo.16 Both narra-
tor and narratee(s) figure prominently, and this description could give
the impression of an ongoing dialogue between the narrator and his
narratee, similar to Hesiod’s (→) Works and Days. This, however, is not
exactly the case because the narrator does not use ‘second-person nar-
ration’ throughout the hymn (as the narrator does in hh. 24 and 29,
and similarly in hh. 22 and 30). Apollo is both the main narratee and
the main ‘hero’ of the hymn, in which guise he is regularly referred
to in the third person. As a consequence, the hymn contains a curious
mixture of ‘you’s’ and ‘he’s’ which have the same referent—sometimes
with rather awkward transitions from ‘you’ to ‘he’ (e.g. 129/130: ‘The
fastenings no longer held you back, but all this came undone. At once
Phoebus Apollo spoke among the goddesses’) and vice versa.17 The refer-
ences to the addressee’s ‘you’ stop after 282 (until the epilogue in 545–
546). That is to say, from a narratological point of view the bipartition
is less between a Delian and a Pythian part of the hymn than between
ll. 1–285a and 285b–546. This same bipartition is also suggested by the
relative distribution of simultaneous iterative narration and subsequent
singulative narration.18 Whereas the first part is dominated by simulta-
neous iterative/durative narration (2–13, 20–24, 30–44, 140–164, 182–
15 The unity of the text transmitted as Hymn to Apollo in the manuscripts is a
notorious zetema. A majority of scholars argue that two hymns, one to Delian Apollo
(1–178), another to Pythian Apollo (179–546), have been connected in a rather clumsy
way (e.g. West 1975, Janko 1982), but there are defenders of the unity (e.g. Miller 1986,
Clay 1989).
16 References to the narrator’s ‘I’: ll. 1, 19, 166, 171, 177, 207, 208, 546. References
to the addressee’s ‘you’: ll. 14 (the addressee being Leto), 19–22, 25, 29, 120, 127–129,
140–149, 166 (addressees are the Delian girls who perform a song), 167 (idem), 171 (idem),
179–181, 207–209, 215–225, 229–230, 239–246, 277–282, 545–546.
17 It is noteworthy that this mixed style is used both in the Delian and in the Pythian
part of the hymn. In other words, if the hymn originally consists of two separate hymns,
the ‘imitator’ also took over this unusual style.
18 For the distinction between simultaneous iterative (and durative) narration and
subsequent singulative narration see chapter on Hesiod.

206, 231–236), this type of narrative is virtually absent from the second
part of the hymn, where the more standard subsequent singulative nar-
ration dominates as in the other long hymns.19
In addition to the bare references to the ‘I’ of the narrator, his
function as the one ‘who is in control’ is further emphasized by the
two questions, each followed by a list of possible topics, which stress the
wealth of material from which the narrator must and does choose.
How shall I hymn you, fit subject as you are in every respect? Shall I sing
of you as a wooer and lover, of how you went to court the Azantid maid
(sc. Coronis)? or …? or …? or …? (h.Ap. 207–215, cf. 19–29)20
The narrator expressly mentions a number of different directions where

his story could go. As for the controlling function of the narrator, this
resembles ‘presentation through negation’ and ‘if not-situations’ (see
above), but it is much more discernible in the present case. At the same
time, these questions recall the opening question of the Iliad (1.8) and at
other transitional points (e.g. Il. 5.703–704).
Probably the most important difference of the Hymn to Apollo con-
cerns the person of the narrator himself. Whereas the narrators of the
other hymns are covert and say virtually nothing about themselves,
the Hymn to Apollo contains a remarkable self-advertisement that made
Thucydides (3.104) quote the passage.
But now, may Apollo be favourable, together with Artemis, and hail, all
you Maidens! Think of me in future, if ever some long-suffering stranger
comes here and asks, ‘O maidens, which is your favourite singer who
visits here, and who do you enjoy most?’ Then you must all answer with
one voice(?). ‘It is a blind man, and he lives in rocky Chios; all of his
songs remain supreme afterwards.’ And we will carry your reputation
wherever we go as we roam the well-ordered cities of men, and they will
believe it, because it is true. And myself, I shall not cease from hymning
the far-shooter Apollo of the silver bow, whom lovely-haired Leto bore.
(h.Ap. 165–176)
19 Short passages like h.Ap. 393–396 are only apparent exceptions and in accordance
with Homer’s narrative technique. The Hymn to Aphrodite begins with simultaneous
iterative narration (ll. 2ff., cf. above on ‘presentation through negation’) before it leads
into the (more common) subsequent singulative narration of the main narrative (ll. 45–
291).
20 Miller 1986: 21 explains the two passages as aporetic questions, elaborated by a
summary priamel. Both passages combine the rhetorical functions of Hindernis- and
Leichtigkeitsmotiv (Nünlist 1998: 33, 136). In accordance with standard priamel technique
and with the ‘continuity of thought’ principle, it is the last item on the list that is
eventually chosen.
This indirect form of self-praise is remarkably concrete and without

parallel in any of the other Homeric hymns.21
To conclude: one of the leading experts in the field has recently for-
mulated the following hypothesis regarding the relation of the Homeric
hymns to the Homeric epics: ‘the familiarity of the material (i.e. the
myths: RN) and its smaller scale may have invited experimentation and
innovation in both diction and narrative technique’.22 In terms of nar-
rative technique, the present chapter seems to indicate that ‘experimen-
tation and innovation’ do not completely alter the Homeric model but
are of a comparatively small scale. The sole remarkable exception is the
Hymn to Apollo, but it appears to be the exception that proves the rule.
21 The closest parallels in early Greek poetry are Hesiod (→) and Theognis’ sphragis
(19–30), but unlike them the present poet does not actually name himself. However,
the ‘blind man of Chios’ is of course a thinly veiled periphrasis for ‘Homer’. In this
connection it is perhaps worth mentioning that Thucydides, in contrast to modern
scholars, does not seem to doubt the authenticity of this self-identification.
22 Clay 1997: 492.
chapter four
APOLLONIUS OF RHODES
M.P. Cuypers
By virtue of being an epic, Apollonius’ Argonautica invites direct com-

parison with Homer. The poem’s narrator relates a story about the
remote past at length in hexameter verse; he is external, omniscient,
omnipresent, and anonymous, and he uses virtually every Homeric
narrative technique. However, whereas Homer operates largely in the
background, Apollonius directs his narrative in an altogether overt and
self-conscious manner. In his constant interaction with his narratees he
not only uses Homeric ‘interactive’ devices with un-Homeric frequency
and in novel ways, but he also covers topics and adopts narrative strate-
gies which are at home in other genres. This results in a Protean nar-
rative persona, an amalgam of (at least) the Homeric singer of epic, the
hymnic and Pindaric singers of praise, the Herodotean historian, and
the Callimachean scholar—these last two already complex personali-
ties themselves, who, just as the narrator of the Argonautica, are tugged
between the roles of epic storyteller and historian.
Narrative goals, main narrative, and digressions
Whereas the Homeric epics start with Muse-invocations, the opening

of the Argonautica1 rather recalls the Homeric hymns (→)—in particular
the Hymn to Apollo, which stands out among the hymns for its overt
narrator. In his first clause, ‘starting from you, Phoebus, I shall recall
the famous deeds of men of long ago’,2 Apollonius draws attention to
his role as a narrator by opening with the word ‘starting’ (arkhomenos),
by addressing the god after whom he is named (Phoebus Apollo),
and by using a first-person verb form (mnēsomai). At the same time
1 E.g. Albis 1996: 17–26; Clare 2002: 20–32; Clauss 1993: 14–25; DeForest 1994: 37–
46; Goldhill 1991: 286–294; González 2000; Hunter 1993a: 119–129; Wheeler 2002.
2 My translations of the Argonautica are based on Hunter 1993b, with adaptations.
44 part one – chapter four
he evokes the suggestion that the Argonautica should at least on one

level be read as a hymn to Apollo, the god who is the cause of the
poem’s action (by giving the oracles which instigated the Argonauts’
quest for the golden fleece), the cause of its narration (as the god of
poetry and eponymous god of this poet), and the divine model of
its main hero (Jason is associated with Apollo throughout the poem).
Since the Homeric hymns were traditionally sung as proems to epic
recitation, Apollonius’ hymnic opening also conjures the fiction of an
epic singer performing before a live audience.3 Finally, by assuming the
role of a hymnic narrator, Apollonius underscores that the Argonautica’s
narrative goal is praising—praising both the gods and the ‘famous
deeds of men of long ago’ (palaigeneōn klea phōtōn). This double goal—
simultaneous celebration of human exploits and the gods—is shared
with the victory odes of Pindar (→), whom we may regard as another
model for the Argonautica’s narrator. This is suggested in the proem
by lines 5–17, which provide the background to the Argonautic quest.
Besides owing their content to Pindar’s fourth Pythian ode, these lines
also reflect Pindar’s allusive narrative style of ‘reminding’ his narratees
of a story which they already know (i.e. which is already famous) instead
of telling it in full.
The narrator engages in open dialogue with his predecessors in the
last section of the proem (18–22), which explicitly demarcates the main
story and also contains a late address to the Muses:
As for the ship, this is still celebrated in the songs of earlier singers, who
tell that it was built by Argos on the instructions of Athena. Now I shall
narrate the lineage and names of the heroes, their voyages over the vast
sea and all the deeds they accomplished on their wanderings. May the
Muses be hypophētores [discussed below] of my song.
In the c. 6,000 lines that follow, the narrator faithfully follows this
programme. He first provides a catalogue of the Argonauts as they
gather in Iolcus. Then, after a series of departure scenes, the heroes sail
out on their ‘voyage over the vast sea’ and we hear ‘all the deeds they
accomplished on their wanderings’, as the narrator transports them in
a linear and episodic fashion from Iolcus to Colchis and back again, in
a narrative that covers a full sailing season (from spring till fall), singling
out some parts of the story for scenic development4 and summarizing
3 This fiction is supported by a verbal allusion to a performance of the epic singer
= Odyssey 8.73 klea andrōn).

Demodocus in Homer (klea phōtōn ~
4 Notably the Argonauts’ adventures in Colchis (3.1–4.240). Though clearly devel-
m.p. cuypers – apollonius of rhodes 45
others. This setup implies a departure from the organization of the

Homeric epics (centred on Achilles’ wrath and Odysseus’ homecoming
respectively) in favour of the ‘historical’ organization of the so-called
Cyclic epics (criticized in Aristotle’s Poetics) and, indeed, historiography.5
In 4.1773–1781 the narrative closes with an apostrophe of the heroes:
Be propitious, heroes, blessed race, and may these songs be from year
to year sweeter to sing for humans. For now I have reached the famous
conclusion of your struggles, since no other labour confronted you as
you sailed up from Aegina (…); and on the shores of Pagasae you gladly
debarked.
This epilogue takes the narrative programme to an explicit and abrupt

end.6 The narrator had set out to recall the Argonauts’ deeds during
their journey. Although this journey did not end at Aegina, there are no
more deeds to be celebrated after this point, and therefore the narrative
ends. Another narrative goal has also been reached. In the address ‘be
propitious’ (hilate) and the expressed hope that ‘these songs be from
year to year sweeter to sing for men’, the epilogue resumes the genre
of the proem: hymn. However, what started out as a hymn to Apollo
now ends as a hymn to the Argonauts. For the expression translated
as ‘blessed race’ (makarōn genos) can mean ‘descendants of gods’ (almost
all Argonauts had a divine ancestor), but it may also be taken as ‘race
of gods’.7 The Argonauts, it is suggested, have become immortal; not
because, as their one-time companion and all-time exemplum, Heracles,
they have gained a seat on Olympus, but because Apollonius has
immortalized them with his epic, which he prays will be ‘sung’ forever.
Finally, just as the first word of the Argonautica (‘starting’) drew attention
to the act of narrating, so does the very last. For although eisapebēte
must in this context be translated as ‘(you) debarked’, it also evokes
the frequent use of the verb apobainō as ‘to end’. Thus the end of the
Argonauts’ journey coincides with the end of the path of song.8
oped with a view to their thematic potential, the longer episodes stand out as semi-
independent stories (→ Herodotus).
5 Hunter 2001.
6 Contrast the unmarked endings of Homer (→), which, however, show implicit
signs of closure. For similar signs in Arg. 4 see Theodorakopoulos 1998; on the epilogue
further Albis 1996: 39–42, 118–120; Clare 2002: 159–162, 283–285; Goldhill 1991: 294–
300; Hunter 1993a: 119–129; Wray 2000: 240–247.
7 As suggested by ‘for humans’, added in enjambement.
8 On the Argonauts’ journey as a metaphor for the path of song see Albis 1996;
Clare 2002; Wray 2000.

The story of the Argonauts’ exploits constitutes only part of the Arg-
onautica’s narrative. A large percentage of the poem is taken up with
descriptions and digressions whose argument function, if it is not the
narrator’s and narratees’ love of information for its own sake, is to visu-
alize and/or authenticate the events of the main story, but which also
help to create the complex web of thematic connections which unifies
the poem. These descriptions and digressions stand out by their weight
in comparison to the main story, notably in books 1, 2, and 4, where
they take up as much as half of the narrator-text and exacerbate the
fragmentation of the narrative which naturally arises from the episodic
plot. Some have Homeric ancestry, such as the Catalogue of Argonauts
(1.23–233), the ekphrasis of Jason’s cloak (1.721–767), the many extended
(‘epic’) similes, and external prolepses. However, the Argonautica also
contains numerous geographical and ethnographical excursions, such
as that on the customs of the Mossynoeci in 2.1015–1029 (as the Arg-
onauts sail past them without landing). Passages of this type are rela-
tively rare in Homer but very frequent in Herodotus (→). Entirely un-
Homeric are the poem’s numerous aetiological asides, which explain
the ‘origins’ (aitia) of phenomena still extant at the time of narration,
such as the Etesian winds which delay the Argonauts in the Bosporus
(2.498–528). These to some extent resemble the historical digressions
of Herodotus, but they first and foremost recall Callimachus’ Aetia (→),
which is entirely organized around this theme.
The narrator
Following Pseudo-Longinus’ qualification of Herodotus as ‘most Home-

ric’ of all historians (Subl. 13.3; → Herodotus), we may perhaps qual-
ify Apollonius as ‘most Herodotean of all epicists’. It is tempting to
describe the Argonautica’s narrator as ‘Homer gone Herodotean’ or
‘Herodotus trying his hand at epic’. Nearly all his narrative modes and
devices have their roots in either Homer or Herodotus. However, their
implementation and combinations are unique to the Argonautica. This
applies first and foremost to Apollonius’ negotiation of the seemingly
incompatible rhetorical strategies of the epic storyteller, who knows and
states, inspired by the Muses, and the historian, who argues from evi-
dence. The narrator of the Argonautica is both and does both (sometimes
invoking Muses and historiē in one breath). In the following survey I will
focus on his ‘Homeric’ interaction with the Muses, other divinities, and
his characters (1), and on his ‘Herodotean’ manifestations as organizer

(2), ‘researcher’ (3), and commentator (4) of the narrative.
(1) The narrator’s interaction with the Muses is more elaborate and
complex than in Homer. We have seen how at the end of his proem he
asks the Muses to ‘be hypophētores of his song’. In this debated phrase9
‘hypophēt(or)’ seems to be the opposite of ‘prophet’. This disquietingly
suggests that the divine Muses provide insight in the past in the way
that divinely inspired prophets provide insight in the future: uncertainty
remains. Prophecies usually require intellectual activity from mortals,
which results in an interpretation that may or may not be correct. This
analogy invites taking the narrator’s discourse with the Muses as a trope
for Apollonius’ critical dialogue with his sources. But other passages
resist such a reading. In his story of the origin of the name Drepane
(‘Sickle’), the narrator piously apologizes to the Muses for telling a
discrediting story about the gods (4.982–986):
At the head of the Ionian strait, set in the Ceraunian sea, there is a large
and fertile island, under which, as you know [dē], they say [phatis] lies the
sickle—your gracious pardon, Muses! it is against my will that I relate a
story told by men of earlier generations—the sickle with which Cronus
pitilessly cut off his father’s genitals.
Here the Muses are imagined as divine overseers, who are listening in,
and might step in, on the narrator’s communication with his narratees.
The argument that the story must be mentioned ‘because it is out there’
recalls Herodotus, as does the fact that it is followed by an alternative
explanation.10 The rhetoric of this passage is especially remarkable
because the narrator elsewhere seems to use the Muses largely to excuse
parts of the narrative which might seem improper or incredible. In
2.844–845 the words ‘and if, under the Muses’ influence, I must also tell
this without constraints’ introduce a discrepancy between the story and
present-day cult which might lead to disbelief. In 4.1381 a reference to
the Muses opens a tale in which the Argonauts display a strength and
perseverance which sits ill with their overall characterization (4.1381–
1390):11
9 E.g. Albis 1996: 20–21; Clauss 1993: 17–18; Fusillo 1985: 365–366; González 2000;
Hunter 1993a: 125, all with further discussion of Apollonius’ Muses.

10 Ironically, this alternative is a very pious one, which could easily have been
presented as the authoritative aition. For further discussion see e.g. Clare 2002: 266–
267.
11 The structuring Muse-invocations that open books 3 and 4 fit the pattern (3.1–4:
This tale is the Muses’, and I sing obedient (hupakouos) to the maidens
of Pieria. This report too I heard loud and clear that you—indeed [dē]
much the greatest sons of kings!—by your own strength, by your own
excellence placed your ship and all that your ship contained aloft upon
your shoulders, and carried it for twelve days and an equal number of
nights through the sandy deserts of Libya. Who would be able to tell of
the suffering and wretchedness which was the fate of those men as they
laboured? For sure they were of the blood of the immortals, such was the
task which the violent constraint of necessity forced them to undertake.
Again it is attractive to read ‘Muses’ as ‘sources’. This passage also con-
tains the only apostrophe of the Argonautica’s heroes beside the epilogue,
with which it shares its emphasis on the Argonauts’ divine status (‘of the
blood of the immortals’). Apostrophe of individual characters is equally
rare.12
Problematic events are also framed by addresses to other divine
agents. Thus, the narrator ‘mitigates’ the guilt of Jason and Medea in
murdering Medea’s brother Apsyrtus by scolding the god of love, Eros,
as the first cause of their atrocious deed (4.445–451):13
Wretched [skhetli’] Eros, great curse, greatly loathed by men! From you
come deadly strives and grieving and troubles, and countless other pains
on top of these swirl up. May you rear up, divine spirit, against my
enemies’ children as you were when you threw hateful folly in the heart
of Medea. For how then [dē] did she slay Apsyrtus with bitter death as he
came to see her? This was in fact [gar] the next part of my song.
(2) The last sentence of this passage shows the narrator as organizer of
the text—with a phrase that suggests a realization that with this sudden
emotional evaluation of an act yet untold, he may have moved too fast
for his narratees. We have seen him direct his narrative in the first-
person singular in the prologue (‘I shall recall’, ‘now I shall narrate’,
‘my song’) and in his Muse-‘invocations’ (‘if … I must also tell this
the decidedly un-epic subject of Medea’s love requires a ‘specialist Muse’, Erato; 4.1–
5: an embarrassing conclusion as to why Medea accompanied Jason—voiced by the
narrator himself despite ‘now you yourself, goddess, relate …’). Cf. furthermore 4.552–
556 (introducing a geographically impossible journey).
12 The apostrophe of Canthus in 4.1485 accomplishes the transition from the Arg-
onauts’ search for Heracles to the story of Canthus’ death; that of Theras in 4.1763
allows a wordplay hinging on the formal identity of the vocative of his name and the
name of the island called after him (Thera).
13 Compare also 4.1673–1675, where the narrator invokes Zeus to witness his bewil-
derment over the method by which Medea kills Talos (‘Father Zeus! My mind is all
aflutter with astonishment!’). The apostrophe of Hera in 4.1199 mirrors the characters’
invocation of Hera in a wedding.
without constraints’, ‘it is against my will that I relate’, etc.). After

digressions he sometimes redirects with a capping phrase: ‘well, this
is the story that is told among these people’ (4.618; cf. 2.528); ‘these
things, however, happened in a long course of time’ (4.1216; cf. 1.1309,
4.1764; → Herodotus). Elsewhere he breaks off in the middle of a story
(→ Pindar), apparently deciding that, on second thoughts, it is too long
and/or irrelevant to be finished: ‘but why should I tell these stories
about Aethalides all the way through?’ (1.648–649); ‘but these things
would make me stray far from the path of my song’ (1.1220).14
Reminiscent of Herodotus, but also in line with the narrator’s over-
all piety (which in turn reflects his narrative goals), are explicit silences
motivated by religious propriety (themis), for example when the Arg-
onauts are initiated in the cult of the Cabiri of Samothrace, learning
‘ordinances about which one may not speak’ (arrhētous themistas, 1.917):
About these I shall tell nothing more—but hail to the island itself and to
its indigenous gods, the guardians of those mysteries—about which it is
not proper [themis] for us to sing. (1.919–921; cf. 4.247–250)
In the same spirit the narrator apologizes for statements that might
offend a god. Whoever is bitten by the snake that killed the Argonaut
Mopsus in Libya cannot escape death, ‘not even if Paean, if it is proper
(themis) for me to speak openly, should administer drugs’ (4.1511–1512;
cf. 2.708–710).
Finally, it should be noted that, for all his ‘Herodotisms’, Apollonius
stays true to Homer in avoiding cross-references of the type ‘as I have
said earlier’/‘as I will tell later’—with one exception. In 2.1090–1091 he
uses a phatic question to remind his narratees of an earlier (riddling)
part of the story: ‘what was Phineus’ intention in making the divine
expedition of heroic men put in here?’
(3) Calling upon the Muses does not prevent the Apollonian narrator
from embracing ‘historiographical’ authentication strategies as well.
This leads to an ironic paradox: in his attempts to persuade his narratees
of the veracity of his story, the narrator constantly undermines his
authority as an inspired epic bard. What are we to think, for example,
14 Readers are surely invited to wonder why among so many ‘irrelevant’ digressions,
these specific stories are suppressed. In the case of the second story, how young
Hylas became the protégé of Heracles, the reasons must be that it was told in full
in Callimachus’ Aetia (from which the reader may supply the details) and that it is
potentially discrediting to Heracles.
when we hear that ‘Lynceus had the sharpest eyes of any mortal, that is
to say, if the lore is true (ei eteon ge pelei kleos) that with ease he could see
even down beneath the earth’ (1.153–155). The narrator, if instructed by
the Muses, should know if this is true. Even more striking is Apollonius’
introduction, directly after the Muse-invocation of the proem, of the
Argonauts’ own bard as ‘Orpheus to whom Calliope herself is said to
have given birth near the Pimpleian height, after she had shared the
bed of Thracian Oeagrus’ (1.23–25). One thinks that Calliope, being
the muse of epic, should remember whether and where she gave birth
to Orpheus, and whether she shared Oeagrus’ bed or that of someone
else. She might have enlightened our bard.
References to sources abound in the Argonautica. In some cases these
are quite elaborate. In Herodotean style, the narrator presents differ-
ent versions of a story in 4.597–617 (two explanations for the presence
of amber in the river Danube, one uncredited, the other told by ‘the
Celts’) and 4.982–992; but unlike Herodotus, he does not weigh the rel-
ative merits of the competing stories or declare a preference. In general,
the narrator does not discuss sources, but merely acknowledges their
existence, usually with phrases of the type ‘x is (so) told/called’.15 His
spokesmen remain anonymous or are identified as a group of people,
typically ‘locals’; information is never credited to specific individuals.
With the anonymous ‘is said’ in the Catalogue entry on Orpheus com-
pare for example ‘one mouth (of the Istrus) they call Narex, the other on
the south the Lovely Mouth’ (4.312–313) and ‘this route … is now called
Jason’s Path’ (1.988; cf. e.g. 2.929, 4.309). Local spokesmen are invoked
in ‘the spring which those who live nearby call Pēgai’ (i.e. ‘Springs’,
1.1221–1222; cf. e.g. 1.941; 2.506–507); ‘herdsmen’ are the source of a
name in 3.277, ‘hunters’ in 4.175.16
Although the narrator often evokes the possibility of autopsy, he never
claims to have actually seen evidence himself or to have interviewed
witnesses in person, unlike Herodotus. Very much like Herodotus, how-
ever, he never gives references that imply written sources (contrast →
Callimachus). Just as he consistently presents himself as a speaker, he
strictly maintains the illusion that all information he relates derives
15 Verb forms used are kaleousi/ontes/ontai, kiklēskousi/ontai/etai, (meta)kleiousi/ontai, klēï-
zetai, phasi kaleesthai/kleiesthai, pephatai, phatizetai, pephatistai, enepousi, hudeontai.

16 Interesting are also 3.1323, where ‘Pelasgian akainē’ stands for ‘the instrument
which the Pelasgians call akainē’, and passages with a verb of speaking in the past tense
which dates a source to the time of the story (such as ‘the people of the area called all
these heroes Minyans’, 1.229–230; cf. 4.1149, 1514).
from, and belongs to, the oral tradition or ‘collective memory’ of the
Hellenic world. Noteworthy are two references to ‘singers’, in the pro-
em (‘as for the ship, this is still celebrated in the songs of earlier
singers’, 1.18) and in the Catalogue (‘for singers tell that … Caeneus
was still alive’, 1.59–60). Although Apollonius here certainly refers to
poets whose work was available in writing, it would be misleading to
translate the Greek text’s aoidoi as ‘poets’. In the fictional discourse
of the poem, poets are still ‘singers’ and their poetry exists in perfor-
mance. The Argonautica itself, according to the rhetoric of the epilogue
(‘may these songs be from year to year sweeter to sing for humans’,
4.1773–1775), will become part of the same oral tradition.
The narrator does not engage in the elaborate arguments that we
find in Herodotus. Yet many small signs show that his tale is the
result of critical thinking. In 1.196 oiō ‘I think’ (not found in narrator-
text in Homer) conveys the narrator’s estimation of the potential of
young Meleager as a personal opinion: ‘he would have surpassed all
the others, I think, but for Heracles, if he had remained for only one
more year to come of age among the Aetolians’. He also regularly uses
the particle pou ‘I suppose’ (also absent from the Homeric narrator-
text) to mark statements as assumptions, temporarily ‘forgetting’ his
omniscience—either regarding the words, thoughts, and feelings of his
characters (‘after their release from chilling fear I suppose [pou] the
Argonauts breathed more easily’, 2.607) or regarding the facts of his
story (1.972–975):17
He [Cyzicus] too, I suppose (pou), was just sprouting the first beard of
manhood. In any case (nu) he had not yet been blessed with children, but
his wife … was untouched by the pains of child-bearing.
Such ‘micro-arguments’ conducted with interactive particles (here: pou,
nu) are common in the Argonautica.
(4) The narrator evaluates characters and events much more frequently
than Homer: the narrator-text of the Argonautica contains numerous
evaluative terms, including many words that are (almost) exclusively
used by characters in Homer.18 Apart from offering the occasional
17 For pou of assumed thoughts, words, facts cf. also 1.636, 996, 1023, 1037, 1140, 1222,
2.1028, 3.926, 4.557, 1457, 1397 (with 1436, where a character states what the narrator
assumed …). It also appears in similes to suggest the ‘arbitrariness’ of a chosen vehicle
or its details: 1.537, 3.758, 1283, 1399.
18 Hunter 1993a: 105–111. Apollonius also adopts the ‘generic’ qualifications that
‘unbiased’ judgment (Caeneus was ‘a noble warrior but not better

than his father’, 1.58; the killer of Canthus ‘was by no means a lesser
man’, 4.1489), he dispenses praise and blame in accordance with his
‘hymnic’ goal: the Argonauts are praised, criticized is whoever sins
against the laws of the gods and opposes the Argonauts (crimes which
often coincide).
Throughout the poem the narrator presents the Argonauts as mod-
els of religious observance. He lets them build altars and perform sac-
rifices and rites wherever they go, and constantly comments that their
behaviour is ‘proper’ (themis). When celebrating the wedding of Jason
and Medea ‘they prepared a mixing-bowl of wine in honour of the
blessed gods, as is proper (hē themis), and with correct ritual (euageōs)
brought sheep to the altar’ (4.1128–1129); after murdering Apsyrtus,
Jason ‘cut off the dead man’s extremities; three times he licked the
blood, and three times he spat the pollution out from his teeth, as is the
proper way (hē themis) for killers seeking to expiate murder by treachery’
(4.477–479). In the last example, the emphasis on religious observance
is particularly remarkable: the killing of Apsyrtus and the mutilation of
his corpse are not beyond reproach. The narrator, however, insists on
evaluating his heroes’ behaviour in a positive way. Where this would be
hard to do, he withholds comment.
The narrator’s preoccupation with themis also appears from his inter-
est in customs (themistes) which are at variance with those practised by
Greeks (and sanctioned by the gods). If the Argonauts do not suffer
harm from such customs, the narrator presents them without moraliz-
ing judgments, as in his description of the Mossynoeci (2.1018–1022):
The customs of these people are quite at variance. Everything that it is
proper to do openly, whether in the public assembly or in the market-
place, all of this they carry out at home; everything that we do in our
houses, this they do out in the streets and without incurring censure for
it.
When different ethics pose a danger to the Argonauts, the narrator

does not spare his critique, as in his introduction of the Bebrycian king
Amycus (2.1–9):
There were the stalls and lodge of Amycus, the arrogant king of the
Bebrycians (…), most outrageous of men, who imposed upon his guests
Homer does use, e.g. nēpios/oi ‘the poor fool(s)!’ (2.66, 137, 4.875), skhetlios/oi/ē ‘the
wretched man/men/woman!’ (1.1302, 2.1028; 3.1133; 4.445, 916, 1524).
an improper ordinance (…). At that occasion too (…) he insolently scorned to

inquire of the Argonauts’ mission and identity (…)
The Lemnian women and Amazons receive similar introductions

(1.609–619, 2.985–989).
Occasionally, narratorial comments take the form of gnomic utter-
ances. These all breathe a pessimistic spirit. A man in a simile wanders
far from his own land ‘as indeed we wretched men often do wander’
(2.541–542). The wedding of Jason and Medea is both a joyful occasion
and a sad one, because ‘necessity forced them to lie together at that
time. It is a fact that we tribes of suffering men never plant our feet
firmly upon the path of joy, but there is ever some bitter pain to keep
company with our delight’ (4.1165–1167).19
The narratees
Although the narrator does not explicitly identify his audience, it is

clear for example from his ‘sociological’ comments (above) that he
assumes that his narratees adhere to Greek customs and respect the
Greek gods. This broadly defines them as virtuous inhabitants of the
Hellenized world, an audience ‘bound’ to admire, and be proud of, the
exploits of the famous Greek heroes of the past.
First-person plurals may create an effect of ‘shared focalization’.
Apart from 2.1021, 542, 4.1166 (above), see e.g. ‘everything that we do
in our houses’ (2.1021); ‘as indeed we wretched men often do wan-
der’ (2.542). The narrator also often invites his narratees to cross the
distance between here and now and there and then, and to imagine
themselves on the scene with the poem’s characters on that long ago
day. This is sometimes achieved by simple means: the boxers Amycus
and Polydeuces are ‘neither in physique nor in stature similar to be-
hold’ (2.37); the wings of the Boreads are ‘a great marvel to see’ (1.220).
Sometimes focalization is transferred to characters who, like the nar-
ratees, witness the events of the story: ‘on that day all the gods looked
from heaven upon the ship and upon the generation of half-gods who
sailed the sea, best of all men’ (1.547–549). The narratees may also rec-
ognize themselves in ‘anonymous eye-witnesses’ and ‘anonymous inter-
19 For similar comments see 1.82, 458–459 (only apparently more optimistic), 1035–
1036; 4.1504.
locutors’: ‘if someone were to count all branches (of the river Ther-
modon), he would find them four short of a hundred’ (2.974–975; →
Herodotus); ‘this is what each citizen said as he saw (the Argonauts)
rushing forward with their weapons’ (1.240–241; → Homer). Finally,
there is the ‘indefinite second-person’ device. When the Argonauts face
a giant wave at the entrance of the Bosporus, the narratees are made to
go through this ordeal (2.171–176):
You would say that there was no escape from a miserable fate, as the
violent wave hangs like a cloud over the middle of the ship. But it drops
if you happen to have an excellent pilot. So the Argonauts too came
through by the skill of Tiphys—unharmed but terrified.
Elsewhere they are invited to imagine what no character actually expe-
riences, such as the spectacle of the Planctae when Hephaestus’ furnace
was still clouding up the air: ‘you would not have seen the rays of the
sun’ (4.927–928). Here the narratees are made to consider how badly
the Argonauts’ passage of the Planctae might have ended if Hephaestus
had not extinguished his fire20—compare the ‘if not’-passages of Homer
(→), which likewise suggest what might have happened (if not …). Such
passages are also common in the Argonautica: the sons of Boreas would
have killed the Harpies ‘if swift Iris had not seen (…) and checked
them’ (2.284–287); the Argonauts would have delayed in Mariandynia
even longer ‘if Hera had not put great boldness into Ancaeus’ (2.864–
866).21
The narrator most pervasively engages his narratees’ expectations,
and anticipates their reactions, in digressive passages. The Catalogue
of Argonauts, for example, provides innumerable examples (1.23–233).
When the narrator introduces Eurytion and Erybotes, he seems to
realize as he is speaking that his words may need clarification (1.71–74):
In the group too were Eurytion and bold Erybotes, one the son of
Teleon, the other of Irus son of Actor—that is to say (ētoi): glorious
Erybotes was the son of Teleon, Eurytion the son of Irus –
and not the other way around, as ‘the one … the other’ (ho men …
ho de) suggests. The clause ‘not even the son of mighty Pelias himself
(…) wished to remain in his father’s house’ (1.224–226, ‘presentation
through negation’), evokes and contradicts the expectation that Pelias’
20 Other second-person forms occur in 1.726, 765/767; 3.1265 (cf. 3.1044); 4.238
(simile), 429 (ecphrasis), 997. See Byre 1991.

21 Cf. 1.493, 863, 1300, 2.993, 3.1142, 4.22, 905, 1653; similarly figures e.g. in 4.638–
639, 1305–1309.
son would not have wanted to take part in what his father had devised
as a deadly undertaking. Heracles enters the poem in the same way:
‘not even the mighty Heracles (…), so we are informed, scorned Jason’s
needs’ (1.122–123). Here it seems hard to deny that the expectations
addressed are not only based on reasoning (the great Heracles might
have been too busy or unwilling to join lesser heroes for an expedition)
but also on literature: in earlier Argonauticas Heracles was not always
part of the crew. ‘We are informed’ (peuthometha) acknowledges the
existence of such sources—and implicates the narratees in the sup-
pression of alternative accounts.22 One step further is the narrator’s
explanation that the prophet Idmon ‘was not really a son of Abas but ra-
ther a son of Apollo himself ’ (1.142). This makes no sense at all unless we
already know that Idmon is sometimes said to be a son of Abas, for the
narrator has not told us. In other words, Apollonius tends to construct
complex communicative situations. Compare also, for example, the
following digression on the tomb of Idmon, an Argonaut killed on the
way to Colchis (2.841–855):
And as you know (dē toi), the tomb of this man rises in that land. On
it is a marker that is visible also to people of later generations: a ship’s
roller made from wild olive, green with leaves. It lies just below Cape
Acherusia. And if, under the Muses’ influence, I must also tell this
without constraints: Phoebus instructed the Boeotians and the Nisaeans
to pay honours to this man as ‘Protector of the City’ and to establish a
city around this roller of ancient wild olive. But rather than the descen-
dant of god-fearing Aeolus, Idmon, it is Agamestor whom they glorify to
the present day.
So who else died, then (dē)? Because once more at that time the heroes
raised up a tomb over a lost companion—for two markers of those men
may still be seen. It was the son of Hagnias, Tiphys, who died, so they
say [phatis]; he was not fated to sail any further.
Notable in this and other aetiological passages are expressions that

imply a great distance in time and space between the world of the
narrator and narratees and that of the story. One of these expressions,
‘ancient wild olive’ (palaigeneos), recalls ‘the famous deeds of ancient men’
(palaigeneōn) in the poem’s first line. Distance is also created by the
temporal adverb tote, ‘at that time’, and by the pronoun for remote
22 Compare 1.133–138, where 135 idmen, ‘we know’, makes the narratees accomplices
to the invention of four otherwise unattested generations of Danaids, in a passage where

the narrator apparently ‘sets the narratees straight’ as in 1.71–74 (cf. esp. 134 ē gar).
deixis, keinos, in the phrases ‘in that land’ and ‘those men’ (→ Homer).23
Other expressions explicitly link the remote past with the time of narra-
tion (the ‘continuance’ motif): ‘to the present day’, ‘visible also to peo-
ple of later generations’, and ‘may still be seen’.24 Furthermore, the
narrator asks a question, reasons, adduces the Muses and sources, and
engages his narratees’ expectations with interactive particles. The first
words of the passage, kai dē toi, translated as ‘and as you know’, mark
what follows as an elaboration upon what went before (i.e. a digres-
sion, kai), call for the narratees’ special attention (toi), and suggest that
what follows should not come as a surprise to them (dē).25 An even
more complicated rhetoric of anticipation is implied by the second
paragraph. ‘So who else died, then?’ (tis gar dē thanen allos).26 Why this
question should (dē) occur is explained in the next two clauses, which
lead back to the evidence, which has an implication, which evokes
a question—evidence: there are two tombs at the site > implication:
two Argonauts died > known: Idmon died; question: who else died? >
answer: Tiphys died.
The expressions ‘to be seen’ and ‘so they say’ create an interest-
ing problem of perspective: to whose perception do these verbs refer,
and when, where, and how does it take place? In other words, what
communicative situation is imagined here? I would suggest that there
is no single answer, but that Apollonius ‘contaminates’ a number of
fictional communicative situations. In this aetiological context it is hard
not to be reminded of the narrative framework of the first half of Calli-
machus’ Aetia (→), where the narrator interviews the Muses, asking
them to explain the origin of phenomena (‘why, Muses, is it that …’).
But the lines also display the ‘rhetoric of enargeia’ (‘placing before the
eyes’), the illusion that the narratees are travelling the route of the
Argo with the narrator as their guide. In this setup the attention of
the narratees is fixed on the tombs because the narrator is pointing
them out (‘to be seen’) and telling the story that goes with them, for
which he relies on a source (‘they say’). Also evoked is the ‘rhetoric of
23 Compare keinēi aiēi ‘in that (far) land’ at 4.534; ēmati/ēmar keinōi/o ‘on/during that
day (in the past)’ at 1.547, 1070–1071; 2.760, 1097; 3.850, 922.
24 Greek eiseti nun and kai opsigonoisin idesthai (cf. 1.1062; 4.252) and eti phainetai. Typical
expressions linking past to present are (eis)eti (nun), exeti (keinou), ex hou, (es)aiei/aien.
25 Interactional particles regularly used by the narrator include mēn, kaitoi, ē, ētoi, ē
gar, and mentoi.

26 For similar questions see 2.1090–1092 and (explicitly addressed to the Muses)
4.552–556.
personal observation and communication’ of Herodotus—though it is

not taken to the point where the narrator actually claims personally
to have visited Heraclea Pontica (autopsy) and asked questions about
the tombs. It is certainly suggested, however, that the narratees could go
and verify the narrator’s words. Finally, there is the rhetoric of shared
tradition: narrator and narratees both know about the tombs because
they know the same stories—according to the fiction of the Argonautica:
from hearing poems. In the actual context, however, in which the Argo-
nautica was produced and first received, ‘they say’ means ‘they write’,
‘to be seen’ is entirely hypothetical, and the whole digression is a
bonding exercise for Apollonius and his readers, who together indulge
their encyclopedic interests and assert their erudition. Yet one wonders
how many of Apollonius’ contemporary readers actually knew about
the two tombs before they encountered this passage (the ‘rhetoric of
shared reading’ is surely also to some extent a fiction).
Secondary narrators and narratorial alter egos
Characters speak 45 per cent of the Iliad and 67 per cent of the Odyssey,
but only 29 per cent of the Argonautica, of which the largest part is taken
up by dialogue. In other words, the poem’s narrator tends to speak
more in his own voice than Homer, and less often yields the floor to
secondary narrators.27 Interesting in this respect are several passages
where the narrator ‘usurps’ a potential embedded narrative, such as his
report of Jason’s tale to king Lycus, which summarizes the entire narra-
tive up to this point (2.762–771). It is attractive to think that Jason is here
robbed of a chance to act the role of Homer’s Odysseus, who is allowed
to narrate his own adventures at length, and that Apollonius is hinting
at the possibility of an Argonautica focalized by its main hero. Details
in the narrator’s report may indeed be attributed to the focalization
of Jason, who is trying to impress his host Lycus and presents embar-
rassing facts in a favourable light. However, other embellishments must
be attributed to the narrator who, as we have seen, is also inclined to
give a positive spin to the Argonauts’ deeds.28 This convergence of the
27 The resulting loss of dramatization is compensated by the interaction between
narrator and narratees, and by a significant increase in embedded focalization and

indirect speech (Fusillo 2001).
28 The clearest example is the statement that Jason told Lycus ‘all they had done at
focalization of narrator and character is illustrative of a general ten-

dency: virtually all secondary narrators of the poem to some extent
reflect the primary narrator. In the passage under discussion this mir-
roring is especially clear: Jason reiterates the main narrative, his first
words echo the poem’s proem, and his narratee reacts as if he had
been listening to an inspired bard (‘as Lycus heard Jason’s story unfold
in due order, his spirit was bewitched’: 771). Other character-narratives
recall the primary narrative in other ways. Lycus tells Jason the aition of
games celebrated at Heraclea (2.780–785); Jason describes Thessaly to
Medea in Herodotean style (3.1085–1095); and the greatest scholar of all
is Argus, son of Phrixus, who, setting out to explain an alternative route
from the Black Sea to Greece, launches into a long mytho-geographical
excursion that leads back to the dawn of human civilization (4.257–293;
cf. also 2.1141–1156, 1200–1215).
However, the narrator’s true text-internal alter egos are the Argonauts’
own inspired singer, Orpheus, and the seer Phineus, who echo Homer’s
(→) Demodocus, Phemius, and Tiresias also in this metanarrative role.
Orpheus, son of the Muse Calliope, is significantly the first Argonaut
listed in the Catalogue, directly after the Muse-invocation that closes
the proem (1.23–32). The position and exceptional length of Orpheus’
‘lemma’ may be read as a metanarrative indication of the ‘primacy of
the narrator’ in the poem. Moreover, it is suggested by the emphasis
on Orpheus’ power to cast a spell, and by the indication of his moral
qualities implied by the fact that Jason recruits him ‘in obedience to
the urgings of [his teacher] Chiron’, how the poem’s narrator would
like to see himself honoured by his narratees (represented by ‘the hard
mountain rocks’ of barbarian Thrace moved by Orpheus’ lyre, i.e.
not the easiest audience): as a spiritual guide with enchanting powers.
Moreover, Orpheus’ mother is the Muse of epic, and Orpheus himself
is, like Apollonius, a ‘man of Apollo’ by profession and birth (because
according to one tradition he was Apollo’s son). All this suggests that
with Orpheus, the narrator is the first man to step on board the Argo
to be Jason’s trusted helper—on the path of his quest, the path of song,
and the road to immortality. On this journey he will, hopefully, likewise
enchant his narratees.29
Dolionian Cyzicus’: Jason cannot possibly have covered up the embarrassing fact that
at Cyzicus the Argonauts had killed their royal host.
29 Meanwhile there is a striking contrast between the ‘naïve’ spell, which the poem
here and elsewhere presents as the (desired) effect of narration, and the intellectual
Though Orpheus’ metanarrative role corresponds to that of Phemius

and Demodocus, the narrator ‘handles’ him in a different way. This is
signalled, for example, by the way in which the he presents Orpheus’
first song (1.496–511). The narrator of the Odyssey always introduces his
bards’ words in indirect speech (‘he sang how …’), but after a few lines
shifts to the free style—a setup that allows the voices of narrator and
text-internal singer to blend naturally (→ Homer). Apollonius, however,
maintains the indirect speech for no less than eleven lines before finally
switching to the free style. This is probably not merely an experiment
with indirect speech, but a metanarrative statement. The narrator of
the Argonautica does not want his voice to converge with that of the
intratextual singer. He retains control over the narrative while the
singer retains responsibility for his song. This arrangement at one point
causes an amusing authority crisis (2.705–710):
[Orpheus sang] how once at the foot of the rocky ridge of Parnassus
the god [Apollo] killed the monster Delphyne with his bow, when a
young boy still in his nakedness, still rejoicing in long curls—be gracious!
Eternally, lord, your hair is uncut, eternally it remains unravaged. So does
holy law (themis) proclaim, for only Leto herself … may hold it in her
dear hands.
Who is responsible for the sacrilege of suggesting that Apollo’s hair was
at some time cut (‘still rejoicing in long curls’)? Does the narrator cor-
rect his own clumsy representation of Orpheus’ song or is the song itself
at fault? It seems impossible to decide, but what is clear is that both
ideas are equally inconceivable in Homer. The context of this song is
also illustrative of Orpheus’ position within the poem’s cast of charac-
ters. He functions as a mediator between the Argonauts and the gods,
acting as their master of ceremonies in contexts that require religious
action—in this case an epiphany of Apollo. In his song Orpheus pro-
vides the origin (aition) of the name of Apollo’s main cult site, Delphi,
and of the custom of hymning Apollo with the word hiepaian, thus mir-
roring the primary narrator as a singer of aitia and of a hymn to Apollo.
In the ‘normal’ action of the story Orpheus takes no part. Like the nar-
rator, he is not an actor but an observer, commentator, and spiritual
guide.30
involvement which Apollonius the author requires from his readers if they are to be
equally impressed.
30 In this respect Orpheus shows closer resemblance to the ‘wise adviser’ self-images
of Herodotus (→) than to Homer’s Phemius and Demodocus, who are primarily
Unlike Orpheus, the narrator’s other major alter ego, the seer Phine-
us,31 receives ample opportunity to voice his own words. Phineus’ first
words and his invocation of Apollo in 2.209–214 echo the narrator’s
proem, proving his ‘Apollonian’ omniscience to primary and secondary
narratees alike. As he describes his own fate (2.215–239), he invokes
all his rhetorical skills to secure the Argonauts’ assistance—and con-
veniently omits that the Harpies who plague him are a punishment
incurred for abusing his omniscience. His secondary narratees sense the
truth which the primary narratees already know, because the narrator
has just given them his account of the story (2.178–193). Repetition of
information is in fact very common in the Argonautica. Many stories and
facts are related more than once, by different narrators, with a different
focalization, and with a different goal. Where, as here, the character’s
version follows that of the narrator, the primary narratees are led to
believe that they can judge the character’s words against the ‘facts’. Yet
this only holds true to a certain point, because the narrator’s account
is not necessarily entirely objective either: he has his own agenda and
pre-occupations (which imply a negative view of sacrilegious persons
such as Phineus).32
Phineus’ potential as a narratorial self-image is fully developed in the
long monologue (2.311–425) in which he gives the Argonauts instruc-
tions for their journey to Colchis, exploiting his ‘Apollonian’ omni-
science. In exploring the device of prolepsis by an omniscient character-
narrator, Apollonius’ goes far beyond his Homeric model, Tiresias’
prophecy in Odyssey 11. While Tiresias provides a bare outline of events
to come, Phineus, after he has provided detailed instructions for passing
the Clashing Rocks, offers an equally detailed description of the people
and places which the Argonauts will pass along the Black Sea coast, in
a style which is almost undistinguishable from that of the primary nar-
rator. In fact, his ethnographical, geographical, and aetiological excur-
sions are so adequate that when the Argonauts reach the places he
described, the narrator sometimes provides less detail, merely rephrases
entertainers of their superiors. For other reported songs of Orpheus see 1.496–515,
569–579; and 4.903–911. See further Busch 1993; Clare 2002: 231–260; Clauss 1993:
26–32, 66–95; Fusillo 1985: 60–63; Hunter 1993a: 120–121, 148–151; Nelis 1992; Pietsch
1999.
31 E.g. Clare 2002: 74–83; Feeney 1991: 60–75; Hunter 1993a: 90–95; Manakidou
1995.
32 Cf. the double presentation of the Lemnian women (1.609–639: 657–707) and
Amycus (2.1–10: 11–18).

Phineus’ words or even verbally repeats them. There is indeed little

honour to be gained in retracing the steps of a narrator of the same
interests and authority. The close similarities between the two accounts
have the effect of mutually confirming their reliability. Such confirma-
tion is welcome, because the authority of both narrators is open to
question. The primary narrator is as much a fallible historiographer as
he is an inspired epic singer; and Phineus explicitly forewarns the Arg-
onauts that the rules of prophecy prevent him from telling exactly what
will be (2.311–316):
Listen to me now. It is not permitted (ou themis) for you to know every-
thing precisely, but as much as is pleasing to the gods I will not conceal. I
have already before made a mistake regarding the purpose of Zeus by
recklessly prophesying it in order and to the end. For this is the way he
himself wants it: that to humans the god-speak of prophecy be revealed
imperfectly, so that they will still be ignorant of part of the gods’ purpose.
The metanarrative implications of this passage are hard to deny, if
difficult to pin down. At any rate, it is tempting to see the prophet’s
desire and, at the same time, inability to ‘tell everything precisely’, ‘in
order and to the end’, as reflecting a dilemma faced by the ‘hypophet-
backed’ narrator of the past.33
Conclusion
We have seen that the narrator of the Argonautica engages in an overt,

self-conscious, pervasive, and variegated dialogue with his narratees,
his sources, and the Muses, in which he displays a distinct personal-
ity. This personality can be understood from the poem’s narrative goal,
the literary tradition that it reflects, and the social-intellectual milieu in
which it was created. Congruous with his aim of ‘hymning’ the Arg-
onauts and Apollo, the narrator shows a strong awe for the gods and
for the heroes of the past about whom he narrates, and an outspo-
ken disapproval of those who oppose either. He shows characteristics
of the divinely inspired epic bard and hymnic and Pindaric singers of
praise, which evoke the illusion that the Argonautica is conceived and
performed in front of an enchanted live audience. But at the same time
the poem operates other communicative fictions, which are to a vari-
33 Compare also Phineus’ words in 2.388–391, which (beside Call. Aet. fr. 1.3) echo
the narrator’s words in 1.648–649 and 1.915–917.

ous degree incongruous with the fiction of epic performance and with
each other, and are indebted to other models. In treating the Muses
as his interlocutors, the Apollonian narrator resembles the narrator of
Callimachus’ Aetia (→). He appears as a Herodotean ‘oral historian’ in
his frequent references to physical evidence and (oral) sources, and in
general in his critical attitude and use of ‘historiographical’ authenti-
cation strategies (diametrically opposed to those of the epic bard). He
also appears as an ‘Alexandrian’ scholar who seizes every opportunity
to parade his learning in front of his narratees, whom he engages in
scholarly discourse—with the crucial difference that there is no men-
tion of reading and writing, but merely of hearing, seeing, speaking,
and singing. This emphasis on face-to-face communication creates an
intimacy that compensates for the writer’s physical distance from his
readers in the actual reception context for which the Argonautica is ulti-
mately intended.
chapter five
CALLIMACHUS
M.A. Harder
Apart from the hymns and epigrams the work of Callimachus has been
preserved only in fragments and the corpus studied for this chapter is
comparatively small.1 Even so, Callimachus’ work presents an impor-
tant step in the history of ancient Greek narrative, because it devel-
ops and modifies techniques of earlier authors, engages in a dialogue
with contemporary poetry, and altogether shows a high degree of self-
conscious sophistication.
Narrators
What kind of narrators do we find and how do they manifest them-

selves in the text of Callimachus? The situation varies according to the
genre in which they appear.
In the hymns there is usually an overt narrator who gives informa-
tion about his motivation for singing about a certain god (hh. 1, 3, and
4) or about an occasion in which he/she is actually taking part while
telling the story (hh. 2, 5, and 6; the so-called mimetic hymns). As is
pointed out in the introduction to this volume, in the second group we
may postulate a suppressed primary narrator and consider the narrator
as secondary.
In h. 4 there is an external primary narrator, who at the beginning of
the poem is briefly characterized as a singer-poet who feels obliged to
sing of Delos in order to win Apollo’s favour:
1 The corpus consists of the hymns and the Aetia (particularly the larger fragments).
Other narrative texts, like the Hecale and Iambi fr.191 have been adduced when relevant
(for a general narratological analysis of the Hecale see Lynn 1995: 7–117). As the corpus
is small and fragmentary, conclusions must be treated with some caution. For the text of
Callimachus see Pfeiffer 1949–1953; Lloyd-Jones–Parsons 1983 (fragments quoted from
this edition are indicated with SH); Hollis 1990.
64 part one – chapter five
When, my soul, will you sing of holy

Delos, the nurse of Apollo? Truly all the
Cyclades, which are the most holy islands in the sea,
are well served with songs, but Delos wants to carry the first prize
from the Muses, because Phoebus, the master of song,
was first washed and wrapped and praised by her.
Just as the Muses hate the singer who does not sing
of Pimpleia, Phoebus hates the man who forgets Delos.
I shall now give a portion of my song to Delos, in order that
Cynthian Apollo will praise me because I care for his dear nurse.
(h. 4.1–10)2
In this passage Callimachus gives his narrator more of a persona than
was usual in the Homeric hymns (→), where we usually find only one
formulaic line to announce the narrator’s intentions or a request to
the Muse to sing, although the address of the Delian chorus in h. Ap.
165ff. provides an example how the figure of the hymnic singer may be
elaborated and may have inspired Callimachus. In his poems the rudi-
mentary persona of the Homeric hymns seems to have acquired a more
developed personality, constructed in such a way that it draws atten-
tion to the responsibilities of the hymnic singer towards his patron-god
and to the function of the genre in celebrating the gods. In the actual
narrative (about Leto’s search for a place where she can give birth to
Apollo) the narrator is occasionally visible, as in the various apostro-
phes of characters in his story (mainly Delos, as the hymn’s subject, in
27–55, but also Hera in 106–107 and 215–216), and particularly in the
invocation of the Muses in 82–85, where we get a glimpse of the narra-
tor asking and receiving information which is strictly speaking outside
the scope of the hymn:
Tell me, Muses, my goddesses,
is it true that the oaks were born when the Nymphs were born?
‘The Nymphs rejoice when rain feeds the oaks,
but the Nymphs cry when the oaks have lost their leaves.’3
Apart from this digression the narrator is practically omniscient and
his narration is subsequent (as can be seen in the references to the
past in 39–40, 49 and 253–254), apart from 166–190, where Apollo as
2Unless indicated otherwise the translations are my own.

3The punctuation as established by modern scholars implies that in 84–85 the
Muses’ answer is given in direct discourse and that, as in Aetia 1–2 (on which see
below), the idea of a dialogue between a poet which the Muse-invocations in Homer
and Hesiod suggest, was taken literally. See further Mineur 1984: 117–118; Harder 1988:
3–14.
m.a. harder – callimachus 65
a secondary narrator tells about the war with the Celts in the future.
At the end of the hymn the narrator briefly reappears in the hymnic
farewell in 325–326, which again recalls the conventions of the Homeric
hymns.
In h. 3 we are dealing with a similarly overt external primary nar-
rator: 1 gives a brief motivation for the song; next, we find explana-
tions (47–48, 172, 244–245), evaluative and metanarrative comments
(64, 136–137, 255), apostrophes of Artemis, which form part of an
extended passage of second-person narration (72–190), and instances
of the ‘reference to the narrator’s own time’ motif (77, 145). As in h. 4,
there are also brief ‘dialogues’ in which the narrator asks for informa-
tion: a series of three questions and answers in 113–135 (where the first
answer may be attributed to Artemis, but the second and third, which
refer to the goddess in the second person, most likely derive from the
narrator himself) and a series of brief questions in 183–186, a passage
which comes close to the ‘aporia’ motif:
Which island, which kind of mountain pleased you most?
Which harbour, which kind of town? Which nymph did you love
most of all and which kind of heroines did you take as companions?
You must tell me that, goddess, and I shall sing about it to the others.
You loved …
At the end of the hymn there is a conventional hymnic farewell and
request to receive the song favourably (268), but the farewell is preceded
by two other occurrences of khaire in 225 and 259, each followed by a
series of brief references to (other) stories from Artemis’ career.
In h. 1 we are again dealing with an overt external narrator: the
motivation for the song in 1–2 includes a slight hint of the occasion
at which one should celebrate Zeus (libations), but that occasion is
not referred to again in the rest of the poem. Other signs of the
narrator include frequent apostrophes of Zeus, rhetorical questions (1–
3, 62–63, 75, 92–93), the ‘aporia’ motif (4–5), anonymous spokesmen,
including poets (14, 39, 45, 51, 59–60), instances of the ‘reference to the
narrator’s own time’ motif (18–20, 40–41), evaluative and metanarrative
comments (63–64, 65, 68–70, 85–87), a quotation from Hesiod (79),
and a typical hymnic farewell (91–96). Taken together, the narrators
of Hymns 1, 3, and 4 display an interesting mixture of epic, hymnic, and
historiographical (→ Herodotus) aspects.
The ‘mimetic’ hymns hh. 2, 5 and 6 have a dialogical frame, and
the stories of Tiresias (h. 5.57–136), Erysichthon (h. 6.24–117), and—
perhaps—the episodes from Apollo’s career (h. 2.58–112) are recounted
by a kind of ‘master (or mistress) of ceremonies’,4 who at the beginning

and end of the poems gives instructions to young men (h. 2) and women
(hh. 5 and 6), who seem to be participating in a ritual event, e.g.:
You bathpourers of Pallas, come out everyone,
come out. The mares just now began to neigh,
I heard the sacred mares; the goddess too is ready to come.
Hasten, fair women of Argos, hasten now. (h. 5.1–4)5
Athena really comes now, so make welcome
the goddess, girls, whose duty the task is,
with acclamation, with praying, with joyous answering cries.
Hail, goddess, look after Inachian Argos.
Hail as you drive out, and as you drive back in again
your horses, and protect the whole Argive estate. (h. 5.137–142)
These secondary narrators are less visible than in h. 4: there are no
apostrophes, apart from the pathetic address of Erysichthon’s mother
in h. 6.83 (the address of the bathpourers in h. 5.134 and to Demeter in
h. 6.116–117 signal the return to the ritual frame). Sometimes, however,
there are brief evaluative comments, as in h. 5.78 ‘poor fool’ (about
Tiresias unwittingly seeing the naked Athena) and 6.36 ‘shamelessly’
(about Erysichthon’s men rushing into Demeter’s sacred grove).6 In
h. 2 the situation is more complex, in that it is not entirely clear
whether the voice that tells about Apollo must be attributed to the
(secondary) narrator or to the chorus of young men functioning as a
tertiary narrator. In 8 the narrator urges the young men to sing, in
16 he praises them for having started the music, in 17–21 he asks the
other participants to be silent, and then in 25 to shout the ritual cry
hiē hiē for Apollo (the addressee of this request is not entirely clear),
and in 28–31 he concludes with a promise that Apollo will honour
the chorus, who in return will sing about him for more than one day.
Although a change of voice in 32, where the praise of Apollo begins, is
not made explicit, it is at least hinted at by the preceding passage and
the effect is that the voices (of narrator and young men) merge. In 97
(‘we hear hiē hiē paiēon’) it seems most likely that the voice is that of the
secondary narrator again, but the ‘we’ could also be the chorus hearing
4 For this description of the poetic persona in these hymns see Hopkinson 1984: 3 n.
2. On the ‘mimetic’ hymns in general see e.g. Hopkinson 1984: 11 n. 4; Bulloch 1985:
5–8; Harder 1992: 384–394; Depew 1993: 57–77. One may compare the mimetic poems
in Theocritus (→).
5 Translations from h. 5 are taken from Bulloch 1985.
6 Cf. also h. 6.56 and 68.
the other participants joining in. The effect of this merging of voices
may be to suggest an enthusiastic crowd taking part in the celebration
of Apollo, and it recalls the indeterminacy of the encomiastic voice in
Pindar and Bacchylides (→). In contrast with hh. 5 and 6 the narrative
parts of this hymn contain several apostrophes of Apollo (particularly
in 65–104, about the foundation of Cyrene and the origin of the ritual
cry).
In all the hymns the narrators draw attention to the fact that they
depend on a long tradition: in h. 4.28–54 the narrator states that there
are already many songs about Delos, asks the island what it would like
to hear, and then suggests that he should tell how, after a long period of
floating in the sea, the island was eventually allowed to settle in a fixed
position because it received Leto for the birth of Apollo. This passage
recalls the ‘aporia’ motif, which, as in the Homeric Hymn to Apollo (→),
suggests the quantity of stories available to the narrator. In h. 5.55–56
the narrator explicitly states that he has heard from others the story of
the blinding of Tiresias, which he will tell to the women while waiting
for Athena’s epiphany (‘and meanwhile I shall speak to these women;
the tale is others’, not mine’); in h. 6.18–23 the narrator lists a number
of possible stories about Demeter, from which he eventually selects the
edifying story of Erysichthon. Thus the narrators present themselves as
a trustworthy and erudite medium for the transmission of stories that
were already told elsewhere.7 The references to anonymous spokesmen
in h. 1 (including the ‘ancient singers’ in 60), to the richness of material
in h. 2 and to the narrator’s role in the transmission of knowledge in h.
3 fit in with this picture.
In the Aetia there is a great variety of overt narrators,8 and the pre-
sentation of the stories is more complicated than in the Hymns. In
Aetia 1–2 the aetiological stories are told within the framework of a
dialogue: there is an internal primary narrator, ‘the old Callimachus’,
who tells, to a primary narratee of whom we find no explicit traces
in the fragments, how he has dreamt that as a young man he was
brought to Mt. Helicon, where he asked the Muses the origins of rit-
uals and related matters.9 The embedded narratives are therefore pre-
7 This technique may be related to Callimachus’ famous statement ‘I sing of noth-
ing which is unattested’ (fr. 612).

8 E.g. Harder 1990: 287–309; Cameron 1995: 351–354; Fantham 1998: 17–18.
9 Cf. Schol. Flor.15ff. (in Pfeiffer 1949: 11) (‘he [sc. the old Callimachus], who had
just spoken the prologue to the Aetia, told) how in a dream he met the Muses on Mt.
Helicon and received from them the explanation of the aitia, being a young man …’
sented as ‘answers’ by the Muses, functioning as external secondary

narrators, who respond to the questions of ‘the young Callimachus’,
who thus acts as the external secondary narratee. This practice recalls
the way in which the Muses are used in Homer (→) to explain the
narrator’s omniscience. However, in the Aetia these roles are not fixed,
as ‘the young Callimachus’ sometimes inserts long digressions in his
questions, in which he displays his own knowledge (such as the cata-
logue of Sicilian cities in fr. 43.28–55, which represents in nuce a series
of foundation-stories), and perhaps adds stories on his own account (as
has been thought about the story of Heracles and Thiodamas in frr.
24–25).10 Thus both ‘the young Callimachus’ and the Muses function
as secondary narrators as well as secondary narratees, and the old con-
vention of the Muse-invocation is reworked creatively.11
In Aetia 3–4 the framework of the dialogue is no longer used, but
instead there is a great variety of primary narrators, each with a distinct
persona, which sometimes evokes a certain literary genre. Some of these
narrators are external, like the ‘scholar-poet’ in frr. 67–75, who tells the
love-story of Acontius and Cydippe and is characterized by his garrulity
(fr. 75.4–9, where he rebukes himself for almost telling an unsavoury
story about Hera and cuts himself short in a Pindaric manner),12 and
by the fact that he quotes his source at length (fr. 75.54–77). This nar-
rator is obviously overt, as is shown most clearly from fr. 75, which is
the longest and best preserved fragment of the Aetia. Narratorial inter-
ventions which reveal the narrator’s presence include: apostrophes of
Acontius (40, 44–48, 51, 53, 74–75), an instance of the ‘reference to the
narrator’s own time’ motif (51), a gnomic statement (8–9), and eval-
uative and metanarrative comments (13–14, 44–49, 74–77).13 Another
external narrator is the ‘epinician poet’, who offers a victory song to
Berenice in SH 254–268 and includes a myth about Heracles.
Other narrators are internal and in a voice that recalls the ‘I’ in
dedicatory or funeral epigrams recount their own experiences, such as
the ‘lock of Berenice’, which tells of its catasterism in fr. 110 and shows
a delicate mixture of pride and regret (cf. especially fr. 110.75–78), or the
10 See Hollis 1982: 118.

11 On the complex interaction between Callimachus and the Muses see also Lynn
1995: 154–164, who observes that, strictly speaking, the Muses ‘are his mouthpieces, the
product of his own dreaming mind’ (155).
12 On this passage as a Pindaric Abbruchsformel see Harder 1990: 296.
13 For further discussion see Harder 1990: 287–309; and on the narrator in fr. 75 in
general Lynn 1995: 203–238.

dead poet Simonides in fr. 64, who tells how his tomb was destroyed by
a tyrant of Acragas.14
Secondary narrators seem to have been rare in Aetia 3–4, but there
is one instance of a reported narrator, Xenomedes in fr. 75.54–77.
Here we find the Cean history of the fifth century BC prose-author
Xenomedes presented in indirect discourse by the primary narrator,
who takes care to remind the primary narratee that he is summarizing
another author by repeatedly inserting markers of indirect discourse
(cf. 56 ‘beginning to tell how …’, 58 ‘and how …’, 60 ‘and how
…’, 64–66 ‘he put in hubris and death by lightning and …’, 70 ‘and
[he told] how …’, 74–75 ‘and he told about …’). Thus the primary
narrator stays in control, but the way in which he begins and ends his
summary shows that he also poses as some kind of ‘reader’-narratee
of Xenomedes’ story: 53–54 ‘we heard about your [sc. Acontius’] love
story from ancient Xenomedes, who once preserved the whole island
in a book of stories’ and 76–77 ‘the old man, devoted to the truth, told
of your passionate love, from where the boy’s story came quickly to
our Calliope’. As elsewhere, Callimachus creates a picture of a narrator
who is part of a chain of transmission.
A special case is fr. 178, where the convention of stories being told at
a symposium is used, and ‘Callimachus’ tells how he met the merchant
Theogenes of Icus at a symposium in Alexandria, where he heard from
him about the cult of Achilles’ father Peleus at Icus. It has been sug-
gested that this fragment may be from Aetia 2,15 and in that case the
dialogue with the Muses may have been even more complex than indi-
cated above: the passage may have interrupted the dialogue, since ‘the
old Callimachus’ may have reminisced about an earlier event, or ‘the
young Callimachus’ may have told the Muses about this symposium
and quoted Theogenes as a tertiary narrator.
The motivations for narrating in both Hymns and Aetia are sometimes
given explicitly, and, as with the narrator’s persona, here too one may
observe differences according to the genre at hand. The hymnic genre
presupposes a focus on the gods, and, indeed, we see that in h. 4 the
decision to sing about Delos is motivated by a wish to please Apollo.
In hh. 5 and 6 the motivation for telling the stories of Tiresias and
14 On Callimachus’ play with the conventions of the epigram see further Harder
1998: 96–99
15 See Zetzel 1981: 31–33, whose suggestion has been received favourably by a
number of scholars.
Erysichthon respectively is related to the ritual context and also shows

the need to respect the gods: in h. 5.51–56 the narrator indicates that
narrative has to kill time before the epiphany of Athena, but at the
same time warns the men of Argos, who should beware of seeing
Athena naked; in h. 6.18–21, the narrator, who has first rejected the
story of Demeter’s search for Persephone as too sad, considers a few
other possibilities and then settles for the story of the punishment of
Erysichthon ‘so that one may avoid transgression’ (h. 6.22).
In the learned Aetia curiosity and a wish for knowledge are presented
as an important factor, but one should bear in mind that, as in other
authors, e.g. Pindar (→), the explicit motive need not be the only or
even the main reason for telling a story. In Aetia 1–2 there are some
passages in which it is suggested that the stories are being told in order
to satisfy the curiosity of the secondary narratee, i.e. of ‘the young
Callimachus’, who asks the Muses for explanations, as in the case of
the Theodaesia in Haliartus:
Thus she finished her story, and I wanted to know this also
– for my amazement was secretly fed –
why near the water of Cissusa the Cretan festival of the Theodaesia
is celebrated by the town of Cadmus, Haliartus. (fr. 43.84–87)
In Aetia 3–4, too, some aitia are introduced by questions (e.g. fr. 79),
which may suggest a similar motivation, and in fr. 178.21–30 ‘Calli-
machus’ says that he longs to hear about the ritual at Icus:16
and tell me everything that my heart desires
to hear from you, in answer to my questions:
why it is an ancient custom for you to honour the leader of the Myrmi-
dons,
Peleus, how it is that Thessalian matters are connected with Icus,
for what reason a girl carrying an onion … (?). (fr. 178.21–25)
Other aitia are presented without an explicit motivation, such as the
story of Acontius and Cydippe in frr. 67–75, where, however, the sum-
mary of the narrator’s source in fr. 75.54–77 may create an impression
of the ‘scholar-poet’ eagerly collecting his material.17
16 In Hec. fr. 253.1–6 (= 40 Hollis) a similar question by Theseus seems to have
motivated Hecale’s telling of her life-story, which took up at least 100 lines; see Hollis
1990: 175–177.
17 The rather fragmentary source-indication at the end of the story of Melicertes in
fr. 92 (‘if the old Leandrian stories say something …’) and the reference to an inscribed
pillar in the first line of the story of Androgeus in fr. 103 (‘o hero at the stern, because
I conclude that there is a predominance of overt narrators in Cal-

limachus, which may be related to a Hellenistic awareness of taking
position in a long and rich literary tradition. In this respect one may
point to some recurrent features of these narrators: (1) primary as well
as secondary narrators are often presented as telling the stories because
they are in some way knowledgeable, either as Muses, or because they
rely on an earlier tradition of song, stories and the scholarly work of
earlier authors, or report as ‘eye-witnesses’ about their own country or
experiences; (2) the narrators are often given a persona and a motiva-
tion for telling their story that seems to be intended to underline the
text’s genre or generic pretentions; (3) the narrators may be part of
complex patterns of interaction in which the roles of narrator and nar-
ratee may change and overlap and are not always clearly defined, so
as to suggest an intricate chain of transmission. A specific aspect of
all Callimachean narrators, regardless of their status as primary, sec-
ondary or tertiary narrator, is the way in which the text may create
an ironic distance between poet and narrator by means of allusions to
other authors, which invite the readers to look at that text from another
angle than the one suggested by the narrator, to supply information
from other sources,18 or in the case of the future narrative in h. 4 (on
which see below) to realize that the narrator’s future has become their
own past. This kind of distance between poet and narrator may also be
observed in other Hellenistic poets, such as Apollonius of Rhodes (→),
Theocritus (→) and Lycophron (→).
The more scanty remains of the Hecale and Iambi fr. 191 fit in with
these observations. In the Hecale the knowledge of the primary nar-
rator is not explicitly accounted for, but in the embedded stories the
secondary narrators are presented as trustworthy: Hecale is knowledge-
able because she tells her own life’s story (frr. 40–49 Hollis), and the
old crow, which tells about Erichthonius and the early history of Attica,
because it witnessed the events as a young bird.19 In Iambi fr. 191 we
find as it were the ‘first attestation’ of a story: Hipponax comes from
Hades to tell the moralizing story of the golden cup of Bathycles (32–
77), which each of the Seven Sages modestly passed on, when he was
offered it, to the quarrelling philologists of Alexandria and he asks them
the pillar sings this …’) may be part of other explicit references to the activities of the
‘scholar-poet’ consulting his sources.
18 On this use of intertextuality in the Aetia see Harder 2002: 189–223.
19 Hollis 1990 on Hec. fr. 73.13–14.
to write it down (31), i.e. he provides them with a story which they may
pass on to others so that it will become part of the tradition. As to
the adaptation of the narrator’s persona to the genre, the fragments of
the epyllion Hecale suggest a covert external primary narrator, perhaps
intended to recall the narrators of epic texts; cf. e.g. the poem’s opening
in fr. 230 (= 1 Hollis) ‘Once there lived an Attic woman in the hills of
Erechtheus’; an instance of the ‘there was a place/person X’ motif.20
Ia. fr. 191 is presented as a speech by Hipponax, who was part of the
iambic canon and, therefore, well suited to be the narrator of the first
poem of Callimachus’ Iambi.
Narratees
The visibility of the primary narratees varies according to the genres

used by Callimachus. In hh. 1, 3, and 4 the hymnic narrator does not
explicitly refer to a primary narratee: in h. 4 the first line is addressed
to the narrator’s own heart and other addresses are directed towards
characters in the poem. In hh. 1 and 3, too, the ‘you’ is always the
god, though h. 3.186, ‘I shall sing about it to the others’, suggests that
there will be human narratees in the future. In hh. 5 and 6 the ‘mimetic’
form of the hymns brings along secondary narratees, who are addressed
several times in the ritual instructions and addresses which precede
and follow the narrative (h. 5.1–4, 13–17, 27–32, 134–138; 6.1–2; 118–
119). The addresses underline the identity of these narratees as female
participants in a ritual, and in h. 5.33–34 (‘Come out, Athena; here
is company to satisfy you, daughters of the mighty Arestorids’) a brief
description of them for the benefit of the goddess helps to complete the
picture. Sometimes there is a complex pattern of addresses of secondary
narratees and apostrophes of characters, as can be observed at the
beginning of the story of Tiresias in h. 5:
Pelasgian men,
beware lest unwitting you see the queen.
Whoever should see Pallas, the city’s guardian, naked
shall look on this city of Argos the very last time.
20 E.g. Il. 2.811; 6.152; Od. 4.844–847; 9.508–510; 15.417–418; 20.287–288 and de Jong
(2001: ad 3.293–296).
Lady Athena, you come out, and meanwhile I shall speak

to these women; the tale is others’, not mine.
Girls, … (h. 5.51–57)
Here the narrator first addresses the men of Argos, for whom the tale’s
warning seems to be intended, then the goddess, whose appearance
is eagerly awaited, and then the women who are present at the ritual
and must be entertained with a story to shorten the time of waiting.
Because in the rest of the hymn only the women are addressed, one
may assume that the men, though part of the fictional Argive world of
the poem, were not supposed to be present at the ritual scene. At the
beginning of the story of Erysichthon in h. 6.22, too, a wider audience
may be implied by the impersonal phrase (‘so that one may avoid
transgression’),21 but this is not certain and ‘one’ could also apply to
the women who are the primary narratees. In h. 2 the situation is more
complex as there are several addressees within the text: the sinners who
must leave the scene in 2, an anonymous eye-witness in 4, the bolts of
the doors which must open of their own accord in 6–7, the young men
who must sing in 8, the people who must be silent in 17, the people
who must cry hiē hiē in 25, and finally Apollo in 65–79. As with the
narrators of this hymn, it is hard to pin down a specific narratee and
in fact there seems to be a deliberate merging of addressees, including
even the doors of the temple, so as to create an impression of general
involvement of all concerned.
The most elaborate characterization of narratees is given in Iambi
fr. 191.26–30, where ‘Hipponax’ describes how the philologists swarm
around him and orders them to be silent, and in 33, where he tells one
of the philologists not to turn up his nose at him. As far as we can
see, there is no breaking of the dramatic illusion in this poem, but it
has been observed that the philologists of Alexandria would include the
historical Callimachus,22 and readers of Callimachus’ Iambi could enjoy
the notion that Callimachus had obeyed Hipponax’ instruction to write
down his tale (31).
In the Aetia we have no signs of primary narratees in Aetia 1–2, where
the framework implies that the emphasis is on the Muses and ‘the
young Callimachus’ as secondary narrators and narratees. In Aetia 3–
4, however, there are some instances of primary narrators addressing or
referring to a primary narratee, e.g. fr. 64.5–6, an instance of the ‘indef-
21 Translations from h. 6 are from Hopkinson 1984.

22 See D’Alessio 1996: 1, 9.
inite second person’ device (‘if you have heard of a certain Phoenix’), fr.
75.13–14, where a statement in the first person plural about ‘the illness
we send to the wild goats’ involves both narrator and narratees, and fr.
75.48–49, where the narrator calls to witness those who have experi-
ence of Eros, in order to support his view of Acontius’ delight in his
wedding-night, which could be considered a variant of the ‘anonymous
witness’ device. These passages seem to appeal to the narratees’ erudi-
tion23 and knowledge of the world, and the reader of the poem may feel
that this invitation to cooperate includes himself.
A special case is found in the Victoria Berenices:
Let him find out for himself and cut away some of the poem’s length,
but what he said to him in answer to his questions, that I shall tell:
‘Old father, the other matters you will hear when we are at dinner,
but now you will hear what Pallas [said to me?] …’ (SH 264.1–4)
Here the primary narrator promises to tell what Heracles—who has
just returned from killing the Nemean lion—said to Molorcus, but
invites the primary narratee to find certain information (presumably
about the story of Heracles killing the Nemean lion)24 for himself ‘so
that he may shorten the poem’, and Heracles neatly cooperates with
the narrator, because he too does not tell Molorcus the story of the
killing. Thus the primary narratee, though present in the text, is here
temporarily refused the role of ‘narratee’, but, instead, is asked to
lighten the task of the narrator.
In all these texts the question of the narratees is complicated by
the fact that Callimachean narrators are much given to apostrophe
the gods (e.g. fr. 18.6; 67.5–6), themselves (e.g. fr. 75.4–9, reminiscent
of Pindar [→]), or the characters in their stories. In the last case
the apostrophes may even give rise to long sections of second-person
narration, as in e.g. h. 4 (Delos), fr. 23 and 24 (Heracles) or fr. 80
(Pieria).25 In such passages the primary or secondary narratees are
reduced to accidental listeners.
Sometimes secondary narratees are visible too, and they may be
intended to steer our reception of embedded narratives. In the Aetia
23 On Callimachus’ technique of appealing implicitly to the reader’s cooperation in
decoding his learned texts as a captatio benevolentiae, see now Schmitz 1999: 152–178.
24 For a discussion of the question what was left out see Fuhrer 1992: 71–75. There is
a similar case of omitting a central fact of the story, i.e. the cleansing of Augeas’ stables,
in Theocritus 25 (→).
25 A number of other aitia in Aetia 3–4, of which only the beginning is preserved,
begin with an address, e.g. fr. 90 ‘There, Abdera, where now … leads … the scapegoat’,
there are some examples of this, as in fr. 7, where Calliope addresses

‘the young Callimachus’ as follows:
First of all make yourself remember Aegletes and Anaphe,
the neighbour of Laconian Thera, and the Minyans,
beginning with the time when the heroes sailed from Cytaean Aeetes
back to ancient Haemonia. (fr. 7.23–26)
Here the secondary narratee seems to be invited to cooperate with the

narrator: he must activate his memory and find the right starting-point
for the story that will explain the scurrilous ritual for Apollo at Anaphe.
This technique recalls that of the activation of the primary narratee in
SH 264.
Two instances of more elaborate introductions of aitia seem to serve
a similar purpose of steering the reception of the primary narratee.
The first is fr. 43.28–55, where a question of ‘the young Callimachus’
includes a long catalogue of well-known foundation stories of Sicil-
ian cities, which serves to draw attention to the unusual and obscure
nature of the ritual at Zancle (a founder’s cult in which the founder
is not called by his name) and to the speaker’s urgent need for expla-
nation. The primary narratee may thus be steered towards curiosity
and grateful acceptance of the story when it is finally told by Clio to
‘the young Callimachus’ as secondary narratee (fr. 43.56–83). The sec-
ond is fr. 178.1–34, where the primary narrator, ‘Callimachus’ (who will
become the secondary narratee), tells how he has met Theogenes at a
symposium, how pleased he was that they shared a preference for talk
over much drink, and how he asked him, as a native of Icus, about the
ritual for Peleus on that island. This long introduction of the story of
Theogenes again seems to prepare the primary narratee for something
special, as it invites him to share the secondary narratee’s curiosity as
well as his faith in Theogenes as a narrator. Furthermore it may serve
to remind the reader of the worth of stories, which surpasses that of
the transient material joys of the symposium (for a similar notion cf. fr.
43.12–17).
In conclusion one may say that the Callimachean narratees are
presented as active participants in the process of storytelling, either as
or a question, like fr. 79 ‘Why do they call …?’; in frr. 84–85 such a beginning
introduces a story told as second-person narrative, but in the other cases the evidence
allows no conclusions as to the narrative situation. See further Harder 1998: 109.
In contrast with Callimachus, Apollonius of Rhodes (→) does not address the main
characters of his stories.
the respectful audience of a hymnic singer or a fierce iambic poet or as

the learned and curious readers of a scholarly work.
Narratives
Callimachus’ narratives are often moralizing (as in hh. 5 and 6; Ia.

fr. 191) or aimed at the explanation of present situations or rituals (as
in h. 4 and the Aetia). In hh. 1, 2, and 3 they contain a series of episodes
from the career of the gods celebrated in those hymns.
The narratives are presented as subsequent narration and the way in
which the stories begin sometimes helps the narratees quickly to get a
picture of the main issues, locations and characters. In hh. 5.51–56 and
6.22–23 the moralizing purpose is explained beforehand and then the
actual narratives begin with an introduction of the situation (h. 5.57–
59: Athena’s love for Chariclo, the mother of Tiresias, who will suffer
a great deal when the goddess punishes her son) or setting of the story
(h. 6.24–30: a description of Demeter’s grove, which will be destroy-
ed by the evil Erysichthon). In a similar way in h. 4 the need to sing
about Delos (1–10) is followed by a description of the island (11–18),
which leads to the story of how Delos, formerly called Asteria, became
a ‘fixed’ island, because it allowed Leto to give birth to Apollo on its
soil (28–274). In the episodes from the gods’ careers that we find in
hh. 1, 2, and 3 the narratives are presented in a less formal manner:
in hh. 1 and 3 there is a chronological scheme, starting at the birth
and early youth of Zeus and Artemis respectively and following their
career and the acquisition of their attributes (the first stage of which,
h. 3.4–40, is presented in the form of a dialogue between Zeus and
Artemis). However, whereas h. 1 has a strong focus on Zeus’s career
and his establishment as the king of the gods, which prepares the
narratees for the praise of his human equivalent Ptolemy, h. 3 briefly
refers to a number of other narratives about Artemis, such as e.g.
the stories of Britomartis (189–207), Agamemnon (228–232), Lygdamis
(251–258) or Oineus (260–262). Thus the narratees of h. 3 are referred
to a wealth of other material, which could be adduced in order to
enhance the goddess’s fame even further. In h. 2 the episodes are
narrated without regard to chronology: 47–54 relate Apollo’s stay with
his beloved Admetus, 55–64 his building activities as a four-year-old
child, 65–96 his role in the foundation of Cyrene, 97–104 his early
defeat of the serpent at Delphi, and 105–112 a brief dialogue with
Momus in which he appears as the authoritative god of poetry. The

lack of ‘order’ in the arrangement of the narratives may be related to
the idea of a ‘non-continuous’ poem in the Aetia-prologue (fr. 1.3).
There are two instances of narratives that are not in the past tense.
In h. 3.142–169 the traditional episode of the young god’s first arrival
on Olympus is replaced by a description in the present tense of the
way in which Artemis is ‘now’ (145) received on Olympus, whenever
she arrives there with her hunting-spoils, eagerly awaited by the ever-
hungry Heracles; we are dealing with the simultaneous iterative narra-
tion also found in Hesiod (→). In h. 4.166–195 the unborn Apollo as a
secondary narrator tells Leto about the war with the Celts which will
take place in the future (but which is already the past for the primary
narratees).
Patterns of introductions and beginnings being tuned to each other
can also be observed in the Aetia. In Aetia 1–2 the questions which pre-
cede the narratives inform the reader beforehand of the story’s sub-
ject and purpose, as in fr. 7.19–20 ‘and why, goddesses, do the peo-
ple of Anaphe sacrifice with shameful words and those of Lindus with
words of bad omen …?’ and in the motivation at the end of the
extended question in fr. 43.54–55 ‘… for in none of these towns does
the man who once built its wall go to the customary feast without
being called by his name’ (as in Zancle, the anonymous founder-cult
of which had raised the speaker’s curiosity). Then the narratives begin
with a quick survey of the main characters and locations, as in fr. 7.23–
26 (quoted above) and fr. 43.58–60 ‘people from Cumae and Chalcis,
led by Perieres and the arrogant Crataemenes, went to Trinacria …’
In Aetia 3–4 some of the narratives seem to have a similar indication
of the contents at the beginning, as they start with requests or ques-
tions (e.g. frr. 76 and 79), while other narratives start without such an
introduction, but with a brief survey of the main facts, as in fr. 67.1–8:26
Eros himself taught Acontius his art, when
the boy burned with love for the beautiful girl Cydippe,
– for he was not very clever—, in order that he would be called
by the name of lawful husband all through his life (?),
because, lord, he came from Iulis, she from Naxos,
Cynthian, to your Delian sacrifice of oxen,
he a descendant of Euxantius, she of Prometheus,
both beautiful stars of the islands.
26 Similarly e.g. fr. 84 ‘when you came from Pisa, Euthycles, having defeated men
…’ about the Olympian victor Euthycles.

This introduction was followed by an elaborate description of the

beauty of Cydippe and Acontius and its effects on the people around
them (fr. 67.10–15 [?]; frr. 68 and 69); it is not clear how many lines
were occupied by this and at which point the description of the events
at Delos began.
In Ia. fr. 191.26–30 (which follow the lacunose 12–25) the description
of the unruly crowd of philologists may suggest the moral of the fol-
lowing tale, and the fact that the ancient summary of the poem27 states
explicitly that ‘Hipponax’ told them not to be jealous of each other,
suggests that something to this effect has been in the text. The narrative
then begins with a description of the rich and fortunate Bathycles, who
enabled the Seven Sages to show their generosity towards each other.
The Hecale starts with a simple description of the situation of Hecale in
fr. 230 (= 1 Hollis) (quoted above).
The narratives’ endings often refer back to their beginning so that a
sense of closure is achieved; thus h. 4.273 mentions the fact that Delos is
now a ‘fixed’ island; h. 5.131–133 draws attention once again to Athena’s
power; h. 6.116–117 finishes the story of Erysichthon with a moralizing
remark. In the last two hymns the end of the narrative is also marked
by a return to the ritual scene (hh. 5.137–142; 6.118–138). In the Aetia,
some ends of narratives are preserved and show varying markers of
closure. In fr. 43.78–79 the explanation is given for the anonymity of the
founder-cult at Zancle, and thus the end of the narrative refers back to
the questions at the beginning. In some other aitia there are traces of
hymnic closure, as in fr. 7.13–14, which contain a farewell and request
for lasting fame to the Charites, who were the subject of the preceding
narrative (frr. 3–7), and fr. 23.19–20, which is part of a farewell to
Heracles, whose adventures with a Lindian farmer had been the subject
of frr. 22–23. The story preceding that about Zancle concluded with a
short programmatic passage about the worth of stories surpassing that
of the ephemeral pleasures of the symposium (fr. 43.12–17). The story
of Acontius and Cydippe seems to end twice: in fr. 75.50–52, where
Acontius’ offspring is mentioned, the love-story ends, but it is followed
by an extensive summary of the narrator’s source, the Cean history of
Xenomedes of Ceos, which ends with a reference to the ‘boy’ Acontius
in fr. 75.77,28 thereby recalling the narrative’s beginning in fr. 67.2. The
27 Printed in Pfeiffer 1949: 163.

28 The way in which Acontius is here called a ‘boy’ also suggests that the narrator
has been dissociating himself gradually from his involvement with his main character:
summary may be read as an elaborate footnote providing background

information about Acontius and his native Ceos.29
The narrative of the cup of Bathycles may end in Ia. fr. 191.76–84,
where the cup comes back to Thales, to whom it was given in the
first place, but the text is too fragmentary to be certain. The end of
the Hecale is lost, but the ancient summary (printed in Hollis 1990: 65)
indicates that it ended with an aition.
The way in which the narratives are presented often shows a cer-
tain concern with issues of communication. Generally speaking, Calli-
machus’ narratives are part of poems which were of a literate rather
than an oral nature,30 but within these poems the oral transmission of
stories is frequently enacted and the narrator poses as an oral narra-
tor (the ‘feigned orality’ motif), as in hh. 5 and 6, where the narratives
are presented as orally performed in front of an audience. Sometimes
Callimachus even seems to create a certain tension between the real-
ity of written communication and the fiction of oral communication
within his poems. Thus in Aetia 1–2 the fiction of a dialogue between
the Muses and ‘young Callimachus’ recalls the tradition of early oral
epic, which at the same time is being modified because ‘the old Calli-
machus’ has written it all down. The same applies to fr. 178, where the
oral convention of storytelling at a symposium is presented in writing.
In the Victoria Berenices (SH 254–268), the encomiastic effects of a promi-
nent position in a written work are combined with an evocation of oral
praise, as the poem derives part of its impact from its position at the
beginning of the third book, but at the same time recalls the Pindaric
epinician with its convention of oral performance.31 Similarly, in Ia.
fr. 191, the poem evokes an oral situation, but in 31 ‘Let there be silence,
and write down my tale’, oral and written transmission are linked and
the poem as a whole seems to have been placed emphatically at the
in frr. 75, 40, and 44, where Acontius finally wins Cydippe and enjoys his wedding
night, the narrator apostrophizes him as ‘Acontius’, in 53 and 74 just before and after
the summary of Xenomedes he addresses Acontius as ‘Cean’, as if drawing attention to
his public position as ancestor of the ruling family at Ceos, and in 77 the aition ends, as
it began, with a third- person reference to Acontius as a boy.
29 See further on this passage Harder 1998: 103–104
30 For a general discussion see Bing 1988. This view has recently been challenged
by Cameron 1995, but his arguments for performance of Callimachus’ poems are not
compelling and do not detract from the primarily literate nature of the poems.
31 For a detailed comparison of the Victoria Berenices and the conventions of the
Pindaric epinician see Fuhrer 1992: 86–134.

opening of the collection of Iambi. Again we may observe an ironic

distance between poet and narrator, reader and narratee.
The status of the stories about the gods in the hymns is traditional
and seems to affect the way in which they are introduced, as in h.
2.30–31 where a large amount of material for songs about Apollo is
indicated; h. 5.56, where the story of Tiresias is said to be derived from
others; h. 4.28–29, where the story of Delos is presented as one of a
large corpus of songs; and h. 6.18–23, where one of many stories about
Demeter is chosen. In the last two instances the great diversity of the
tradition is also emphasized.
The stories in the Aetia range from early myth (e.g. the story of
the Argonauts in frr. 7–21 or the stories of Heracles in frr. 22–25), via
the archaic and classical period of Greek history (e.g. the colonization
stories in fr. 43,32 the love-story in frr. 67–75, or the story of the tomb
of Simonides in fr. 64), to the Ptolemaic court in third-century BC
Alexandria (as in fr. 110). Here too we have evidence that the status
of these stories could affect the ways in which they are accounted for:
in fr. 7.19–22 stories about the Argonauts and Heracles are attributed
to the Muses; in fr. 75.54–77 the story of Acontius is attributed to the
historian Xenomedes of Ceos; while in fr. 110 the Ptolemaic fiction
of Berenice’s lock turned into a star is told by the lock itself which
adduces the authority of the astronomer Conon (fr. 110.7–8). In fr.
43.28–55 the catalogue of Sicilian cities and the subsequent question
creates the impression that the young scholar-poet has learned a great
deal from his researches, but the fact that in the end only the Muses are
truly omniscient is brought to our attention when they have to supply
additional information about the ritual at Zancle.
Conclusion
The narrators and narratees in Callimachus’ texts are overt and self-
conscious, highly interested in narratives, and hence the process of
storytelling and the transmission and reception of knowledge receive
full attention. Important matters are the role, credibility and motivation
of the narrator, the cooperation of the narratees at various levels, the
steering of their reception of the stories, and issues of communication
32 Similarly the anecdote in Ia. fr. 191.32–77.

(including the notion of feigned orality) and the status of the narratives.
The character of Callimachus’ narrators, narratees and narratives well
reflects his position as a scholar-poet in third-century BC Alexandria
and his need to carve out a position for himself with respect to the long
literary tradition of which he forms part.
chapter six
THEOCRITUS AND MOSCHUS
R. Hunter
The Theocritean corpus, as it appears in modern editions, does not

reproduce ancient collections of bucolic poetry,1 and may cover as
much as five centuries of Greek poetry. Although there is no good rea-
son to doubt that the majority of poems are indeed the work of The-
ocritus of Syracuse in the first half of the third century BC, the corpus
offers us the best available glimpse into the workshop of poetic forms
which Hellenistic poets inherited and fashioned. Narrative is at the
heart of this variety. The corpus contains narrative poems which use
traditional modes inherited from hexameter epic and hymns, embed-
ded narratives by the primary narrator which depend upon a framing
dialogue between primary narrator and addressee, embedded narra-
tives by secondary narrators, and poems which are wholly ‘mimetic’
and dialogic; as in Plato (→), this last category may be thought to pre-
suppose a suppressed primary narrator, particularly if the poems are
read rather than ‘performed’.
If all the poems of the corpus are thus, in some sense, ‘narratives’,
it is nevertheless the case that, for example, the meetings of Battus and
Corydon in Idyll 4 and of Comatas and Lacon in Idyll 5 are not nor-
mally thought of as ‘narrative poems’, though both of course allude to
past and current ‘narrative situations’. To label such poems ‘narratives’
is not, however, merely a terminological subterfuge; rather it is a help-
ful way of recording an important literary fact. The genuine bucolic
poems of Theocritus inscribe within themselves a sense of tradition, of
an already known world of rustics engaged in hexameter song and dis-
putation, and hence of a creator of that artificial ‘natural’ world whom
we may in different contexts call ‘narrator’, ‘poet’, ‘author’. The artifice
of metre, dialect, and poetic allusion never allows us to imagine that we
have unmediated access to a rural reality, and that sense of distance is
1 Cf. Gutzwiller 1996.

84 part one – chapter six
in part our consciousness of the poetic presence of a creative (though

sometimes suppressed) narrator. In post-Theocritean bucolic, our sense
of familiarity, of stepping into a textual world frozen since we last left it,
is more overtly aroused by poets who pay homage to ‘the founder’ by
imitation of, and variation from, the base which his usage had autho-
rized and by constant allusion to Theocritean situations and characters;
bucolic is thus a remarkable instance of the very rapid invention of tra-
dition.
External narrators
External narrators appear throughout the corpus.2 Idyll 18 reports the

performance at Sparta of a maiden choir celebrating the wedding of
Helen and Menelaus. Though formally a narrative, and one in which
the narrator is overt in the manner with which the poem begins (‘Once
then [en pok’ ara]3 in Sparta at the palace of golden-haired Menelaus
…’), this poem is very largely devoted to the wedding-hymn (vv. 9–58);
there is no return to the narrative frame at the end. Such a structure
has analogues in earlier narrative forms as widely separated as choral
lyric (cf. Bacchylides 20, another Spartan wedding-song, Bacchylides
5.56–175, the story of Heracles in the Underworld)4 and Platonic dia-
logue (the Symposium), though it is clearly the former that is evoked here
(cf. esp. vv. 7–8). In recreating a lyric form and telling of a mythical
event, Theocritus relies, as do the lyric poets themselves, on an audi-
ence able to contextualize the narrated moment within a larger, and
in this case very famous, story. In the case of Idyll 6, however, a rather
similar form is used for an apparently less familiar occasion:
One day, Aratus, Daphnis the cowherd and Damoetas
Gathered their herd into one place. Damoetas’ chin
2 Idyll 23, which is certainly post-Theocritean, is the completely uncontextualized
story of a hopeless love: ‘a passionate man loved a cruel youth …’ The central section
of the poem is given over to the final speech of the ill-fated lover, but the guiding
presence of the narrator is highly visible through the use of emotive and judgemental
language (e.g. vv. 1–2, 60) and the repeated ‘moral’ of the story (the power of Eros,
vv. 4–5, 63). The poem is also unusual in that no motive is given for this non-mythical
narration: the most ‘natural’ context for it would, in fact, be as part of an attempt by
the narrator to win over his own beloved. Cf. further Hunter 2002.
3 For ‘then’ (ara) as marking the narrator’s choice of where to begin cf. 22.27, Hunter
1996: 149–150.
4 Both Bacchylidean narratives begin with pote, ‘once upon a time …’
r. hunter – theocritus and moschus 85
Was red with down, while the other’s beard was just coming.
It was noon in summer, and the two sat down by a spring.
This was their song. Daphnis had made the challenge, so he began.
(6.1–5, trans. Verity)5
The bulk of the poem is devoted to the exchange of songs between

Daphnis (impersonating, as an internal secondary narrator, a ‘friend’ of
the lovesick Cyclops) and Damoetas (responding as the Cyclops him-
self); the songs are divided by a verse from the external primary nar-
rator (v. 20) and a five-verse conclusion, corresponding to the open-
ing, seals a frame around the poem.6 Critics dispute as to whether the
Daphnis of this poem is the ‘mythical’ Daphnis of (for example) Idyll
1 or rather to be understood as a contemporary herdsman taking his
name from the legendary figure. In either case, however, the abrupt
beginning of the narrator, here made overt by the address to Aratus,
exploits the narratees’ foreknowledge and draws it into collusive plea-
sure; whereas, however, in Idyll 18 our pleasure derives from our famil-
iarity with the imagined mythical world which the narrator conjures
up for us, in Idyll 6 it derives from our familiarity with the world of
bucolic poetry: we do not need to be told about ‘bucolic song contests’
because we are entering a literary world created as timeless and ever-
present.
The narration of the grim fate of Pentheus in Idyll 26 also begins
in mediis rebus (with, on this occasion, no inceptive particles), ‘Ino and
Autonoe and Agaue of the white cheeks, three themselves, led three
thiasoi to the mountain’, and the narrative ends by reversing the move-
ment of the opening verses and sealing the story with a closural pun in
two highly alliterative verses: ‘they came to Thebes, all splattered with
blood, bringing from the mountain not Pentheus but grief (penthēma kai
ou Penthēa)’, vv. 25–26. The pun is a marker of the narrator’s arrange-
ment of his material, but his overt presence in this opening section of
the poem had already been signalled in a number of ways. First, there is
the manner in which the women’s ritually correct practice is described
(vv. 8–9) and the explanatory glosses of v. 11 (‘an ancient mastich-bush,
a plant growing in the area’) and vv. 13–14 (‘the holy things of fren-
zied Bacchus, which the uninitiated do not see’ [though the omniscient
5 Cf. Verity-Hunter 2002.

6 Idyll 27, the dialogue between ‘Daphnis’ and a girl he is seducing, has a similar
closural frame, apparently by an external primary narrator (vv. 67–72), but the opening
of the poem is lost, so we cannot say more.
narrator may do]),7 which suggest a contemporary significance for this

narrative set in the distant mythic past; secondly, the violent language
with which the dismemberment of Pentheus is described (‘they divided
out the rest of his flesh’) reveals the narrator making lexical choices
among possible modes of expression, just as verbal echoes of Euripi-
des’ Bacchae8 reveal the learned narrator as a negotiator between texts.
Nevertheless, despite these clear signs, what follows the narration still
comes with a shocking suddenness:
I do not care. Nor let anyone else have a thought for one hated by
Dionysus, not even if he suffer more grievous things than this and be
in his ninth year or even entering his tenth. May I myself be holy and
pleasing to the holy; thus has the eagle honour from aegis-bearing Zeus.
A better fate attends the children of the pious, not of the impious.
Farewell to Dionysus, whom Zeus most high set down on snowy Dra-
canus when he had unbound his great thigh. Farewell to lovely Semele
and her sisters, Cadmean women honoured by many heroines, who at
the instigation of Dionysus, performed this deed in which there is no
fault: let no one object to the things of the gods. (26.27–38)
Whatever the import of these mysterious (and textually very difficult)
verses,9 the sudden first-person revelation of the narrator10 and the
hymnic farewell to Dionysus confer upon the narrative that has pre-
ceded an aetiological and exemplary status.
Idyll 24, the story of the infant Heracles strangling the snakes and
of Tiresias’ prophecy of his future greatness, seems also to have been
structured as ‘once upon a time … (poka)’ narrative, which is then
concluded by a first-person hymnic envoi in which the narrator asks
Heracles to grant him ‘victory’; unfortunately, the end survives only in
the most tattered scraps, and we can say nothing in detail about how
the narrative may have been contextualized at its conclusion.11 In this
7 The power of writing to evoke mental images within us (enargeia, phantasia) is
obviously relevant here, where such power is seen to promote an illicit and voyeuristic
desire. A related phenomenon is the experience and fate of Tiresias in Callimachus’
Hymn to Athena, another poem in which our emotional ambivalence about a narrated
divine tale is set against the demand for piety.
8 Cf. Cairns 1992: 5–9.
9 Some reference to the story of the daughters of Minyas is not improbable, cf. Gow
ad loc.; White 1999: 54–56.

10 Cairns 1992 argues that this is not the narrating poet, but the ‘choir’ that is
imagined to be performing. If the eagle of v. 31 evokes what ancient readers at least

took to be a familiar image for the poet in Pindar, this would certainly reinforce the
strength of the first-person intrusion (cf. Cairns 1992: 22–23).
11 Cf. Gow 1952: II 436.
poem, the presence of a learnedly innovative narrator is most strikingly

revealed (before the conclusion) by a virtuoso handling of traditional
epic technique (see esp. the ‘waking and dressing’ scene of vv. 34–53)
and by an elaborate textuality which reworks scenes of (particularly)
Homer and Pindar at the level of close detail: the narratee of this poem
is closely familiar with earlier literature.12 So too, the concentration
upon homely detail and the witty domestication of epic characters,
whose effect depends crucially upon our acquaintance with what is
‘normal’ in hexameter narrative, focus our attention upon the narration
as a literary act and hence upon the narrator as a poiētēs. Not for a
moment can we imagine that we are being offered unmediated access
to ‘narrative fact’.
The act of narration itself is also a focus of attention in Idyll 22, the
Hymn to the Dioscuri, but here the principal technique is the juxtaposition
of contrasting narrative styles in the telling of two ‘duel’ stories, a jux-
taposition which foregrounds the choices open to every self-conscious
narrator.13 Unlike the poems we have just considered, Idyll 22 opens
with an extended first-person introduction (‘We hymn the two sons of
Leda and aegis-bearing Zeus …’), and then, in a surprise variation on
the traditional epic and hymnic ‘aporia’ motif, ‘Where shall I start?’ (cf.
Idyll 17.11), the narrator reveals that he intends to celebrate the brothers
separately:
O pair of helpers for mortals, beloved pair, horsemen, lyre-players, ath-
letes, singers, with Castor or with Polydeuces first shall I begin to sing?
In hymning both, I will sing of Polydeuces first. (vv. 23–26)
There is an obvious danger for any mortal in choosing one brother
ahead of the other, so the narrator takes the evasive steps of stressing
their unbreakable ‘twoness’ (vv. 23–24) and implicating the twin gods
themselves in the chosen order: whether or not we imagine a ‘pause’
after v. 25, v. 26, in which the choice is made, is to be understood to be
spoken ‘under the inspiration of ’ the Dioscuri themselves.14
The story of Polydeuces’ boxing victory over the brutish Amycus
(vv. 27–134) is a virtuoso example of narrative inventiveness—a locus
amoenus, the ekphrasis of a man who resembles a work of art, (an unpar-
alleled) stichomythia and no other direct speech, extensive rewriting of
Homer and (probably) Apollonius of Rhodes, and an archly expressed
12 Cf. Hunter 1996: 11–13; Fantuzzi-Hunter 2002: 275–286, 344–359.

13 Cf. Hunter 1996: 58–59; Sens 1997: 14–15.
14 The ordering is therefore not (represented as) ‘arbitrary’, pace Sens 1997: 95.
claim of complete dependence upon the knowledge of the goddess

Muse (vv. 115–117), immediately before the most elaborate and ‘tech-
nical’ boxing description of the whole poem,15 all play their part. Cas-
tor’s killing of the hapless Lynceus (vv. 137–211) is also narrated by an
external primary narrator, but the contrast of tone with the Polydeuces
narrative could not be more marked—the apparent moral certainties
of the latter (Amycus is a ‘gluttonous’ [v. 115] bully) give way to the
ambiguities which attend recognition of man’s weakness in the face of
the divine; the hymnic and encomiastic apostrophes to the god of the
earlier narration (vv. 85, 131–132) are not repeated in the bleak tale of
the death of Lynceus and his brother. There is also a clear contrast of
structural form. It is debated whether the central section of the second
narrative is devoted to a speech by Lynceus or to a pair of pre-fight
speeches (à la Homer) by Lynceus and Castor,16 but in either case the
form is markedly more traditional than the poikilia of the Polydeuces
narrative.
One remarkable example of narrative experimentation is Idyll 25,
a probably non-Theocritean poem, which presents three ‘scenes’, all
apparently told by a primary external narrator, from the story of Her-
acles and King Augeas.17 The poem opens in mediis rebus, ‘Him did the
old man address …’, with an abruptness (and an initial de) that have
been taken by most readers as signs that the opening of the poem is lost.
In the first section (vv. 1–84) the newly arrived hero learns from a hum-
ble rustic about the king and his estates; in the second section (vv. 85–
152) Heracles and the king admire the extraordinary royal herds, and
in the third (vv. 153–281) Heracles, as an internal secondary narrator,
narrates the story of the Nemean lion to the king’s son, Phyleus, as the
two of them leave the royal property. The narrative form of the scenes,
which together form a crescendo of confrontations between Heracles and
animals (dogs, a marvellous bull, and finally the Nemean lion), is care-
fully varied. In the first, Heracles asks questions and receives a long and
full answer; in the second, there is no direct speech, and in the third
Heracles’ narration is characterized by the full panoply of epic devices,
including similes, while it remains bound by the non-omniscience of an
internal narrator (cf. 195–200).
15 Cf. Gow on 119ff.; Zanker 1987: 87.

16 For the arguments cf., e.g. Sens 1997: 190–191.
17 On this poem cf. Kurz 1982; Zanker 1996; Hunter 1998b; Fantuzzi-Hunter 2002:
286–291.
The narrative form in the third scene is of a very common type.

Phyleus suspects that his travelling partner is the great hero of whom
he has heard (from an anonymous tertiary narrator, vv. 164–173) and
he asks the stranger to satisfy his curiosity: ‘Come tell me then first,
so that I may know in my mind … and also tell me how you killed
that dreadful beast …’ (vv. 177–182). The narratee’s thirst to know is
perhaps the most common of all motives for narration, and in return
Phyleus receives from Heracles as full an account as he can give (‘I
will give you a detailed account of what happened concerning this
monster, since you desire to hear it …’, vv. 195–196). Heracles’ narrative
is also marked by strong final closure which clearly demarcates the
‘boundaries of narrative’: ‘This then, my friend, was the death of the
beast of Nemea …’ (v. 280).
The very clarity of narrative form and content, however, throws into
relief the novelty of the poem as a whole. The three scenes can be
fitted, easily enough, into the outlines of the story of Heracles and
Augeas familiar from our main sources (Apollodorus 2.5.5, Pausanias
5.1.9–10), but the central ‘fact’ of that story (the cleaning of the stables)
is never mentioned (contrast the full narrative of the killing of the
Nemean lion), and any unifying narrative is left entirely implied, to
be supplied, with whatever level of detail, by each narratee. This is
a much more radical technique than, say, the typical lyric practice of
allusive narrative. There is in fact no real parallel in Greek narrative
poetry for such a chronologically linear account in which we are merely
given excerpts from ‘the full story’, each of which is, however, itself
detailed and coherent; this exploration of narrative continuity and
disjunction foregrounds the role of the poet, the creating intelligence
which turns ‘events’ into narratives. Mimetic and diegetic forms were
drawing closer together in certain areas of poetic practice, just as their
distinction was being theorized and hardened in critical discussion.
A different kind of narrative exploration is on show in Idyll 13. The
poem begins with a personal address to Nicias of Miletus, a figure who
occurs elsewhere in the corpus as a friend of the poet:
Love was not born for us alone, as once we thought,
Nicias, whichever god it was who fathered him.
We were not the first to be beguiled by beauty,
We who are mortal, and cannot see tomorrow. (13.1–4, trans. Verity)
The exemplification of this sentiment is the story, told by the same

poetic voice but now acting as an external primary narrator, of the
loss of Heracles and his beloved squire Hylas from the Argonautic
expedition. There is no return to Nicias at the end of the poem, which
arrives at its destination (‘Phasis’, the river of Colchis, is the last word,
v. 75) at the same time as does Heracles. In this it is contrasted with the
otherwise apparently similar Idyll 11, where another gnomic thought
addressed to Nicias (that ‘the Muses’ are the only remedy for love) is
illustrated by the story of the lovesick Cyclops and, for the bulk of the
poem (vv. 19–79), the song he sings (as an internal secondary narrator)
to the beloved Galatea. In the final two verses of Idyll 11 the primary
narrator returns with a teasing reference to Nicias’ medical profession:
So by singing the Cyclops shepherded his love,
And more relief it brought him than paying a large fee.
(11.80–81, trans. Verity)
In Idyll 13, Theocritus’ highly descriptive version of the Hylas story
very likely reacts to the epic version of Apollonius of Rhodes,18 and the
use of a personal narratee (Nicias), in place of the traditional absence
of specified primary narratees for epic narrative, as in Homer (→), is
part of the transformation of epic narrative material into a different
poetic mode. So too, the primary narrator reveals himself at work in
covering the Argonautic journey to the Phasis twice over: first in a rapid
survey of ‘the bare facts’ (vv. 16–24) and then again at a more leisurely
pace (vv. 25–75), with the loss of Heracles and Hylas to the expedition
occupying the centre of attention (vv. 36–75).
Internal narrators
Narratives by an internal narrator are common in Greek drama, and

two of Theocritus’ poems with strong links to the tradition of mime,
offer internal narrators who tell their stories within a quasi-dramatic
structure. In Idyll 14 Aeschinas tells his friend Thyonichus the story
of his unhappy affair with a girlfriend, Cynisca, whose affections have
now gone elsewhere. This vivid tale of a disrupted drinking-party may
be related to on-stage narratives of elaborate banquets, which seem to
have enjoyed remarkable popularity in the comedy of the fourth and
third centuries.19 Aeschinas’ unbroken narrative occupies the greater
18 The matter is very disputed, and detailed arguments cannot be rehearsed here;
much bibliography is readily available in Köhnken 2001.

19 Cf. Fraenkel 1912: 9–32.
part of the poem (vv. 12–56), but variety of tone and effect is achieved
by the quotation of direct and colloquially lively speech in the mouth of
more than one character (both male and female); so too, the external
narratee, Thyonichus, is drawn directly into the narrative (‘Then I
hit her on the temple, once, then again—you know me, Thyonichus’,
vv. 34–35). Moreover, at the conclusion of his speech, the telling of
past events is not neatly demarcated from deliberation with Thyonichus
about the future (vv. 50–56). All of these techniques are designed by the
poet to minimize the distinction between conversational exchange and
‘narrative’, understood in a narrow sense. The use of hexameters for
such a mimetic exchange creates a tension between form and content,
a tension that is advertised rather than concealed by the opening verses,
in which frequent use of verse-splitting (antilabē) sets the colloquial tone
for the whole poem. Theocritus thus accommodates his narrative form
to the traditions of popular mime, but in such a way that the formal
novelty of his undertaking remains visible. Such a procedure, which is a
form of literary history within creative poetry itself, finds many parallels
in the literature of the third century.
In Idyll 2 there is only one speaker, the internal primary narrator,
but the associations of the poem are again with the mime tradition. In
the first part of the poem a woman called Simaitha performs a magic
ritual, with the help (or hindrance) of a slave-girl, to try to win back
her faithless lover Delphis (vv. 1–63). When the slave goes to perform a
rite at Delphis’ door, Simaetha is now free to tell her story: ‘Now that I
am alone, from what point shall I bewail my love? At what point shall I
begin? Who brought this misery upon me?’ (vv. 64–65). There is a cer-
tain realism in only laying bare her soul once the slave is out of the way,
but the emphasis upon a connection between being alone and narra-
tion is part of Simaitha’s self-presentation as a heroine from drama and
‘literature’, for it is unhappy lovers and, above all, characters in drama20
who relate their stories to the heavenly bodies—the moon-goddess is in
fact the external narratee of Simaitha’s story, as the repeated refrain,
‘Take note, lady Selene, whence came my love’, makes clear. It is per-
haps not unfair to ask why Simaetha tells her story. On the one hand,
it is important that avenging divinities such as the moon-goddess, who
is intimately connected with magic, should be fully acquainted with
the wrong which has been done (cf., e.g. Electra’s lamentation to the
20 Cf. Plautus, Mercator 3–5 ‘I won’t do what I’ve seen love make others do in
comedies—tell their woes to Night or Day or the Sun or the Moon.’

stars in the parodos of Euripides’ Electra), but there is also a clear self-
positioning by Simaetha within what she conceives to be a familiar pat-
tern: Simaetha must narrate her story because (as she imagines) that is
what women in her position do. It was, above all, Euripidean drama
(→) that had licensed lengthy female narration and complaint. As in
Idyll 14, Theocritus exploits the tension between the formality and liter-
ary affiliations of unbroken poetic narrative and the relatively low status
of mime; in these poems ‘narration’ is denaturalized, so that it becomes
a specific literary form associated with the higher registers of poetic
expression. The sly acknowledgment of the presence of an audience (of
watchers or readers) in ‘Now that I am alone …’ reinforces this sense.
Thus, in adopting a self-conscious role fashioned by tradition, Simaetha
apes narrators as paradigmatic as Odysseus in the Odyssey in wondering
where to begin (the ‘aporia’ motif, for which cf. Od. 9.14, → Homer),
perhaps the most crucial choice facing any narrator.21
One Theocritean poem, however, is entirely devoted to the recollec-
tions of an internal primary narrator: this is the famous Thalysia (Idyll
7). The narrator, Simichidas, who introduces himself at first only as ‘I’,
recalls an occasion when he went with two friends from Cos town into
the countryside to join in the harvest festival of an old Coan family. On
the way they meet, ‘with the aid of the Muses’,22 a goatherd called Lyci-
das whom Simichidas invites to an exchange of ‘bucolic song’. Lycidas
sings a song about his passion for Ageanax, and in reply Simichidas
sings of the hopeless passion of his friend Aratus for a boy called Phili-
nus. Simichidas and his friends reach Phrasidamus’ farm where the
celebration takes place in a marvellous locus amoenus. The opening of
this extraordinary poem, ‘There was a time when I and Eukritos were
going from town towards the Haleis …’, finds its closest analogues in
the opening of certain of Plato’s dialogues (cf. Lysis, Republic), but the
poet and his very self-consciously ‘literary’ narrator23 seem specifically
to evoke the Phaedrus—a walk in the countryside in the heat of the day
and an exchange of performances designed to win over a beautiful boy;
it is as though the Phaedrus has been transposed into a narrative related
by Phaedrus.
21 On Simaetha as narrator cf. esp. Andrews 1996.

22 One implication of this is that the meeting will be memorialized in song, as of
course it was. For the limited role of the Muses in Theocritus’ bucolics cf. Fantuzzi
2000.
23 Cf. Hunter 2003.
Embedded narratives
Embedded narratives, by both external and internal (e.g. Idyll 5.41–42)

narrators, play important roles in the bucolic poems, and the founding
myth of the bucolic world is ‘the sufferings of Daphnis’ (algea Daphnidos)
which are sung in Idyll 1 by a shepherd called Thyrsis to an unnamed
goatherd for the price of a milking-goat and a marvellously decorated
wooden bowl; the exchange of narrative is here materialized in the
barter economy of humble rustics. The song itself, punctuated by the
folkloristic technique of refrains, is represented as a traditional one, or
perhaps merely on a traditional theme (cf. vv. 19, 23–24); the external
narrator is thus the bucolic equivalent of a very skilled rhapsode per-
forming a celebrated passage from Homer, and he proudly identifies
himself (‘I am Thyrsis of Etna, and sweet is the voice of Thyrsis’, v. 65),
not perhaps unlike the narrator of the Homeric Hymn to Apollo (→).
Thyrsis begins in medias res, with an indignant question to the
Nymphs about their whereabouts when Daphnis was wasting away.
The reasons for Daphnis’ condition and subsequent death are never
made explicit. Such an allusive narrative mode is familiar from the
victory odes of Pindar and Bacchylides (→) and choral narration in
Aeschylus (→) and Sophocles (→), but the mode itself becomes central
to the concerns of the passage in a way that might be thought typical
of Hellenistic self-consciousness:
First of all came Hermes from the mountain and said, ‘Daphnis, who
is tormenting you? For whom, my friend, have you conceived so great
a passion?’ … The cowherds came, the shepherds and the goatherds
came: all asked what misery had befallen him. Priapus came and said,
‘Poor Daphnis, why are you wasting away? In her search for you the girl
roams by every spring, through every glade …’ (vv. 77–83)
Here the different characters embody different levels of knowledge
and curiosity: Hermes, who seems to believe that Daphnis himself
is in love with an unknown person, the human rustics who know
only that Daphnis is suffering terribly, and Priapus who thinks he
knows the whole story—Daphnis is suffering for no real reason. This
dramatization of ignorance, desire to know and probably delusory self-
satisfaction at the head of the story is, in part, an anticipation of, and
hence stimulus to, the reception of the story by the primary narratees.
It also foregrounds, however, the strikingly oblique mode of a narrative
at the heart of which lies a bitter confrontation between the only two
characters who certainly do know the true situation, Aphrodite and
Daphnis himself; we the audience never learn ‘what has happened’,

except perhaps that Daphnis’ suffering and death may be viewed either
as a victory for erōs or a victory over erōs (vv. 97–98, 103). What matters
is what an audience or a community, whether of goatherds or readers,
makes of the ‘the idea of Daphnis’.24
Conclusion
The poetry of Theocritus and his imitators is important evidence for

a self-conscious concern in third-century and later Greek poetry with
narrative experimentation. Thus, Apollonius (→) explores (inter alia) the
rhythms of a long epic narrative and the limits of epic ‘repetitiveness’
and authorial discretion, how episodes of varying length play off against
recurrent thematic patterns; so too, in the Hymns, Hecale, and Aitia, Cal-
limachus (→) experiments with the inherited modes of form and voice
imposed by genre and metre, and seeks to recuperate within hexame-
ter and elegiac poetry some of the narrative techniques of archaic and
classical choral lyric, notably that of Pindar and the choruses of Attic
tragedy, which had long since died out. Particularly for scholar-poets
such as Apollonius and Callimachus, it is tempting to see in these devel-
opments the influence of, and a reaction to, Platonic and Aristotelian
theorizing about narrative form, as well as of the intensive study of
the great narrative works of the past which was carried out at Alexan-
dria and elsewhere. For Theocritus, however, two contemporary phe-
nomena are of particular importance. One is the dominance of (rela-
tively) short poetic forms, which worked against fullness of expression
and narrative explanation; this tendency reaches its peak in ‘narrative’
epigrams and epigrams consisting merely in snatches of conversation.
Secondly, the ever-increasing expectation of reception through reading
blurred the boundaries of, to use the traditional Platonic terminology,
‘mimesis’ and ‘diegesis’ and allowed poets to create potentially elabo-
rate instances of ‘feigned orality’ (cf. the addresses to Aratus in Idyll 6,
to Nicias in Idylls 11 and 13, and to Diophantus in the post-Theocritean
Idyll 21) and to produce a new range of ‘mixed’ narrative forms.
24 Cf. further Hunter 2003: 230–231.

The Europa of Moschus
The Europa of the Syracusan poet Moschus (mid-second century BC)

is the only surviving example of what was probably a common poetic
form in the later Hellenistic period: a self-contained mythic narrative
of relatively short compass (in the case of the Europa, one hundred
and sixty-six hexameters) in the voice of an external primary narrator.
No motive is given for the narration and there is no invocation of
the Muses: the poem is presented, in the in medias res mode already
familiar from Theocritus (‘Once upon a time the Cyprian sent a sweet
dream over Europa …’), as a single incident within the broad tapestry
of received story, but the opening word ‘Europa’, the girl who gave
her name to the land mass, reveals at once that this is a story whose
ramifications are very much still present.
The Europa tells the story of Zeus’s abduction and seduction of the
beautiful Europa; the god took the form of a bull, tricked the girl into
sitting on his back, and then carried her westwards across the sea from
Phoenicia to Crete, where, as Zeus tells her, she is to conceive his
‘glorious sons, who will all hold the sceptre of power among men’ (160–
161). And so it came to pass: Zeus made love to her, thus procreating
the people of ‘Europe’. The poem’s simple structure, which moves
from Europa’s prophetic and erotic dream to another, but shared,
bed,25 a structure which suggests unmediated, progressive narrative, is
reinforced by a narratorial voice which, on the whole, remains largely
covert. The (omni)presence of the omniscient narrator, who knows, for
example, the details of Europa’s dream and her emotional state during
it, the divine history of her basket, and the fact that Zeus fell in love
‘at first sight’, emerges perhaps less frequently than one might have
expected: cf. 5, a ‘learned commentary’ on the timing of the dream; 7,
‘Europa who was still a virgin’, suggesting the narrator’s knowledge of
what is coming; 38, mega thauma, ‘a great marvel’ describing Europa’s
basket; 72–73, a narratorial prolepsis with mellein of a familiar epic
type, ‘Not for long was Europa to amuse herself with flowers …’; 74,
the interactional particles ē gar dē; 76, Aphrodite ‘who alone can even
conquer Zeus’; 80–83, the mannered, quadruple denial that this was
any ordinary bull; 84, the interactional particles dē toi; 97–98, describing
the divine bull’s lowing, ‘you would say that you were hearing the
25 The movement is marked, as often, by verbal repetition (lekheon ‘bed’ 16 ~ lekhos

164).
sweet sound of a Mygdonian pipe’, an instance of the ‘indefinite second

person’ device; 113, the bull travels over the water ‘like a dolphin’.26
Another aspect of the narrator can be gathered from the thickly allu-
sive textuality of his writing, his re-writings of Odyssey 6 (Nausicaa) and
Book 3 (Medea) of the Argonautica of Apollonius. This marks the nar-
rative as, in one sense, a replaying of previous narratives and identi-
fies the narrator as a poiētēs, self-consciously making choices within the
ocean of tradition. In perhaps no other passage of the poem is the con-
trolling power of the narrator so strongly marked as in the meeting of
Europa and the divine bull on the seashore. Europa and her compan-
ions have come to the shore to gather flowers, as Nausicaa and her
friends went to wash clothes. In Homer a naked male appears, whom
the poet compares to a lion (Od. 6.139–147), and his terrifying appear-
ance causes all the girls except Nausicaa to run away. Nausicaa is won
over by his pleas, and a bath and new clothes (and a little help from
Athena) restore the hero’s handsome grace that turns female heads (Od.
6.227–237). Moschus unpacks the powerfully unspoken possibilities of
this scene by having not a rough and haggard man appear, but the very
embodiment of male sexual power, a glorious bull; in explicit contrast
to the Homeric scene, ‘the appearance of the bull did not cause the
maidens to flee, but all felt a desire (erōs) to approach and touch the
lovely bull …’ (89–91).27 Such games with the literary tradition are very
strong markers of narratorial presence.
In contrast to these overt games, there is a very striking narratorial
discretion in the Europa concerning the double aetiological focus of the
narrative—the founding of the Cretan royal house and the origin of the
name of Europe; these are given little prominence. The second theme
plays over the narrative of the dream in which Europa saw herself
being fought over by two women, who were in fact continents (the
‘mainland’ opposite Asia is still nameless, 9), but the idea is never really
developed. Aetiology, teleology, and narrative consequence are replaced
by the portrait of a (paradoxically) universal experience (the passage
of a young girl towards physical sexuality and motherhood). A certain
epic grandeur and amplitude is indeed lent to the poem by the devices
of the opening dream and the ekphrasis of Europa’s marvellous basket,
26 Hunter 1993: 132–133 (with further bibliography).

27 There seems to be a verbal echo of Theocritus 13.48, the effect of Hylas’ beauty
upon the water-nymphs. It is tempting to believe that Moschus wants us, in the light of
his reworking, to activate the sexual sense of mixesthai at Od. 6.136.
for these suggest pattern and importance: only great events achieve
this double mirroring. Both devices, however, emphasize retrospective
interpretation: significance emerges ‘after the event’, a pattern which
allows the events themselves to seem ordinary and everyday, almost
‘unepic’. In the contrast between such a tone and the remarkable
nature of what is being described, it is fair to see once again the
self-conscious concern in third-century and later Greek poetry with
narrative experimentation.28
28 For a much fuller account of the Europa cf. Fantuzzi-Hunter 2002: 291–301.
part two
HISTORIOGRAPHY
chapter seven
HERODOTUS
I.J.F. de Jong
In discussing the narrator of the Histories, I take as my point of depar-

ture the Homeric narrator. For the Herodotean narrator, as his proem
makes amply clear, places himself in the tradition of the Homeric nar-
rator, in that, he, too, is the guardian of the klea of men from the past.
At the same time, this proem, in replacing a Muse-invocation by a ref-
erence to Herodotus’ own investigation (historiē), signals a crucial differ-
ence between this narrator and his eminent predecessor. Indeed, the
Herodotean narrator is an intriguing mixture of an epic storyteller and
a historian.
Resemblances to the Homeric narrator
The Herodotean narrator1 resembles the Homeric narrator, in that he

is external, omnipresent, and omniscient. Let us examine these three
characteristics in turn. The main story of the Histories covers the years
560–478, which means that Herodotus (c. 484–425) could theoretically
have been an internal narrator for at least part of his story, as are
Thucydides and Xenophon. Nowhere, however, does he play a role
in the events he recounts, and he is therefore an external narrator.
Whereas the Homeric narrator occasionally refers to the ‘now’ of his
own time in order to stress the difference with the heroes of the past,
the Herodotean narrator likewise stresses differences between ‘then’
and ‘now’, using the ‘reference to the narrator’s own time’ motif (e.g.
1.173.2),2 but more often points out the continuance of customs or mon-
uments (e.g. 1.52: Croesus sent two gifts to the oracle of Amphiaraus.
1 De Jong 1999: 220–229.

2 Cf. 1.50.3.
102 part two – chapter seven
‘Both of these still lay until my time in Thebes’).3 What is remarkable

in these passages is the frequent use of the imperfect, even though the
narrator is referring to the present of his own time: e.g. ‘For many states
that were once great have now become small; and those that were great
in my own time were small in the past’ (1.5.4). Is the narrator, true to his
own maxim of the instability of fate, anticipating that his own present
may once become the past?4
The omnipresence of the Herodotean narrator appears most clearly
in such private scenes as Gyges watching the naked wife of Candaules
in her bedroom (1.10) or Darius lying in bed and discussing his military
strategies with his wife (3.134). In later books he has equally easy access
to all camps, Greek and Persian, Athenian and Spartan.
The omniscience of the Herodotean narrator is clear from his knowl-
edge of how events will end, which he shows in prolepses (e.g. 1.8: ‘After
a little while Candaules—for he was doomed to end badly—said …’),
and from the access he has to the thoughts of his characters, in the
form of embedded focalization (e.g. 1.86.9: ‘Cyrus, realizing that he,
being himself a human being, was burning alive another human being,
who had once been not less fortunate than himself, fearing retribution,
and knowing that there was no stability of human affairs, ordered the
fire to be quenched.’). We may add here his ability to quote verbatim
the words of his historical characters, even when they were spoken in
remote times or places, or privately.
So much for the similarities to the Homeric narrator, which confirm
pseudo-Longinus’ qualification of Herodotus as ‘most Homeric’ (Subl.
13.3). However, there are also differences, which in fact considerably
modify the picture just sketched.5
An overt narrator
The first difference is that, in sharp contrast to the covert Homeric

narrator, the Herodotean narrator is overt: in his proem he ‘seals’ his
work, referring to himself in the third person (‘This is the publication
3 Cf. 1.66, 92.1, 93.3, 173.3; 3.183.3; 4.12.1, 204; 5.88.3, 115.1; 6.119.4; 7.107.2, 178.2;
8.121.1. Here and elsewhere I quote the translation of Godley, with minor changes.
4 For the imperfects see Rösler 1991 and Naiden 1999.
5 Just as Herodotus in many other respects distances himself from Homer; Marin-
cola 1997: 6–7, 225–226 and Boedeker 2000: 103–105.

i.j.f. de jong – herodotus 103
of the research of Herodotus of Halicarnassus …’), and in the ensuing

nine books repeatedly steps forward in propria persona (using first-person
forms), though more often in the early books than in the later ones. I
distinguish three types of narratorial interventions: the narrator speak-
ing as (1) narrator, (2) historian, and (3) commentator.6
Ad 1. The Herodotean narrator regularly reveals his own presence
by referring to his activity as organizer of the text: ‘I will now tell’,7
‘let so much be said about’,8 ‘I return to my logos’,9 ‘as I have said
earlier’, or ‘as I will tell later’10 (no such cross-references appear in the
Homeric epics, where links between different sections of the story are
marked by verbal repetition). His cross-referencing is not systematic:
in places where one would expect a cross-reference, we do not find
one (e.g. 7.114, where mention is made of the Persian habit of burying
people alive, an instance of which is recounted in 3.35); conversely, a
back-reference sometimes follows after just a few sentences (e.g. 1.61.1).
Ad 2. Most narratorial interventions refer to the narrator’s role as a
historian. We see him at work while interviewing witnesses or locals,
e.g. 2.54.2:
The priests of the Theban Zeus told me that two priestesses had been
carried away from Thebes by Phoenicians; one of them was taken away
and sold in Libya, the other in Hellas. When I asked them how it was
that they could speak with so much certainty, they replied that their
people had sought diligently for these women and had never been able
to find them, but had later learnt the tale which was now told me.11
inspecting sites or monuments, e.g. 2.44.1–2:

Wishing to get clear knowledge of this matter from a place where it was
possible to do so, I took ship to Tyre in Phoenicia, where I heard that
there was a very holy temple of Heracles …12
reasoning, e.g. 7.238.2:
6 Schepens 1980: 46–51; Dewald 1987; Marincola 1987; 1997: 6–8; and Fowler 1996:
70–71.
7 Cf. 1.15, 75, 95, 177; 2.102, 105, 147, 155; 3.6; 4.14, 38, 145, 5.65; 8.55.
8 Cf. 1.92.4, 140.3; 2.34.2, 35.1, 76.3, 117.6; 3.3.1, 113.1, 119.1, 136, 138.4; 4.31.2, 15.4,
32.2, 36.1, 45.5, 96.2, 199.2; 5.62.1; 6.55, 100.1, 153.1; 7.100.1, 152, 153.1. See Lateiner
1989: 44.
9 Cf. 1.95.1, 140.3; 4.82; 5.62.1.
10 Cf. 1.18.2, 61.1, 75.1, 85.1, 140.3, 169.2; 2.38; 3.106.2, 159.2; 4.79, 145.1; 5.36.4, 62.1;
6.19.2, 3; 7.108.1, 110, 115.2, 213; 8.2.1, 55, 95.

11 Cf. 2.3.1, 91.5, 104.1, 113.1.
12 Cf. 2.12.1, 29.1, 75.1, 106.1, 131.3, 148.1, 5, 150.2; 3.12.1, 4; 4.195.2; 5.59; 6.47.1.
[Xerxes has Leonidas’ head cut off and impaled.] It is plain to me by this

proof especially that while Leonidas lived king Xerxes was most incensed
against him of all people; else he had never dealt with his dead body so
outrageously. For the Persians are of all men known to me the most wont
to honour valiant warriors.13
and evaluating stories, e.g. 3.56:
When the Lacedaemonians had besieged Samos for forty days with no
success, they went away to the Peloponnesus. There is a foolish tale
which says that Polycrates bribed them to depart …14
In short, we see him performing the historiē, akoē, opsis, and gnōmē which
he lists in 2.29.1 and 99.1 as the tasks of the historian, and to which he
had referred under the general header of historiē in his proem. Drawing
attention to his activity as a historian in such an explicit manner is one
of the main ways in which he establishes his authority.15
A special characteristic of this narrator is that on occasion he explic-
itly refuses to narrate something, e.g. 2.171.1–2:
On this lake they enact by night the story of the god’s suffering, a rite
which the Egyptians call the Mysteries. Although I know more about
these matters, how each of them is, let me be silent. About the rites of
Demeter, which the Greeks call Thesmophoria, let me be silent about
those, too, except for what is allowed to say.16
Most of these instances of praeteritio are occasioned by religious pru-
dence (and one is reminded of the reticence of the narrator in the
Homeric Hymn to Demeter [→] with regard to the Eleusinian mysteries:
479–482),17 the other ones by various considerations: damnatio memoriae,
fear of not being believed, a desire to emphasize what he does recount,
or the fact that something has already been told by someone else. Simi-
larly, the narrator occasionally assures that he does not recount all, but
concentrates on what is ‘worthy of telling’, or apologizes for including
a ‘digression’ (parenthēkē).18 Although there is some discrepancy between
13 Cf. 1.75.3–6; 2.27, 104, 120; 5.3.1; 7.137.1–3, 238.2; 8.8.3, 13, 53.1, 73.3, 77, 129.3;
9.65.2, 71.2, 100.2.

14 Cf. 1.51.3, 182, 214.5; 2.5.1, 12.1, 28.1–2, 45, 106.5, 120.1, 131.3; 3.3.1, 9.1–2, 16.7,
56.2, 80.1; 4.5.1; 5.10, 86.3; 6.121–124; 7.214.1; 8.8.3.

15 Marincola 1997: 63–67.
16 Cf. 1.51.4, 193.4; 2.123.3, 170.1, 171.1–2; 5.72.4; 6.55; 8.85.2; see Lateiner 1989: 65–
69, 73–75.
17 Harrison 2000: 182–191.
18 1.177, 194.1; 3.125.3; 5.65.5; 6.43.3; 7.171.1. On the Herodotean notion of ‘digres-
sion’, see Cobet 1971: 45–82.

the theory of those assurances and the practice of his text (which, at
least to modern tastes, seems to contain a great deal of digressional
material), they make clear that the narrator has a clear notion of what
constitutes the subject matter of his narrative: everything related to his
theme (the conflicts between Asia and Europe, culminating in the Per-
sian Wars of 490–479) and ‘amazing deeds’ (erga … thaumasta).19
The narrator does not, however, suppress stories that he does not
believe. As he says more than once,20 he feels it is his duty to recount all
versions.21 He even occasionally turns to the writing of virtual history,
of what might have been, e.g. 1.191.5:
Now if the Babylonians had known beforehand or learnt what Cyrus
was planning, they would have let the Persians enter the city and killed
them. For they would have shut all the gates that opened on the river
and themselves mounted the walls that ran along the river, and caught
the Persians as in a trap.22
Ad 3. More often than the Homeric narrator, the Herodotean narrator

explicitly comments on the deeds of his characters. He calls customs
‘very wise’ (1.63.2), advice ‘very useful’ (1.170.1), the death of Polycrates
‘unworthy of himself and his designs’ (3.125.2), and the blinding by the
Thracian king of his own sons ‘a monstrous deed’ (8.116.1).
Like the Homeric narrator, the Herodotean narrator is not ham-
pered by nationalism when attributing praise or blame:23 the words
spoken by Xerxes during the scourging of the Hellespont are called
‘barbarian and godless’ (7.35.2), but that same Persian king is called
‘the most worthy to hold command, because of his beauty and stature’
(7.187.2). Whether he is equally impartial where the Greek poleis, nota-
bly Athens and Sparta, are concerned is a matter of much discussion
and a subject too complex to be discussed here.
Sometimes the narrator couches his opinion in a general saying, e.g.
8.3.1:
The Athenians waived their claim [to be leaders of the fleet], deeming
the safety of Hellas of prime importance and seeing that if they quar-
relled over the leadership Hellas must perish, which is a right thought; for
an internal war is as much worse than a united war as war is worse than peace.
19 For the latter category, see Barth 1968.

20 2.123.1, 130.2; 4.173, 195.2; 6.137.1; 7.152.3.
21 Occasionally he does leave out alternative versions: 1.95.1; 2.70.1.
22 Cf. 7.139.
23 For Herodotus’ impartiality see Marincola 1997: 164.
Alongside the explicit narratorial interventions, the Herodotean narra-

tor has many more implicit ways of making clear his view. One is to
insert at the moment of a person’s success a prolepsis about that same
person’s end, which, in accordance with the principle of the instabil-
ity of fate, is usually a negative one (e.g. 7.213, where, at the moment
Ephialtes is ingratiating himself with the Persian king, the narrator
reveals that he will later be killed). In this subtle and implicit way the
narrator makes clear that crime does not pay, or that too much prosper-
ity leads to ruin. Another technique is that of suggestive juxtaposition
(→ Homer): e.g. in 3.40–43 there is a telling contrast between Poly-
crates, who is advised to throw away what is most precious to him and
has to search a long time for it among his many possessions, and the
simple fisherman, who spontaneously offers his king the biggest (and
most valuable) fish he has ever caught.
A partially omniscient narrator
A second difference between the Homeric narrator and the Herodo-

tean narrator is that the former ‘relies on’ the Muses, eyewitnesses of
the past (and thereby ‘authenticates’ his omniscience, which is of course
no more than poetic licence), whereas the latter relies on autopsy,
eyewitness reports, traditions, and his own reasoning. This means that
in point of fact he is not always omniscient: indeed, in many places
he indicates that he is not able to tell something (e.g. 1.49: ‘As to the
reply which the Lydians received from Amphiaraus … I cannot say
what it was, for nothing is recorded of it, except that Croesus held that
from this oracle too he had received a true answer’),24 or makes clear
that he is only guessing (e.g. 3.13.4: ‘Cambyses received in all kindness
the gifts of the Libyans; but he seized what came from Cyrene and
scattered it with his own hands among his army. This he did, I think,
to mark his displeasure at the littleness of the gift’).25 Also, he is not
always omnipresent: there are moments where he reaches the ‘borders’
of his story (e.g. 4.16.1: ‘As for the land of which my history has begun
24 Cf. 1.5.3–4, 57.1–2, 160.2; 2.3.2, 28.1–2, 103; 3.121.1; 4.16.1–2, 81.1, 96.1; 5.66;
6.14.1, 82.1; 7.26.2, 60.1, 133.2, 153.3, 187.1; 8.8.2, 87.1,2, 112.2, 128.1, 133–136; 9.8.2,
18.2, 32.2, 84. Cf. also the ‘X did y, either because … or …’ passages in 1.61.2, 86.2,
191.1; 2.181.1; 7.2; 8.54.
25 Cf. 1.51.4, 193.4; 2.123.3, 170.1, 171.1, 2; 6.95.2, 98.1.
to speak, no one knows exactly what lies northward of it; for I can learn
from none who claims to have been an eyewitness’).26
An epideictic narrator
Having examined the prominent presence of the Herodotean narrator

in his narrative, we are left with the question of the origin of this new—
in comparison to his main model Homer—style of presentation. As
regards his frequent comments on the characters and actions of his
story, Bowie (2001: 64) has suggested that in devising his narratorial
persona Herodotus may have been influenced by elegiac and iambic
historical narrative. An even greater influence was the work of the
scientists of his time (doctors, sophists, and physicists): their epideictic
speeches display all the typical features of the Herodotean narrator’s
style, such as his many first-person interventions, conclusions of the
type ‘let so much be said about’ (discussed above) and second-person
forms, rhetorical questions (to be discussed below in the section on
narratees).27 In this context it is relevant to note that the Herodotean
narrator presents himself as a speaker rather than a writer (as does
the Thucydidean narrator): almost without exception, he refers to his
narrating activity as legein, not graphein. This may be a relic of the
Histories as lectures,28 but we could also be dealing with a variant of
the ‘feigned orality’ which we also find in Pindar (→), Apollonius of
Rhodes (→) and Callimachus (→): although he is obviously writing a
text, the Herodotean narrator behaves as an epideictic speaker.
In sum, the Herodotean narrator has the persona of a historian, poses
as an epideictic speaker, and allows himself the liberties of an epic
singer.
Secondary and reported narrators
The Histories feature fewer secondary narrators than the Homeric epics.
There are a handful of external secondary narrators, such as Solon,
26 Cf. 4.45.2; 5.9.1. Nesselrath 1995.

27 De Jong 1999: 227–229 and Thomas 2000: 235–248.
28 For a discussion of the question whether the Histories started as lectures, see
Johnson 1994.
who tells Croesus the story of the Argive youths Cleobis and Biton
(1.31); Socles, who recounts the stories of the Corinthian tyrants Cypse-
lus and Periander to the assembled Greeks (5.92); and the Tegeans and
Athenians, who recount the story of the Heraclidae (9.26–27). There is
a slightly larger number of internal secondary narrators who—usually
quite briefly—recount events in which they and often their narratees
have participated (e.g. 7.10, where Artabanus recalls in the presence
of Xerxes his advising Darius not to undertake an expedition against
the Scythians).29 In most cases the embedded narrative functions as
a—persuasive or dissuasive—paradigm, a function that is often made
explicit (e.g. 9.26: the Tegeans and Athenians ‘claimed that they should
hold the second wing of the army, justifying themselves by tales of deeds
old and new.’).30
An intriguing internal secondary narrator is Cambyses, who in 3.65
tells an audience of Persians about the murder of his brother Smerdis
and the dream that led him to it. He now knows that he interpreted the
dream incorrectly and killed his brother without reason, and this insight
leads to a highly emotional story, which is interspersed with narratorial
comments: ‘When I was in Egypt, I saw in my dream a vision, which
I wish I had never seen … I acted with more haste than wisdom; for
[as I know understand: ara] no human power can turn fate aside. But
I, foolishly, … did wholly mistake what was to be …’ Here we are close
to tragedy (note hamartēn in 69.4), as a character ruefully looks back on
what he has done.
The Histories abound in what we might call reported narrators, i.e.,
narrators whose stories are presented in indirect speech; e.g. the story
of Arion and the dolphin as told by the Corinthians and Lesbians
(1.23–24) or the account of Miltiades’ vicissitudes after Marathon as
told by the Parians (6.134). Herodotus is particularly fond of includ-
ing local inhabitants among his reported narrators. Reported narra-
tives may also be marked not by indirect speech, but by the simple tag
‘as the X say’, inserted before or after the narrative or in both places.
In fact, the entire Histories is based on the logoi of others, even when
the narrator does not explicitly say so (for obvious reasons; this would
have resulted in a text entirely in indirect speech). The Herodotean
narrator introduces reported narrators when there are different ver-
29 Cf. 1.117.3–5, 120.2; 3.36.1–2, 73; 7.8.a, 10.a.1, 104.2, 209.2.

30 Gould 1989: 55–57.
sions of a story (e.g. the stories about Io in 1.1 and 5),31 or when he
wants to stress the source of a story (e.g. the well-known ‘Rhampsinitus
and the thief ’ story in 2.121, which forms part of a series of Egyp-
tian logoi). In itself, the use of reported narrators does not automati-
cally mean that he is distancing himself from these stories.32 Indeed,
in the case of the Egyptian logoi of Book 2, the repeated references to
Egyptian priests as his source are intended to increase the authority of
what he is recounting. When he is sceptical or critical, he usually—
although not always—makes this clear in a narratorial intervention
(e.g. in 3.56, ‘There is a foolish story …’). In these instances we see
the birth of one of the most distinctive aspects of ancient historiog-
raphy, the polemic with predecessors.33 A kind of shorthand variant
of the reported narrators are the anonymous spokesmen: whereas the
former are specific (‘the Persians/the Lesbians say …’) and responsi-
ble for an entire story, the latter are anonymous and invoked only to
modify—negatively or positively—certain details of a story, usually in
the form of legetai, e.g. 1.159 (‘But while he did so, a voice it is said/they
say came out of the inner shrine calling to Aristodicus’).34 This device
appears to be a continuation and expansion of the occasional anony-
mous spokesmen in Homer (→), introduced in the form of phasi, ‘they
say’.
The primary narrator regularly intervenes in the reported narratives,
in order to add comments addressed to the primary narratees, e.g. ‘The
Persians say that next certain Greeks (they cannot tell who) landed at Tyre
in Phoenicia and carried off the king’s daughter Europa. These Greeks
must, I suppose, have been Cretans. So far, then, the account between them
stood balanced. But after this the Greeks became guilty of a second
wrong’ (1.2). All in all, one gets the impression that, although formally
presented by other narrators, the reported narratives are very much
the product of the primary narrator. This observation is confirmed by
the fact that the indirect speech is often dropped after a few sentences
or, conversely, the primary narrator suddenly slips into indirect speech
(both phenomena are exemplified in the episode of Cyrus and Croesus
after the fall of Sardes: 1.86–91). In other words, primary and reported
31 Groten 1963 and Lateiner 1989: 76–90.

32 Harrison 2000: 25–30. The other view (indirect speech suggests distance) is in fact
the communis opinio, see e.g. Gould 1989: 50–51.
33 Marincola 1997: 218–226.
34 Westlake 1977: 361–362. Sometimes the subject of legetai is specified, e.g. 1.87.1, just
as, conversely, sometimes the subject of legousi is not specified, e.g. 9.120.4.
narrative in the Histories are less fundamentally distinct from each other
than would appear at first sight, and the primary narrator exercises a
tight control on all levels of the story.
Narratees
The Herodotean primary narratees are not as overt as the narrator: in

some places they are specified as Greeks (e.g. 3.80.1 or 103), but there
are no further indications as to what kind of audience the narrator
envisages for himself.35 However, their presence is acknowledged in
various ways:36
– the ‘indefinite second person’ device (e.g. 1.139: ‘You shall find, if
you search, that not some but all Persian names alike end in this
letter’).37
– passages featuring an anonymous interlocutor (often tis) with
whom the narratees can identify (e.g. 5.54.1: ‘If someone desires
a more exact measurement, I will give that too’).38
– passages featuring an anonymous witness with whom the narra-
tees again can identify (e.g. 4.19: ‘But to the east of these farming
Scythians, for someone who has crossed the river Panticapes, the
nomadic Scythians live …’).39
– rhetorical questions (e.g. 1.75.6: ‘Some say that the ancient channel
was altogether dried up. But I do not believe this; for how did they
then pass the river when they were returning?’).40
– gar-clauses, which provide explanations to questions which the
narrator assumes the narratees will have (e.g. in 1.10, where having
recounted that, after the Lydian queen was seen naked by her
husband’s servant Gyges, she was ashamed and decided to take
revenge, the narrator has tens to add ‘for among the Lydians
and most of the barbarians it is held a great shame that even a
man should be seen naked’, thus answering his narratees’ implied
question as to why the queen would be upset).
35 Flory 1980; Gould 1989: 15–17; and Malitz 1990: 327–328.

36 Lateiner 1989: 31–33.
37 Cf. 1.199.4.
38 Cf. 2.6.2, 146.1 (bis), 148.2, 179; 3.6.2, 38.1, 122.1; 5.45.2; 7.126, 139.5.
39 Cf. 1.51.1; 2.5.1.
40 Cf. 2.11.4 (bis), 15.2, 22.2, 45.2,3; 57.2; 125.7; 4.46.3; 7.21.1 (bis).
– ‘presentation through negation’-passages, which contradict the

narratees’ expectations or create new ones (e.g. 1.10: when Can-
daules’ wife realizes that she has been seen naked, ‘she did not cry
out nor let it be seen that she had perceived anything’; this is an
unexpected reaction and raises the question of what the queen is
up to).
To this list of devices the Herodotean narrator adds the use of inter-
actional particles: kaitoi (e.g. 8.86: ‘The barbarians, no longer fighting
orderly or using their heads, were bound to end as badly as they did.
And yet they proved themselves that day much better than off Euboea
…’; this counters a conclusion which the narratees might otherwise
have drawn on the basis of the preceding, viz. that the barbarians are
fighting as badly as never before),41 mentoi (e.g. 1.172.1: ‘The Caunians, to
my mind, are aboriginals; they themselves, however, say that they came
from Crete’; the narratees are invited to note the contrast),42 dēn (e.g.
1.59.4: ‘Wounding himself and his mules, Pisistratus drove his carriage
into the market place, pretending to have escaped from his enemies,
who allegedly had wanted to kill him when on his way to the fields’;
the narrator draws his narratees’ attention to the deceit),43 and ara (e.g.
6.110.1: ‘The Eretrians … asked help from the Athenians. These did
not refuse the help. … But it turned out that the designs of the Eretri-
ans were not sound at all …’; the narratees are alerted to the fact that
things are different from what one might expect).44
The main narrative and the digressions
As in the Homeric epics, the beginning of the main narrative is promi-

nently marked (with both a proem and an indication of the starting
point: Croesus, the first barbarian of whom the narrator knows for sure
that he conquered Greeks). The end is not signalled in the same way,
and the question whether closure is effected in the Histories is a matter
of dispute.45
41 Cf. 2.142.2, 148.2; 3.15.3, 152; 4.77.1; 8.71.1, 112.3; 9.71.3.

42 There are some 86 instances in narrator-text (as against 30 in speech).
43 Cf. 1.73.5; 3.74.4, 136.2, 156.1; 6.1.1, 39.1; 8.5.1, 6.2; 9.66.3, 80.3, 99.3.
44 Lateiner 1989: 31–33.
45 Dewald 1997.
In the course of his main narrative the narrator tends to vary his
pace: some parts are recounted in summary fashion, others scenically,
i.e., with much detail and speech. Some of the latter parts are real
gems and, featuring ring-composition, closure, and the leitmotif style,
stand out as self-contained units. An example is the ‘Atys and Adrastus’
story in 1.34–45.46 At the beginning of the story Croesus considers
himself the most fortunate of all mortals. But then he has a dream
in which the death of his favourite son is foretold. He takes very
concrete steps to avert the danger. At that point Adrastus, a man in
the grip of misfortune (leitmotif ), comes to his palace, because he has
unintentionally killed his brother. It is this man who then accidentally
kills Croesus’ son and makes the dream come true. The story ends
with Adrastus’ suicide (closure), because he realizes that he is the most
unfortunate of men (leitmotif and contrastive ring-composition with
beginning). Though forming a self-contained unit, the story is firmly
integrated into the main narrative, in that it provides the primary
narratees with a first proof that Croesus’ fortune, like that of all men,
can change, just as Solon had warned him (1.32); for Croesus himself
this moment of insight will not come until later (1.91).
The main narrative is frequently interrupted by digressions.47 These
are of two types: ethnographical/geographical or historical. The his-
torical ‘digressions’ are in fact analepses or prolepses, which are usu-
ally marked off by means of ring-composition, narratorial interventions
which announce or conclude a section (see above), or by anaphoric
and cataphoric pronouns (hōde, hōs, toionde, etc.). These anachronies are
almost invariably functional (hence my use of the term ‘digression’,
with its connotation of irrelevance, between inverted commas): they
either provide background information which helps the narratees to
understand the unfolding of the main story (e.g. the analepsis on Athe-
nian history in 5.55–96, culminating in the increased tension between
Athens and Persia, explains why the Athenians accept Aristagoras’
request for help against the Persians), or they clarify the main story
by providing an analogy (e.g. the analepsis on the reforms of Clisthenes
the Sicyonian in 5.67–68, which is intended to explain the behaviour of
46 Cf., e.g. ‘Polycrates and the ring’ in 3.39–43, ‘Xerxes and the wife of Masistes’ in
9.108–114. Discussion of these semi-independent stories told by the primary narrator

in Long 1987; Erbse 1992; Van der Veen 1996; and de Jong 1999: 242–251. For reasons
why these passages should not be labelled ‘novellae’ see de Jong 2002: 257–258.
47 Cobet 1971 and de Jong 2002.
Clisthenes the Athenian).48 The ethnographical/geographical ‘digres-

sions’, which in narratological terms are descriptions, have often been
seen as relics of Herodotus’ early career as an ethnographer and ge-
ographer. However this may be, the Herodotean narrator has thor-
oughly integrated them into his narrative, inserting them, for exam-
ple, at the moment an oriental king undertakes a military expedition
against another people, and thereby providing his narratees with the
information necessary to understand what follows. Thus the fact that
the Scythians are nomads (4.46.3) explains why Darius will not be able
to defeat them, and the description of Babylon (1.178–187) clarifies the
ensuing account of the siege of this city. Where the narrator gives more
information than his story demands, as is often the case, this is in line
with the announcement in his proem that his historiē encompasses ‘what
is wrought by human beings’; he describes the world as it was then
known to mankind.49
Narratorial alter egos and the function of the Histories
Like the Homeric narrator and his singers Phemius and Demodocus,
the Herodotean narrator uses the device of the narratorial alter ego:
certain characters in his story seem intended as images of himself. In
the first place, we may think of Solon,50 Artabanus, and Demaratus.
All three are wise advisers who in some cases have at their disposal
information about unknown people and places or about the past; thus
Artabanus informs Darius about the Scythians, and Demaratus tells
Xerxes about the Greeks. This provides us with an important clue to
the function which the Herodotean narrator sees for his own work: in
addition to the typically epic function of preserving the glorious past,
it provides information about that past and about other people, on
the basis of which political decisions can be taken.51 An alter ego of a
somewhat different nature is Periander in the story of the poet Arion
and the dolphin (1.23–24). Having been forced to jump into the sea
by sailors and then miraculously saved by a dolphin, Arion returns to
48 Gray 2002.
49 Cobet 1971: 85–140.
50 Shapiro 1996.
51 Dewald 1985 and Christ 1994. Whether Herodotus has a specific warning in mind
for the Athenian readers, is a matter of discussion; see most recently Moles 2002.
Corinth and tells Periander what has befallen him. The latter does not
believe his tale and arranges for a confrontation with the sailors, who
first claim that they safely transported the poet to Taras. When they
then see him alive in front of them, they are forced to admit their crime
(and, we may conclude, Periander now believes the tale of the dolphin).
There are in fact many incredulous characters in the Histories, who dis-
believe reports of facts, past or present, but who eventually revise their
own views by collecting information. Thus Periander seems the alter
ego of both the narrator, who likewise investigates the stories told to
him (like him, Periander is said to historieisthai: 24.7), and the narratees,
whose scepticism with regard to the narrator’s stories must be aban-
doned when they are confronted with his proofs (autopsy, spokesmen,
or logical reasoning).52 The briefest but most pregnant self-image of the
Herodotean narrator is that found in 5.36: ‘All the rest favoured revolt,
except Hecataeus the historian; he advised them that they would be
best guided not to make war on the king of Persia, recounting to them
the tale of the nations subject to Darius and all his power’, in short, the
tale which is told in—part of—the Histories.
52 See Packman 1991.

chapter eight
THUCYDIDES
T. Rood
Thucydides’ history of the Peloponnesian War (431–404 BC) has tra-

ditionally been seen as a model of scientific history writing. But many
readers in recent years have reacted against idealized readings of Thu-
cydides as an objective historian. They see his unmediated presentation
of events, and his reluctance to allow competing versions into his text,
as designed to blind readers to his role in shaping his version and per-
suade them that it is true. But Thucydides’ narratorial style has also
been seen in more positive terms. Scholars have increasingly presented
a Thucydides who is not so restrained and objective after all.1 And
Hanson has polemically celebrated a Thucydides who is ‘selective and
guarded, not promiscuous, in his disclosures’, and whose personal voice
is ‘mysterious and subtle’, ‘possess[ing] a certain dignity by the restraint
in its use’.2
Thucydides’ role as a paradigm of impersonal scholarship makes
closer exploration of his narrative persona desirable. This chapter will
discuss first the primary narrator, and his interaction with his narratees,
and then other narratives included in the work. It will raise questions
about Thucydides’ place in the development of Greek narrative, about
the move from works composed for oral performance to works written
to be read, and about the political implications of narrative modes.
Narrator
‘Thucydides, an Athenian, wrote up the war of the Peloponnesians and

Athenians …’ (1.1.1). The first event Thucydides’ history records is its
1 Note especially Connor 1977 and 1984: he makes the case for a ‘postmodernist’
Thucydides who is very different from the positivist historian praised by earlier genera-
tions.
2 Hanson 1998: 139.
116 part two – chapter eight
own composition. Who is the narrator? As the preface continues, the

initial third-person form gives way, as in the prefaces of Hecataeus
and Herodotus, to first-person forms (‘from the evidence which I can
trust … I think that previous events were not great’, 1.1.2). The run of
the sentence shows that the speaker of these first-person utterances is
Thucydides. So it is Thucydides who starts by narrating his own act
of composition, and Thucydides who narrates the Peloponnesian War
itself.
What sort of a narrator is ‘Thucydides’? ‘Thucydides’ is a contem-
porary (he ‘wrote up the war … beginning when it was first breaking
out’) whose home city was one of the participants in the war (‘an Athe-
nian … the war of the Peloponnesians and Athenians’). But he is also
an internal narrator in as much as he later appears as an agent: ‘they
sent to the other general in Thrace, Thucydides the son of Olorus, who
wrote this …’ (4.104.4). Herodotus (→), by contrast, finds a place in his
own text only through his activities as traveller, researcher, and writer:
that is, he is an external narrator.
That Thucydides the character is ‘the son of Olorus’ and Thucy-
dides the writer is ‘an Athenian’ demands explanation. By giving a
patronymic on his first appearance as general (as he does with other
high-ranking characters,3 and by using third-person forms for his own
actions, he suggests that he treats himself in the same way as he does
other characters. He anticipates the technique of Xenophon’s Anaba-
sis [→] and Caesar’s Commentaries—except that, unlike Xenophon and
Caesar, he states within the text that the character is the writer. By call-
ing himself ‘an Athenian’, on the other hand, he looks back to his pre-
decessors, who introduce themselves as ‘Hecataeus, a Milesian’ (FGH 1
F1a) and ‘Herodotus, a Halicarnassan’. It is not (as the common trans-
lation ‘Thucydides the Athenian’ suggests) that ‘Thucydides’ is classify-
ing himself as someone famous (contrast the use of the definite article
in the cultural notice about Thucydides at D.S. 12.37.2, and at e.g. Th.
1.126.12, 138.6). Rather, these openings give information that, later in
the history of the book, is given paratextually, on the title page rather
than in the text proper. But ‘Thucydides’ reference to himself as ‘an
Athenian’ does not simply attach him to his predecessors. The allu-
sion to the author’s nationality has a methodological point that it lacks
in Hecataeus and Herodotus. By revealing that he belonged to one of
3 Hornblower 1994: 161–162.

t. rood – thucydides 117
the protagonists in ‘the war of the Peloponnesians and Athenians’, and

by calling the war by that neutral formulation, ‘Thucydides’ suggests
that he had access to good sources, and that he was not parochial. (So
too his statement that he began at the start of the war, and foresaw
its greatness, boosts his authority as narrator and thinker.) In the sec-
ond preface, by contrast, where he writes that ‘the same Thucydides,
an Athenian’ has written up to the point when ‘the Spartans and their
allies put an end to the rule of the Athenians’ (5.26.1), the narrator’s
identification as an Athenian seems emotive.
Thucydides also refers to himself in another way: by name alone,
without either patronymic or nationality. He does this in the formula
found at the end of many war-years (e.g. ‘the second year of this war
which Thucydides wrote up ended’, 2.70.4). Presumably it is because
he is closing smaller sections of the work, rather than opening one
of its major divisions, that he omits his nationality. But the fact that
Thucydides does frequently use first-person forms confirms that the
third-person form is a framing device.4
How often, and in what contexts, do Thucydides’ first-person inter-
ventions appear? They are rare enough for it to be practical to give
a complete list here. He uses them to refer back to earlier passages
(5.1.1; 6.94.1), to explain why he has included a section (1.23.5, 97.2),
to introduce a new section (2.48.3; 3.90.1; 5.26.6) or announce what a
new section will prove (2.48.3; 6.54.1), and to express an inability to say
something (3.113.6; 5.68.2; 6.2.1). He also uses them ‘for statements of
opinion, reasoning, inference, autopsy, and methodology, that is, any-
thing that affected the history qua history’:5 that is, as metanarrative
narratorial comments. Methodological instances of the first person are
particularly interesting because they show how ‘Thucydides’ separates
his role as agent (where third-person forms are used) from his role as
enquirer. He uses the first person to state that he was present at some
of the speeches and events he describes (1.22.1–2), and that his exile
helped him to gather information from the Peloponnesian side (5.26.5).
More complex is his statement about the plague at Athens: ‘I will show
the symptoms, I who myself fell ill and myself saw others suffering’
(2.48.3). Gribble claims that here ‘the first person refers to the narra-
tor as an actual agent in the narrative’ (he contrasts the use of third
4 Though hardly a seal to separate books, pace Hemmerdinger 1948.

5 Marincola 1997: 184 n. 52.
persons in 4.104).6 But the passage is methodological. The use of par-

ticiples (rather than main verbs) for the illness and for the claim of
autopsy suggests that the passage is primarily explaining why Thucy-
dides is able to give such a good description of the plague. Besides these
occurrences, first-person forms are used to limit the chronological scope
of narratorial judgments (7.86.5; 8.24.4, 68.2, 97.2; also 1.1.4, 13.4, 18.1,
7.87.5, 8.41.2, where the first-person plural is used) or foreground their
subjectivity (1.23.6; 2.17.2; 3.89.5: three contentious statements of cau-
sation, and 7.87.5), to speculate about the future (1.10.2; 2.54.3), to mark
uncertainty about characters’ motivation (1.93.7; 8.56.3, 87.4) or even
(apparently) about a fact (8.64.5), to mark a story as one handed down
by tradition (2.102.6: another first-person plural), and for reasoning and
statements of opinion in ‘digressions’ dealing with past history, such as
the Archaeology and the account of the fall of the Pisistratid tyranny at
Athens (1.3.1, 2, 3, 9.1, 3, 10.4; 6.55.3).
The relative paucity of Thucydides’ first-person appearances has led
to the suspicion that Thucydides manipulates his narratorial stance so
as to impose his judgment on narratees without their being aware of it.7
It has even been claimed that ‘the third person narrator of the opening
sentence yields to a new speaker, the war itself ’.8 But Thucydides’
activity as writer is underlined by the year-end formulae, his activity
as narrator by the first-person forms listed above. It is underlined,
too, by techniques such as ‘if not’ situations and ‘presentation through
negation’; by small temporal shifts as well as by the larger temporal and
intellectual range of passages where Thucydides analyses Athens’ defeat
(2.65), civil war (3.82–83), or the Athenians’ complex dealings with
Alcibiades (6.15.2–4); by other explicit judgements (about characters,
for instance, before they make speeches [1.139.4; 3.36.6]; compare the
introduction of Nestor at Iliad 1.247–252); and by generalizations about
human nature (4.108.4) or ‘what a crowd tends to do’ (2.65.4; 4.28.3;
6.63.2; 8.1.4).9
6 Gribble 1998: 48 n. 46.

7 E.g. Loraux 1986: 149.
8 Connor 1984: 29, also Kurke 2000: 150, alluding to 1.21.2: ‘this war will show
(dēlōsei) those who look at it from the facts themselves that it was greater than previous
ones’. But note that dēloun is commonly used, as here, in contexts where recipients’
response to what is shown is at issue (1.73.1; 2.48.3), and of letters or inscriptions (1.129.1,
134.4, 137.4; 7.10, 16.1); and that ‘this war’ implies the war that Thucydides is going to
narrate—a narration that begins when ‘the war begins’ (2.1.1).
9 Hornblower 1994: 148–160; Gribble 1998; and Rood 1998, index, s.v. ‘narrator’ on
Expressions of uncertainty further undermine claims that Thucy-

dides’ story tells itself. ‘Thucydides’ fails to express himself with full
confidence more often than is sometimes realized (and notably when he
is discussing Spartan or Persian actions and motives: 2.18.5, 20.1, 57.1;
3.79.3; 5.54.1, 65.3; 8.46.5, 50.3, 94.2). He also sometimes blurs numer-
als (e.g. ‘two or three times’ at 2.4.2; cf. 3.76.1; 4.38.3; 5.10.9; 6.101.6;
7.79.6; 8.74.2). Such expressions of uncertainty are often thought to
reinforce the narratee’s belief in statements of fact about which no
doubt is expressed. That is, scholars have it both ways. Both Thucy-
dides’ apparent certainty, and his occasional uncertainty, are taken as
means of boosting his authority.
How plausible are such readings? Certainly, expressions of doubt
may affect some narratees’ belief in other statements. But variants can
also have distinctive effects of their own. Phrases such as ‘two or three
times’ tend to appear in vivid sections of narrative like the night-time
attack on Plataea (2.4.2). At 5.65.3, the alternative ‘either … or’, unique
in Thucydides, give a further touch of Herodotean colouring to the
account of the uneasy peace.10 When the Thucydidean narrator reports
what ‘is said’ (legetai) to have happened in a place in the distant past,
he exploits the mythical aura of that place—of the strait between Italy
and Sicily, say, ‘the so-called Charybdis past which Odysseus is said
to have sailed’ (4.24.5). The historicity of Odysseus’ travels is scarcely
at issue here. Or is Thucydides subtly alluding to a famously embed-
ded narrative? Hornblower notes that ‘Homer notoriously distances
himself from the Charybdis story … by having Odysseus recount his
own adventures to the Phaiakians. … [Thucydides] says, in effect,
that Homer says that Odysseus, etc.; legetai is thus doing a kind of
double duty.’11 If this is right, we have an (early) example (in prose)
of the ‘Alexandrian footnote’—‘the signalling of specific allusion by a
poet through seemingly general appeals to tradition’.12 But perhaps
this passage is just an illustration of the narrator’s qualified omni-
science.
narrative interventions in Thucydides; also, on counterfactual statements, Flory 1988

and Rood 1998: 278–280.
10 Rood 1998: 106, with n. 100.
11 Hornblower 1991–1996: II 182.
12 Hinds 1998: 1–2. On Thucydides’ use of legetai, see Westlake 1977.
Particularly worth discussion is the sole passage where Thucydides

reveals the different stories that his sources told him about an event
in the Peloponnesian War: ‘this is what the Thebans say, and they say
that the Plataeans swore an oath; the Plataeans do not agree that they
promised to give the men back at once, but only if after discussion
they came to some agreement, and they deny that they swore an oath’
(2.5.6; the nearest parallels are 1.138.4, 6, from an excursus about the
past). This passage is interesting for four reasons: ‘Thucydides’ names
as his source not individual eyewitnesses, but ‘the Thebans’ and ‘the
Plataeans’; he reports not what they said, but what they say, that is,
what they still say at the time of narration; he does not judge which
side is right; and the disputed episode comes right at the start of the
war. Why did ‘Thucydides’ include this discrepancy, and no others,
even though they too may have contained (or even been) important
facts? It would be facile to regard this passage as a cynical ploy to instil
belief in the rest of Thucydides’ narrative. Rather, the citation of what
‘the Thebans’ and ‘the Plataeans’ say points (as often in Herodotus) to
the continuing political importance of stories about the past—at a time
when Plataea itself had been destroyed by the Spartans to gratify the
Thebans (note how the Plataean story is explicitly cast as a response
to the Thebans). At the same time, Thucydides’ refusal to decide
between the two accounts itself emphasizes the hostility of Plataea and
Thebes. The dispute also prepares for the complexities of the Plataean
debate (3.52–68), where the difficulties of moral judgment about the
past are a major theme. And one of the messages of the Plataean
debate, that perceived self-interest drives out considerations based on
past behaviour, may itself explain why the political use of stories about
the past are not going to be a major theme of Thucydides’ history.
‘Thucydides’ nods, in the symbolically rich Plataean setting, towards
a Herodotean manner of writing history—a style that he then moves
away from.
The Thucydidean narrator’s expressions of uncertainty cannot be
reduced to tools for persuading narratees of his reliability and objec-
tivity. Even to talk of the ‘objectivity’ of the rest of Thucydides’ nar-
rative is anachronistic. In those parts of his work where his narratorial
persona is covert, he is following in the footsteps of Homer, and also,
to a lesser extent, of Herodotus (→), whose first-person interventions
become much less frequent the closer he gets in space and time to
his own time. The Herodotean influence is greatest, however, in those
parts of Thucydides’ work where his narratorial presence is overt, as
when he interprets the distant past in the Archaeology with the help of
very Herodotean modes of reasoning.13
There do remain important differences between the narrators in
Herodotus and in Thucydides. Marincola (1997: 9) notes of the Archaeol-
ogy, for instance, that ‘the narrator is just as present … but he is not as
intrusive as in Herodotus’. Thucydides is more critical than Herodotus
of other people’s attitude towards received information (1.20.1–3;
6.54.1). The way he opposes his own rigorous methods to the slack
pleasure-seeking of other people mirrors, indeed, the opposition be-
tween the unwavering Pericles, Athens’ supreme leader, and the volatile
Athenians.14 Hence the temptation to take Thucydides’ narratorial per-
sona as far more elitist and autocratic than Herodotus’, and to relate
it to his work’s (alleged) status as a written, and so more autonomous,
text, a possession for private study.15 But other explanations can also
be suggested for Thucydides’ reserve. In some parts of his work, he
was aiming not at ‘objectivity’, but at vividness—a vividness that would
convey something of the suffering caused by the Peloponnesian War.16
To interfere too often in his own person, to say who (if anyone) told him
that their bodies shook as they watched the final sea battle at Syracuse
(7.71.3), would have been to spoil some of his greatest effects. But at the
same time the narrator does occasionally intervene at the end of vivid
narratives to make ‘pathos statements’17 about the scale of suffering (e.g.
7.30.3)—statements that recall the summaries found in tragic messenger
speeches. The only safe conclusion is that Thucydides creates a greater
sense of a controlling and single-minded purpose in his narrator than
does Herodotus. Whether this sense of control becomes apparent in his
interaction with his narratee will be examined in the next section.
Narratees
To whom does ‘Thucydides’ address his account of the war? He does

not say. There are no second-person addressees in the narrative. ‘Thu-
cydides’ does address the narratees in two rhetorical questions (7.44.1;
13 Fowler 1996: 76–77; also Marincola 1989a on the phrase dokei moi, ‘seems to me’.
14 Crane 1998: 38 on the analogy between Pericles and Thucydides, with further
bibliography in his n. 7 (and add Murari Pires 1998).
15 E.g. Crane 1996.
16 Rood 1999: 166; Kurke 2000: 151–152; and more broadly Walker 1993.
17 Immerwahr 1985: 447; cf. Lateiner 1977.
8.96.2), and again when he argues that the smallness of Mycenae

does not mean that it could not have once been powerful: ‘it is not
reasonable to be distrustful and to look at the appearance rather than
the power of cities’, but one should think that the expedition from
Mycenae against Troy was, in its time, the greatest ever, ‘if it is right
to trust Homer’s poetry’ (1.10.3). Elsewhere he employs the ‘anonymous
interlocutor’ device (often tis) with whom the narratees can identify (as
when ‘Thucydides’ refers to calculations that ‘one’ could make on the
basis of facts or arguments that he has given: 1.10.1, 21.1; 5.20.2–3, 26.2–
3; 6.55.1). What ‘Thucydides’ makes explicit, however, is his idea of his
ideal readers: ‘it will be enough if my work is judged useful by those
who want to have a clear picture of the past and of the similar things
which will happen again at some time’ (1.22.4; cf. also 1.21.2, ‘those who
look at the war from the facts themselves’). He also makes clear that he
foresees readers in the future for the ‘possession for all time’ that he has
written (1.22.4).
How does the teller of the narrative relate to its recipient? Towards
the end of his preface, Thucydides says that he has written an account
of the causes of the war ‘so that no one should ever have to seek how
so great a war arose for the Greeks’ (1.23.5). Thucydides’ explanation
of why he wrote the causes of the war has been taken as a sign that
he wanted to prevent any future research on the war (‘it is difficult
to imagine a more unHerodotean remark’).18 One might contrast his
attitude towards the causes of the plague at Athens: ‘let each man,
doctor and layman, say about it as he thinks’ (2.48.3; cf. the similar
expression at 6.2.1). He goes on to suggest that his description of the
plague is worth studying because it reveals, if not the ultimate cause
of the disease, at least some of the causes of its spread. With the war,
by contrast, Thucydides thought it helpful to state explicitly his own
view of its cause—not least because it was ‘so great a war’, that is,
so long and disastrous a war (as the reader realizes after 1.23.1–3). He
is not preventing future research (translations like Gould’s, ‘so that no
one need ever again investigate’, are misleading), but making sure that
narratees should at least have an explanation to hand.19
Thucydides’ statement on the usefulness of his history also points
18 Gould 1989: 111.

19 For the topos of preventing narratees having to search for something, cf. Plb. 1.3.9,
12.6 (where the importance of the topic is also stressed), 3.57.2; 4.40.1; 6.11.4, 14.1; D.S.
1.1.4, 6.1.
to the narratee’s active engagement in the work. Scholars have long

noted the fact that he does not say what sort of advantage the narratee
is meant to gain: is it simply a better understanding of the world,
or is it an understanding that has practical use? Thucydides’ lack of
precision is itself revealing. He leaves it to readers to make of his history
what they will. ‘The political lessons and the utility of the Histories …
derive not from Thucydides’ explicit comments or implicit theorizing,
but from the reader’s own involvement in the work.’20
Secondary narrators
Further light can be shed on the authority of the narrator in Thucy-

dides by examining how the primary narrative relates to the narratives
embedded within it. We have already seen that Thucydides was reluc-
tant to allow one type of secondary narrator into his text (oral sources
giving variant stories—the reported narrators of Herodotus [→]), and
that this reluctance to admit competing voices creates a sense of strong
narratorial control. How do other competing voices appear in the His-
tory?
It is as rare for Thucydides to cite written evidence such as inscrip-
tions as it is for him to cite oral sources. Inscriptions and letters, many
of which are narratives, are more common in digressions dealing with
the past.21 In his excursus on Themistocles’ life in exile, for instance,
Thucydides includes (in direct speech) a letter Themistocles writes to
Xerxes in which he tells, rather allusively, how he has harmed and
helped Xerxes in the past, then argues that he deserves to be repaid
for the good he has done—‘writing about the warning to retreat from
Salamis and the non-destruction of the bridges over the Hellespont,
which he falsely pretended was due to him’ (1.137.4). The primary nar-
rator here interferes with the text of a secondary narrator, Themistocles
(for the technique, cf. 8.50.2, 4; Herodotus 1.86.5 [→]). That this is the
only place where Thucydides edits Themistocles’ letter is revealing. He
suppresses the details of the narrative, doubtless because he expected
the primary narratees to be familiar with Herodotus, and leaves the
moral appeal made on the basis of that narrative. From the war narra-
20 Connor 1985: 11.

21 For narrative inscriptions, see 1.132.2; 6.54.7, 59.3; on the narrative element in
decrees, see Rood 1998: 92.
tive itself there is a letter from the Persian king to the Spartans which
is intercepted by the Athenians: ‘much else was written, but the chief
point for the Spartans was that he could not understand what they
wanted: for many envoys had come, but none said the same thing’
(4.50.2). Even though most of his letter is omitted, the Persian king
appears here as a narrator, and a useful one too. His letter fills a gap in
the primary narrative’s treatment of Spartan–Persian diplomacy. That
diplomacy was doubtless difficult for Thucydides to hear about, and
difficult in its own right. Introducing the intercepted letter explains how
Thucydides learnt about secret discussions, and shows that the more
detailed discussions that he omits were futile.
Whereas the letter sent by the Persian king to the Spartans adds
details not found in the primary narrative, the much longer letter that
the Athenian general Nicias writes to the Athenians at home about
their growing problems in Sicily (7.11–15) covers ground already famil-
iar from the primary narrative. The extent of the repetition focuses
attention on Nicias’ reliability—especially as he wrote the letter so that
his message would not be distorted, and the Athenians could delib-
erate about the truth (7.8.2). (Thucydides reports the letter as it is
read out to the Athenian assembly: hence the narrative in the letter
is doubly embedded.) Nicias gives some details not found in the earlier
narrative; and he glosses over his own responsibility for some recent
developments.22 Exploring the contrast between Nicias’ letter and the
surrounding narrative is important for what it reveals about Nicias,
and especially about his interaction with his secondary narratees. The
defensive tone taken by Nicias as he tells his story to the Athenians is
a telling contribution to Thucydides’ analysis of the troubled relation
between the Athenians and their leaders. The gap between Nicias’ pre-
sentation in his secondary narrative and Thucydides’ primary narrative
could also be interpreted as a critique of democratic knowledge—or at
least as a comment on the difficulties involved in any form of decision-
making.23
As with Nicias’ letter, the interest of the narratives made by Thucy-
dides’ speakers derives not least from how they supplement, comple-
ment, or clash with the primary narrative. First, it will be helpful to
make some general points about the speakers themselves. Often they
are not individualized (‘the Athenians’ or ‘the Thebans’ speak). Unlike
22 For a more detailed discussion, see Rood 1998: 189–191.

23 Contrast Shrimpton 1998: 74–75.
the primary narrator, they are all very conscious of their narratees,
whom they are attempting to win over (and who may even be gods, as
with Archidamus at 2.74.2). As in Homer, characters use far more emo-
tive language than the primary narrator: talk of ‘freeing’ and ‘enslav-
ing’, for instance.24 They tend to use first- and second-person forms
even when they are talking about past events in which neither the
narrators nor the narratees could have been involved. That is, Athe-
nian speakers, addressing Spartans, assimilate past and present Athe-
nians and Spartans (a usage common in actual deliberative speeches
like Andocides 3). Thucydides’ speakers are mostly making delibera-
tive speeches, and so focussing on the future, not the past (cf. Arist.
Rhet. 1358b13–18). But they often support their plans for the future by
telling stories about the past. Their narratives tend to be short but mul-
tiple. They may begin by explaining how they come to be making a
speech (e.g. 1.32.2, 120.1; 4.17.1); later narratives may be introduced by
gar (e.g. 1.34.2, 140.2; 3.54.2).25 Typically, these later narratives are inter-
laced with analysis. But speakers who are not (primarily) offering advice
about the future, like the Plataean speakers whose lives are threatened
(3.52–59), have to confront the past more urgently.
How do their narratives relate to Thucydides’ primary narrative?
Often speakers tell a story already told by ‘Thucydides’; sometimes they
even use the same words (1.108.3 ~ 4.95.3). At other times the story
speakers tell has been told by Herodotus.26 In such cases, ‘Thucydides’
presumably expects his narratees to take the accuracy of the story
for granted, and to be concerned rather with the use these speakers
make of the past. But, as with Nicias’ letter, secondary narrators may
add details about recent events not contained in Thucydides’ earlier
narrative of those events. The Plataeans, for instance, say that the
Theban attack on Plataea occurred at a sacred time of the month
(3.56.2)—a delay which ‘categorises the item as one relevant to the
rhetoric of praise and blame (the point in Book 3), not one that affected
the Theban decision to attack … still less one that might explain, as it
might have done in Herodotus, why the Theban attack failed’.27
24 Rood 1998: 238 n. 50.

25 See de Jong 1997 on this technique.
26 Hornblower 1991–1996: II 133, who takes such passages as evidence of Thucy-
dides’ familiarity with Herodotus.

27 Pelling 2000b: 69. For other examples, cf. e.g. 1.40.5, 41.2 (the Peloponnesian
League meeting in 440 BC, with the contrasting explanations of Badian 1993: 139 and
Rood 1998: 217–219); 2.13 (detailed financial information given by Pericles, including
The use of familiar stories about the past is self-consciously examined

in the longest, and most clearly delineated, of all the narratives in
Thucydides’ speeches: the account of the Athenian role in the Persian
Wars given by Athenian speakers at Sparta. To dissuade the Spartans
from declaring war, the Athenians turn to the past—but not the very
distant past: ‘What need is there to speak of remote events, which are
attested more by the voice of tradition than by the experience of our
audience? We must speak of the Persian Wars and events you know
of yourselves, although we are rather tired of continually raising the
subject’ (1.73.2–4). The Athenians then tell the stories of Marathon and
Salamis, stories that are all too familiar not just to the speakers, but also
to their narratees, the Spartans, and to ‘Thucydides’ own narratees.
Why, then, does Thucydides include this familiar story? At one level,
the embedded narrative has a structural role within his History. It
prepares for the Pentecontaetia (1.89–118), Thucydides’ account of the
period between the Persian and Peloponnesian wars, during which the
Athenians rose to power. The link between the two narratives is not
just a matter of textual and chronological succession: echoes within
the Pentecontaetia of the Athenians’ earlier self-characterization suggest
that the story they tell about their performance against Persia is itself
one of the drives to power.28 That is, here Thucydides does hint at the
historical importance of stories about the past. Further evidence for
Thucydides’ interest in the ideology of narrative is supplied by Pericles’
Funeral Oration (2.35–46). The story told by the Athenian speakers
at Sparta is one that was commonly told in Funeral Orations. But
Pericles refuses to tell the story of the deeds of their fathers’ or their
own generation ‘to people who know it’ (2.36.4). Here ‘Thucydides’
and Pericles’ narratees coincide: ‘Thucydides’ audience does not need
to know what has already been told in the Athenian speech at Sparta
and in the Pentecontaetia, Pericles’ audience does not need to be told
what it knows anyway.29
an iterative narration about tribute paid to Athens); 3.62.3 (the Theban constitution
in 480, with Pelling 2000b: 264 n. 31); 6.16.2 (Alcibiades’ performance in the chariot
race at Olympia: perhaps a correction of the epinician poem written by Euripides for
this victory, quoted by Plu. Alc. 11.2–3, who noted the divergence from Thucydides);
6.38 (alleged oligarchic plot at Syracuse: but the speaker, Athenagoras, is shown to be
unreliable in other ways).
28 Rood 1998: 244–246.
29 Similar coincidences between primary and secondary narratees occur in Homer
(→).
To view Pericles’ refusal to narrate as driven only by the demands

of Thucydides’ own narrative is to underplay the significance of his
gesture. Pericles refuses to narrate not only the story of the Persian
Wars, but also the Athenians’ great mythical achievements—deeds that
the Athenian speakers at Sparta pointedly passed over (1.73.2, quoted
above; contrast dismissals of the past by way of a transitional formula,
as at Hdt. 9.27.5; Livy 9.34.14). It is, moreover, not just the Atheni-
ans’ refusal of mythical narrative that demands explanation. Xenophon
includes mythical narrative in diplomatic speeches (e.g. Hell. 6.3.6), and
there is no reason to think that similar speeches were not made in the
Peloponnesian War. The absence of mythical narrative in Thucydides’
speeches may be because he makes his speakers share his own scepti-
cism about knowledge of the distant past (cf. 1.73.2). But it also points
to a theoretical interest in, and suspicion of, the use of narrative within
speech.
An interest in the role of narrative in speech also emerges from
variations between diplomatic speeches. In the debate at Athens over
how to punish the Mytilenaeans, allies whose revolt has just been
crushed, neither of the two Athenian speakers, Cleon or Diodotus, has
much use for narrative. But Diodotus does include one narrative (apart,
that is, from his impersonal history of the ideology of punishment at
3.45): ‘if you destroy the demos of the Mytilenaeans, who had no share
in the revolt, and when they got weapons, they willingly handed over the city,
you will do an injustice to your benefactors by killing them’ (3.47.3).
This coincides with his only appeal to justice. His deliberative speech
becomes, briefly, forensic. Most telling as an examination of the use of
narrative in speech is the Plataean debate. As in the Mytilene debate,
the introduction of narrative is linked with Thucydides’ exploration
of justice. The Plataeans have surrendered to the Spartans, and have
been asked whether they have done the Spartans any good in the
war (3.52.4). Rather than answer this question, they ask permission to
make a speech. They then tell a story about their past relations with
Athens, Sparta, and Thebes. Their problem is that they are telling
it to people who know it already (3.53.4)—and that it is a story that
must end (3.59.3). Unlike Scheherazade, they cannot postpone death.
The Spartans repeat their question, the Plataeans answer ‘no’, and are
killed. But by telling a story they do at least expose the hollowness of
the Spartan notion of justice.
Conclusion
We have seen that, unlike, say, Nestor and Phoenix in Homer or

Solon and Socles in Herodotus, characters in Thucydides are not much
given to sustained storytelling: they tend rather to insert snippets of
narratives dealing, rather generally, with their own or their opponents’
past behaviour. Nor does Thucydides, unlike later Greek novelists, play
with different narrative levels through elaborate framing devices. Such
experimentation would perhaps have been at odds with the seriousness
of his historical analysis. And it is the (emotional and moral as well
as political) seriousness of this analysis that explains why Thucydides
adopts so engaged, yet so elusive, a narrative voice. He explains how
he has gone about the task of collecting information, and then tells his
story. When he does intervene, his analysis reinforces the impression of
intellectual and emotional commitment.30 At the same time, readers are
left with much interesting work to do by the infrequency of Thucydides’
narratorial interventions, and by the absence of a narratee within the
text to guide their responses.
30 Cf. Gribble 1998: 43.

chapter nine
XENOPHON
V. Gray
Ancient historians claim to present a true account, and the historical

narrator has a special need to persuade his narratees that he is telling
the truth. Hellenica and Anabasis engage their narratees with various
kinds of narrative devices, some of them from Homer, others new,
which are designed mainly to inspire their belief in what they say.1
An anonymous narrator
Part of the historian’s strategy to secure a positive reception was the

projection of a persuasive persona. The preface was the place for this.
Hecataeus’ preface claims superior judgment (FGH 1 F 1a); Herodotus
(1.1–5) asserts the greatness of his subject and displays his research
method in his account of an early conflict; Thucydides (1.1–22) goes
much further, identifying himself, revealing his experience, demonstrat-
ing the greatness of his theme, displaying his research method, and
asserting his effort and his greater reliability over others (→ Thucy-
dides); and later historians followed his lead.2 The narrator of Hellenica
and Anabasis on the other hand discards the preface;3 he suppresses his
identity and plunges the narratees directly into the events. His virtual
absence as a personality from his text not only makes the events appear
reliable because unmediated,4 but projects an unspoken impartiality;
1 Arist. Rh. 1356a1–4, 1.2.3 indicates that speakers persuade their audiences by
projecting a persuasive character and making the audience emotionally disposed to

accept the persuasion, as well as by presenting them with logical proofs. Xenophon’s
narrators also adopt these modes of persuasion. Marincola 1997: 128–129.
2 Marincola 1997: 34–62 (greatness), 63–86 (method), 128–174 (experience, effort,
impartiality etc.).
3 The idea that Hellenica originally did have a preface has not met with acceptance:
MacLaren 1979.
4 Gribble 1998: 41.
130 part two – chapter nine
for he makes himself appear like one of those historians praised by

Lucian, who says that the historian’s absence from his text projects his
‘justice’ where praise and blame are concerned. Lucian indeed named
Xenophon as such a historian, and went on to describe the objective
historian as ‘a stranger in his text, a man without a city’, who has no
interest in the glorification or vilification of himself or his characters,
who neither spares friends nor grudges enemies.5
The narrator of Hellenica needed to assure his narratees of his impar-
tiality because he often intervenes to praise and blame contemporary
characters. This was even more true of the author of Anabasis, who is
a main character in the events recounted.6 ‘Xenophon, an Athenian’
first appears during the battle of Cunaxa, as an interpreter (1.8.15); but
from the third book he takes on a leading role, when the Greeks have
lost their leaders (he is mentioned no fewer than 230 times). No mean
device was needed to persuade the narratees to believe that the man
who writes about his own adventures does not falsify them to his own
credit.
Xenophon therefore not only removes himself from Anabasis, but
actually attributes the narration of his own deeds to another, who
describes ‘Xenophon’ in the third person. This narrator remains un-
named within the work, but Hellenica 3.1.2 summarizes the first four
books of Xenophon’s Anabasis and names as its narrator ‘Themisto-
genes of Syracuse’. Since no such person has been found as a candi-
date for authorship,7 most commentators follow the view that Plutarch
advances, namely that Xenophon created Themistogenes as a fiction
in order to make the record of his own achievements more persuasive.8
The same strategy is found in Isocrates, who adopts it to the same end.9
However, it is the first time it appears in historical writing. And as if
5 Hist. Conscr. 38–41; for the naming of Xenophon: 39; cf. D.H. Pomp. 5; Gray 1990;
Marincola 1997: 158. Modern historians accuse Xenophon of prejudice, particularly

toward his Spartan characters.
6 Hellenica in contrast does not directly describe Xenophon’s own achievements,
although 3.2.7 may refer to him as ‘the commander of the mercenaries of Cyrus’.
7 Krentz 1995: 157. His name, ‘Born of Themis’, suggests a narrator who tells the
truth.
8 Plu. Mor. 345 E expresses this view, in a discussion of whether the larger glory goes
to the characters or their narrator. Cf. Marincola 1997: 186; MacLaren 1934; Anderson
1974: 81.
9 Isocrates addressed one work to Nicocles of Cyprus in his own voice: Ad Nic. He
addressed another to the subjects of Nicocles, calling this work Nicocles. Instead of using
v. gray – xenophon 131
to confirm his objectivity, the narrator of Anabasis does not hesitate at

times to criticize the character Xenophon, e.g. in the account of how he
consulted Socrates about the wisdom of seeking the friendship of Cyrus
the Persian (An. 3.1.4–9). Socrates warned him that the Athenians might
blame him, because Cyrus had helped their enemies defeat them in the
recent war, and he told him to consult Delphi, but Xenophon asked not
whether he should go to Cyrus, but what gods would secure him a safe
return. This presents Xenophon as unwise, since his decision eventually
leads to his exile. The narrator Themistogenes can also praise the
friends of the Xenophon-character, such as Cyrus (An. 1.9), without
the impression of bias that would be attached to the author, and he
can show further objectivity by both praising and blaming Xenophon’s
friend Proxenus (An. 2.6.16–20).
The anonymous narrator of Hellenica reinforces his persuasive per-
sona when he continues Thucydides’ unfinished history unannounced,
beginning with a phrase that usually connects events within a contin-
uous work: ‘not many days after this’.10 Narrators of historical texts
traditionally desire to appear superior to their predecessors,11 but too
much competition could destroy the impression of trustworthiness and
make them appear unreliably self-promoting.12 The narrator of Hel-
lenica avoids this impression by not even bothering to indicate that
he has taken over from his great predecessor; this narrative ploy has
provoked speculation that the work was actually based on Thucy-
dides’ notes.13 Hellenica’s conclusion confirms that the narrator sees the
record of events as a continuum (‘let this be the limit of my writing;
what happened after this will perhaps be of interest to another’), but
again modestly characterizes him as just one in a line of writers. He
may show another aspect of his persona in his proleptic reference to
the period of even greater ‘confusion and indecision’ that followed his
his own voice, he introduces Nicocles as narrator, and makes him say that the king will
instruct his own subjects more persuasively than the author could.
10 The continuation Hell. 1.1.1–2.3.9 is much discussed: Gomme–Andrewes–Dover
1981: 437–444; Krentz 1989; Gray 1991.

11 Marincola 1997: 57–62. D.L. 2.57 notes that Xenophon could have claimed Thu-
cydides’ unfinished books and published them as his own work, but published them
instead under the name of Thucydides ‘for the glory of ’ Thucydides.
12 Plb. 12.7–11. Marincola 1997: 218–236 notes Xenophon’s lack of polemic self-
definition, but accepts that there are covert polemics (227).

13 Marincola 1997: 237–257 on the characterizing force of continuation.
chosen end-point. Hornblower notes that a refusal to deal with ‘guilt

and misery’ can show the narrator superior to others who delight in
them.14
An overt narrator and his narratees
Even if the narrators of Hellenica and Anabasis are not dramatized, i.e.
lack an explicit personality, they are overt narrators, in the sense that
they comment abundantly on their own narrative, interventions which
are all aimed at equally undramatized narratees.
By far the largest group consists of first-person interventions. The
‘continuation of Thucydides’ (Hell. 1.1.1–2.3.9) has no examples, but
one appears at Hellenica 2.3.56, another at 4.3.19, and they become
regular thereafter (4.4.12, 4.5.4, 5.1.4, 5.3.7, 5.4.1, 6.2.32, 6.2.39, 6.5.50,
7.2.1, 7.5.8, 19). Different periods of composition have been held to
account for this distribution, but it might also be interpreted in terms of
the narratologist Prince’s ‘persuasive rhythm’, the narrator increasingly
expressing his dominance over the narratees.15 The battle of Mantinea
(7.5), which is the climactic event of Hellenica, is more heavily marked
with narratorial interventions than any previous event. The same pat-
tern of engagement is observable in Anabasis 1, which reaches a climax
in the account of the battle of Cunaxa and the death of Cyrus, where
we find the first—laudatory—narratorial intervention (An. 1.9.24). In
general, Anabasis has fewer first-person comments than Hellenica, but
sometimes we hear the Xenophon-character evaluating in the narra-
tor’s place, as when he puts the rhetorical question to his army: ‘To
cross a difficult ravine and put it behind you when you are about to
fight, is not that an opportunity worth seizing?’ (An. 6.5.18). This is like
the rhetorical question of the Hellenica narrator about the folly of Iphi-
crates (Hell. 6.5.52—‘To being up many, but still fewer than the enemy,
how is this not complete folly?’).16
14 Hornblower 1994: 156. D.H. Pomp. 3 does not admire Thucydides because he
writes mainly about disasters.

15 Prince 1980: 22. Henry 1966 dismisses the ‘composition problem’, but it has been
a major feature in scholarship.

16 Another example: An. 7.7.23, where the Xenophon-character evaluates the virtue
of being true to one’s word. Other examples of characters sharing the narrator’s
evaluations: Hell. 4.8.4 (Dercylidas uses the narrator’s evaluative comment of 7.2.2);
Scholars have tried to identify a coherent programme of ‘what is

worthy of report’ in the first-person interventions in Hellenica.17 In fact,
we find that their function is to validate praise and blame, more specif-
ically to characterize the narrator as a person who is able to discrimi-
nate between what is and is not worthy of praise, and who in this proves
superior to his narratees. This makes them quite unlike the narrators in
Herodotus (→) and Thucydides (→). He acknowledges that what he
says is not immediately praiseworthy in their eyes, but will prove to be
praiseworthy on reflection. An example is a passage on Theramenes
(Hell. 2.3.56), where the narrator says: ‘I know that these remarks (he
has just recounted some jokes that Theramenes told as he was led off
to die) are of no worth, but this seems to me to be worthy in the man,
that with death at hand, neither wit nor playfulness departed his soul’.
This proves that the apparently unworthy is in fact worthy.18 When we
recall Hiero 2.3–5, where it is said that the ability to discern what the
majority do not marks the philosopher, this characterizes our narrator
as a philosopher.19 This does not exclude him from being a historian.
Indeed, in Hellenica 7.2.1 he puts himself in their company when he
declares that ‘all the other historians (sungrapheis)’ remember the single
deed of the large polis, but he thinks it still worthier to record the many
fine deeds of the small polis.
The narrator of Hellenica may intervene to mark the beginning or the
end of narratives, in the manner of Herodotus (→), but mainly when he
is about to narrate events that occurred at the same time as a preceding
set of events. This does not always involve a first-person remark (3.2.21,
31; 3.3.11, 3.5.25, 4.2.1, 4.2.23 etc.), but the first person marks prolepses
and analepses where the content is of special significance (6.5.1, 7.3.4,
7.4.1, and see below). Sometimes, in addition to bringing attention
to the process of narration, this type of intervention authorizes the
account by characterizing the narrator as a man of discrimination, as
at 4.8.1: ‘This is how the land war went. Events at sea and on the coast
2.3.52–53 and 56 (Theramenes uses the same words of Critias’ vice as the primary
narrator uses of Theramenes’ virtue).
17 Cf. the discussion in Tuplin 1993: 36–40.
18 Cf. Hell. 7.5.8, 7.5.19 and 7.2.1, where the ordinary expectation of the narratees is
represented as what the other historians normally praise; An. 1.9.24.

19 D.L. 2.48 sees Xenophon primarily as a philosopher and writes ‘foremost among
philosophers, he also wrote history’. D.H. Pomp. 5 calls Theopompus a philosophic

historian, because he uncovered virtue and vice and revealed things that are not easy
for the majority to know.
that occurred while these were going on, I will report now, and I will
record what is worthy of report and pass over those unworthy.’
Apart from first-person interventions the Xenophontic narrators
have many more devices that mark their presence and above all engage
their narratees. These, and the implicit forms of evaluation to be dis-
cussed later, often validate events that strain ordinary belief and per-
suade the narratees to believe them They also guide the narratees’
interpretation of events:
– instances of the ‘reference to the narrator’s own time’ and ‘con-
tinuance’ devices. The Hellenica narrator makes three references
to his own times, the first time to show that the Athenians have
remained loyal to the oaths they swore at the end of their civil war
(2.4.42), the second time to note the continuance of Tisiphonus in
power in Thessaly (6.4.37). The first implicitly praises the Atheni-
ans, the second seems to show that the dynasty was long-lasting
and therefore in some sense better than what preceded. Their
functions in Cyropaedia (→) are comparable. However, the descrip-
tion of a battle as unique ‘at least among those in our times’ (Hell.
4.3.16) uses the contrast between past and present to show that the
narrator is not willing to over-exaggerate (i.e. prove unreliable) by
proving it greater than battles of the ancient and venerable past.
– lessons which the narrator draws for the narratees, which may be
shared by the characters. An example is Hell. 5.2.7 (‘men acquired
wisdom in this respect at least, not to let a river run through their
walls’); the characters were victims of their lack of this wisdom,
as well as the narratees and the larger ‘mankind’ that they repre-
sent.20
– prolepses/analepses. The narrator of Hellenica usually narrates
events in sequence, but uses prolepses and analepses (the story
of the disaster at Lechaeum: 4.5.11–17; of Jason’s dynasty: 6.4.33–
37, Phlius: 7.2.1–46, of Euphron: 7.3.4–12) to mark the importance
of a theme (in these cases the remarkable insecurity of tyrannical
houses, remarkable loyalty, remarkable treachery, great disaster);
the narrator thus shows the narratees that he will allow impor-
tant events to spill over the time-frame. The analepsis at 4.5.11–17
secures a dramatic effect by delaying an account of a calamity
20 Cf. Hell. 6.2.19; 7.1.32 (cf. 7.2.9); An. 3.4.19–20 generalizes in the present tense
for the narratee what the Greeks as characters learned from their sufferings about
marching in the formation of the ‘square’.
until after the (praiseworthy) reaction to it. The main prolepsis in

Anabasis is the narrator’s account of what happened to Xenophon
after the end of the expedition, which recounts how he returned
from Asia, settled in Scillus, and fulfilled his vows to the god (An.
5.3). This confirms that he did obtain the safe return home that
the god had promised in the episode in the third book (above), and
thus acquits both of their obligations. Another prolepsis reveals
Xenophon’s future exile to the narratees, in precisely the man-
ner of Herodotus (→), being strategically placed at the very time
Xenophon is in all innocence preparing to return home (7.7.57);
the effect is pathos.
– tekmēria. The narratees sometimes need logical proofs to produce
belief, as in Hell. 6.4.13: ‘the fact that Cleombrotus and his men
were at first victorious in the battle may be known from this proof:
they would not have been able to take him up and carry him
off still living, had not those who were fighting in front of him
been holding the advantage at that time’.21 The narrator’s appeal
to tekmēria appears first in Herodotus, but whereas this aligned
the narrator of historical texts with the—medical and physical—
scholars of his time, by now it marks him more as a law court
pleader persuading his jurors.
– hōs eikos. The appeal to ‘what is natural’ often validates by contrast
responses that are unnatural. The recognition that the narrator
knows the difference engages the sympathy of the narratees to
accept the truth of what is unnatural: e.g. Theramenes ‘as was
natural’ complained about the injustice he suffered, in contrast to
the boulē, his jurors, who were terrified into unnatural silence (Hell.
2.3.55). An. 2.2.19 describes an uproar ‘as is natural, when people
are in great fear’, in order to contrast the abnormal self-possession
of their generals.22
– qualification. The Hellenica narrator engages his narratees in the
common belief that the gods cause the inexplicable, but he proj-
ects a persona properly aware of the limits of his own knowledge,
and of the scepticism of the narratees in this area. To this end
he qualifies the gods’ agency: the seer was driven ‘as if by some
divine fate’ (2.4.19; cf. 7.5.10); ‘the god, it seemed, was leading them
21 Cf. 5.2.6; An. 1.9.29–30.

22 Cf. Hell. 6.4.16; 7.2.15, 5.21; An. 3.2.24. Other passages contrast a first natural
reaction with a later unnatural one without this device: Hell. 2.2.3–4; 4.3.13.
on’ (6.4.3; cf. 7.5.13). He attributes the massacre at Corinth to the

gods: ‘the god brought it about’, but adds a rhetorical question as
if he is not sure that the narratees’ will agree: ‘how could a man
not think it divine agency?’ (4.4.12). He offers double motivation
‘either a god or men’ for the narratees to choose (7.4.32, 7.5.12–
13). Herodotus (→) uses praeteritio to convey this sense of delicacy.
Hellenica 5.4.1 is more forthright because the overthrow of tyrants
by a mere seven men, and the defeat of the previously undefeated
Spartans, are so inexplicable that even the narratees must think
the gods responsible, but the narrator still feels the need to sup-
port it further by presenting it as one among many other instances
where the gods have punished wrongdoers. Hellenica 7.4.3 suggests
that Lycomedes’ death was caused by a god because the coinci-
dence involved was too bizarre to explain otherwise. Not unex-
pectedly, because he says that the outcome confounded the expec-
tation of the entire world, the narrator abandons equivocation in
the climax of the work: the god did do it (Hell. 7.5.26).
The narrator of Anabasis has a more secular persona.23 He credits
little to divine agency even in a qualified manner (An. 1.4.18 has
characters give a divine explanation of an event which proves
false; 5.2.24 combines divine and human: ‘someone’ set fire to a
house, causing the enemy to run, and this ‘fate’ teaches Xenophon
what to do; the unknown agent is a sign of divine intervention:
Hell. 4.5.4; An. 4.7.26). His characters however, particularly Xeno-
phon, acknowledge the role of the gods in their affairs instead (An.
3.2.6–13).
– gnomic utterances. Hellenica 4.5.6 comments on how conquerors
provide a great spectacle to the viewer, which makes the narra-
tees appreciate the contrast when the erstwhile conquerors are
defeated and travel at night to avoid being a spectacle of another
sort (4.5.18). At 7.3.12, juxtaposition of the condemnation of
Euphron by the Thebans and the contrasting honour done to his
dead body by his citizens evokes the interpretation: ‘Thus, so it
seems, most people judge their benefactors to be good men’. This
sends a message about the power of benefaction, which characters
recognize elsewhere, and how it proves greater than other virtues
(cf. the sentiment at 4.8.4).
23 Tuplin 1993: 215 notes that this is unusual in an author who in his other works is
so devoted to the gods.

– rhetorical questions. The rhetorical question is used to engage

the narratees in an agreement with the narrator, on the things
that lead to good morale for example: ‘Wherever men do this,
how is it not likely that they (would) not (have been) be heart-
ened?’ These matters are usually incredible in some way and
require some logical thought from the narratees. Other rhetorical
questions attribute incredible slaughter to the gods (Hell. 4.4.12),
incredible folly to Iphicrates (6.5.52), incredible courage to cavalry
(7.5.16).
– the ‘presentation through negation’- device. This is used to antic-
ipate and confound the narratees’ expectations (and thereby ac-
tively engage them in the evaluation). This happens not only in
passages of praise and blame, e.g. Hell. 5.1.4, where the narra-
tor confounds their expectation that he should have written about
some achievement that is great by ordinary standards and thereby
makes his own judgments appear controversial and more percep-
tive,24 but also in other passages: e.g., Hell. 5.3.20, where Agesi-
laus did not feel pleasure in the death of his rival ‘as a man (i.e.
you, the narratee) would have expected’; the effect is of course
to make the judgment of Agesilaus exceptional in its compas-
sion.25 The confounded expectation is often shared by the char-
acters. Tissaphernes, for example, shares the expectation of Hell.
3.4.21 that Agesilaus will deceive.26 Frequently characters also con-
found the expectations of the narratees when they ‘do not delay’
(Hell. 3.2.10). Hellenica 4.1.18 is expectation based on inferior num-
bers.
– the ‘anonymous interlocutor’ device, which occurs in Hell. 5.1.4,
19; 5.2.7; 5.3.7.
– visualization. Rhetoric persuades law court judges that alleged
events did really occur by producing witnesses (as well as tekmēria
and proofs of plausibility). The narrator uses the technique of visu-
alization to secure belief by using the ‘anonymous witness’ device,
which draws in the narratees. For example, after the Spartan
defeat at Leuctra, the ephors forbid all forms of public mourn-
24 Cf. Hornblower 1994: 152–158, who reads negative presentation as polemic
against (real) audience expectations.

25 Other examples use the same device to various other ends: Hell. 2.4.19; 7.5.8–13;
An. 3.4.14.
26 The expectation that Lysander will not take action is shared by the Spartan
authorities (Hell. 3.5.18; cf. 3.5.3; 4.8.35–37).

ing: ‘And on the following day one could see those whose relatives
had been killed going about in public with bright and cheerful
faces …’ (Hell. 6.4.16).27
– ‘if not’-situations. Another means to engage the narratees is by
confronting them with what might have happened, as e.g. in Hell.
7.5.10 above, Epaminondas ‘would have’ taken Sparta, ‘had it not
been’ that a Cretan ‘by some divine luck’, warned the absent
Spartan army.28
The narratorial interventions therefore organize the narrative, mark
the narrator’s areas of interest, reveal his discrimination, establish his
lack of prejudice, and persuade the narratees to develop their ordinary
perceptions to match the narrator’s more philosophic ones. Two sets
of narratorial interventions are particularly rich in their implications,
which balance praise and blame for two individuals: Teleutias the
Spartan and Iphicrates the Athenian.
Example I: Teleutias the Spartan
The first narratorial intervention concerning Teleutias takes the form

of a summary judgment, which evaluates the enthusiasm of his men
for Teleutias as their commander at the moment he is leaving them
and sailing home (Hell. 5.1.4): ‘I know that in these (events) I am nar-
rating neither an expenditure nor perilous undertaking nor any mem-
orable stratagem, but I swear that this seems to me to be worthy for
a man of worth (andri), to consider, what Teleutias did to create such
dispositions in those he ruled. This is an act of a man of worth, most
worthy of narration, more worthy than great expenditures or perilous
undertakings.’ ‘Presentation through negation’ represents the ordinary
perception of the narratees, which the narrator can understand, but he
then reveals the hidden worth that the narratees have not seen, invit-
ing them to be ‘real men’—one who after reflection might appreciate
the achievement of Teleutias, another ‘real man’. The distance closes
between narrator, narratees, and character. The summary judgment on
Teleutias prefaces the ensuing narrative, guiding the narratees to read
27 Cf. Hell. 3.4.16–20; An. 1.5.8, 9; 2.3.11; 4.7.13–14.

28 Cf. Hell. 5.2.41; 6.2.23; 7.5.10; An. 4.1.11 for other expectation-raising near-misses.
it in terms of the evaluation: the commanders who replace Teleutias

provide negative models (5.1.13). Teleutias then returns to the fleet to
confirm the narrator’s earlier positive evaluation. The first proof is in
character-text: Teleutias announces his intention to secure provisions
for his men and share their hunger until he does so (5.1.14–17). The
second is in narrator-text: his attack on Piraeus (5.1.19–24). This pas-
sage begins with a very similar summary judgment involving the nar-
ratees: ‘if anyone supposes that he was foolish/unwise to sail against
those who had many ships with only twelve, let him consider the calcu-
lation he made’; the use of the ‘anonymous interlocutor’ device suggests
that the narrator is in fact addressing a large public. Whereas they only
find folly, the narrator detects hidden reason. Teleutias’ own discern-
ment is then evaluated, first in the account of his reasoning (5.1.20),
and then in the narrative, where the ‘presentation through negation’
device highlights the foolish course that he did not take, in order to
endorse the wise course that he did take: when he neared Piraeus
he took no action, but moved at dawn; he did not allow anyone to
harm a cargo ship, but encouraged them to harm the triremes; he thus
secured provisions, kept the ships manned and the men willing (5.1.21–
24).
The same material does not always evoke narratorial comments.
Thus, the enthusiasm that their men showed for Hermocrates (1.1.27–
31) and Agesilaus (4.2.1–5) is not evaluated. This makes Teleutias spe-
cial, and could suggest prejudice in the narrator, but he evaluates
another act of Teleutias in terms of blame within a very short space
(5.3.5–7). Summary judgments again preface and close this section.
They do not use presentation through negation but engage the nar-
ratees in a different way. The narrator recounts how Teleutias in anger
pursued the enemy close to their walls, and then interrupts his account
to make the following comment: ‘Many others too have pursued the
enemy closer to the walls than is opportune and got back badly, and
these too … were compelled to retreat in complete confusion.’ The
generalization (‘many others …’) leads the narratees to anticipate a spe-
cific outcome and thus creates a (horrified) suspense. Indeed, Teleutias’
death and the slaughter of his men follows. A closing summary judg-
ment then draws the larger lesson: ‘From such sufferings as these I say
that mankind is instructed most of all that it is not right to punish even
servants in anger (for often masters have suffered more harm than they
inflicted in anger), but to attack an enemy in anger without calcula-
tion is a complete blunder. Anger does not look ahead, whereas reason
looks to avoiding suffering no less than inflicting harm on the enemy.’

This passage confirms that the narrator poses as a philosopher, who
considers the whole of mankind as his narratees.
Example II: Iphicrates the Athenian
The Athenian Iphicrates is subject to the same evaluation as Teleutias

the Spartan. The narrator thus proves willing to praise and blame both
Spartans and Athenians, indicating that he has no partiality for poleis
or individuals. The narrator first praises his campaign against Corcyra.
The Athenians had appointed him because his predecessor had wasted
time in trying to find trained crews (Hell. 6.2.13). Iphicrates solves
this problem by sailing and training at the same time (6.2.27). The
narrator ends the section with a summary judgement, which involves
the narratees (6.2.32): ‘I know that all this training and exercise is done
when men think they are about to fight at sea, but I praise this, that
when he had to proceed in a hurry to where he thought he would fight
the enemy at sea, he found a way to make his crews not inexperienced
of things relating to fighting at sea in spite of the journey, but not
to arrive there more slowly on account of the practice.’ Once again,
the narrator suggests that he has greater powers of perception than his
narratees, who can see only the surface meaning. Another narratorial
intervention follows in a section on Iphicrates’ selection of colleagues
for this campaign (6.2.39). ‘I praise this campaign in particular of all of
Iphicrates’ campaigns, and further his instruction to choose in addition
to himself Callistratus the demagogue, who was not on good terms with
him, and Chabrias, who had a big reputation as a general …’ This
evaluation disrupts the natural time sequence, since the selection of
colleagues happened before the campaign began, whereas we are now
dealing with its first phase. The effect of this analepsis is to strengthen
the impression of Iphicrates’ speed, which is a vital element of his
praise, because the praise of his selection of colleagues at the beginning
of the campaign would have delayed the narrative, whereas its location
here, after his speedy arrival is achieved, does not.
The criticism that balances the praise (6.5.51–52: ‘If he was a good
general on any other occasion, I do not blame him. But his actions on
that occasion, I find to have been all useless and even positively unfortu-
nate’) makes a new and direct assault on the question of narrator prej-
udice, by denying that the narrator is incapable of praising Iphicrates
when the occasion arises. The negative ‘I do not blame him’ confronts
the narratees’ expectation that the narrator might be so prejudiced
that he is incapable of anything but criticism. (They do not apparently
remember the narrator’s previous praise, in a case of contextualized
amnesia.) The narrator strengthens the justification of his criticism in
this case by appealing to ‘anonymous spokesmen’ who share his criti-
cal view: ‘They said that many (of the soldiers involved) turned up for
the campaign before Iphicrates’ in the midst of the general enthusiasm,
and when he delayed at Corinth ‘many blamed him epsegon [the word
the narrator uses of his own blame] for delay at first’ (6.5.49).29 The
narrative then supports the narrator’s criticism through ‘presentation
through negation’ (though he wanted to prevent the Thebans escap-
ing, he left the best pass unguarded), narratorial comment, and appeals
to the narratees via the particle kaitoi and a negative rhetorical ques-
tion: ‘Wishing to find out where they were, he sent all the cavalry as
scouts. Yet a few are no less able to see than many, and if there is a
need to retreat, it is easier for few than many. And to bring up many
who are still inferior to the enemy, how is that not complete madness?’
(6.5.52).
Implicit forms of evaluation
Apart from the explicit narratorial interventions, the Xenophontic nar-

rators use a variety of narrative devices that implicitly validate events,
guide the narratees’ interpretations and characterize their ordinary
beliefs:
– similes and metaphors. The narrator reveals his own world and
that of his narratees in similes and metaphors, which are almost all
from nature and ordinary life. These evoke emotional engagement
in the narratees (cf. Oec. 17.15), but mainly make familiar to them
what otherwise is difficult to believe or conceive, and are thus
persuasive. Hellenica 4.4.12 makes the familiarity explicit when it
29 The narrator often voices his criticisms of commanders through their men, in
keeping with his interest in commanders’ relations with their men: e.g. Hell. 7.1.17–18
first sets out the advantageous course of action that a commander rejected, and then
cites a ‘majority’ to whom the adopted course ‘seemed’ disadvantageous.
defines unimaginable numbers of corpses in a small space: ‘men

accustomed to seeing heaps of grain, wood, stones, then gazed on
heaps of corpses’. 7.5.10 captures the unimaginable vulnerability
of the polis when it describes Sparta (in the absence of her military
men) as a nest of chicks, without the adult birds to defend them.
This guides the narratees to fear for the chicks, grieve with the
absent parents, or admire Epaminondas as the bold killer bird,
and expect that the nest will be taken.30 Anabasis uses images to
explain sights that the Greek narratees have not seen before; this
fits its foreign content: 1.5.1 compares a completely level plain to
the sea, a smell to spices; 1.5.3–4 says that the flesh of strange
birds is like venison and that ostriches use their wings like sails;
1.7.8 compares the dust raised by a huge army to clouds.31
– descriptions. The narrator of Anabasis gives an account of a jour-
ney and on the way describes—much in the manner of Herodotus
(→)—the breadth of rivers and bridges, the stages of the march
and rest periods, mentions ‘large and prosperous’ cities, names
the rivers that run through them and narrates stories about them
(e.g. 1.2.5–9). He describes the customs and flora and fauna too
of the lands through which the journey goes (e.g. 1.4.9, 1.5.1–3).
Such descriptions are found less commonly in Hellenica because
it is set in environments dominated by Greek culture (including
Greek Asia). However, the narrator does express the love of natu-
ral beauty through the eyes of characters, for example the pathetic
desire of the feverish Agesipolis for the shade and the cool water
of a shrine (Hell. 5.3.19). Descriptions also often have a narrative
function. Anabasis informs the narratees about Persian military
arrangements that prove crucial to the understanding of events
30 Other examples: the Spartan campaign against Elis is an ‘harvest’ for the Pelo-
ponnese, to encapsulate the richness of the plunder and the unbelievable lack of mil-
itary resistance (Hell. 3.2.26); the complete devotion of the demos to their champion is
captured in the image of bees swarming around a king bee (3.2.28); the description of
Ephesus as a workshop of war captures the unimaginable activity (3.4.17); the surprising
behaviour of Pharnabazus, constantly changing his camp-site, is like that of nomads
(4.1.25); the Spartans compare their allies’ excessive fear of the peltasts to children’s fear
of giants (4.4.17); Agesipolis as pentathlete shows the huge range of his competition with
Agesilaus (4.7.5); the comparison of Epaminondas’ battle formation with the prow of a
trireme makes sense of his unusual dispositions (7.5.23).
31 Cf. 1.5.6; 2.3.15; 2.4.13; 5.4.28.
(1.8.22; cf. 3.4.35). Hellenica 4.1.15–16 describes Pharnabazus’ fine

palace estate in Phrygia because it is about to be ravaged.
– juxtapositions. The organization of the narrative itself can also
evaluate, as for example when the prolepsis on the remarkable
loyalty of the Phliasians (Hell. 7.2) is juxtaposed with the prolepsis
on the remarkable treachery of Euphron (7.1 and 3).
Secondary and reported narrators
The Xenophon-character in Anabasis 5.7.13–26 and 5.8.8–11 recounts

narratives to the men for the sake of information, in these cases to
correct their impressions. The second example is interesting because
the primary narrator has already told of Xenophon’s admirable role in
the march through the snow, but has not mentioned this incident (4.5),
which ‘Xenophon’ now tells in his own defence. The primary narrator
had no need for such defence and this perhaps shows the degree to
which he is objective about Xenophon. Reported narrators in Hellenica
take the form of anonymous reporters of victories or defeats; their
being eye-witnesses is a required quality (4.3.2), like that of messengers
in tragedy (→ Aeschylus, Sophocles, Euripides). The point of these
narrations seems to be to evoke a reaction from other characters,
which is then evaluated. Thus, the messenger’s reported speech of
the disaster of the battle of Cnidus (Hell. 4.3.10–12) is followed by a
description of Agesilaus’ reaction to it. This is the case also with the
shorter report of 7.1.32, which allows an assessment of the reaction of
the Spartans to their first victory in a long time. In another instance
(4.5.7–17), a messenger briefly reports a disaster without details and the
primary narrator describes the immediate reaction of Agesilaus, but
then he gives his own more complete report of the precise nature of the
disaster. It is as if the narrator does not want to delay Agesilaus’ worthy
reaction by letting the messenger give him too full a report in the first
instance. It is significant that the reaction is the point of most embedded
narratives also in Xenophon Cyropaedia (→). Hellenica 1.1.23 contains a
laconic secondary letter-narrative: ‘Ships gone, Mindarus dead, men
starving, what to do?’, but here too it is Pharnabazus’ reaction to the
defeat that takes centre stage. This focus is in keeping with the primary
narrator’s interest in the evaluation of action.
Historians persuade narratees to believe the unbelievable, such as
great numbers, by referring to their own autopsy or another source of
information.32 Xenophon agrees that witnesses are authoritative,33 and

even though his primary narrators make no reference to autopsy, they
do use anonymous spokesmen, characters involved in the action as eye-
witnesses, a source of persuasion known from Herodotus (→) onwards.
They are mostly unnamed, and they confirm the primary narrator’s
account on details that defy ordinary belief. They have special interest
for modern historians because they look like sources, but they are once
again also a rhetorical feature.34 Hellenica 7.5.12 points this way when it
substitutes ‘it is possible to blame … it is possible to say’ for the more
concrete ‘they blame … they say’.
The only named spokesman is Ctesias. The narrator recounts the
charge of Cyrus at Cunaxa and the wounding of his brother, King
Artaxerxes, but uses Ctesias to confirm the power of the blow Cyrus
gave Artaxerxes ‘right through the breastplate’ (An. 1.8.26–27).35 Ctesias
‘says’ that he treated the wound. His description as ‘the doctor’ points
to his special value as an eye-witness, for a doctor is close to the
casualties; he later ‘says’ how many died on the king’s side. We have
to go outside the text to discover that Ctesias ‘speaks’ as an historian as
well as a doctor (Ctesias did declare that he was an eye-witness: FGH
688 T 8). The narrator uses this same pattern of shared authority when
he describes in his own voice how Cyrus routed those around the king,
but introduces anonymous spokesmen (legetai) to verify his sensational
slaughter of their commander ‘with his own hand’ (1.8.24).36
32 Marincola 1997: 80–83.

33 Marincola 1997: 69. Agesilaus tells Dercylidas that he would best report his news
to the allies because ‘you were present’ (Hell. 4.3.1–2); Ctesias is cited for the wounding
of Artaxerxes because ‘he was at his side’ (An. 1.8.26–27). Characters for this reason
dramatize their encounters with protagonists in their speeches (Hell. 3.3.4–11; 6.1.2–
17). Clearchus dramatizes the trial of Orontas by Cyrus (An. 1.6.5) as the only Greek
eye-witness present; the narrator says that there was ‘no ban’ on him telling the story,
to highlight the openness of Cyrus’ judicial procedure. The obituaries of Cyrus and
Clearchus also appeal to ‘what was said’ by those with personal experience of their
leadership (An. 1.9.1; 2.6.1).
34 Tuplin 1993: 39 n. 91 sees them as a stylistic quirk of the later Hell. This needs
some refinement; see Gray 2003.

35 There is a pattern in which the narrator vouches for some stages of an action, but
leaves the most sensational to the spokesman; An. 1.8.18 has the narrator vouch for two
phases of the action, leaving spokesmen to verify the culmination, that they clattered
on their shields to frighten the horses.
36 Anonymous spokesmen verify also: the enormous numbers who opposed Cyrus:
deserters before the battle and survivors afterwards (An. 1.7.13); the almost unbelievable
lack of Greek casualties at Cunaxa (An. 1.8.20), a long retreat (1.10.1), a large number of
wagons (1.10.18).
The narrator makes a similar use of spokesmen in Hellenica too, e.g.,

when he describes the campaign of Derdas in his own voice, but lets an
anonymous spokesman verify the enormous number of cavalrymen he
personally killed over a ninety-stade pursuit (5.3.2).37 The other kinds
of material that attract spokesmen in Hellenica are sensational in other
ways. The dramatic strangling of Mania by her son-in-law is one good
example.38 Sometimes, spokesmen mark the first part of a remarkable
contrast, such as the depth of the Spartan line, to contrast with the
remarkable depth of the Theban line (Hell. 6.4.12).
The actions of the gods also attract spokesmen: they vouch for
Jason’s sensational intentions concerning the festival of Delphi; the god
alone ‘is said’ to have known his intentions toward the sacred treasures
(Hell. 6.4.29–30); they also report the four ominous events preceding
Leuctra (that the girls killed themselves after being raped, the temples
opened of their own accord, the gods were heralding victory, Heracles’
arms were missing) and their interpretation that these were all human
contrivances (Hell. 6.4.7–8).
The equivalent of spokesmen in some cases is ‘seeming to be’, which
is used to mark superlative reputations, as in Cyropaedia (→). Dercylidas
is described as ‘a man of worth seeming very much to be resourceful’
(Hell. 3.1.8), Thrasybulus as ‘seeming very much to be a good and fine
man’ on his death (Hell. 4.8.31). The equivalence between ‘seeming’
and ‘being said to be’ (which is also used of reputations: Hell. 3.2.27) is
spelled out in the case of Dercylidas: ‘for he was called (evidently ‘said
to be’ by others) “Sisyphus”’.
Xenophon’s narrators occasionally give us unresolved alternative
accounts or motivations, some of them attributed to spokesmen (Hell.
3.5.19; 5.4.7; 6.2.39; 6.4.37; 7.4.32; 7.5.12; An. 1.2.25, 1.8.29–30). While
these could express uncertainty and characterize them as less than
omniscient,39 they also often give the narratees two angles on the cen-
tral characterization to which they point, and thus through dissonance
on detail produce consonance on the substantial point. The killers of
the Theban tyrants gained entry to their drinking party disguised as
respectable citizen women or as clowns—but both point to the low-life
37 Hell. thus verifies: a large quantity of cash (6.2.16); enormous numbers of sacrificial
beasts (6.4.27–28); vast numbers of helots (6.5.29).

38 Such as remarkable sayings (Hell. 1.6.32; 2.3.56; 4.4.10; 7.1.30 and 32; 7.4.40);
unbelievable beauty or luxury (Hell. 3.2.27; 3.3.8; 3.2.10; 5.4.57; 6.2.6; 6.4.8); remarkable
actions or reactions (Hell. 3.5.21; 4.8.36, 6.4.37, 6.5.49; An. 1.2.12, 1.2.25, 1.10.7).
39 So Breitenbach 1950: 23–26.
interests of the killers as the reason for their downfall; Alexander of

Pherae is killed by his wife either because he is abandoning her for her
infertility or because of his violence toward the boy she has asked him
to spare—but both point to the lack of friendship toward her that made
him kill her; the courage of the Spartans could come from the gods or
sheer mad desperation—but both indicate how superhuman it was; and
Menon’s men perish either because they plundered friends or because
they became lost—but both point to their essential lack of discipline.
Conclusion
The narrators of Anabasis and Hellenica characterize themselves by their

absence as personalities and at the same time their presence as nar-
rators. They use explicit narratorial interventions, but also orchestrate
the reception, above all the acceptance of their at times sensational
information, through more implicit devices. Hellenica reveals in partic-
ular the narrator’s ability to perceive behind appearances, which is the
mark of the philosopher, in matters of praise and blame. The narratees
of both works have limited discernment of virtue and vice, which need
evaluation to be appreciated, and limited expectations, in which any-
thing excessive (numbers, behaviour, conditions, achievements, specta-
cles, customs etc.) must be made familiar (through visualization, con-
trast with ‘what is natural’, qualifications, images), supported with evi-
dence (proofs), or attributed to reliable first-hand witnesses (anonymous
spokesmen). They have to be persuaded and educated by a superior
narrator.
chapter ten
POLYBIUS
T. Rood
Hamilcar brought the mercenaries to such a pass that … they were at

last driven by famine to eat each other: the divine power was bringing
on them a fitting retribution for their violation of all law human and
divine in their treatment of their neighbours. (1.84.9–10)
Polybius, historian of Rome’s rise to universal rule, and connoisseur of

punishment,1 makes everything explicit. ‘Es gibt keine Polybiosfrage’—
‘Polybius presents no problems.’2 But there are problems in assess-
ing the place of this explicit Polybius in the history of Greek narra-
tive. It used to be conventional to claim that Ephorus, writing soon
after Xenophon, inaugurated a ‘new phase of Greek historiography’ by
‘attempt[ing] the fusion of rhetoric and history’; that ‘a still lower level
was reached by Duris of Samos in the theory that history must affect
the emotions’; and that ‘it was left for Polybius two centuries later to
affect the redemption of Clio from the bondage of fiction’.3 No one
nowadays would be satisfied with this simplistic story. It rests on an
uncritical assumption that the Thucydidean way of writing history is the
way to write history, and on a naïve belief that that mode of historiogra-
phy is itself free from rhetoric. And it founders on the almost complete
loss of the historians writing between Xenophon, in the first half of
the fourth century BC, and Polybius, in the second half of the second
century BC. We have to reconstruct lost historians from later writers
who used them (such as Diodorus, Plutarch, and Arrian), from unrep-
resentative excerpts in writers such as Athenaeus, and not least from
the hostile comments of Polybius himself. The dangers of circularity
in assessing what sources later historians used, and of underrating the
capacity of those later historians to adapt their sources, remain great.4
1 For Polybius’ (occasionally sadistic) moralizing on punishment, cf. e.g. 2.60.7;
4.18.7, 81.5; 5.28.9, 56.13, 111.7.

2 Howald, quoted by Walbank 1948: 157.
3 Barber 1935: 159.
4 Important accounts of other historians between Xenophon and Polybius that
148 part two – chapter ten
The problems of writing a narratological history of Greek histori-

ography between Xenophon and Polybius are greater still. One can
(imagine that one can) get an idea of the tenor of some authors from
later historians who drew on them, and from the comments of later
critics. A stronger narratorial presence would presumably be demanded
by the stress on praise and blame associated with historians such as
Ephorus (Polybius himself praises Ephorus for his ‘expressions of his
personal judgments’, 12.28.10 = FGH 70 T23), and by the methodologi-
cal, and often polemical, prefaces that became standard once book divi-
sions were introduced.5 A less intrusive narrator might conversely be
expected in the visual and emotive mode of historiography associated
with Duris of Samos. But this assumption would apply only to narra-
tive passages: we know that Duris himself intruded into his narrative
by making methodological criticisms of his predecessors (FGH 76 F1,
from the preface, and naming Ephorus and Theopompus). Our prob-
lems stem from the facts that authors who used these historians were
interested in their content rather than in their style; and that one needs
their actual words to form an idea of how they presented themselves
as narrators. Thus while we know of some historians who were char-
acters (i.e. internal narrators) in their own works—Alexander’s general
Ptolemy (FGH 138), for instance, or his admiral Nearchus (FGH 133),
we do not, on the whole, know how they presented themselves.6 In
some cases, however, the actual words do survive, and fruitful parallels
can be drawn with features of Polybius’ narratorial style. Thus there
are a number of first-person forms, as well as rhetorical questions, in
Theopompus’ famous description of the debauched court of Philip II,
a passage cited by Polybius himself (8.9.6–13 = FGH 115 F225a; cf. also
F210).
touch on some narratological aspects include Flower 1994 on Theopompus; Pearson

1960 and Bosworth 1988 on the Alexander historians; Hornblower 1981 on Hierony-
mus; Pearson 1987 on Timaeus. For Athenaeus as a source for historians, and for Poly-
bius himself, see Pelling 2000a and Walbank 2000; and for the problems posed by Poly-
bius’ polemics, see Walbank 1962 and Schepens 1990. For Diodorus, see the contrasting
views of Sacks 1990, who stresses Diodorus’ own thematic shaping, and Stylianou 1998:
1–139, who, e.g. sees shifts in the style of moralizing judgment as reflecting shifts of
source (pp. 5–10).
5 Book divisions were perhaps introduced by Ephorus: cf. FGH 70 T10; also F7 for
the prefaces of Ephorus and Theopompus as similar, and FGH 115 F24 for the abusive
prefaces of Anaximenes and Theopompus.
6 We can, however, observe the use of first-person plural forms in geographical
t. rood – polybius 149
There are also problems in attempting a narratological analysis re-

stricted to Polybius alone. Only the first five (of forty) books of his
history survive in full. For the remaining books, we rely on the use of his
work by later historians and on Byzantine excerpts that do not always
keep to Polybius’ own phrasing. It is also unfortunate that we cannot
compare his narrative persona in his other works, all of them lost
(a monograph on the Numantine War, a biography of Philopoemen,
a tactical treatise that would presumably—like other such treatises—
have offered narrative examples). But analysis of the existing parts of
Polybius’ history is still worthwhile for the light it can shed on the
development of a scholarly mode of historical narrative, and on the
political implications of such a mode of narration. In this chapter, I will
examine first the character of the primary narrator in Polybius, then
the various narratees addressed in the work, and finally the shifting
weight given to different types of secondary narrative.
‘Polybius’ is an external narrator in the earlier parts of his work, and

an internal narrator in its later stages. He begins with two preliminary
books (the proparaskeuē), which treat events from the First Punic War
down to 220 BC, that is, before he was himself born. In books 3–30, he
fulfils the promise made at the start of the work: to tell of ‘the means
and system of polity through which the Romans in fewer than fifty-
three years succeeded in subjecting nearly the whole inhabited world to
their sole government’ (1.1.5). He first appears as a character towards
the end of this second part. It is in the final section of the history—
where he reports ‘the condition of each people after all had been
crushed and had come under the dominion of Rome, until the con-
fusion and disturbance that afterwards ensued’—that Polybius appears
as a character most often. Indeed, he claims that he was induced to
write about this final period of confusion and disturbance ‘as if starting
on a fresh work … chiefly because I not only witnessed most but took
part and even directed some of the events’ (3.4.12–13).
While the character Polybius only appears towards the end of the
work, the narrator ‘Polybius’ is a dominating presence from the start.
works such as Ps.-Hanno’s account of an expedition down the west coast of Africa
(GGM 1.1–14).
As Marincola says, ‘the Polybian narrator combines a largely unobtru-

sive narrative of the deeds with a highly intrusive explicator of that nar-
rative’.7 A good example of this fondness for commenting on the narra-
tive comes in the account of the Second Punic War, where the Romans
are besieging Capua, which is held by the Carthaginians, while Han-
nibal is outside the town trying to force the Romans to abandon the
siege: ‘It seems to me that events at that time would puzzle not only
the Carthaginians, but anyone who heard of it. For who could believe
that the Romans, who had been beaten in so many battles by the
Carthaginians, and did not yet even dare to face the enemy in the field,
nevertheless refused to retire or to abandon the open country?’ Here
the comment, and the rhetorical question, are responses to a problem
felt both by characters and by narratees—a problem that threatens the
cohesion of Polybius’ narrative. Hence the need for an explanation: ‘It
seems to me that the reason of this conduct on the part of both was that
both had perceived that Hannibal’s force of cavalry was responsible for
the Carthaginians’ victories and the Romans’ defeats. So … events at
that time around Capua happened for both sides with good reason’
(9.3.5–11). The strong narratorial presence offers the reassurance that
events are after all explicable.8
The intrusiveness of the Polybian narrator is also seen in his use of
evaluative words within the narrative, and in the tone in which those
evaluations are expressed. He is ready to speak of how the Illyrian
queen Teuta acted ‘with a womanly temper and irrationally’ (2.8.12,
cf. 4.8), and of how the Spartans ‘were liberated through Antigonus
and through the generous zeal of the Achaeans’ (4.16.5). Often a polit-
ical bias can be readily detected in the narrator’s judgments. It is Poly-
bius’ own Achaeans, and not the Macedonian king Antigonus, who
are credited with a ‘generous zeal’; and evidently the ‘liberation’ they
procured is itself evaluative.9 Conversely, it is the Achaeans’ enemies,
the Aetolians, who are castigated as ‘innately unjust and aggrandizing’
(2.45.1). Polybius’ generally earnest narrator can at times seem surpris-
ingly crude.
What is striking about the way in which ‘Polybius’ expresses these
evaluations is that he often uses in the narrative phrases that tend to
7Marincola 1997: 10.

8Cf. Davidson 1991: 11–12 on Polybius’ fondness for correcting mistaken views.
9 Cf. Walbank 1957–1979: ad loc.: ‘to the Achaeans Cleomenes was a tyrant; but to
many Spartans “liberation” obviously had a different look’.

be used only in speeches by earlier historians such as Thucydides and

Xenophon. Frequent in Polybius’ narrative as well as in his speeches
is the exclamation nē Dia, ‘by Zeus’, which is common in comedy and
oratory, but found only four times in Xenophon’s Hellenica, each time in
direct speech, and absent altogether from Thucydides.10 Polybius also
frequently supports his judgments with rhetorical questions: when he
describes how the Aetolians elect as general Scopas, ‘the cause of all
the aforementioned acts of injustice’, he first remarks that ‘I do not
know how to express myself ’ (the common aporia-motif), then says that
it ‘seems to me the very height of villainy’: ‘for how can we characterize
otherwise such base conduct?’ (4.27.2–3). Here the question suggests
some uneasiness about offering, within a historical work, judgments in
a heightened manner more redolent of the law courts.11
The Polybian narrator is no less intrusive when he justifies the meth-
ods of his narrative.12 He explains principles of history-writing: the
function of geographical information (3.36–38, 57–59), the use of ter-
minology for causation (3.6–7), the usefulness of history (e.g. 1.35, 3.31–
32)—and this despite the fact that he states in the very first sentence
of the work that he does not need to explain why history is useful.
His sense of a historiographical tradition is far stronger than anything
that can be found in earlier historians. He defends the merits of history
against such varied competitors as barbershop gossip (3.20.5), invective
(12.14. 2–7), political theory (6.5.1–3), philosophy (12.26c), and epideic-
tic (12.28–28a), while criticizing Ephorus’ unfair castigation of music as
deceptive (4.20.5 = FGH 70 F8: from Ephorus’ preface, so presumably
from a contrast between music and history).
The ordering of the narrative is also subject to the same rigorous
narratorial control. The ends of digressions, for instance, are clearly
marked as the story returns to the point ‘from which we digressed’
(parexebēmen, e.g. 2.36.1; 4.9.2). Such signalling is also found in ear-
lier historians (Ephorus, FGH 70 F191, is very similar), and also in
geographical periploi (coastal accounts), where the narrator returns to
a mainland from an island with the formula epaneimi … hothen …
exetrapomēn (‘I will go back to the point from which I turned aside’: e.g.
10 De Foucault 1972: 313 with n. 2 lists examples.

11 For Polybius as judge, cf. Vercruysse 1990: 35–36; Darbo-Peschanski 1998: 177–
189.
12 For detailed analyses, see Pédech 1964; Sacks 1981; and on causation, Derow 1979:
9–13.
Scylax 29). Further examples of the precise control exercised by Poly-

bius may be seen in his consciousness of book divisions. Polybius tends
to define in advance the limits of each book, and to proclaim at the
end of the book that he has fulfilled his promises (the Sicilian univer-
sal historian Diodorus followed this practice).13 He also includes book
numbers in some of his frequent cross-references, for instance when he
makes a back-reference to a forward-reference (18.28.1). Perhaps it was
the vast scope (both temporal and spatial) of the work that called for
this degree of definition. Polybius offers at the start of his work a ‘table
of contents’ for the work as a whole (3.2–6); he ends it by ‘recalling its
beginning and the plan which I laid down at the commencement of
my history, and then giving a summary of the entire subject, establish-
ing the connection between the beginning and the end’ (39.8.3); and
then, after reiterating the utility of his work and the uniqueness of his
theme, he announces that he is appending the periods embraced by
the history, the number of books and what he calls the arithmos of the
whole work, whatever that was (39.8.8; none of this final section—book
40—survives). All that was missing was a bibliography.
Polybius’ judgments on the methods of his narrative can be seen
as no less ideologically laden than his judgments on the events of the
narrative. His obsessive concern for order in the world of the narra-
tion parallels his concern for order in the world at large. Just as it is
fitting that barbarians, mercenaries, women, the masses, and dissolute
young aristocrats should be subjected to strong control,14 so it is fitting
that narratives should have a certain order, and that narrators should
explain why this is so (e.g. 6.2.1–7). Elsewhere Polybius’ methodologi-
cal intrusions can be more directly related to his Achaean sympathies.
He illustrates, for instance, the tragic style of history that he opposes
by quoting from Phylarchus’ pathetic account of the sack of Mantineia
(2.56)—a sack in which the Achaean League was involved. And in dis-
cussing how the Peloponnese came to be united through the Achaean
League, Polybius rejects chance as an explanation (‘one must rather
seek a cause’, 2.38.5). Yet elsewhere he does assign chance causal force.
13 There are other similarities between these two intrusive narrators: e.g. the use of
the participle proeirēmenos (‘aforementioned’)—which appears more than 500 times in

Polybius’ extant sections, in places where earlier writers would have, at most, houtos or
ekeinos; cf. Palm 1955: 76, who sees in this use a pedagogical striving for comprehensibil-
ity; and de Foucault 1972: 319.
14 Cf. Eckstein 1995, esp. 118–160, on such ‘threats to the social order’, and their
part in Polybius’ ‘moral vision’.

Other narratorial interventions could be read as part of the rhetoric

of narratorial authority. By stressing at the start of the main part of his
work how uncertain is its completion (3.5.7), he makes its completion
the more impressive.15 And the word he uses to define his undertaking
is the same as that used of the Romans’ project of universal domina-
tion (epibolē: 1.3.9, 4.2): the ambition of the work matches the deeds it
embraces (but the same word is used of Theopompus’ project at 8.11.1).
It remains to consider Polybius’ techniques of self-reference. Polybius
follows earlier historians like Herodotus (→) and Thucydides (→) by
using the first person for his appearances in methodological contexts:
as an eyewitness (22.19.1; for another claim of autopsy, 39.2.2, we have
only Strabo’s words), and as a researcher who undertakes the perils of
travels (3.48.12, 59.7–8)—a latter-day Odysseus (12.27.10; cf. also Cato’s
quip at 35.6.4)16—and consults inscriptions (3.33.17–18). It is for his
appearances as a character that Polybius departs from the precedent
of earlier writers (so far as we can tell from our evidence).
Assessing Polybius’ references to himself as character is difficult be-
cause the parts in which he appears as character are the parts that are
preserved only in excerpts. So far as we know, he never explicitly identi-
fied the character Polybius with the writer (or identified the prominent
Achaean statesman Lycortas as the writer’s father).17 But he did make
explicit his criterion for using first- and third-person forms for his own
actions:
It should cause no surprise if at times we use the proper name in
speaking of ourselves, and elsewhere use general expressions such as
‘after I had said this’ or again, ‘when we agreed to this’. For as we were
personally involved in the events that are now about to be chronicled,
it is necessary to change the phrases used to allude to ourselves, so that
we may neither offend by the frequent repetition of our name nor again
by constantly saying ‘when I’ or ‘because of me’ fall unintentionally into
an ill-mannered habit of speech. What we wish is by mixing these modes
of expressions and always using the appropriate form to avoid as far
as possible the offence that lies in speaking constantly about oneself, as
15 But Polybius’ conviction that others will finish the task if he does die (3.5.8), a
more straightforward assertion of his task’s importance, is perhaps evidence that some
of the earlier books were published before the completion of the whole work. Walbank
1972: 17–29 offers a good review of arguments about the composition and publication
of the work.
16 Walbank 1948: 172; Marincola 1997; Clarke 1999: 100–101.
17 Note that he explains at 24.6.5 that he was chosen as an envoy, though too young,
because of his father’s connections.

such a manner of speaking is naturally unwelcome, but is often necessary

when the matter cannot be stated clearly in any other way. Luckily we
have been assisted in this by the fortuitous fact that no one as far as we
know, up to the time in which we live at least, has received from his
parents the same proper name as ours.18
Why does this explanation of Polybius’ procedure only appear at this

rather late stage in the work? In all excerpts from earlier portions
of the work, Polybius appears as character in the third person. The
explanation itself follows a passage where Polybius has shifted from the
third to the first person:
When a letter reached the Peloponnese … saying that the Achaeans
would do well to send Polybius the Megalopolitan … the Achaeans voted
to send him … And we, thinking that for many reasons we ought to obey
the Romans, … set sail; and arriving at Corcyra … and thinking that the
war was over and that there was no further need of us, we sailed back to
the Peloponnese. (36.11.1–4)
Marincola has plausibly argued that this explanation occurred at the

point where Polybius’ appearances as character were going to start to
become particularly frequent.19 The shift in the immediately preceding
passage is not enough in itself to support the view that this is the first
time Polybius used the first person: it matters that the phrase ‘Polybius
the Megalopolitan’ appears in a letter. Unfortunately, we do not get a
chance to see how Polybius alternated between first and third persons.
Subsequent allusions to Polybius’ actions come from later historians,
not from excerpts. The exception comes right at the end of the work:
18 36.12.1–4: I have here modified Paton’s translation to keep to Polybius’ own

variations between first-person singular and plural forms. Clarke 1997: 96 states that
‘Polybios considered the use of the first person singular and particularly self-referential
phrases as alien to his project’ (that is, his project of universal history). But it is
because narratees are suspicious of self-glorification that Polybius is wary of excessive
reference to the character Polybius, and so keen to vary his modes of self-allusion.
There was nothing wrong with self-referential phrases in themselves. Her further claim
that ‘Polybius and Diodorus … wrote of themselves in the first plural, as a general
rule’ (p. 97) is also misleading. Plural forms are more common for introducing and
closing books, but singulars are extremely frequent: in back-references, Polybius uses
the singular forms eipa and eipon (and compounds) eight times as often as the plurals
eipomen and eipamen; he writes dokei moi far more than the (more emphatic) dokei hēmin; he
uses singulars for assertive expressions such as egō de phēmi (‘but I say’: e.g. 3.6.3, 7, 9.5)
as well as for weaker parentheses such as legō dē or oimai (‘I mean’); and he sometimes
combines the two forms in a single sentence (e.g. 3.5.7; 5.105.9; 31.23.1, 38.8. 14; cf. D.S.
3.38.1; de Foucault 1972: 85).
19 Marincola 1997: 189–192.
‘having achieved this, we returned home from Rome, having, as it were,

capitalized the results of previous political action … Therefore we pray
to all the gods that during the rest of our life all may remain in the same
condition and on the same terms, seeing as we do how apt Fortune is
to envy men’ (39.8.1–2).20 The first person is doubtless apt here because
of the personal nature of the prayer, and as a bridge to the conclusion
of the work.
There is no space here to analyse in detail Polybius’ self-presentation
as a character,21 but the most startling appearance of Polybius in his
own narrative demands discussion—the passage where he appears to
narrate his own death: ‘each city now took every means to confer
the highest honours on him during his life and after his death. And
everyone thought that this was fully justified: for had he not perfected
and drawn up the laws on the subject of common jurisdiction, all would
have remained uncertain and in the utmost confusion. So one should
consider this to be the most brilliant achievement of Polybius among
all those hitherto mentioned’ (39.5.4–6). Polybius does pose problems.
‘This passage derives from the posthumous editor of the Histories’22—or
rather, it is the only evidence for this posthumous editor. And this editor
is also a narrator.
The problem posed by this new narrator is the greater because of the
style and content of the narrative. It is not just that this narrator adopts
typical Polybian phraseology such as ‘one should think’ and ‘aforemen-
tioned’: the repetitive Polybius is all too imitable. What is odd is that the
narrative of Polybius’ death is so thematically rich at both the intratex-
tual and the intertextual level. The counterfactual claim that ‘all would
have remained uncertain (akrita) and in the utmost confusion (tarakhēs)’
evokes Polybius’ earlier characterization of the ‘confusion and distur-
bance’ of the period that followed Rome’s assumption of the universal
hegemony (tarakhēs kai kinēseōs, 3.4.13). And while counterfactual state-
ments elsewhere often emphasize how the chance for a decisive end
was missed,23 here a counterfactual secures the closure of a return to
20 For the narrator making a wish, cf. 4.32.9.

21 Especially interesting episodes include 28.7.8–13, 12–13 and 29.23–25 (where Poly-
bius and his father are given a favoured treatment by the narrator in various ways);
31.24.9–11 (details on Polybius’ friendship with Scipio); 31.11–15 (Polybius as adviser on
an escape-plot); and 32.3.14 (an Achaean appeal on his behalf).
22 Walbank 1957–1979: III. 735. Shuckburgh bracketed the passage in his translation
as a ‘note by a friend of Polybius’.

23 E.g. 2.45.5; 3.50.4, 53.1; 4.12.13, 61.3, 87.10. For other counterfactual statements,
civic order. It also sets this strongly closed ending against the ending
of Xenophon’s Hellenica, where the position in Greece after the battle
of Mantineia in 362 BC was described as one of even greater ‘uncer-
tainty and confusion’ than before (akrisia … kai … tarakhē, 7.5.27, a pas-
sage already alluded to by Polybius himself at 2.39.8, on the position in
Greece after the battle of Leuctra in 371 BC). This new narrator con-
trasts the position of Greece under Roman rule with the position in the
past when states like Sparta and Thebes were striving for hegemony.
He also puts a close to Polybius’ own story. When narratees come to
Polybius’ Herodotean anxiety about whether his prosperity will survive
to the end (39.8.2, quoted above), they know that Polybius’ prosperity
(unlike Croesus’) did endure to the end (and beyond). The narrative
of Polybius’ death contributed by his ‘posthumous editor’, it emerges,
interacts richly with the themes of the history as a whole.
We have seen that Polybius presents an intrusive narrator who is
ready to meet at every stage any possible bewilderment on the nar-
ratees’ part. First-person forms can also, however, be used to assert a
link between the narrator and the narratees. They can, for instance,
describe supposedly universal properties shared by all ‘us’ humans
(4.21.1, 31.4–5; 5.75.4–6). As with some other techniques, this usage
is not found in Thucydides or Xenophon, but it can be paralleled in
Ephorus (FGH 70 F9, 20, 63, 122a—where it means either ‘we’ humans
or ‘we’ Greeks) and Phylarchus (FGH 81 F66).24 Closer to the usage of
his predecessors is the first-person plural found in the agonistic insis-
tence on the greatness of the First Punic War—‘the longest, most con-
tinuous, and greatest war that we know of by hearsay’ (1.63.4). Here
both the claim of greatness and the qualification ‘that we know of by
hearsay’ (hēmeis ismen akoēi) recall the manner of the Herodotean nar-
rator, while the criteria of length and continuity recall those used by
Thucydides to stress the greatness of the Peloponnesian War (1.23.1,
2.1). Indeed, Thucydides himself had adopted that Herodotean ‘that
we know of ’ in his highly agonistic depiction of the Sicilian expedition
(7.87.6).25
see e.g. 1.18.11, 20.16, 28.11; 2.33.8, 68.5, 70.3; 3.9.8, 14.4, 68.3; 4.11.7–8, 87.10; 5.11.7–9,
97.6, 110.9–10.
24 Ephorus FGH 70 F97, by contrast, where ‘we’ are opposed to Boeotians, is for
that reason assigned to his Epikhōrios Logos rather than to his universal history.
25 For similar claims, cf. e.g. 2.57.8 (‘I do not know if ’ with a comparative) and 58.4;
and for a first-person comment limiting a superlative, 2.14.7.

Polybius also uses first-person plural forms in the phrase kath’ hēmas
to situate himself (and his contemporaries) in time (e.g. 3.26.2; an
instance of the ‘reference to the narrator’s own time’ motif) and space.26
The phrase ‘the sea by us’ (hē kath’ hēmas thalatta: 1.3.9; 3.37.6, 9, 10,
39.4; 16.29.6; 34.8.7; cf. also the use of kath’ hēmas with hē oikoumenē, ‘the
inhabited world’, at 3.37.1; 4.38.1) is particularly interesting because it
becomes common after Polybius (e.g. in Strabo), and because it has
been thought to suggest a Roman perspective (cf. the Latin phrase mare
nostrum).27 This interpretation (which could be supported by the use of a
similar phrase at Ps.-Scylax 40 to refer to the Saronic gulf, implying an
Athenian perspective) would give the phrase a political charge. Yet the
phrase ‘the sea par’ hēmin / peri hēmas’ could also be used in opposition
to the Red Sea or the Ocean (Theoph. HP 1.4.2, 4.6.1; cf. also Pl. Phd.
113a8, also quoted at Arist. Meteor. 356a). So Roman usage, and Roman
power, are not prerequisites for Polybius’ usage.
This example raises questions about the audience for which Polybius
was writing his Greek account of a Roman achievement. Polybius
himself makes pronouncements about his intended audience, and I
turn now to look at these, as well as at other ways in which the narrator
draws the narratees into the work.
Narratees
The Polybian narrator interacts most conspicuously with the narra-

tees through the hundreds of rhetorical questions that can be found in
the history (several have already been quoted above). The frequency
of such questions gives Polybius’ work a very different texture from
Thucydides’ or Xenophon’s—and even from Herodotus’ work, which
does have a handful of rhetorical questions. Perhaps, while Thucy-
dides had avoided elements that seemed too ‘oral’, rhetorical questions
had come by Polybius’ time to be an acceptable part of the profes-
sional style—an amiable way of introducing a new paragraph. Several
such questions can be found in the fragments of Xenophon’s near-
26 Cf. the use of this phrase in sections of Diodorus derived from Posidonius, where
it has been thought that Diodorus simply took it straight from his source (Hornblower
1981: 27–28, 263 n. 4); but it has also been argued that the vague phrase was still
appropriate for Diodorus (Sacks 1990: 83–93).
27 See Burr 1932: 115; and Dubuisson 1985: 172 for the Roman connection.
contemporary Theopompus (FGH 115 FF124, 225, 263—where there

are three in a row); but in Theopompus the tone seems rather more
indignant.
A far more rare mode of interaction used by the Polybian narrator
is the generalizing second-person address: ‘the current bears towards
Byzantion even if you do not want’ (4.44.2). The example may perhaps
be explained by the geographical context: compare the second-person
singulars at Ps.-Scylax 67 (‘until you come to …’) and 100 (‘if you go
…’). But Polybius also uses second-person forms in a non-geographical
generalization at 1.81.8.
On the whole, however, ‘Polybius’ adopts a fairly impersonal man-
ner of address. He uses forms such as the dative singular of the par-
ticiple (e.g. 1.35.7); and, in particular, impersonal forms such as khrē
(‘one should’) and hēgēteon (‘one should consider’)—forms common in
scientific writings such as those of Aristotle. Forms in -teon, by contrast,
are found eight times in Herodotus and fourteen times in Thucydides,
and on each occasion they express the words or thoughts of characters
(the same pattern is found with the rhetorical question pōs ou, ‘how not
…?’). Once more, the historian whose narrative manner is closest to
Polybius is Diodorus.28 The phrase ‘one need not be surprised’ (ou khrē
thaumazein), for instance, which is found in earlier texts in representa-
tions of speech (e.g. Pl. Prot. 326e4, Leg. 793d3; [Arist.] Ath. Pol. 15.5),
is used five times in Polybius (and four times in Diodorus, all from the
first five books).
In what ways does Polybius identify his narratees? He starts by claim-
ing that his story is of universal interest: ‘the very element of unexpect-
edness in the events I have chosen as my theme is sufficient to challenge
young and old alike to peruse these pages. For who is so worthless or
indolent as not to wish to know’ how the Romans conquered the world
(1.1.5)? And later he states that his work is a possession for all time: ‘it
will be clear to people now whether Roman rule is acceptable or the
reverse, and to people in the future whether one should think their gov-
ernment was worthy of praise and admiration or of blame; and the util-
ity of my work for the present and for the future will chiefly lie in this’
(3.4.7–8). A little later, Polybius makes it clear that it is for people ‘eager
for learning’ (philomathountes)—whom he equates with statesmen (3.7.4–
5). And at 9.1.4–5, he defines three types of history, and three types
28 Palm 1955: 91–92; Hornblower 1981: 264.

of audience: genealogical history, which appeals to those who are fond

of a story (philēkoos); accounts of the foundations of cities and colonies,
which attract the curious (polupragmōn kai perittos); and the type that Poly-
bius himself writes—history which treats ‘the doings of nations, cities,
and dynasts’, and attracts ‘the student of politics’ (politikos). But else-
where Polybius uses what seems here to be a derogatory expression,
philēkoos, as a general term for reader (e.g. 4.40.1).
Polybius defines his narratees further through his comments on how
they will benefit from his work. The work is said to have a general
utility. Through the observations it allows of others’ misfortunes, it
offers lessons on human instability and the fickleness of fortune (1.35.7,
cf. 1.1.2). Polybius further stresses that his work (as a universal history) is
useful to statesmen because it allows causal connections between events
to be recognized (3.31–32). Often, however, he offers more restricted
advice: practical advice on the use of ladders in sieges (5.98, 9.18.5–9),
for instance, or on techniques of fire-signalling (10.43–47). And some-
times he gives advice to, or about, particular Greek states (such advice
is far more common in the early part of the work): the Arcadians should
keep up their music (4.21.10; cf. 32.10, 33.11–12, advice to Messenians
and Arcadians); the Acarnanians are good people to form an alliance
with (4.30.4–5). In general, Polybius’ stress on the usefulness of his work
recalls Thucydides (→)—indeed, Polybius’ key discussion of the utility
of history at 3.31.13 closely echoes Thuc. 1.22.4. Yet ‘Polybius’ inter-
ests are more particular; there is not the same easy movement as in
Thucydides between the specific incident and the larger awareness of
the human condition’.29 But Polybius’ more particular interests do at
least make it easier for us to grasp the audience for which he was writ-
ing primarily—an audience of active politicians and generals.
That Polybius’ narratees are conceived as Greek is shown in various
ways. He offers advice to (Peloponnesian) Greek states, as we have
seen. He translates the Latin trans (2.15.9), and promises to improve
on earlier accounts of the geography of the west and ‘make these
parts of the world also known to the Greeks’ (3.58.8). And he can
be seen as aligning himself with a tradition of Greek historiography
dealing with Greek resistance to barbarians: ‘I consider that the writers
who have chronicled and handed down to us the story of the Persian
invasion of Greece and the attack of the Gauls on Delphi have made
29 Rutherford 1994: 56.

no small contribution to the struggle of the Hellenes for their common

liberty’ (2.35.7: note the shift from first-person singular to plural—to
‘us’ Greeks).30
Polybius was also aware that Romans would be interested in his
story. He expresses some concern about how his account of the Roman
constitution will seem to those used to it (6.11.3); and the fact that ‘the
writer knew’ that ‘Romans above all’ would read an account of ‘their
most splendid achievements’ is offered as a mark of its truth (31.22.8–
11). Here Polybius seems to define one narratee (the Roman reader)
as a way of swaying another (and a more privileged) narratee—the
Greek audience. But he does also offer advice to Rome—albeit in a
less direct form: ‘this happened to them on that occasion and on many
others, and it will happen until they correct this daring and violence’
(1.37.10). Unlike when he advises Greek states, Polybius here adopts a
more detached, Thucydidean pose (for the form of the statement, cf.
Th. 3.82.2).
That Polybius regarded his narratees primarily as Greek is further
confirmed by the content of the most notable of the secondary nar-
ratives presented by speakers in the history. It is to the role of these
narratives that I now turn.
Secondary narrators
Polybius reports a conference at Sparta in 210 BC at which two speech-

es with large sections of narrative are reported: the Acarnanians urge
the Spartans to join an alliance that involves Philip V of Macedon, the
Aetolians urge them to join one that involves the Romans (9.28–39).
What is telling about these speeches is that the secondary narrators
look back beyond the start of Polybius’ history to focus on examples
from earlier Greek history (the Persian Wars, the campaigns of Philip II
and Alexander of Macedon, the Gallic invasion) where the issue is ‘the
preservation of the Greek cities in the face of the threats posed by suc-
cessive kings and dynasties’.31 Perhaps here the content of the secondary
narratives reinforces the hint offered at 2.35.7 about Polybius’ concep-
tion of his narratees. He seems to be writing for a Greek audience,
about the problems of the Greeks’ confrontations with powerful out-
30 Cf. Millar 1987: 7; Lehmann 1989–1990: 68–69; Clarke 1999: 100.

31 Millar 1987: 16.
siders. And he is offering them some classic examples (cf. the speakers’
stress on how their internal narratees know the exempla: 9.28.8, 29.1–2,
5, 30.1).
With so much of Polybius’ work lost, it is unsafe to be too categorical
about the role of narrative within the speeches, but the debate at Sparta
does seem unusual in the space devoted to stories about the past.32 As
one would expect, however, many of the speeches in Polybius’ history
do contain narratives. But these narratives tend to be as brief and
allusive as those told by Thucydides’ (→) speakers (see, for instance,
Eumenes’ speech at 21.20).33 There are times, however, when a definite
resistance to secondary narratives can be detected.
How does Polybius shy away from including secondary narratives?
At times he describes people saying ‘what was appropriate’ (e.g. 5.53.6,
60.3), when what was appropriate may well have included a narra-
tive. Elsewhere he merely gives a summary of a narrative section of
a speech, while giving in full the non-narrative component. When
Scipio addresses his men, for instance, Polybius writes that ‘most of
what he said related to the exalted position of their country and the
achievements of their ancestors; what concerned the present situation
was as follows’ (3.64.2); and this non-narrative part occupies some 35
lines. When Aemilius addresses the troops before the battle of Can-
nae, ‘the greater part of his speech was devoted to accounting for his
former reverses, for it was particularly the impression created by these
that made the soldiers disheartened and in need of encouragement’
(3.108.3). Here Polybius does at least proceed to give a version of this
narrative component of the speech (20 lines). But it is only after this
that he switches from indirect to direct speech (3.109); and again it is
this non-narrative section that is longer (50 lines). More conventional
is the stress of the victorious general Hannibal on the greater value
of deeds over words at 3.111.6. Polybius is also intolerant of seemingly
32 The ‘Polybian’ portions of Livy offer some examples of secondary narratives that
may be derived from Polybius: e.g. the Macedonian envoy’s speech at 31.29.4–16. But
comparison of Plb. 23.11.1–8 and Livy 40.8.11–16 shows how Livy transforms his model
by introducing Roman exempla in a speech by Philip V of Macedon (Chaplin 2000:
80–81; cf. further 23–25 on Livy’s independence).
33 Pédech 1964: 281–289 reviews the use of historical allusions in the speeches. Note
that many of them can be found in the excerpts from the later books describing
embassies at Rome, and we are often not in a position to compare them with the
primary narrator’s account of the same events. Polybius’ spatially organized chronolog-
ical scheme itself created a problem for the narratee: as Polybius himself noted, he often
had to narrate the dealings of embassies at Rome before their dispatch (28.16.9–11).
irrelevant narrative—of mythical paradigms, for instance. Thus he crit-

icizes Timaeus’ version of a speech of Hermocrates in which the advan-
tages of peace were supported by an account of how Heracles founded
the Olympic games—a sign of his preference for peace. Again, there
are occasions where intertextual parallels raise the expectation of a nar-
rative. The famous speech by Agelaus of Naupactus in 217 warning of
‘the clouds that now loom in the west’ (5.104.10)—that is, Carthage and
Rome—recalls the speeches in Xenophon’s Hellenica warning of threats
to Greece from the north (5.2.11–19, 6.1.4–16). But Agelaus does not
have a narrative: Polybius has already provided that.
It is not just in speeches that secondary narratives are found. There
are also letters such as military communiqués (e.g. 3.75.1, 85.8); inscrip-
tions such as the ones put up by Hannibal (3.33.18, 56.4); and mythical
stories introduced by ‘they say’ (4.43.2; cf. 2.16.13, of a habit derived
from myth) or even by ‘the stories (muthoi) say’ (4.43.6, 59.5). Particu-
larly interesting is the frequency with which Polybius cites, and criti-
cizes, the stories told by earlier historians (e.g. 1.58.5: ‘not, as Fabius
[Pictor] says, …, but’; cf. e.g. 3.8.1–8, 26.3). There are, indeed, some
quite extensive citations from earlier historians in Polybius’ polemical
sections, especially in book 12. Earlier historians had reported stories
told by sources. Polybius once more shows that he is writing more con-
sciously as a ‘pro’.34 Yet sometimes it seems to be for political reasons
that he includes, and rails against, the accounts of other historians (cf.
above on his criticism of Phylarchus’ account of the sack of Mantineia).
The most pregnant of all the secondary narratives in Polybius’ his-
tory is a narrative of the future told indirectly, through a citation from
a poet. It will be appropriate to close on this episode, since it illustrates
well the difficulties in making a narratological analysis of a partly frag-
mentary author such as Polybius.
Polybius has earlier expressed his admiration for one narrative of the
future that has come true—Demetrius of Phalerum’s prediction of the
fall of the Macedonian empire (29.21). Now he describes how Scipio
wept as he looked at the destruction of Carthage:
After being wrapped in thought for long, and realizing that all cities,
nations, and authorities must, like men, meet their doom; that this hap-
pened to Troy, once a prosperous city, to the empires of Assyria, Media,
and Persia, the greatest of their time, and to Macedonia itself, the bril-
liance of which was so recent, … he said: ‘A day will come when sacred
34 Derow 1994: 84.

Troy shall perish, and Priam and his people shall be slain’ [Il. 4.164–
165, 6.448–449]. And when Polybius speaking with freedom to him, for
he was his teacher, asked him what he meant by the words, they say
that without any attempt at concealment he named his own country, for
which he feared when he reflected on the fate of all things human. Poly-
bius actually heard him and recalls it in his history.
(38.22.1 = Appian, Punica 132)
The episode is fascinating not just for what it implies about the his-
torical perspectives of empire, but also for Polybius’ portrayal of him-
self. The character Polybius seems to match the narrator Polybius in
his love for the explicit. Scipio’s Homeric allusion seems too indirect
for the insensitive Polybius. Or is the character Polybius so insensitive?
Appian’s version of the story may be contrasted with Diodorus’ version,
which survives thanks to a Byzantine excerptor (32.24). And in this ver-
sion Scipio quotes the Homeric prophecy when Polybius asks him why
he is weeping. So we have a Polybius who still has that demand for the
explicit—but one who is slightly less insensitive and obtuse.
Whichever of these versions is more true to Polybius’ lost original,
Polybius’ presentation of himself in the episode of Scipio’s tears also
raises questions about the relationship between his character as narra-
tor and his character as author. Polybius himself is aware of the danger
of associating author and narrator: he warns against ‘pay[ing] regard
not to what Fabius writes but to the writer himself and taking into con-
sideration that he was a contemporary and a Roman senator, at once
accepting all he says as worthy of credit’ (3.9.4). Yet he also reports
that ‘Timaeus says that poets and authors reveal their real natures in
their works by dwelling excessively on certain matters’ (thus he says
that ‘Homer, by constantly feasting his heroes, shows that he was more
or less of a glutton’)—and naturally turns this insight against Timaeus
himself (‘his pronouncements are full of … craven superstition and
womanish love of the marvellous’, 12.24.1–5). Anyone exasperated by
the nannying intrusiveness of the Polybian narrator may be tempted to
apply this pattern of thought to Polybius himself.35 But the advances of
narratology, and the key separation of author and narrator proposed
by Genette and others, suggest that we should resist this temptation. It
may be better to interpret the character Polybius conversing with Sci-
35 Contempt for Polybius’ ‘schoolmasterly temperament’ (p. 24) mars the most de-
tailed analysis of his narratorial persona, Ibendorff 1930.

pio at Carthage as a narratorial alter ego: a man who instructs and

speaks freely with powerful statesmen and who is ever quick to search
for explanations.
Conclusion
We have seen that Polybius, for all his very foregrounding of the ‘gaze’36
of participants and narratees, does not share Thucydides’ concern
to preserve that clarity of narrative that seems to offer the narratee
unmediated access to events. His narrator is much more obtrusive even
than the Herodotean narrator. With a persona close in some ways to
that of some modern academics, Polybius emerges as a professional
historian writing for narratees committed to the enquiry into historical
causation. At the same time, the controlling voice of the narrator blends
easily with the voice of social control.
The excessively didactic and explicit narratorial persona that Polybius
adopts is confirmed even by a passage where the narrator advises
against taking his narrative at face value: ‘finally the king—if one
should call the opinions he then delivered the king’s; for it is not
probable that a boy of eighteen should be able to decide about such
grave matters. But it is the duty of us writers to attribute to the supreme
ruler the expression of opinion that prevailed at his council, while the
readers should suspect that such arguments and decisions are due to
his associates’ (5.24.1–2; note the anacoluthon whereby Polybius cuts
the king off without a verb). And yet Polybius here seems to risk
destabilizing his narrative by inviting narratees to consider how writers
are constrained not just by convention, but also by the power structures
that those conventions uphold. Some of his later readers have also
suspected that Polybius’ account of the relations between conquering
Rome and conquered Greece can be read in some sense against the
grain. It is no accident, perhaps, that it is its analysis of how states and
individuals respond to shifts in power, as much as the ‘universal’ interest
of the subject matter itself, that has given Polybius’ didactic mode of
historiography its enduring interest.
36 The term adopted by Davidson 1991.

chapter eleven
ARRIAN
T. Hidber
‘You must know who I am’
And as to who I am that I make this judgement in my favour, I do not

inscribe my name, for it is not at all unknown to men, nor what my
native land is, nor my family, nor if I have held any office in my own
land; but this I do inscribe, that my native land, family, and offices are
this work, and have been even from my youth. And for this reason I
count myself not unworthy of the first place in Greek speech, just as I
hold Alexander to have been in arms. (1.12.5)1
In the ‘Second preface’ to the Anabasis there is quite a unique recusatio
of the historiographical topos whereby a narrator provides some infor-
mation on himself, such as his name, his country of origin, his offices,
and his social standing. The narrator claims that he does not need to
mention his identity and his background since they are so well known.
As this anonymity calls attention to itself (in contrast to Xenophon’s,
Herodian’s, or, indeed, Homer’s), it is to be regarded as an expression
of striking self-confidence rather than modesty, and the very praeteri-
tio clearly shows an awareness of fame and distinguished social status.
However, as for the narrator’s identity, only his skills and achievements
as a writer as well as his almost lifelong devotion to this work are
claimed to be of relevance.
The narrator’s self-definition, then, is based on the identification of
himself both with his work (i.e. with the Anabasis) and with his topic
(Alexander the Great). The dramatic setting of the ‘Second preface’
is instructive: having crossed the Hellespont and being now ready to
launch his campaign against the Persians (i.e. to achieve his greatest
deeds), Alexander pauses at the tomb of Achilles in Ilium, thereby
1 Translation by Moles 1985: 163.

166 part two – chapter eleven
blessing the hero ‘on the ground that he had obtained Homer as herald
for his future commemoration’ (1.12.1). It is at this crucial point that the
narrator intervenes:
And indeed from Alexander’s point of view Achilles really was to be
counted blessed, not least for this reason, that, for Alexander himself,
not in accordance with his general good fortune, this area happened
to be left free and Alexander’s deeds were not published to mankind
worthily, neither certainly in prose nor did anyone compose them in
verse [worthily]. But Alexander was not even sung in lyric … so that
Alexander’s deeds are much less known than the most. (1.12.1–2)2
This claim is underpinned by the following sunkrisis between the march

of the ‘Ten thousand’ and Alexander’s achievements, culminating in
the emphatic assertion of Alexander’s superiority:
But there is no other single man who performed so many or so great
deeds in number or magnitude among Greeks or barbarians. Wherefore
I declare that I myself set out on this history, not judging myself unworthy
to make Alexander’s deeds known to men. (1.12.4)3
Here, the narrator implicitly sets himself in line with Herodotus to

whose preface he clearly alludes, and explicitly competes with Xen-
ophon whose Anabasis he claims to surpass because of the unrivalled
greatness of his topic.4 There is no sense of inferiority to the old
masters whatsoever. Rather the narrator is quite confident to do for
Alexander what Homer did for Achilles. And indeed, the very lack
of a worthy presentation of the king’s unparalleled achievements, as
deplored by Alexander himself, is claimed to be the narrator’s motive
for writing and making those deeds known to men. Thus, Alexander,
styling himself as a Homeric hero, is calling on the narrator to become
his Homer. It is also made clear that the Anabasis is expected to equal
the king’s achievements in warfare in terms of literature. This parallel is
made explicit in the proud last sentence of the work, where the narrator
states that he embarked on this history ‘like Alexander not without
God’s help’ (7.30.3).

4 Recent research has made clear many aspects of Arrian’s exceptionally sophisti-
cated ‘self-definition by continuity and contrast’ (Marincola 1997: 253). Cf. Moles 1985;
Marincola 1989b; Gray 1990; Marincola 1997: 253–254. Moles 1985: 163 additionally
draws attention to the allusion in 1.12.2 (to chorion touto eklipes xunebē) to Thuc.1.97.2
t. hidber – arrian 167
Superior sources
In pointed contrast to the narrator’s own anonymity, the names of his

sources are given at the outset of the work with the very first words of
the preface:
Ptolemy, son of Lagus, and Aristobulus, son of Aristobulus—whatever
these two say in common about Alexander the son of Philip, these things
I have written as entirely true, but where they differ I have selected the
version I regarded as more trustworthy and also better worth telling.
(1 praef. 1)5
This opening with the names in the nominative is a clever play with
convention, since the founding fathers of ancient historiography had
chosen this formula at the beginning of the work to advertise their own
name and country of origin.6 The first sentence of the work makes clear
that the reliability of the historical account is based on the records
of the only two trustworthy sources: Ptolemy and Aristobulus who
were both eyewitnesses of the events as they took part in Alexander’s
expedition, but wrote after the king’s death, and therefore ‘under no
constraint or hope of gain to make him set down anything but what
actually happened’ (1 praef. 2).7 Furthermore, Ptolemy’s authority and
trustworthiness is stressed by a reference to his distinguished social
standing: ‘as he himself was a king, mendacity would have been more
dishonourable for him than for anyone else’ (1 praef. 2). As the Anabasis
basically reconstructs the remote past, the handling of sources is of
utmost importance. Thus, it is not only the narrator’s literary ability,
but also his consistent restriction to the evidence provided by the only
trustworthy witnesses of the events that guarantees the reliability and
superiority of his account over those of his predecessors. The latter,
then, are dismissed out of hand in a bold and direct call on the
narratees:
Anyone who is surprised that with so many historians already in the field
it should have occurred to me too to compose this history should express
his surprise only after perusing all their works and then reading mine.
(1 praef. 3)
(eklipes touto … to chorion—marking the beginning of the Pentecontaetia excursus). Marin-

cola 1989b: 186–187 suggests that the Second preface might also have been intended to
outdo Theopompus’ claims in his Philippica.
5 Translation taken from Marincola 1997: 275 and Brunt 1976.
6 Hec. FGH 1 F 1a; Hdt. 1.1; Th. 1.1.
7 Translation (here and in the following) taken from Brunt 1976.
The earlier Alexander historians are no match for Arrian and, there-
fore, are not even worth mentioning: Indeed, it is only his Anabasis
that provides the first adequate and worthy account of the king’s great
deeds.
Explicit narratorial interventions
The presence of this exceptionally sovereign and self-confident narrator

is not felt with invariable intensity throughout the work (comprising
seven books). Many long stretches of narrative are devoid of explicit
narratorial interventions and the story seems ‘to tell itself ’. This is
often the case with records of particular campaigns or with elaborate
battle descriptions narrated in one stretch and without a break. Such
passages, then, are given a rather ‘objective look’ and often offer an
absorbing read.8 They are, however, counterbalanced by quite a large
number of often rather extensive narratorial interventions framing such
narrative stretches or occurring in digressions. Thus, the narratees are
from time to time reminded that this text is presented to them by
the tough-thinking and self-assured narrator whose persona has been so
impressively established in the two prefaces.
Most of the explicit narratorial interventions deal with sources and
source-criticism, showing the narrator at work as a historian of the dis-
tant past—a distant past that is only accessible through written sources.
Indeed, sources are mentioned more often than in most other histo-
riographical works of antiquity.9 Usually Aristobulus and Ptolemy are
mentioned as authorities for particular accounts or pieces of informa-
tion,10 whereas alternative versions are reported via anonymous spokes-
men (‘others say’, ‘some have also recorded’, ‘a story of the following
sort is told’).11 Occasionally other sources are referred to by name,
mainly in digressions.12 In a few instances differences between the
8 Cf. e.g. Arr. An. 1.1.4–1.1.8 (campaigns in Greece); 1.20.2–1.23.4 (siege of Halicar-
nassus); 2.18.1–2.24.6 (siege and capture of Tyre); 3.13–15 (battle at Gaugamela).

9 Cf. Bosworth 1988: 39.
10 Cf. Arr. An. 1.2.7; 1.8.1; 2.11.8; 3.3.3; 3.11.3; 3.17.6; 3.26.2; 3.28.6; 3.30.7; 4.6.1;
4.8.9; 4.13.5; 4.14.1; 4.15.8; 5.14.3–4; 5.15.1; 5.20.2; 5.20.8; 6.10.1; 6.22.4–7; 6.28.3; 6.29.4;
6.29.6; 6.29.10; 7.17.5; 7.18.1; 7.19.3; 7.20.5; 7.22.5; 7.24.1; 7.28.1.
11 Cf. Arr. An. 1.1.1; 1.1.5; 1.11.7; 1.12.1; 1.12.10; 2.2.2; 2.4.7; 2.5.4; 2.7.8; 3.2.1; 3.5.7;
3.10.1; 3.27.1; 4.10.5; 4.12.3; 6.11.2; 6.22.8; 6.24.1; 6.28.1; 7.1.2–3; 7.2.3; 7.5.3; 7.6.2; 7.11.9;
7.14.4; 7.20.1; 7.22.1; 7.24.4.
12 References are made, among others, to Homer (4.1.1; 5.6.5; 6.1.3), Herodotus
accounts of Aristobulus and Ptolemy are noted and (once) are critically
commented on:
Thus not even those whose narratives are entirely trustworthy and who
actually accompanied Alexander at that time agree in their accounts of
events which were public and within their own knowledge. (4.14.4)13
There are more instances of conflicting versions, whereby more often
than not the matter is explicitly left undecided; only by way of excep-
tion is one or the other version given preference, on account of its plau-
sibility. This procedure is clearly designed to demonstrate the narrator’s
circumspection and prudence:
Alexander was unable to untie the knot but unwilling to leave it tied
… some say that he struck it with his sword, cut the knot, and said it
was now untied—but Aristobulus says that he took out the pole-pin, a
bolt driven through the pole, holding the knot together, and so removed
the yoke from the pole. I cannot say with confidence what Alexander
actually did … (2.3.7–8)
That some divine help was given him I can confidently assert, because
probability suggests it too; but the exact truth of the story cannot be told;
that is precluded by the way in which different writers about Alexander
have given different accounts. (3.3.6)14
In a few instances, however, divergent versions are included or alluded
to only to be dismissed as falsifications and to prove the narrator’s
full knowledge of the tradition as well as his soundness of judgment
(sometimes displayed in rather lengthy deliberations):
Many other stories have been written by the historians about the misfor-
tune, and tradition has received them as the first falsifiers told them, and
still keeps them alive to this day, nor indeed will it ever cease handing on
the falsehoods to others in turn, unless it is checked by this history.
(6.11.2)
My own obligation has been adequately discharged by including the
story, unreliable though it is. (6.28.2)
So much for stories which I have set down to show that I know they are
told rather than because they are credible enough to recount. (7.27.3)15
(2.16.3; 3.30.8; 5.6.5; 5.7.2), Nearchus (6.13.4–5; 6.24.2–3; 7.3.6; 7.20.9), Eratosthenes
and Megasthenes (5.6.2), Ctesias (5.4.2), Aristus and Asclepiades (7.15.5), Callisthenes
(4.10.1–2), and the ‘Royal journals’ (7.25.1; 7.26.1–2).
13 Cf. also Arr. An. 3.3.5–6; 3.4.5; 3.30.5; 4.3.5; 4.14.3; 5.20.2.
14 Cf. Arr. An. 2.4.7–11; 2.12.6–8; 3.30.5; 3.30.8–9; 4.9.2–4; 4.14.1–2; 4.14.4; 5.4.5;
6.26.1; 7.1.1–4; 7.15.5–6.

15 Cf. also Arr. An. 5.4.3–4; 5.14.3–6; 6.11.1–8; 7.12.5; 7.13.2–3.
A similarly critical, though more indulgent, attitude is displayed to-

wards the mythical tradition, referred to in some digressions:
For my part I cannot gather whether the Theban Dionysus, starting from
Thebes or from the Lydian Tmolus, led an army against the Indians,
after assailing so many warlike peoples, unknown to the Greeks of that
time, and yet subduing none of them by force but the Indians; still,
one must not be a precise critic of ancient legends that concern the
divine. For things which are incredible if you consider them on the basis
of probability appear not wholly incredible, when one adds the divine
element to the story. (5.1.2)16
All this is perfectly in line with the claim made at the outset, that
the Anabasis is to deliver the first truly worthy account of Alexander’s
deeds, since it is exclusively based on the only trustworthy sources and
presented by an exceptionally gifted and sharp-witted historian of great
literary skills. Frankly admitting a few contradictions in his sources, the
narrator strengthens the authority of the rest of his text and of the
Anabasis as a whole. At the same time, the narratees are given the
opportunity, as it were, to participate in the narrator’s deliberations
and, above all, to appreciate his cautious and critical approach to
tradition.
As the Anabasis deals with the distant past, its narrator is clearly
external; hence, the stress laid on his scrupulous handling of written
sources. However, the narrator appears to fill a few occasional gaps in
the tradition with references to his own personal experience:
What I do know is that this part of the mainland is capital pasture to this
very day and rears excellent oxen. (2.16.6)
This too I cannot decide, yet the quickest way of bridging I know is the
Roman use of boats, and I shall here describe it, for it merits description.
(5.7.1)
Such references, well known from Herodotus (→) and particularly fre-
quent in Cassius Dio (→), assure the narratees that the narrator is a
much-travelled man and well-acquainted with the present-day world
too.17 These very moments the narrator figures himself as an eyewit-
ness, a role which he is forced to leave to his sources with regard to the
historical events themselves.
16 Cf. also Arr. An. 2.16.1–6; 3.5.7; 4.28.1; 5.3.1–4; 5.4.5; 7.13.3–5.
17 Cf. also Arr. An. 1.16.4; 2.5.1; 3.16.8; 5.28.7; 6.20.2.
Another type of narratorial intervention reflects the narrator’s role

as organizer and commentator. There are a few reminders of what has
been told earlier18 and some references to future writing, the Indikē,19
but on the whole, cross-referencing is accomplished rather by implicit
parallels and contrasts than by explicit signs. Stretches of narrative may
be introduced by a short narratorial remark pointing at what is going
to be narrated:
A story of the following sort is told, and personally I do not disbelieve
it. (3.2.1)
At this point it will be the moment for me to relate the tragedy of Clitus
son of Dropides and the suffering it caused to Alexander, even though it
actually occurred later. (4.8.1)20
Likewise, passages may be concluded by a narratorial statement. Some
of the closing remarks are short summaries, sometimes adding the year
and month (e.g. 2.11.10: ‘So ended this battle, fought in the archonship
at Athens of Nicocrates and in the month of Maimacterion’).21 Others
follow the traditionally historiographical rhetoric of amplification as,
for example, the elaborate and somewhat lengthy demonstration that
the capture of Thebes by Alexander was the greatest disaster ever to
befall a Greek city (1.9).22 Finally, there are closing remarks in which the
narrator applauds or condemns a particular action or conduct (often in
the first person), which recall the Xenophon-historian (→):
I too do not think that Alexander showed good sense in this action nor
that he could punish Persians of a long past age. (3.18.12)
For my part, I do not approve of this excessive punishment of Bessus;
I regard the mutilation of the extremities as barbaric, and I agree that
Alexander was carried away into imitation of Median and Persian opu-
lence … Nor do I at all approve the facts that … he substituted the dress
of Medes for that traditional with Macedonians and that he exchanged
the tiara of the Persians, whom he himself had conquered, for the head-
dress he had long worn … (4.7.4)
18 Cf. Arr. An. 5.5.3; 6.17.2; 6.30.1.

19 Cf. Arr. An. 5.5.1; 5.6.8; 6.16.5; 6.28.6.
20 Cf. also e.g. Arr. An. 2.6.7; 2.14.4; 2.22.1; 4.10.5; 4.12.3; 4.20.1; 5.23.5; 6.25.2; 6.26.4;
7.12.5; 7.18.1; 7.24.1; 7.25.1.

21 Cf. also Arr. An. 2.24.6; 3.15.7; 3.22.2; 5.4.1; 5.6.8; 5.16.1; 5.19.3; 6.16.2; 7.18.6;
7.23.2.
22 For this sort of argument cf. e.g. Th. 1.2–19; Plb. 1.2; Dion. Hal. Ant. Rom. 1.2–3;
App. Hist. 1 praef. 8–10.

I myself strongly blame Clitus for his insolence towards his king, and
pity Alexander for his misfortune, since he then showed himself the slave
of two vices, by neither of which it is fitting for a man of sense to be
overcome, namely, anger and drunkenness. (4.9.1)
Some of these deliberations and judgments evaluate the behaviour of
other characters,23 but most of them are aimed at creating a picture of
the disposition and the personality of Alexander.24 Whereas throughout
the work the king’s lack of control over his passions and his inclination
to take over barbarian customs is censured several times (once in a
passage which lists examples of these weaknesses 4.7–14),25 the long
obituary at the end of the work in which all these moral judgments
finally culminate (7.28–30) stresses Alexander’s unrivalled excellence in
almost every respect; his errors now appear to be almost negligible:
If Alexander was at all guilty of misdeeds due to anger, or if he was
led on to adopt barbarian practices involving too much pretension, I
do not personally regard it as important; only consider in charity his
youth, his unbroken good fortune, and the fact that it is men that seek to
please and not to act for the best who are and will be the associates of
kings, exercising an evil influence. But remorse for his misdeeds was to
my knowledge peculiar to Alexander among the kings of old times, and
resulted from his noble nature. (7.29.1; cf. also 7.29.2–4)
In fact, anyone who reproaches Alexander should compare his own
actions with those of the king, ‘and then carefully reflect who he himself
is and what kind of fortune he enjoys, that he can condemn Alexander’
(7.30.1). Thus, the narrator has to justify his own attitude:
So, while I myself have censured some of Alexander’s acts in my history
of them, I am not ashamed to express admiration of Alexander himself;
I have made those criticisms from my own respect for truth and also for
the good of mankind: it was for that purpose that I embarked on this
history. (7.30.3)
The narrator’s self-criticism once again shows how much he is com-
mitted to the truth of his portrayal of Alexander, while the references
to divergent versions enhance the authority of the text as a whole.
Together these devices underpin the historiographical claim of the
Anabasis and distinguish the work from an encomium.
23 Cf. e.g. Arr. An. 3.22.2–6 (Darius); 4.7.5–6 (Clitus); 4.12.6–7 (Callisthenes); 7.3.6
(Calanus).
24 Cf. also e.g. Arr. An. 1.17.2; 2.12.8; 3.10.2–4; 4.8.2; 4.9.2; 4.19.6; 5.19.6; 6.13.4;
6.26.1–3; 7.1.4; 7.4.7; 7.8.3; 7.16.7–8; 7.19.6; 7.23.8.

25 Cf. particularly Arr. An. 4.8.2 and 4.12.6–7.
Secondary narrators
There is remarkably little character-text in the Anabasis. In fact, books

3 and 6 do not even contain any speeches. On the other hand, there is
quite a large amount of indirect speech, a form of reporting, in which
the narrator remains clearly visible as the presenter and organizer of
the account. The most elaborate speech is the king’s address to his
mutinous Macedonian troops at Opis (7.9.1–10.7) in which he recounts
as secondary narrator the events from the time of his father to his own
days.26 Clearly, Alexander speaks as an internal narrator. Whereas he is
trying to win over his troops, i.e. the secondary narratees, his speech at
the same time provides the primary narratees with first-hand informa-
tion on the ‘prehistory’ of the Anabasis. The effectiveness of his narrative
is demonstrated by its result: the Macedonians ‘stunned by his speech’
(7.11.2) ruefully give up their mutiny. There is one other case of embed-
ded narrative: Acuphis of Nysa, telling Alexander, how Dionysus once
founded his city (5.1.5–6). Acuphis, of course, is an external narrator,
and the story he tells is mythological rather than historical. And that is
obviously why it is not told by the primary narrator himself. However,
it proves congenial to Alexander, who ‘also wanted Nysa to be founded
by Dionysus, in which case he had already reached the point Diony-
sus reached, and would go even farther’ (5.2.1). Consequently, the story
has the effect hoped for by Acuphis, as Alexander grants freedom and
independence to the settlers of Nysa.
Narratees
The primary narratees are less overt than the primary narrator. Still
various signs indicating their presence are occasionally found. There
are many gar-clauses providing the necessary background information
(e.g. 1.25.8: ‘For the swallow is a domestic bird, friendly to man, and
more talkative than any other bird’),27 some if not-situations (e.g. 1.22.7:
‘The city came indeed came near to capture, had not Alexander sound-
ed the retreat’),28 one instance of the ‘anonymous witness’ device (1.2.1:
26 Cf. Bosworth 1988: 101–113.

27 Cf. also e.g. Arr. An. 3.4.1; 4.1.3; 4.2.5; 4.21.3; 5.2.6; 5.21.3; 5.29.4; 6.3.4; 6.9.5;
6.10.3; 6.12.2; 6.20.1; 7.4.5; 7.5.6; 7.15.2; 7.19.5.
28 Cf. Arr. An. 4.4.9.
‘for someone who approaches the Haemus, it is three days from the
Ister’), and several instances of the ‘anonymous interlocutor’ device (e.g.
5.2.7: ‘if anyone can believe this story’).29 Parenthetical remarks such as
‘as often happens in such cases’ (1.7.3) may also be seen as parts of the
narrator’s interaction with his narratees.30
Conclusion
The narrator of Anabasis is characterized as extremely self-assured,

class-conscious, and even conceited. He assures his narratees that he
presents the ultimate version of the Alexander story, the primacy of
which is due to the unique use of superior sources as well as his own
outstanding intellectual power and narrative skills. Where his presence
can be felt overtly, particularly when dealing with sources and source
criticism, it resembles in many respects that of the Herodotean narrator
(→). On the whole, however, this narrator restricts his overt presence
to few but prominent passages, which suffice to show him in perfect
control of his task as an inquirer and historian.
29 Cf. also Arr. An. 1 pref. 3; 5.3.1; 5.4.2; 5.20.10.

30 Cf. also Arr. An. 4.20.4; 6.26.1.
chapter twelve
APPIAN
T. Hidber
From Alexandria to Rome

Who I am, who have written these things, many know, and I myself have
given an indication. To speak more clearly, I am Appian of Alexandria,
a man who has reached the first ranks in my country, and have pleaded
causes in Rome before the emperors, until they deemed me worthy to be
procurator. And if anyone has a great desire to learn all the rest about
me, there is a treatise by me on that subject too. (praef. 15.62)1
At the very end of the preface to Appian’s Roman History the narrator
unveils his own personality and presents himself as a provincial, a native
of Alexandria, with a distinguished career that has led him from Egypt
to Rome and there into the wider surroundings of the centre of impe-
rial power. But at the same time he claims to be well known anyway,
and—perhaps a singular case in ancient historiography—for further
details refers to his own autobiography. Clearly, the mere existence of
the latter must prove also to those who would not read it that the nar-
rator is an eminent man of honour. However, whereas the narrator in
Appian’s contemporary Arrian (→), whose Roman career culminated
in much higher ranks, in a unique praeteritio both proudly and playfully
passes by the old convention, whereby name, native land, and career
are mentioned, in Appian these data are still recorded.2
Although the Roman History deals with the history of Rome’s expan-
sion from its beginnings to the conquests made by Trajan, i.e. with
1 Translation by Marincola 1997: 145.

2 The similarity of the formulation in Arrian (1.12.5: ‘And as to who I am that
I make this judgment in my favour, I do not inscribe my name, for it is not at all
unknown to men, nor what my native land is, nor my family, nor if I have held any
office in my own land; but this I do inscribe, that my native land, family, and offices
are this work and have been even from my youth’) suggests a relationship between the
two passages, though it is still debated as to which is prior; cf. Moles 1985: 168; Gowing
1992: 10; and Marincola 1997: 274.
176 part two – chapter twelve
non-contemporary history, the narrator’s personal experience acquired

during his professional career in Rome and based on his social standing
is shown to be not irrelevant to the subject chosen:
Possessing the best part of the earth and sea they [sc. the Romans]
have, on the whole, aimed to preserve their empire by the exercise of
prudence, rather than to extend their sway indefinitely over poverty-
stricken and profitless tribes of barbarians, some of whom I have seen
at Rome offering themselves, by their ambassadors, as its subjects, but
the emperor would not accept them because they would be of no use to
him. (praef. 7.26)3
Thus, the narrator, though being external, knows what he is talking
about, when he recounts the story of the successful development and
rule of the Roman empire. Recording exceptional details and figures
concerning Egypt already in the preface (praef. 10.39–42) he also shows
himself to be particularly well-informed about his native land, the
integration of which into the Roman empire is presented as one of the
most important stages of the Roman expansion (praef. 14.60).4
As for the narrator’s motive for embarking upon his History, the
unique greatness and importance of his subject matter is brought for-
ward in the first place. The preface begins with an impressive pano-
ramic survey of the peoples and nations under Roman sway, followed
by an elaborate comparison with earlier hegemonies, which serves to
prove that the Roman empire was the greatest so far in terms of both
size and duration (praef. 1.1–10.42). The greatness of the topic, the wor-
thiness of which is underpinned by a reference to the ‘many writers’ it
had already attracted (praef. 12.45), reflects the enormity of the narra-
tor’s task and calls both for a new focus and a new form of presentation:
Thinking that the public would like to learn the history of the Romans in
this way, I am going to write the part relating to each nation separately,
omitting what happened to the others in the meantime, and taking it up
in its proper place. (praef. 13.49)
The narrator claims originality with regard to his ethnographic ap-
proach and the disposition of his material by nations and theatres of
action. This unusual arrangement, intended to avoid frequent changes
of the geographical focus and perhaps inspired by the Herodotean
3 Translation here and in the following by White 1912–1913.

4 Cf. also App. BC 1.5.21; 1.6.24–25. Unfortunately, the four books on Egypt and its
becoming part of the Roman empire (18–21, still known to Photius [Bibl. cod. 57]) have
not come down to us.
t. hidber – appian 177
model5 as well as by his own provincial outlook, is explicitly linked

to the demands of the narratees. The same holds for the narrator’s
announcement that he will usually omit the full Roman names and
‘superfluous’ chronological details:
It seems superfluous to put down the dates of everything, but I shall
mention those of the most important events now and then.
(praef. 13.50)
For purposes of distinction I shall sometimes mention all the names,
especially of illustrious men, but for the most part I shall call these and
others by the names that are deemed most characteristic.
(praef. 13.52)
In the preface, then, the narrator presents himself as the sovereign
organizer of his comprehensive Roman History. At the same time, he
calls attention to his role as commentator of the deeds he recounts and
provides a general interpretative framework for the successful Roman
expansion:
Through good counsel and good fortune has the empire of the Romans
attained to greatness and duration; in gaining which they have excelled
all others in bravery, patience, and hard labour. (praef. 11.43)
Neither famine, nor frequently recurring plague, nor sedition, nor all
these falling upon them at once could abate their ardour; until, through
the doubtful struggles and dangers of seven hundred years, they achieved
their present greatness, and won prosperity as the reward of good coun-
sel. (praef. 11.44)
It is the narrator’s expressed aim to demonstrate the impact of the
Roman expansion in each region and to compare in each case the
qualities of the conquerors with those of the conquered nations and,
thus, to explain their achievements:
I have made this research also in respect to each of the other provinces,
desiring to learn the Romans’ relations to each, in order to understand
the weakness of these nations or their power of endurance, as well as the
bravery or good fortune of their conquerors or any other circumstance
contributing to the result. (praef. 12.48)
The history of the Roman expansion is presented here as a unique suc-
cess story, made possible by prudence, bravery, endurance, and other
virtues of the Romans on the one hand, and good fortune on the other.
5 Brodersen 1993: 358. But Ephorus, Polybius, and Pompeius Trogus may well have
been sources of inspiration, too; Leidl 1996: 67.

In this view, the narrator’s own (Antonine) time is regarded as the cul-
mination of prosperity, peace, and security:
From the advent of the emperors to the present time is nearly two
hundred years more, in the course of which the city has been greatly
embellished, its revenue much increased, and in the long reign of peace
and security everything has moved towards a lasting prosperity.
(praef. 7.24)6
After the preface the narrator’s presence is felt with varying intensity.
Whereas some stretches of narrative—such as reports of military cam-
paigns, sieges, and battles—are virtually devoid of explicit narratorial
interventions, the narrator is overtly visible in many framing passages.
As in many ancient historiographical texts, mainly three types of narra-
torial interventions can be distinguished: the narrator as organizer, the
narrator as commentator and interpreter, and the narrator as historian
and inquirer.
The unusual organization of events often causes the narrator to step
forward in his role as organizer of the story, explaining his arrangement
and providing connecting links between the different parts of the work.
Some of the most elaborate narratorial interventions of this type can
be found at the beginning of books, when another theatre of action is
introduced. The narrator gives a survey of the topography and briefly
touches upon the prehistory in the form of a praeteritio, stressing that he
only deals with the period and the events relevant to Roman history,
e.g.:
What nations occupied it first, and who came after them, it is not my
purpose to enquire closely, as I am writing only Roman history. However,
I think that the Celts, passing over the Pyrenees at some former time,
mingled with the natives, and that the name Celtiberia originated in that
way. (Hisp. 2.5)7
The most extensive preface introduces the five books of the Civil Wars.8
An outline of the civil conflicts recorded in these books is given and
the organization of the historical material in this part of the work is
6 Cf. also App. praef. 11.44.

7 Cf. also App. Ill. 1 (1)–6 (16); 2.4; Hann. 1 (1)–4 (13); Pun. 1 (1)–2 (10); Mith. 1 (1)–2 (3).
8 App. BC 1.1–6.
explained. Similar prefaces, though on a smaller scale, can also be

found at the outset of all the subsequent books of the Civil Wars.9
Due to the ethnographic rather than chronological approach there is
an unusual number of cross-references pointing from one book to the
other and indicating to the narratees that a certain story which could
be expected to be recorded at that point and in that book, is recounted
elsewhere:
What Hannibal himself and what the other Carthaginian and Roman
generals after him did in Spain, I have related in the Spanish history.
(Hann. 4 [13])
The details, however, of these events are related more particularly in my
Egyptian history. (BC 2.90 [379])10
The cohesion of the narrative is further effected by more casual and
unspecific back-references (type: ‘as I have related earlier’).11 In a few
instances, where the arrangement of the material seemed particularly
problematic or a strictly ethnographical account could not be given,
the narrator explicitly justifies and explains the organization chosen by
him:
It is my intention here to relate this war with Viriathus, which was very
harassing and difficult to the Romans, and to postpone any other events
which happened in Spain at the same time. (Hisp. 63 [265])
It has been impossible to distinguish all these exploits by nations, since
they were performed at the same time and were complicated with each
other. Those, however, which could be separated I have arranged each
by itself. (Mith. 119 [589])12
Finally, the narrator’s presence as organizer is obvious in numerous
(often temporal) transitional phrases providing short summaries and
indications as to what is to follow, thereby linking two stories or events
evolving in different places at the same time:
9 Cf. App. BC 2.1(1); 3.1 (1); 4.1 (1–3); 5.1 (1–3).

10 Cf. also App. Hisp. 14 (53); 14 (56); 36–37 (147); 101 (441); Hann. 1 (2); 4 (13); Pun. 2
(10); 4 (17); Ill. 4 (11); 6 (16); 30 (87); Syr. 2 (5); 51 (260); Mith. 11 (34); 22 (85); 64 (264); 114
(557); BC 1.6 (24–25); 1.55 (241); 2.17 (61); 2.18 (67); 2.90 (379); 2.92 (385); 2.154 (649); 4.1
(3); 4.16 (62); 5.10 (39); 5.65 (276).
11 Cf. also e.g. App. Hisp. 80 (311); Pun. 67 (301); Ill. 14 (40); Syr. 66 (348); 67 (353); 70
(367); Mith. 114 (558); BC 1.34 (154); 1.76 (347); 1.100 (470); 2.52 (214); 2.113 (474); 2.124
(518); 2.149 (620); 2.153 (641); 4.83 (349); 5.91 (383); 5.103 (429).
12 Cf. also e.g. App. Hisp. 66 (280); 73 (310); 76 (322); 80 (350); Pun. 136 (644–645); Ill.
9 (25); Syr. 52 (260); BC 4.57 (243); 5.145 (601–602). Cf. also Leidl 1996: 69–72.
Such was the situation of Hannibal’s affairs; the Romans however …

(Hann. 17 [73])
Such was the third civil strife (that of Apuleius) which succeeded those of
the two Gracchi, and such the results it brought to the Romans. While
they were thus occupied the so-called Social War, in which many Italian
peoples were engaged, broke out. (BC 1.33–34 [150])13
These structural devices on the one hand demonstrate the narrator’s

dexterity in handling his complex material, on the other function as
signposts for the narratees.
After having established his commenting voice as early as in the
preface, the narrator occasionally reappears in the narrative in the
function of interpreter of the actions and events he records. In those
passing remarks the narrator expresses his astonishment (BC 1.16 [67]:
‘It is astonishing to me that they never thought of appointing a dictator
in this emergency, although they had often been protected by the
government of a single ruler in such times of peril’),14 his reasoning
(BC 1.97 [452]: ‘This does not seem to me to be inappropriate, for
one of his names was Faustus’),15 or his judgment (BC 2.58 [241]:
‘Thus did Caesar put his trust in fortune rather than in prudence’).16
At other places he points at divine intervention as the explanation
for otherwise incomprehensible forms of behaviour (Hann. 12 [49]:
‘But divine providence turned Hannibal away toward the Adriatic
again’).17 Large-scale comments, however, are restricted to passages of
transition, which frame stories or episodes. Explanations as to how an
event or action is to be interpreted are rarely given beforehand. Much
13 Cf. also e.g. App. Hisp. 28–29 (114); 76 (322); 82 (358); Hann. 17 (73); 33 (137); 47
(206); 55 (228); Pun. 13 (50); 24 (97); 34 (143); 74 (338); 77 (359); 97 (456); 106 (501); 131
(628); 135 (638); Ill. 18 (52); Syr. 9 (34); 21 (93); 22 (101); 28 (137); 38 (193); 45 (231); 47
(241); Mith. 17–18 (64); 22 (83); 28 (108); 68 (286);75 (326); 76 (327); 86–87 (392); 107 (502);
BC 1.27 (121); 1.32 (141); 1.33–34 (150); 1.49 (219); 1.51 (226); 1.74–75 (345); 1.94 (439); 1.96
(449); 1.107 (501); 2.4 (13); 2.15 (54); 2.44 (175); 2.48 (196); 2.95–96 (401); 3.9 (30); 3.26
(97); 3.32 (124); 3.49 (198); 3.66 (269); 3.75 (305); 3.77 (312); 3.97 (399); 4.4 (14); 4.52 (224);
5.12 (45); 5.104 (430).
14 Cf. also e.g. App. Ill. 15 (44); BC 1.103 (481–482); BC 2.88 (371–372).
15 Cf. also e.g. App. Pun. 66 (295); 71 (324); Syr. 69 (366); Mith. 118 (582); BC 1.39 (175);
1.54 (233); 1.97 (454); 1.103 (479); 1.104 (487–490); 1.115 (538); 2.88 (371–372); 2.111 (463);
5.113 (471–472).
16 Cf. also e.g. App. Hisp. 60 (253); Syr. 41 (212–218); BC 1.104 (484); 2.7 (24).
17 Cf. also e.g. App. Hisp. 82 (357); Hann. 12 (49); 40 (173); 53 (224); Pun. 106 (499);
122 (577–578); 132 (629); Ill. 5 (12); Syr. 28 (139); 28 (141); BC 1.113 (526); 2.62 (259–260);
2.63 (262); 2.67 (279); 2.71 (297–299); 2.116 (489); 4.134 (563); 5.87 (365); 5.112 (466); 5.140
(583); cf. also Goldmann 1988: 27–49.
more frequent are interpretative comments at the end of a story,18 or

elaborate conclusions at the end of a particular phase of the Roman
expansion. In these passages evaluation is frequently achieved through
comparison of the virtues of the conquering Romans with those of the
conquered peoples. The endurance and qualities of the latter are paid
no less tribute to than to successful Roman generals such as Pompeius
or Scipio:
Such was the love of liberty and of valour which existed in this small
barbarian town [sc. Numantia]. With only 8,000 fighting men before
the war began, how many and what terrible reverses did they bring
upon the Romans! What treatises did they make on equal terms with
the Romans, which the latter would not consent to make with any other
people! How often did they challenge to open battle the last general sent
against them [sc. Scipio], who had invested them with an army of 60.000
men! But he showed himself more experienced in war than themselves,
by refusing to join battle with wild beasts when he could reduce them by
that invincible enemy, hunger. In this way alone was it possible to capture
the Numantines, and in this way alone were they captured.
(Hisp. 97 [419–420])19
Furthermore, important leaders such as Mithridates, Sulla, Caesar or
Cassius and Brutus are evaluated in quite extensive obituaries that
recapitulate and judge the life and the achievements of the man in
question.20
On the whole, the narrator’s activity as commentator and interpreter
seems more perceptible in the Civil Wars than in the other surviving
books. This part of the work shows many commenting remarks, most
of which refer to the pathology of party strife and civil war and par-
ticularly to the miseries caused by the proscriptions.21 As for the latter,
numerous instructive examples are recorded for the benefit of the nar-
ratees:
18 Cf. e.g. App. BC 1.84 (379–381); 1.105 (491); 2.103 (426); 3.85 (349); 4.16 (61–64);
4.36 (154); 4.115 (479); 4.116 (488).

19 Cf. also App. Hisp. 98 (425); Hann. 25 (108–110); Syr. 50 (251–254).
20 Cf. e.g. App. Hisp. 75 (318–319): Viriathus; Pun. 106 (499–500): Masinissa; Syr. 63
(331–334): Seleucus; Mith. 112 (540–551): Mithridates; BC 1.17 (71–72): Gracchus; 1.20
(85): Scipio; 1.103–106 (478–500): Sulla (a general evaluation precedes the account of
Sulla’s death); 2.86 (363): Pompey; 2.99 (412–414): Cato Uticensis; 2.149–154 (619–649):
Caesar (the most extensive obituary in the surviving books providing a detailed synkrisis
with Alexander the Great); 4.20 (82): Cicero; 4.132–134 (553–568): Cassius and Brutus.
21 Cf. e.g. App. BC 1.20 [85]; 1.58 (259); 1.99 (463); 1.101 (471); 1.102 (474); 2.7 (24);
4.35 (147); 5.17 (68), 5.18 (72); 5.49 (208).

Such are examples of the extreme misfortunes that befell the proscribed.
Instances where some were unexpectedly saved and at a later period
raised to positions of honour are more agreeable to me to relate, and will
be more useful to my readers, as showing that they should never fall into
despair, but that hope will always remain to them …
(BC 4.36 [149])22
In fact, the period of the civil wars is depicted as the crucial phase of
Roman history leading eventually to the present happy state of which it
is shown to be the antithesis:
These things took place not in an ordinary city, not in a weak and petty
kingdom; but the evil deity thus shook the most powerful mistress of so
many nations and of land and sea, and so brought after a long period of
time the present well-ordered condition. (BC 4.16 [61])23
A third, though clearly less prominent type of narratorial interventions
in the Roman History, refers to the narrator’s activity as historian and
enquirer. Sources are hardly ever mentioned by name,24 though anec-
dotal episodes and numerical data are frequently labelled as simply
being taken over from other authorities (‘it is said’, ‘I have learnt’).25
The veracity of such reports is, however, seldom doubted explicitly.26
Furthermore, a few alternative versions are noted,27 and in some in-
stances, the narrator asserts that he had been in vain looking for more
or clearer evidence:
Thus much I have been able to learn concerning the early history of
the Illyrians and Pannonians, and not even in the commentaries of the
second Caesar, surnamed Augustus, could I find anything earlier about
the Pannonians. (Ill. 14 [42])
Although I have searched, I have not been able to find any clear account
of what Antony wrote in reply. (BC 5.21 [83])28
22 Cf. App. BC 4.5–51 (16–224); cf. also Gowing 1992: 247–269.

23 Cf. also App. BC 1.6 (24); 4.16 (64).
24 Cf. App. Pun. 132 (631); BC 2.9 (33); 2.39 (158); 3.77 (315).
25 Cf. e.g. App. Hisp. 13 (49); 30 (119); 56 (235); Hann. 3 (10); 26 (111); 31 (130); 56 (234);
Pun. 56 (245); 131 (626); 131 (628); 136 (645); Syr. 10 (38); 56 (283); 56 (286); 58 (299); 64
(337); BC 1.20 (84); 1.22 (93); 1.61 (274); 1.65 (297); 1.104 (484); 2.8 (26); 2.25 (97); 2.39
(155); 2.60 (249); 2.62 (260); 2.64 (269); 2.71 (296); 2.77 (324); 2.91 (384); 2.102 (421); 2.102
(425); 2.104 (433); 2.109 (454; 4.20 (81); 4.112 (471); 5.8 (33); 5.59 (249); 5.100 (417); 5.116
(484); 5.132 (547).
26 Cf. e.g. App. BC 2.82 (346); 3.84 (347).
27 Cf. e.g. App. BC 1.20 (83–84); 2.70 (289); 2.116 (488); 3.42 (173); 3.77 (315); 4.118
(498).
28 Cf. also App. Ill. 6 (16); 10 (29); 29 (84); 30 (86); Mith. 8 (24); BC 1.86 (391); 1.100
(467); 2.82 (345).

In one case the narrator also claims to have visited a scene of action in
order to conduct his research:
He [sc. Cicero] fled in a small boat, but as he could not endure the sea-
sickness, he landed and went to a country place of his own near Caieta, a
town of Italy, which I visited to gain knowledge of this lamentable affair,
and here he remained quiet. (BC 4.19 [73])
These occasional remarks, then, quite unobtrusively and incidentally
show the narrator’s familiarity with the tasks of historiographical en-
quiry.
Secondary narrators
As in most historiographical works of the imperial period, secondary

narrators do not play an important role in Appian’s Roman History.
The bulk of the character-text consists of speeches delivered by gen-
erals (Romans and non-Romans alike) before decisive battles, petitions
brought forward or offers made by ambassadors as well as answers
given by the other side, and debates on legal issues. Conflicting views
are thus made plain by the opponents themselves, and the personalities
of the characters speaking are highlighted by their own words. Stories,
however, are only rarely inserted in such speeches. Historical exam-
ples are sometimes cited to back up an argument, or characters reflect
on past events and express their own view of events that frequently
are already known to the narratees.29 There is one independent story
told by a character: Sulla’s recounting of a short parable in the peo-
ple’s assembly, with which ‘he terrified them and thereafter ruled as he
pleased’ (BC 1.101 [472–473]). Otherwise, the primary narrator reserves
the telling of the story to himself.
Narratees
The narratees, who are never addressed in the second person, are
clearly less overt than the primary narrator. But there are still quite
a few traces of their presence to be found in the text such as numerous
parenthetical gar-clauses that provide necessary background informa-
29 Cf. e.g. App. Pun. 63 (278–282); 83 (386–392); 87 (408–412); BC 2.140–141 (581–
591); 3.37 (147–151); 4.91–94 (381–395).

tion (e.g. Hann. 4 [15]: ‘When he came to the Alps and found no road
through or over them—for they are exceedingly precipitous— …’),30
occasional instances of the ‘anonymous interlocutor’ device (e.g. Mith. 7
[23]: ‘If anyone is eager to know the sequel …’),31 one ‘anonymous wit-
ness’ (Hann. 34 [142]: ‘The port of Tarentum is on the north side as one
sails in from the sea …’), some if not-situations (e.g. Pun. 91 [429]: ‘They
would have been torn to pieces had they not said that they must make
their first communication to the senate’),32 as well as a few instances
of presentation through negation (e.g. BC 1.11 [43]: ‘What Gracchus
had in his mind in proposing the measure was not money, but men’).33
More prominent and striking a feature that marks the narrator’s inter-
action with his narratees, however, is the abundance of short explana-
tions inserted throughout the work. Some of them deal with questions
of geography or with foreign customs, but most refer to Roman insti-
tutions, customs, titles and laws, which were introduced or invented at
one point in history but are still in use in the narrator’s own days (the
‘continuance’ motif):
For among the Romans the negative veto always defeats an affirmative
proposal. (BC 1.12 [48])
For it is customary among the Romans for the adopted son to take the
name of the adoptive father. (BC 3.11 [38])
At this time, they say, originated the custom and system of cohorts of
night watchmen still in force. (BC 5.132 [547])34
Clearly, the narratees are supposed to wish to share the narrator’s own
intimate and partly first-hand knowledge of Roman institutions.
30 Cf. also App. Hisp. 5 (18); 9 (35); 12 (44); 14 (53); 20 (77); 20 (78); 54 (229); 79 (340);
80 (349); 95 (414); Hann. 8 (31); 39 (165); Pun. 11 (45); 80 (377); 98 (463); 104 (493); 112
(530); 114 (543); 123 (584); 124 (588); 132 (630); Ill. 15 (43); Syr. 8 (30); 36 (184); 37 (190);
41 (215); 48 (248); 52 (261); Mith. 4 (11); 12 (41); 20 (78); BC 1.15 (66); 1.57 (251); 1.70 (323);
1.74 (342); 1.86 (389); 1.91 (421); 1.114 (533); 1.115 (534); 1.116 (541); 2.11 (37); 2.31 (123);
2.52 (213); 2.68 (280); 2.82 (345); 2.112 (469); 2.123 (516); 2.128 (535); 2.143 (599); 3.7 (23);
3.10 (35); 3.42 (173); 3.46 (188); 3.46 (189); 3.74 (302); 4.19 (74); 4.128 (536); 5.9 (37); 5.37
(154); 5.79 (334); 5.83 (352); 5.127 (525).
31 Cf. also App. Pun. 124 (589);
32 Cf. also e.g. App. BC 2.64 (268).
33 Cf. also e.g. App. Pun. 124 (588); BC 1.104 (488)
34 Cf. also e.g. App. Hisp. 1 (1–4); 12 (47); 75 (319); Pun. 71 (323); BC 1.38 (173); 1.103
(479); 2.2 (55); 2.7 (25); 2.39 (157); 2.44 (177); 2.61 (256); 2.85 (359); 2.86 (362); 2.96 (402);
2.116 (488); 2.120 (505–506); 2.148 (618); 3.43 (178); 3.94 (389–391); 4.7 (27); 5.97 (403); cf.
also Goldmann 1988: 85–115; Van der Leest 1989.
Conclusion
The Appianic narrator is characterized by his presence as a distinct

personality: a proud citizen of Alexandria, living and working in Rome,
who presents himself as a mediator between centre and periphery, and
as an official who is in the position to explain ‘Rome’ and her successful
rise to those in the provinces. The narrator also appears quite overtly
as an investigator, a commentator (particularly in the books on the civil
wars), and most prominently as an organizer of information. The latter,
of course, is due to the quite unusual ethnographic disposition of the
material the narrator so proudly introduces as the achievement most
distinctively of his own.
chapter thirteen
CASSIUS DIO
T. Hidber
A historian by vocation
The Roman History, encompassing the whole period from the mythical
origins of the city up to the reign of Alexander Severus, i.e. a span of
about 1,000 years, and comprising 80 books was a truly monumental
work1 and an outstanding achievement. Indeed, its proud narrator is
showing much concern that its greatness is duly acknowledged by the
narratees. In an excursus, preserved in Xiphilinus’ Epitome, he explains
in detail how he came to write such a large-scale work. After his first
work, ‘a little book about the dreams and portents which gave Severus
reason to hope for the imperial power’, had met with a favourable
response from the emperor, ‘divine power’ (to daimonion) commanded
him in his dreams to write a treatise on the period from the death
of Commodus probably down to the accession of Septimius Severus.2
As this first historical work also won ‘high approval’ ‘not only from
the others, but, in particular, from Severus himself ’, he felt the desire
‘to compile a record of everything else that concerned the Romans’.3
And even after Severus’ death, the late emperor, the narrator says,
appeared to him in a dream encouraging him to continue his history
1 Obviously, this was too voluminous a work to survive entirely: The original text
is preserved only in books 36–55, in substantial fragments of books 56–60, and in

a fragmentary section of books 79(78)–80(79). However, at least the outlines and the
structure of the first quarter of the work can be deduced from books 7–9 of John
Zonaras’ Epitome of Histories, whereas the last quarter can be reconstructed from the
Epitome made of books 36–80 by the monk John Xiphilinus in the second half of the
eleventh century; cf. Millar 1964: 1–4.
2 D.C. 74(73).23.1–2 (Xiph.); translation (here and in the following) taken from
Cary 1914–1927. On these two, now lost, minor works—both of which had certainly
encomiastical undertones—cf. Schmidt 1997: 2605–2618.
3 D.C. 74 (73).23.3 (Xiph.).
188 part two – chapter thirteen
beyond the end of Severus’ reign, ‘Come here, Dio. Draw near, that
you may both learn accurately and write an account of all that is said
and done’(79[78].10.1 [Xiph.]).
But ultimately, it was, the narrator claims, divine power again that
gave him the strength to fulfil this huge task:
This goddess (Fortune—Tuchē) gives me strength to continue my history
when I become timid and disposed to shrink from it; when I grow weary
and would resign the task, she wins me back by sending dreams; she
inspires me with fair hopes that future time will permit my history to
survive and never dim its lustre; she, it seems, has fallen to my lot as
guardian of the course of my life, and therefore I have dedicated myself
to her. (74[73].23.4 [Xiph.])
And in the end, divine inspiration also indicated to the historian in a

dream at which point and with which Homeric words he should finish
the History:
I set out for home, with the intention of spending all the rest of my life in
my native land, as, indeed Heavenly Power (to daimonion) revealed to me
… For once in a dream I thought I was commanded by it to write at the
close of my work these verses:
‘Hector anon did Zeus lead forth out of the range of the missiles,
Out of the dust and the slaying of men and the uproar.’
(80.5.3 [Xiph.])4
Applause from those in power and divine inspiration were not nor-
mally considered as legitimate motives for writing history. Whereas the
former seemed too closely related to panegyrical writing and hardly
compatible with historiographical impartiality, the latter was usually a
privilege reserved for poets.5 The Dionic narrator’s reference to these
incentives, generally regarded as problematic, is remarkable indeed
and there are only very few parallels to this kind of call to history.6
Although, unlike in poetry, divine power is not presented as a source of
knowledge or truth but only as a source of hopes and encouragement
for the historian to pursue his task, the Roman History is still claimed to
be sanctioned, as it were, by divine will and, ultimately, to be a work
of such a scope that it could not have been completed, if its narrator’s
4 Hom. Il. 11.163–164.

5 Cf. e.g. Luc. Hist. conscr. 14; 42.
6 The only parallel seems to be the statement of the narrator in Arrian’s Anabasis,
that he embarked on his work ‘not without god’s help’ (Anab. 7.30.3). Cf. Schmidt 1999:
99–101. For dreams as motives cf. Marincola 1997: 43–51.
t. hidber – cassius dio 189
confidence had not been strengthened by supernatural powers. In fact,

the narrator asserts that he ‘spent ten years in collecting all the achieve-
ments of the Romans from the beginning down to the death of Severus,
and twelve years more in composing the work’, thereby demonstrating
the seriousness of his task.7 The narrator, thus, appears as a man who
was not only guided by divine motivation, but was also well equipped
with the intellectual ability necessary to accomplish so much research
and writing. Indeed, his great efforts to find out the truth by gather-
ing and examining as much evidence as possible is especially stressed
in the case of the imperial period, for which reliable information was
particularly hard to find.8
Another motive comes into prominence as the narrative reaches
Dio’s own lifetime, and the narrator, who so far had been external,
turns internal, now often figuring in the world of the narrated events
either as an eyewitness or even as an active participant. For this period
(from the reign of Commodus, viz. from book 73[72]), the narrator
asserts that he was, in fact, better suited for this task than anyone
else, given both his writing skills and his privileged access to informa-
tion as a distinguished senator close to the centre of power. After a
detailed report of Commodus’ performances in the arena, he therefore
announces a more specific account of the events of his lifetime:
And let no one feel that I am sullying the dignity of history by recording
such occurrences. On most accounts, to be sure, I should not have
mentioned this exhibition; but since it was given by the emperor himself,
and since I was present myself and took part in everything seen, heard
and spoken, I have thought proper to suppress none of the details, but to
hand them down, trivial as they are, to the memory of those who shall
live hereafter, just like any events of the greatest weight and importance.
And, indeed, all the other events that took place in my lifetime I shall
describe with more exactness and detail than earlier occurrences, for the
reason that I was present when they happened and know no one else,
among those who have any ability at writing a worthy record of events,
who has so accurate a knowledge of them as I.
(73[72].18.3–4 [Xiph.])
This very self-confident claim is underpinned by the remarkable abun-

dance of information about the historian’s living-conditions and par-
7 D.C. 74(73).23.5 (Xiph.). In Dionysius of Halicarnassus’ Roman Antiquities the narra-
tor also asserts that he spent 22 years working on his history (Ant. Rom. 1.7.2); Marincola
1997: 151–152.
8 D.C. 53.19.1–6.
ticularly his senatorial career, which is why ‘we have for Dio a more
valuable record of personal experience than for any other ancient histo-
rian’.9
The personality of the narrator is therefore that of a well-educated
senator, governor and consul, who has dedicated much of his life to
his historical work, writing from an exceptionally privileged point of
view, and being constantly guided by divine power. Such a figure is
clearly designed to inspire confidence. At the same time it is made
clear that, where the contemporary part of the History is concerned,
the narrator’s mere presence at particular events is reason enough for
their detailed record, since he claims to be the only man who has both
the knowledge and the ability to hand them down to posterity.10 In
this view, later generations are wholly dependent on this record for
their information on the period in which Dio lived. The narrator’s
extraordinary identification of himself with history is most obvious at
the very end of the work, which is not determined by any important
historical break, but by Dio’s own retirement from political life. The
narrative must come to a close, when the narrator is removed from the
‘battlefield’ of his life as a Roman official near to the centre of power.
The narrator of the Roman History is quite overtly present and, on the
whole, certainly much more intrusive than, for example, the narrator
in Herodian’s History, published only two decades or so later. In this
9 Millar 1964: 7; cf. D.C. 49.36.4; 69.1.3 [Xiph.]; 72.7.2 [Xiph.] (father’s career);
73(72).4.2 [Xiph.] (stay in Rome); 74(73).12.2 [Xiph.] (praetor); 75(74).11.2 [Xiph.] (Priscus,
a fellow-countryman); 76(75).4.3 [Xiph.] (a friend of the consul’s of AD 196); 76(75).15.3
[Xiph.] (native from Nicaea); 76(75).16.2–4 [Xiph.] (member of Severus’ consilium);
77(76).16.4 [Xiph.] (3,000 indictments for adultery in the year of his consulate); 79(78).
8.4 [Xiph.] (member of Caracalla’s consilium); 80(79).7.4 [Xiph.] (curator in Pergamum
and Smyrna); 80(80).1.2 [Xiph.] (proconsul Africae); 49.36.4; 80(80).4.2 [Xiph.] (legatus of
Dalmatia and Pannonia Superior); 43.46.6; 80(80).2.1 [Xiph.]; 80(80).5.1 [Xiph.] (consul
ordinarius); 80(80).5.2f. [Xiph.] (visits to Severus Alexander in Campania and Rome,
return to Bithynia).
10 The narrator’s presence at or participation in events he recounts is stressed
in 48.50.4; 73(72).4.2 [Xiph.]; 73(72).7.1f. [Xiph.]; 73(72).18.3f. [Xiph.]; 73(72).21.1–3

[Xiph.]; 74(73).12.2–5 [Xiph.]; 74(73).14.5 [Xiph.]; 74(73).17.4 [Xiph.]; 75(74).2.1 [Xiph.];
75(74).4.1–5.5 [Xiph.]; 75(74).14.4f. [Xiph.]; 76(75).4.3–7 [Xiph.]; 77(76).1.2 [Xiph.];
77(76).8.3 [Xiph.]; 78(77).9.6f. [Exc. Val. / Xiph.]; 79(78).3.3 [Xiph.]; 79(78).8.4 [Xiph.];
79(78).37.5.
respect, Dio seems actually closer to Herodotus (→) than to Thucy-

dides, whose imitator he is generally said to be.11 However, the narra-
tor’s presence is not felt with equal intensity in all parts of the work.
Narratorial interventions are particularly frequent in passages dealing
with crucial phases of Roman history, such as the transition from the
republic to the principate, and in the last books, when he becomes an
internal narrator. The following survey of different types of explicit nar-
ratorial interventions will focus on books 36–60, which preserve Dio’s
original text.
Frequently, the narrator calls attention to his role as organizer and
narrator of the history. Stretches of narrative as well as digressions are
quite regularly introduced by a narratorial intervention (often in the
first person, and at places providing evaluations of the events that are
going to be narrated):
I will give an account of the so-called camelopard, because it was then
introduced into Rome by Caesar for the first time and exhibited to all.
(43.23.1)
I shall now relate in detail also such of his acts as call for mention,
together with the names of the consuls under whom they were per-
formed. (53.22.1)12
Similarly, narrative stretches (stories, digressions, or the accounts of a
whole year) are frequently closed with a short summarizing remark
(43.23.1: ‘Such was the end of this war’; 48.4.1: ‘This was what took
place then; in the following year …’),13 and the report of the death
of an important figure is often followed by an obituary including an
11 For Dio as an imitator of Thucydides cf. Schmidt 1999: 93, n. 3.

12 Cf. also D.C. 36.20.1; 36.24.1; 36.49.1; 37.10.1–3; 37.20.3; 37.28.10; 37.29.1; 38.30.1;
38.31.1; 39.7.1; 39.12.1; 39.17.1; 39.44.1; 40.15.1; 40.21.2; 40.31.1; 40.54.3; 41.43.1; 42.10.1;
42.19.3–4; 42.34.1; 43.3.1; 43.24.4; 44.1.1–2.; 44.3.1; 44.4.1; 44.11.1; 44.15.1; 44.52.1;
45.10.1; 45.12.1; 46.33.1; 46.34.5; 46.35.1–2.; 47.18.1; 47.42.1; 48.13.1; 48.39.1; 48.45.8;
48.49.5; 48.50.4; 49.19.1; 49.30.1; 50.1.2; 50.2.2; 50.4; 51.16.2; 53.21.1–2; 54.26.3; 55.24.1;
56.18.1.
13 Cf. also D.C. 36.16.3; 37.6.1; 37.9.1; 37.16.5; 37.20.1; 37.22.4; 37.47.1; 37.51.3; 37.
57.3; 37.58.1; 38.7.3; 38.14.1; 38.34.1; 39.1.1; 39.29.1; 39.29.1; 39.46.4; 39.56.1; 40.1.1;
40.6.3; 40.11.1; 40.14.1; 40.16.1; 40.30.1; 40.57.1; 41.1.1; 41.3.3; 41.46.1; 41.58.1; 41.59.3;
42.1.1; 42.26.1; 42.27.1; 42.29.1; 42.34.1; 42.44.1; 42.55.4; 43.38.4; 43.46.1; 43.49.1; 44.1.1;
44.35.1; 44.52.1; 44.53.1; 45.10.1; 46.35.1; 46.46.6; 47.11.1; 47.12.1; 47.14.1; 47.15.1; 47.19.4;
48.1.1; 48.4.1; 48.31.1; 48.23.5; 48.34.1; 48.43.1; 48.46.1; 48.51.5; 49.16.1; 49.19.1; 49.23.1;
49.24.1; 49.31.1; 49.33.1; 50.1.1; 50.8.1; 51.1.1; 51.11.1; 51.16.3; 52.41.3; 53.1.1; 53.1.6;
53.13.1; 53.17.1; 53.22.1; 53.22.5; 53.29.1; 53.33.1; 54.5.1; 54.9.1; 54.12.1; 54.30.1; 55.1.1;
55.3.2; 55.9.1; 56.16.4; 56.47.1–2; 57.1.1.
explicit evaluation of that person’s life and deeds. Such closing remarks
frequently highlight the paradigmatic quality of what has been narrated
by referring to (rather commonplace) universal truths, such as the
instability of fortune:
Such was the end of Pompey the Great, whereby was proved once more
the weakness and the strange fortune of the human race. (42.5.1)14
The transition from the late republic to principate is particularly mark-
ed as an important caesura by an explicit closing statement, later
followed by a positive evaluation of this development:
Such were the achievements of the Romans and such their sufferings
under the kingship, under the republic, and under the dominion of a
few, during a period of seven hundred and twenty-five years. After this
they reverted to what was, strictly speaking, a monarchy … (52.1.1)
In this way the government was changed at that time for the better and
in the interest of greater security; for it was no doubt quite impossible for
the people to be saved under a republic. (53.19.1)
Some stories or reports are introduced or closed with the narratorial
remark that these events had, in fact, been announced preliminarily by
certain prodigies. Such comments imply that the course of history is in
one way or other determined by supernatural powers:
And of this, I think, the sea-monster had given them full warning before-
hand. (37.10.2)
All the inhabitants who resisted for a time were finally subdued, as
indeed, Heaven very clearly indicated to them beforehand. For it rained
not only water where no drop had ever fallen previously, but also blood
… (51.17.4)15
The narrator also intervenes regularly in order to make casual and
unspecific cross-references to what has been narrated earlier (‘as I said
before’)16 or, though much less frequently, to what is to be recounted
14 Cf. also D.C. 42.5.1–7; 42.16.1; 51.15.1–4; 54.29; 55.7; 55.7; 56.43.4–56.45.3; 57.14.1;
63.15.1–2; 72.34.2–36.4.
15 Cf. also e.g. D.C. 37.58.2–3; 40.47.3; 44.4.4; 45.17.4–5; 46.33.1–46.34.1; 47.1.2–3;
47.40.2; 48.14.6; 48.43.4–6; 50.8.4–6; 53.33.5; 55.1.5; 67.16.1 [Xiph.]; 79(78).30.1.

16 Cf. e.g. D.C. 36.13.1; 36.21.1; 37.22.1; 37.52.3; 38.7.1; 38.9.4; 38.13.3; 39.63.5; 40.
66.1; 41.14.5; 41.17.3; 41.49.1; 41.60.4; 42.3.1; 42.23.2; 42.50.2; 42.56.2; 43.13.3; 43.20.1;
43.33.2; 43.44.1; 44.13.1; 46.55.5; 47.20.1; 48.14.1; 48.20.3; 48.21.1; 48.22.2; 48.22.3;
48.45.8; 49.3.3; 49.8.2; 49.44.3; 51.23.2; 52.13.2; 52.41.3; 53.13.2; 53.14.6; 53.15.2; 53.20.1;
53.25.2; 54.16.7; 55.5.1.
later on (‘as the progress of the narrative will show’).17 Similar interven-
tions mark changes of scene (e.g. 37.47.1: ‘while this was the course
of affairs in the city, the Allobroges were devastating Gallia Narbo-
nensis’),18 and make plain chronological deviations from the annalistic
scheme:
But these things were done later. At the time he sent a part of the fleet in
pursuit of Antony and Cleopatra. (51.1.4)
Now that I have once been led into giving an account of the legions, I
shall speak of the other legions also which exist today … my purpose
being that, if one desires to learn about them, the statement of all the
facts in a single portion of my book may provide him easily with the
information. (55.24.1)19
Another frequent type of narratorial intervention refers to the narra-

tor’s investigating activity as a historian: in an attitude known from
Herodotus (→) the narrator often confesses that he does not know the
reasons for a particular deed or how something came to be,20 refers
to alternative versions,21 or stresses that he simply records what he has
heard or read, thereby shifting the responsibility for those reports to his
(almost always unnamed) sources.22 Arriving at the time of the prin-
cipate beginning with Augustus—a time, in which ‘things that hap-
pened were kept secret and concealed, and even though some things
are perchance made public they are distrusted just because they can
not be verified’ (53.19.3)—he sets out his method of investigation in
some detail:
Hence in my own narrative of later events, so far as they need to be
mentioned, everything that I shall say will be in accordance with the
reports that have been given out, whether it be really the truth or
otherwise. In addition to these reports, however, my own opinion will be
given, as far as possible, whenever I have been able, from the abundant
17 Cf. e.g. D.C. 53.12.8; 53.29.3; 54.9.1; 74(73).6.3 [Xiph.].

18 Cf. also e.g. D.C. 38.31.1; 39.6.1; 37.52.1; 39.4.1; 39.45.1; 39.54.1; 40.1.1; 40.11.1;
41.25.1; 41.40.1; 42.15.1; 42.17.1; 46.45.1; 47.31.1; 48.36.1; 48.41.7; 48.45.1; 53.29.1; 53.31.1;
54.5.1; 54.9.1; 56.11.1.
19 Cf. also e.g. D.C. 37.24.1; 39.17.1; 40.42.1; 44.4.1; 45.16.1; 45.23.7; 46.4.1–2; 48.15.4;
48.16.1; 48.43.4; 48.45.4; 48.15.4; 49.14.6; 49.22.1; 49.35.1; 50.1.2; 51.1.4; 51.19.1; 53.31.1;
56.46.1; 59.15.6.
20 Cf. e.g. D.C. 36.11.1–2; 37.17.1; 38.13.5; 39.38.4; 43.24.4; 47.3.3; 47.24.4; 48.50.4;
49.4.1; 49.4.4; 50.12.6; 51.22.9; 53.1.2; 53.22.3–4; 54.9.9.

21 Cf. e.g. D.C. 39.61.1; 39.64.1; 40.27.3; 41.14.4; 41.49.2–3; 51.14.1; 54.4.2; 54.28.4.
22 Cf. e.g. D.C. 38.13.5; 39.38.5; 40.54.3; 42.2.5; 48.14.4; 49.25.5; 50.12.5; 51.27.2;
53.22.3–4; 54.9.9; 54.35.3–4.; 57.3.3–5.

evidence which I have gathered from my reading, from hearsay, and

from what I have seen, to form a judgment that differs from the common
report. (53.19.6)
As for the account of plots against emperors, to follow the published
versions of the affairs the narrator announces:
It is not possible, of course, for those on the outside to have certain
knowledge of such matters; for whatever measures a ruler takes, either
personally or through the senate, for the punishment of men for alleged
plots against himself, are generally looked upon with suspicion as having
been done out of spite, no matter how just such measures may be.
For this reason it is my purpose to report in all such cases simply the
recorded version of the affair, without busying myself with anything
beyond the published account, except in perfectly patent cases, or giving
a hint as to the justice or injustice of the act or as to the truth or falsity
of the report. Let this explanation apply to everything that I shall write
hereafter. (54.15.1–4)
This sort of narratorial comment creates the impression of transparen-
cy, as the narrator elucidates gaps, contradictions, and possible bias in
the sources, which prove his trustworthiness as a scrupulous investigator
and reporter.
Another and quite remarkable feature of the narratorial interven-
tions in Dio is the narrator’s constant reflection on the choices he
makes. Time and again the narrator justifies why he does or does not
include certain material in his History. He often announces that he will
omit certain pieces of information as either not suitable for a historical
work or as serving no purpose:
I shall omit those honours which had either been voted to some others
previously … or which, while novel and proposed now for the first time,
were not confirmed by Caesar, for fear that I might become wearisome,
were I to enumerate them all. This same plan I shall follow in my
subsequent account, adhering the more strictly to it, as the honours
proposed continually grew more numerous and absurd. Only such as
had some special and extraordinary importance and were confirmed will
be related. (42.19.3–4)
I need not enumerate them [i.e. the laws enacted by Augustus] all
accurately one by one, but only those which have a bearing upon my
history; and I shall follow this same course also in the case of later events,
in order not to become wearisome by introducing all that kind of detail
that even men who devote themselves to such studies do not know to a
nicety. (53.21.2)23
23 Cf. also e.g. D.C. 47.10.1 (‘I shall accordingly refrain from giving an accurate and
These and similar omissions are claimed to be made in the interest

of the narratees, as otherwise the narrative would become wearisome.
In this view, detailed enumerations of laws, honours, numbers, dates,
actions in war or list of names are considered to be boring and not
useful in a historical work. Accordingly, very often only one particularly
remarkable or instructive example is provided in order to illustrate
the character of a given historical figure or a whole type of events or
deeds. At the same time, such statements can give the impression that
the narrator has got a complete overview of history, since he seems to
know much more than he actually records. They may also be read as
apologies directed towards those among the narratees who would, in
fact, hope to find the very detail the narrator omits on purpose.
On the other hand, the narrator frequently justifies his record of
certain events or data explicitly (and sometimes at length):
This I have written by way of digression from my history, so that no one
might be ignorant of the stories told about Caesar. (43.43.5)
At any rate, he [sc. Domitian] spent most of his time at the Alban Villa
and did many absurd things, one of them being to impale flies on a
stylus. Unworthy as this incident is of the dignity of history, yet, because
it shows his character so well and particularly because he still continued
the practice after he became emperor, I have felt obliged to record it.
(65.9.4 [Xiph.])24
Such narratorial interventions normally serve to justify the inclusion

of anecdotal material or digressions, which do not seem to bear any
relevance to history at first sight. Again, then, the material is claimed
to be selected according to the benefit of the addressees: digressions
record what is of general interest, whereas anecdotes can be particularly
instructive regarding either the general conditions of human life or the
character and behaviour of a particular historical figure.
detailed description of all such incidents, since this would be a vast undertaking and
there would be no great gain to my history, but shall relate what I regard as most wor-
thy of remembrance’); 48.13.1 (‘The most of these operations, especially those involving
no great or memorable achievement, I will pass over, but will relate briefly the points
which are most worthy of mention’); 37.17.4; 38.7.8; 41.2.2; 43.14.7; 43.22.4; 43.24.2;
43.25.1; 43.46.1; 43.46.6; 44.14.3; 47.13.1; 48.13.1; 51.16.2; 51.19.3; 51.20.3; 54.23.2; 54.
23.8; 55.3.2; 55.9.1; 55.28.2f.; 59.22.5; 60.11.6.
24 Cf. also e.g. D.C. 37.10.1; 37.18.3; 38.13.3; 39.17.1; 40.15.1; 45.16.1; 46.35.1; 51.1.1;
53.12.8; 54.19.4; 54.23.1; 57.24.6; 60.26.5; 62.27.3.

A partially omnipresent and omniscient narrator
Although, as we have seen, the narrator shows himself remarkably con-

cerned about the availability of evidence, he does not abstain from
recording such intimate scenes as a private encounter between Cicero
and Philiscus in Macedonia (38.18.1–29.4), a constitutional debate be-
tween Agrippa and Maecenas, witnessed only by Augustus himself
(52.2.1–41.1), or even a lengthy bedroom-talk between Augustus and
Livia (55.14.2–21.4). Obviously, the narrator’s omnipresence and omni-
science in these and similar instances serve as means to present an
in-depth discussion of topics of a wider bearing, such as the right
behaviour of an exile, the ideal constitution, or the loneliness of the
monarch. In other instances, the narrator also has access to the
thoughts, feelings, and even dreams of his characters by means of
embedded focalization (e.g. 45.2.2–3: Cicero and Catullus dreaming of
Octavius) and thus can make clear the hidden motives of these charac-
ters.
Narratees
The large amount of narratorial interventions indicates the narrator’s

interaction with his narratees, even though he never directly addresses
them in the second person. There are also numerous other traces of the
narratees: instances of the ‘anonymous interlocutor’ motif (e.g. 47.39.1:
‘One would naturally surmise that this struggle proved tremendous and
surpassed all previous civil conflicts of the Romans’),25 if not-situations,26
rhetorical questions,27 or gar-clauses anticipating questions by the nar-
ratees which either provide the necessary background information or
explain a given detail by referring to ‘universal truths’, often presented
in the form of a gnome.28 However, it is hardly possible to detect a coher-
25 Cf. also e.g. D.C. 36.16.1; 41.4.3; 41.63.6; 47.17.2; 47.39.1; 50.33.8.
26 Cf. e.g. D.C. 36.45.5; 36.49.7; 37.16.1; 39.65.2; 40.24.1; 42.26.3; 42.35.3; 43.2.2;
46.40.4; 46.37.5; 54.33.2.
27 Cf. e.g. D.C. 42.27.4; 47.13.2; 51.4.8; 53.22.4.
28 Cf. e.g. D.C. 36.24.3; 37.33.1; 38.12.5; 38.16.6; 39.27.3; 39.35.1; 39.58.1; 39.63.6;
40.23.4; 40.61.3; 41.17.2; 41.53.3; 42.1.4; 42.3.3; 42.34.6; 42.40.4; 43.8.1; 43.27.3; 44.51.1;
44.53.4; 45.4.2; 46.34.2; 46.43.3; 46.43.6; 47.29.5; 47.30.4; 48.3.2; 48.16.2; 48.38.2; 48.
53.4; 49.21.3; 49.22.4; 49.32.1; 49.44.3; 51.26.3; 53.10.3; 53.15.3; 53.22.2; 54.10.1; 54.10.3;
55.2.1.
ent picture of the narratees. Indeed, a fragment from the first book
preserves a statement indicating that the History is intended to provide
information for a broad audience:
It is my desire to write a history of all the memorable achievements of
the Romans, as well in time of peace as in war, so that no one, whether
Roman or non-Roman, shall look in vain for any of the essential facts.
(F 1.1 B p. 12)
The numerous statements explaining why some particular piece of
information is or is not recorded, might also be seen as part of a strat-
egy to take into account diverging expectations among the audience as
to what was to be included in a historiographical work. Thus, it would
be wrong to infer from the regular use of the we-form in the records of
Dio’s experiences as a senator (or even as consul)29 as well as from his
judgment of the emperors depending on the rulers’ attitude toward the
senators,30 that fellow-senators were addressed in the first place. As we
have seen, it is also made clear that he hopes that his work will be read
by future generations too(73[72].18.3 [Xiph.]).
There is an unusual and striking amount of passages pointing to the
time of narration (more than fifty in the preserved books alone), such
as:
… the festival which they celebrate even now around the altar of Augus-
tus at Lugdunum. (54.32.1)
He [sc. Augustus] had the names of all the senators entered on a tablet
and posted; and this practice, originating with him, is still observed each
year. (55.3.3)31
Most of these references are designed to explain Roman customs and
institutions, or the origin of buildings still in use or extant in the narra-
tor’s own day, and thus they further the understanding of many particu-
lar aspects of the Roman Empire for a contemporary audience that has
got less access to the relevant information in question. However, these
29 Cf. e.g. D.C. 43.46.6; 74(73).14.5 [Xiph.]; 74(73).17.4 [Xiph.]; 75(74).2.1 [Xiph.];
75(74).4.6 [Xiph.]; 77(76).1.2 [Xiph.]; 78(77).9.6–7 [Exc. Val./Xiph.]; 79(78).37.5 [Xiph.].

30 Cf. Gowing 1992: 25–28; Hose 1994: 406–417.
31 Cf. also D.C. 36.6.2; 36.11.1; 36.20.1; 36.50.3; 37.15.1; 37.18.1; 37.20.2; 37.28.3;
37.46.4; 39.41.1; 39.49.2; 39.50.5; 40.14.4; 40.54.2; 41.49.2; 41.61.4; 43.44.4; 43.46.6;
43.49.1; 43.51.3; 46.46.3–4.; 46.50.5; 47.18.3; 47.49.2; 48.35.3; 48.49.5; 49.14.5; 49.37.3;
50.12.8; 51.17.3; 51.20.8; 51.27.3; 52.42.6; 52.43.2; 53.12.8; 53.16.3; 53.17.7; 53.18.4; 53.
20.4; 54.18.2; 54.23.8; 54.25.2; 54.34.2; 55.2.6; 55.3.2; 55.8.7; 55.17.7; 55.22.4; 55.23.2;
55.24.1; 55.26.5; 55.27.6; 57.8.4; 57.9.3; 60.7.4; 60.11.5.
explanations can also be seen as having particular importance with a

view to future narratees, since they are based on the narrator’s own
authority as an eyewitness and contemporary. In fact, taken together,
these references create a sort of encyclopedia.32
It is implied that the narratees have a genuine interest in the history
of Rome and the Roman Empire, which occasionally (though not con-
stantly and not following a perceivable pattern) is depicted as ‘ours’33
and, therefore, as the narrator’s and—at least to some extent—the nar-
ratees’ own world.
Secondary narrators
The large amount of character-text and the extraordinary length of

some of the speeches and dialogue-scenes have struck many read-
ers. However, characters normally do not tell stories, but only make
announcements, exhort or condemn, raise hopes or fears, or discuss
issues of a wider impact. Occasionally, though, they insert pieces of
narrative in their argumentation. Thus, Pompey, aspiring to the com-
mand against the pirates, reminds people of his deeds (36.25.1–3), Cae-
sar, trying to incite his officers, refers to the great exempla of the fore-
fathers (38.37.5–38.38.4), or Cicero, trying to convince the Senate to
take measures against Antony, enumerates his enemy’s outrages (45.26–
32).
Conclusion
On the whole, the Roman History is clearly dominated by its overt and
intrusive primary narrator, who presents himself as a circumspect, trust-
worthy, and well-informed guide to a historical work of monumen-
tal proportions. His presence and self-confidence can at one point be
felt even in the words of one of his characters. It is the philosopher
Philiscus, who seems to be speaking as a narratorial alter ego when he
32 Cf. also D.C. 55.24.1.

33 Cf. D.C. F 1.1 [B p. 12] (‘I will begin at the point where I have obtained the
clearest accounts of what is reported to have taken place in this land which we inhabit’).
However, whereas it is e.g. said that ‘Nisibis … now belongs to us’ (36.6.2), the Arabians
are stated to be ‘now subjects of the Romans’ (37.15.1).
gives advice to Cicero and recommends historiographical writing as a

most noble, worthy, and rewarding activity for a former consul and
renowned intellectual:
For if you will take my advice, you will be quite satisfied to pick out a
little estate in some retired spot on the coast and there carry on at the
same time farming and some historical writing, like Xenophon and like
Thucydides. This form of learning is most enduring to every man and
to every state … If, then, you wish to become really immortal, like those
historians, emulate them. You have the necessary means in sufficiency
and you lack no distinction. For if there is any virtue in such honours,
you have been consul … (38.28.1–3)
chapter fourteen
HERODIAN
T. Hidber
An elusive persona
Little is known about Herodian, the man who has left no other traces
of his life but his History of the Empire after Marcus, which deals in
eight books with the imperial history from the death of M. Aurelius
in AD 180 to the accession to the throne of Gordian III in AD 238.
So little in fact, that by some he is even called a ‘mystery’.1 This
is mostly due to the extraordinarily vague way the narrator of the
History presents himself, as he hardly conveys any personal information
regarding himself. Although he establishes his voice in a proem at the
outset of the work, he does not reveal his name (though his cognomen
Herodian was certainly mentioned in the title), nor his country of
origin, nor his social standing, nor the place, nor the time nor any
other circumstances of the composition of his work. What is stressed,
however, is the (partially) internal status of the narrator:
I have written a history of the events following the death of Marcus
which I saw and heard in my lifetime. I had a personal share in some
of these events during my imperial and public service (1.2.5)2
The remarkable lack of biographical detail in the proem as well as in

the narrative may be seen as part of a strategy to demonstrate the
historian’s impartiality and authority: the narrator’s historiographical
persona is that of a completely independent and almost anonymous civil
servant, who has witnessed the events and the deeds of the emperors
of his own time as an impartial observer unbiased by any local pride,
personal partiality or political ambitions. The picture is created of a
modest Greek, who has spent at least part of his professional career
1 Roques 1990: 1; cf. also Müller 1996: 9.

2 Translations in this contribution all taken from Whittaker 1969–1970.
202 part two – chapter fourteen
in Rome and in the west, and was therefore closer—though far less
close than e.g. Cassius Dio—to the centre of power than most of the
inhabitants of the Greek-speaking parts of the empire.
At the beginning of the proem this persona is established competitively
against two groups of predecessors in the historiographical field, both
of which are said to have neglected the truth. On the one hand, those
historians who, dealing with the more distant past, were primarily con-
cerned with a display of vocabulary and style, and introduced fabulous
elements (to muthōdes) in order to please their audience, and to win per-
manent glory for their own learning. On the other hand, those who
wrote history based on hatred or in order to flatter an emperor, a city
or a private individual (1.1.1–2). None of these predecessors is men-
tioned by name, but readers might have guessed that Cassius Dio (→)
was one of them, who had written not only a comprehensive Roman
History from the mythical foundation of Rome to his own days, but also
two panegyrically flavoured treatises on the events before the acces-
sion of Septimius Severus. The Herodianic narrator emphasizes that,
in contrast to these predecessors, he is not seeking any personal benefit,
but only aims at sharing with his readers the knowledge he has care-
fully acquired and compiled. It is the significance of the events them-
selves, compressed into a short time, their astonishing peculiarity and
newness (particularly the quick succession of rulers and the reigns of
some very young emperors) that have called him to history (1.1.3–6).
The argumentation, the formulation, and the tone of the first part of
the proem are clearly influenced by Thucydides’ famous digression on
method (1.21–22). This suggests to the narratees that the work at hand
belongs to the historiographical subgenre of contemporary history nor-
mally regarded as most trustworthy, because particularly reliable meth-
ods such as autopsy and eyewitness-accounts could be used. Thus, the
decision to present contemporary history could also be seen as reflect-
ing the narrator’s sincere and truthful character. At the same time, he
could derive authority from the alignment with Thucydides, who was
almost generally acknowledged to have matched the highest historio-
graphical standards of truthfulness.3
Apart from the proem (1.1.1–1.2.5), there is one other passage in
which the narrator reflects at some length on his narrating activity:
at the end of the second book he justifies the lack of detail in his
3 Cf. e.g. Luc. Hist. conscr. 38–39.

t. hidber – herodian 203
report on Septimius Severus by reminding the narratees of his choice

not to write in a flattering manner about a single emperor (as the
authors of panegyrical accounts of the life and deeds of an emperor
used to), but on the reigns and the most important actions of many
emperors he has witnessed during his lifetime (2.15.6–7). This passage
almost apologetically rejects, as it were, the narratees’ expectations of a
panegyrical account, and again stresses the narrator’s commitment to
truth and to the demanding standards of contemporary history.
Whereas the historical author is hardly discernible behind the narra-

tive, the primary narrator is more overtly present throughout the work.
The presence of the narrator is explicit at the beginning of every book,
where the narratees find a short recapitulating remark concerning the
content of the preceding book, such as:
In the previous book a description was given of the death of Pertinax, the
overthrow of Julianus, Severus’ arrival in Rome and the departure of his
expedition against Niger. (3.1.1)4
Furthermore, explicit, if rather casual, references to what has been nar-

rated earlier can be found at the end of books 3 and 7 as well as in a few
other passages of the narrative.5 Four passages also contain references
to events the narrator claims he will relate later on in the text; two of
these events, however, are not actually mentioned again in the text.6
But, generally speaking, in Herodian—as in Thucydides and many
other historians7—cross-referencing is accomplished by implicit paral-
lels (such as verbatim repetitions) rather than by narratorial comments.
The activity of the narrator is also revealed, when after a character-
speech, the narrative is resumed with a formulaic statement of the type
‘after about such a speech …’ (toiauta tina eipontos autou),8 which make
clear that the wording of those character-speeches is in fact the narra-
tor’s own.
4 Hdn. 2.1.1; 3.1.1; 4.1.1; 5.1.1; 6.1.1; 7.1.1; 8.1.1.

5 Hdn. 1.17.12; 3.7.5; 3.9.7; 7.2.1.
6 The cross-references in question that are not followed-up concern details (Hdn.
4.14.2; 5.3.9), whereas the other two (1.14.6; 7.12.9) are much more unspecific.
7 Gribble 1998: 64. This method goes back to Homer (→).
8 Cf. Hdn. 2.2.9; 2.10.9; 2.3.11; 2.8.6; 3.6.8; 3.12.3; 4.5.7; 4.14.8; 6.4.1; 7.5.7; 7.8.9;
More important is the structuring function of the narratorial voice.

Herodian’s narrative can be seen as a series of often quite distinct
episodes interlaced with character speeches, digressions, and descrip-
tions. Whereas digressions are only rarely explicitly introduced as such
(though the break in the main narrative is usually signalled by a general
statement such as e.g. 3.14.6: ‘Most of Britain is marshland because
it is flooded by the continual ocean tide’), the beginning of a new
episode in the main narrative is often highlighted by an introductory
remark, and occasionally, the end of a narrative unit is marked by the
narratorial voice as well. An introductory remark often looks ahead,
e.g.:
At last it became imperative to check the madness of Commodus and to
free the Roman empire from tyranny. (1.16.1)
But it was obviously inevitable that Antoninus’ life would come to an end
and this was how the event occurred. (4.12.3)9
A closing narratorial statement either simply recapitulates the immedi-
ate effect of the narrated event or additionally refers to its consequences
in the future, and its broader paradigmatic quality:
This was the first and foremost reason why young Commodus hated the
senate. (1.8.7)
The fact that there was nobody to take revenge on the perpetrators of
this savage murder of an emperor [i.e. Pertinax] … was a prime cause
in the development of a shameful state of discipline that had permanent
consequences for the future. (2.6.14)
This sort of provocation embittered Maximinus’ feelings, making him
even harsher and more savage than he was instinctively before.
(7.1.12)10
The narratorial voice provides an interpretative framework that con-
firms meanings that are already implicit or explicit in the narrative
itself. In addition, it regularly points out the importance of the episode
for the further evolution of the events, and thus marks the stages
8.3.7; 8.7.7. Similar phrases can also be found after reported speeches (e.g. 3.15.3; 4.3.9;
6.4.6).
9 Cf. also Hdn. 1.9.6; 1.10.1; 1.12.3; 1.14.1; 2.5.1; 3.5.6; 4.9.1; 4.10.1; 4.12.3; 4.13.1;
5.3.1; 7.1.1; 7.9.1; 7.11.1.

10 Cf. also Hdn. 1.13.6; 2.13.12; 3.4.7; 3.7.8; 3.9.12; 4.11.8; 5.4.12; 5.8.1; 6.1.7; 6.2.1;
6.5.10; 6.6.6; 7.1.12; 7.12.7; 8.7.8.

of the particular reigns and elucidates their line of development: the

emperor’s attainment of power, his position gradually becoming more
and more endangered, and, finally, his losing power (and life) to his
successor. Thus, on the whole, a coherent picture is created of the
precarious conditions of imperial rule at the time, and patterns are
shown of how power was won and (often rather fast) lost.
The obituaries that regularly close the report of the death of an
emperor are not very prominent: they are short and rather formulaic
(only one of them exceeding one sentence). And whereas the majority
of them indeed contain an explicit comment on the qualities and flaws
of the ruler (thereby repeating in an extremely abridged, and therefore
often unsatisfactory way, what has been narrated before),11 others sim-
ply state that this was how the emperor died after having ruled for a
certain number of years.12 Again, judgments of the particular reigns are
given implicitly by comparisons, contrasts, and parallels, rather than
by explicit narratorial statements. Of particular importance here is the
short sketch of Marcus Aurelius who is introduced as a model emperor
at the outset of the work (1.2–4). Thus a standard is provided enabling
the narratees to measure the quality of the subsequent reigns.13
As in Thucydides (→), the primary narrator only rarely refers to
alternative versions,14 and hardly ever mentions specific sources at all.15
This is in line with the general invisibility of the historians’ persona:
there are only few places where he critically examines the evidence.16
It is this method of presentation that creates the impression of an
‘objective’ narrative, suspended only at a few places by vague remarks
11 Hdn. 1.17.12; 2.12.7; 3.4.7; 3.7.8; 3.15.2–3; 5.4.12; 6.9.8; 7.9.10; 8.5.9; 8.8.8.
12 Hdn. 2.5.9; 4.13.8; 5.8.10.
13 Cf. Marasco 1998: 2840–2857; Sidebottom 1998: 2803–2812; and Zimmermann
1999c: 24–41.
14 The only passages are Hdn. 1.14.2; 3.2.3; 3.7.3; 4.8.4; 4.12.4; 6.6.1; 7.1.8; 7.9.4;
7.9.9; 8.3.8–9.
15 The only source explicitly referred to is Septimius Severus’ autobiography (2.9.4).
Marcus Aurelius’ speeches and writings (1.3.1) as well as histories of his reign (1.2.5)
are only mentioned, but not referred to as sources. Unspecified historians of the life
of Severus are mentioned in 2.15.6–7; 3.7.3; 3.7.6. Herodian’s main source for books
1–5 is now by most scholars supposed to be Cassius Dio’s History, though he is never
mentioned in the text; cf. Alföldy 1989 and Zimmermann 1999c.
16 Cf. Hdn. 2.9.4; 3.7.6; 7.1.8. Perhaps the most notable exception is 1.11.1: ‘Through
my research (historia) I have discovered why the Romans have an especial veneration for
this goddess …’ This is the only occurrence in Herodian of the word historiē with the
Herodotean sense of ‘research’.
such as ‘it is said’, which shift the responsibility for a particular report
to unspecified and anonymous sources (the ‘anonymous spokesmen’
device).17
On the whole, although intrusions by the narrator are restricted to
the passages of transition between single episodes and narrative units,
the narratorial voice in Herodian is slightly more intrusive and overt
than in Thucydides, but still much less present and intrusive than e.g.
in Herodotus, Arrian or Cassius Dio.
An omnipresent and omniscient narrator
The Herodianic narrator is omnipresent and at times also omniscient,

as he not only recounts intimate scenes, such as the final conspiracy
against Commodus (1.16–2.2), or events going on in the closed quarters
of the imperial palace, such as the end of Plautianus’ plot against Sep-
timius Severus and Caracalla (3.12), but also has access to the thoughts
and feelings of his characters, in the form of embedded focalization: he
reports the sorrows of the old Marcus Aurelius (1.3.1–5), knows what
impact the repeated conspiracies had on Commodus’ soul (1.8.7), or he
can even reproduce the words Marcia spoke to herself (1.17.5).18 Such
signs of narratorial omnipresence and omniscience—not uncommon in
ancient historiography since Herodotus (→)—have often led modern
scholars to the somewhat misleading labelling of Herodian’s History as a
(historical) novel.19
Although the narrator claims in the preface that he has written down
what he had seen and heard in his lifetime, he nowhere figures in the
events he recounts. However, he twice explicitly claims to have been
an eyewitness of events in the city of Rome: he includes himself among
the spectators of Commodus’ performances in the arena (1.15.4: ‘Wild
beasts were brought from all over the world for him to kill, species
which we had admired in pictures but saw for the first time on that
17 Such remarks quite often occur in digressions providing antiquarian or geograph-
ical information (1.11.1–5; 1.14.4; 1.16.1; 3.4.3; 5.6.4; 6.5.2), but can also be found at
places in the main narrative (2.1.6; 3.4.7; 3.7.4; 4.8.8; 6.6.9; 7.1.5; 7.1.7; 8.3.7). Cf. Whit-
taker 1969: LXIII.
18 Cf. also Hdn. 2.8.7; 3.11.1; 3.11.8; 3.14.1; 4.13.1; 5.4.7; 5.5.5; 5.7.1; 5.8.6; 6.1.5; 6.6.1;
7.1.2; 7.8.6; 7.9.11; 8.1.5.

19 Cf. Alföldy 1989. This view is rejected by Hidber 1999 and Zimmermann 1999c:
7–13.
occasion’) and of the Secular Games given by Septimius Severus (3.8.10:

‘In his reign we saw all kinds of different shows in all the theatres at the
same time … these were the ceremonies which were called at the time
the Secular Games …’).20 He also reports that ‘we saw some ludicrous
pictures’ of Caracalla portrayed as a kind of Janus made of Alexander
and himself (4.8.2). A few other passages might additionally suggest to
the narratees that the narrator could have been present at the events or
might have visited the places he describes, e.g.:
Even when one sees the ruins of the wall [sc. of Byzantium] as they are
today, one has to admire the skill of the first builders and the power of
those who later destroyed it. (3.1.7)
It is not possible to describe the scenes of celebration that day after the
arrival of the messengers … (8.6.7)21
Thus, is created the picture of a widely travelled Greek and civil servant
who has acquired first-hand knowledge of Rome, and of other parts of
the empire during his professional career—a knowledge he now readily
shares with his narratees.
Narratees
The narratees are never directly addressed in the second person

throughout the work. But in the proem two groups of narratees are
mentioned: the narrator’s contemporaries, in whose ‘recent memory’
the time described in the History falls, and a future audience which ‘will
derive some pleasure from the knowledge of events which are impor-
tant and compressed within a brief span of time’ (1.1.3). The first group
of narratees guarantees, as it were, the truthfulness of the account for
the second group. In fact, there is one passage in which the contem-
porary narratees are indirectly called to witness, ‘But throughout the
many years of Alexander’s rule no one could recall the name of a per-
son executed without trial’ (6.1.7).
Traces of the primary narratees and of interaction between the pri-
mary narrator and them are most clearly perceptible in the digressions.
20 Most scholars now seriously doubt that Herodian was actually present at Com-
modus’ performances in AD 192. The whole passage seems to be a transformation of

Cassius Dio’s account; cf. Zimmermann 1999c: 285.
21 Cf. also Hdn. 3.4.2; 5.3.5; 5.6.7–10; 7.7.1; 7.12.1.
There, the narrator mainly explains Roman customs and religious

institutions but also provides geographic and ethnographic descriptions
of other parts of the world to a Greek audience supposedly not familiar
with them, e.g.:
Through my research I have discovered why the Romans have an espe-
cial veneration for this goddess, and, in view of the lack of knowledge
about this among some Greeks, I have decided to record it at this point.
(1.11.1)
This story about the goddess of Pessinous … contains quite interesting
information for people who are unfamiliar with Roman history.
(1.11.5)22
As in most ancient historiographical narratives, the presence of the nar-
ratees is also perceptible in passages featuring the ‘anonymous inter-
locutor’ device (e.g. 5.3.7: ‘Given this combination of good looks, youth
and splendid dress, one could compare the young man [sc. Elagabalus]
to the magnificent statues of Dionysus’), as well as in many gar-clauses
which seem to anticipate possible questions of the narratees. In Hero-
dian, such explanations often refer to ‘general truths’ such as ethno-
graphic stereotypes or the allegedly typical behaviour of the members
of a particular social stratum, and frequently adopt the form of senten-
tious gnomai:
The product of the age of Marcus was a large number of scholars, since
(gar) subjects always model their lives on the ideals of their ruler.
(1.2.4)
Like a sensible person—he [sc. Saturninus] was a Syrian, and Easterners
are fairly sharp-witted—, he saw the madness which had taken hold of
his commander [sc. Plautianus] … (3.11.8)23
Other, though much less frequent features of the interaction between
the narrator and the narratees include a few rhetorical questions (7.3.1:
‘What is the point in destroying barbarians, if even more people are
being murdered actually in Rome and the subject nations?’),24 and if
22 Roman institutions or customs are also explained at 1.9.2; 1.10.5; 1.14.2–6; 1.15.9;
1.16.1–3.; 3.8.10; 4.2.1–11; 5.4.8; 5.5.7; 5.6.2–3; 6.7.4; 7.5.8; 7.7.1; 7.10.2; 7.12.7; 8.8.5.
Similar explanations of practices or peculiarities from other parts of the empire can be
found e.g. in 3.2.7–8; 5.3.5–7; 5.6.4; 6.5.3; 6.7.1; 6.7.6–7; 7.2.3–6.
23 Cf. also Hdn. 1.3.1; 1.3.5; 1.6.9; 2.7.9; 3.1.2; 3.2.8; 3.9.3; 3.11.9; 3.14.6; 4.11.3; 4.11.9;
5.8.4; 6.5.3; 7.3.5; 7.7.1.

24 Another example can be found in Hdn. 2.13.11.
not-situations (7.3.1: ‘His [sc. Maximinus’] achievement would have won

him reputation, if he had not proved so oppressive and fearsome to his
own people and subjects’).25
The picture drawn of the narratees is not very coherent. Basically,
they are given the character of Greeks not very familiar with Rome
and Roman institutions. But whereas in two passages they are supposed
not to know any Latin (5.4.8; 7.5.8), another one clearly presupposes
some knowledge of the Roman language (1.12.2). The narratees are
also expected to be familiar with at least the more important person-
alities from late Republican and Imperial Roman history such as Sulla,
Caesar, Pompey, Antony, Marius, Germanicus, and the former emper-
ors referred to as exempla and in similes.26 As for their social outlook,
the narratees are assumed to share the narrator’s elite stance of the
rich, which in one case is revealed by the use of the ‘we’ form (3.9.3:
‘After crossing Mesopotamia … Severus hurried on to Arabia Felix.
The name Felix comes from the fact that the country produces aro-
matic herbs, which we use for perfumes and incense’).27
Secondary narrators
Although character-text amounts to about 10 per cent of the History,

there are hardly any secondary narrators. In their speeches, characters
express hopes or warnings, and make promises and announcements,
but do not tell elaborate stories. Snippets of narrative are inserted
only occasionally. But when a character actually ‘refers to the past or
present, he or she is frequently lying, or sometimes mistaken’:28 Laetus
tells the Praetorians that Commodus has died of apoplexy (2.2.6), Septi-
mius Severus lies about Albinus’ intentions (3.6.1–5), or Maximinus lies
about the life of Gordian (7.8.7).29 In these instances the primary nar-
ratees usually know more than the secondary narratees, and, thus, can
easily see through the insincerity of the narratives.
25 Cf. also Hdn. 6.9.8.

26 A list of similes is given by Sidebottom 1998: 2804 n. 151.
27 Passages betraying an elite stance are collected by Sidebottom 1998: 2823.
28 Sidebottom 1998: 2817.
29 For more references cf. Sidebottom 1998: 2817.
Conclusion
In many respects Herodian’s primary narrator is the counterpart of his

immediate predecessor, the very overt, self-assured and class-conscious
narrator in Dio Cassius’ Roman History: whereas the Dionic narrator is
keen to reveal an extraordinary amount of details of his distinguished
identity, the Herodianic narrator remains very much in the shadow.
And while the former’s presence is very overt in his frequent interven-
tions, ‘Herodian’ explicitly (though rather vaguely) refers to his method-
ological principles only in two passages at the outset of his work (cited
above), then tells his story without stepping in very much. Thus, a
coherent analysis of the quick succession of reigns seems to be pre-
sented not from a specific, individual viewpoint (such as that of a sena-
tor in Dio), but rather from a very general and seemingly unrestricted
point of view.
part three
CHORAL LYRIC
chapter fifteen
PINDAR AND BACCHYLIDES
I.L. Pfeijffer
The victory ode is not a narrative genre. Unlike epic, historiography,

biography, the novel and most poetry from the Hellenistic era, the main
purpose of the victory ode is not to tell a story. A victory ode celebrates
a victory won by an athlete. Contrary to what we may expect on the
basis of our familiarity with modern sport journalism, the poet of a
victory ode does not tell the story how the victory was won: a victory
ode is no match report (even short descriptions of the contest do not
occur; Pi. O. 8.67–69; P. 8.81–87; B. 5.37–49; 9.27–39, and 10.21–30 are
the exceptions confirming the rule). Nor does he engage in giving a
detailed report of the celebrations occasioned by the victory. His main
interest is to record the factual data that define the victory celebrated
(the victor’s name, his father’s name, his place of origin, the venue of
the games, the athletic discipline and the athletic age category) and
to enhance the glory of the victor by associating his accomplishment
with eternal truths and values and with the exploits of heroes from the
(mythical) past. But, although the victory ode as such is not a narrative
genre, the majority of victory odes do include narrative sections. For in
order to put the victor on a par with heroes, the poet tells stories about
these heroes. Most victory odes include at least one longer or shorter
narrative account of great heroic deeds. Only nine of the 45 surviving
victory odes of Pindar lack such a narrative (O. 5; 11; 12; 14; P. 7; N. 2;
11; I. 2; 3); of Bacchylides’ eleven victory odes (excluding the ones that
are too fragmentary: 7, 8, 12, 14A, 14B) five lack a mythical narrative (2,
4, 6, 10, 14). Often, but not always, the mythical narratives occupy the
central portion of the ode. Some odes have more than one narrative
section. The theme of the narrative is often, but not always, connected
with the victor’s homeland or with the venue of the games. The length
of these narrative sections varies considerably, as does, for that matter,
the length of the odes themselves.
214 part three – chapter fifteen
Demarcating the narrative
Any victory ode is a hybrid text. Apart from narrative, it may include
a variety of different sections, such as direct praise of the victor, praise
of his homeland, clan and family, catalogues of earlier victories won by
the victor and members of his family, praise of his trainer, gnomic state-
ments, invocations and prayers. This raises the question how we can
distinguish narrative from the rest of the ode. Pindar and Bacchylides
do not really help us to answer this question. They very rarely pro-
vide explicit signals marking off the narrative from its surroundings. On
the contrary, their usual practice is to obscure the boundary between
the narrative and what precedes it. Often they use a relative pronoun,
camouflaging the narrative as a mere afterthought to what proceeded,
pretending that it has got out of hand, owing to their spontaneous
enthusiasm for the story. The end of the narrative is, as a rule, not
formally marked either. There are two typical ways of ending a narra-
tive: the poet either simply ends it and then plunges into a new section
without any formal sign warning the audience that he is doing so, or
(a procedure occurring especially in the odes of Pindar) he breaks off
his narrative abruptly, pretends that it is irrelevant, and urges himself
to deal with themes that are more closely relevant to the present vic-
tor.
A typical example of this procedure, presenting the narrative as a
mere afterthought and rounding it off with the pretence of its being
irrelevant, is the story of Heracles setting up the Pillars:
If Aristophanes’ son, being beautiful and doing
what fits his bodily form,
has reached the peak of masculinity, yet to travel further
is not easy over the untrodden sea beyond the Pillars of Heracles,
which the god-hero placed as renowned witnesses
of the limit of seafaring. He subdued the superior beasts in the sea,
and, by himself, searched for the currents
in the shallows, by which way he arrived at his goal,
which send him home,
and he made the earth known. My heart, to what foreign
cape do you lead astray my voyage?
I tell you to bring the Muse to Aeacus and his race.
The highest justice follows the precept ‘praise the noble’.
Desires for what belongs to others are not better for a man to bear.
Search at home. Contributory adornment you have,
to sing something sweet. (Pindar, Nemean 3.19–32)
i.l. pfeijffer – pindar and bacchylides 215
The victor’s ultimate achievement inspires a comparison with reach-

ing the boundaries of the navigable world, the Pillars of Heracles. This
reference is expanded with a relative clause (22–23),1 identifying Her-
acles as the hero who has erected these Pillars and mentioning their
symbolic significance. Then Pindar elaborates upon this relative clause,
briefly telling the story of how Heracles came to erect these Pillars (23–
26). In line 26 he breaks off this story abruptly by intruding into his
own narrative and telling himself that he has let things get out of hand:
he should not have told the story of the Theban hero Heracles sailing
to the cape of Gibraltar; he should rather tell a story from the mythical
arsenal of the victor’s homeland, Aegina, a story about Aeacus and the
Aeacids; and fortunately there is plenty of suitable material to be drawn
from this tradition. Although the narrative certainly has a thematic rel-
evance to the ode as a whole,2 the poet pretends that it is irrelevant and
that he has let himself be carried away by his enthusiasm for the story
rather than focusing on what he should have told instead.
It is now well understood that this presentation entails a fictional
mimesis of extempore speech; we are dealing with the ‘feigned orality’
motif. The poet pretends that his ode takes shape on the very spot, dur-
ing the performance, and that it is a spontaneous outpouring of pure
joy, occasioned by nothing but enthusiasm for the victor’s achievement,
while in fact it is a product that was commissioned and paid for by the
victor. Emphasis on the contractual relationship between the poet and
victor would detract from the praise, since any poet would say some-
thing nice about any victor for the right amount of drachmae. So the
poets of victory odes tend to hide this contractual relationship from
view, covering it up by means of pretended spontaneity. Thus the fic-
tion of spontaneity as such contributes to the aims of the genre.3 On the
level of this fiction, the narrative is not neatly introduced and rounded
off, because it was not planned.
But for our present purposes this fiction presents us with a problem.
For, in order to talk about narratives in the victory ode, one should be
able to identify these narratives. But it is not an easy task to isolate
narrative sections from an ode when the poets themselves tend to
integrate these narratives as closely as possible into the drift of their
1 For Pindar’s use of the relative and relative connection, see Pfeijffer 1999a: 37–41
with further references.

2 See Pfeijffer 1999a: 202–206, 224–228, with further references.
3 See Carey 1981: 5, and 1995; Miller 1993; Pfeijffer 1999a: 34–41.
rhetoric and blur the boundaries constantly. There are, however, some
rules of thumb that may help us. Since the narrative sections in the
victory odes nearly always present stories from the (mythical) past, the
narratives are marked by the use of the past tense, which distinguishes
them from most of their surroundings, dealing with the victor’s present
state of bliss and hopes for future continuation of his successes and
the spreading of his glory. Moreover, whereas the shift from praise to
story (and vice versa) is mostly blurred, the shift from the present to
the past (and vice versa) accompanying it is often indicated by means
of adverbs (most notably pote, ‘once’, e.g. O. 3.13; 6.12–13; P. 1.16;
4.10, 20; 8.39; 9.5, 15; 10.31, N. 4.25; 5.9; I. 1.13) or stategically placed
adjectives (e.g. palaios, ‘ancient’, or proteros, ‘early’, e.g. O. 3.10–13; N.
3.13, 32). Another signal helping to demarcate the narrative is the
frequent use of apostrophe accompanying the shift from the past to
the present, either calling upon the poet’s heart to steer the ode into
a direction closer to the present occasion or calling upon the victor’s
attention (after an allegedly irrelevant digression) for a section that
will be closer to his personal interests (e.g. P. 8.33; N. 3.26, 76; 5.48).
Where no apostrophe or temporal adverb is used to indicate the shift
from the past to the present, we sometimes find an emphatic first-
person statement of the poet, reflecting on his encomiastic duties in
a way that is comparable to his urging himself or his heart to steer
the ode in a different direction (e.g. O. 9.35–36; 13.94; P. 1.81; 3.61–62;
10.51; I. 5.51). Finally, narrative sections are sometimes concluded by
means of gnomic statements, reflecting on a prominent theme of the
story and linking it implicitly with the victor’s situation (e.g. N. 5.16–
18).
The primary narrator and narratee
Although the victory odes are performed by a chorus,4 there can be no

doubt whatsoever that the first-person singular or plural occurring in
the odes represents the poet, and more often than not the poet specif-
ically in his professional role.5 He is the primary narrator, being firmly
present in the words sung by the chorus. This primary narrator is overt
4 For the untenable hypothesis of solo-performance see Pfeijffer 1999a: 242–243, n.
8, with references.
5 See esp. Lefkowitz 1991.
and visible. Especially Pindar refers to himself and his narrating activity
in many of his odes, tells us about himself, and comments openly upon
his stories, frequently using emphatic first-person statements to do so
(as in N. 3.26–32). Bacchylides may at first sight seem a less overt and
less visible narrator than Pindar, because he is less fond of referring to
himself in the first person. However, Bacchylides too comments openly
upon his stories, as in the following passage:
‘… What was hateful before is now dear. To die is most sweet.’
Thus he [Croesus] spoke; and he commanded his softly walking servant
to light the pyre. The girls shouted
and raised their hands in the air
to their dear mother. For the death one sees coming
is the most bitter death for mortals.
But when the glowing power
of the horrible fire rushed through,
Zeus raised a black, hiding cloud
and extinguished the fair flame.
Nothing is unbelievable if it is is accomplished by the concerns
of gods. Then Apollo, born on Delos,
bore the old man off to the Hyperboreans and gave him a new home
together with his daughters with their slender feet,
because of his piety, since he had sent
the richest gifts of all men to holy Pytho. (Bacchylides 3.47–62)
This is the climax of Bacchylides’ dramatic narrative about King Croe-

sus, who built a pyre in front of his palace when his city was taken
by the Persians. But he was saved from the flames by Zeus, who sent
a storm to extinguish the flames, and Apollo, who bore him off to
live happily ever after among the Hyperboreans. Although the narra-
tor does not intrude into his own story in Pindar’s manner, by means
of first-person statements, he does comment openly upon the events
twice: first in 51–52, where he gives a general truth on his own account
in order to explain the agony of Croesus’ daughters; then in 57–58,
where he includes a gnomic statement, again on his own account, in
order to add credibility to the unexpected turn his story has just taken.
The former comment is formally set off from its narrative context by
means of the particle gar (51, here rendered as ‘for’), indicating that the
sentence it introduces will clarify a question that may have risen on the
part of the narratees concerning the preceding event. The shift back to
the narrative proper is signalled by the particle alla (53, here rendered
as ‘but’), which signals that the narrator breaks off his preceding argu-
ment and makes a fresh start. After the second authoritative comment
on behalf of the narrator the shift back to the narrative proper is even
more explicitly marked by means of the temporal adverb tote, ‘then’
(58).
Since the narrator tends to anchor himself in the present occasion
of the celebration, often referring explicitly to his professional role and
encomiastic tasks, whereas the stories he tells belong to the realm of
the mythical past, the primary narrator is external, to the extent that
he does not himself belong to the world of the stories he tells. Even
in the rare cases that Pindar or Bacchylides tell about the victory itself
(see above), they do not refer to themselves as spectators or as being in
any other way present at the venue where the victory was won, so as
to remain external even in those cases. The closest thing we find in the
victory odes to an internal primary narrator, is a remarkable passage in
Pindar, where the poet tells about an epiphany:
Rejoicing also myself,
I throw crowns at Alcmaeon; I sprinkle him with song too,
because, as my neighbour and guardian of my possessions,
he encountered me going to the songful navel of the earth,
and he grasped hold of prophecies with his inborn arts.
(Pindar, Pythian 8.56–60)
This passage concludes a mythical narrative about the hero Alcmaeon,
put in the mouth of his father, Amphiaraus. The Pindaric narrator
presents himself as agreeing with the favourable story Amphiaraus has
just told about his son, putting himself on a par with his secondary
narrator (see below), and he motivates his enthusiasm for Alcmaeon by
referring to the fact that he saw him as a vision in a dream or as a
waking epiphany, possibly when he was passing his shrine, while he was
travelling from Thebes to Delphi. On that occasion Alcmaeon revealed
a prophecy, the content of which is not given.
The paradoxical circumstance of performance, featuring a very pres-
ent first-person narrator who presents himself through the mouths of
a chorus which is being referred to in the third person, is mirrored
in the situation of the primary narratee. For the choral performance
itself addresses an audience of the victor’s fellow-citizens partaking in
the public celebrations of his victory. Moreover, the poets sometimes
anticipate more or less official reruns of the first performance elsewhere
in the Greek world. The victory ode derives its success from its impact
on these audiences. The victor pays the poet in order to make his
glory known to his fellow-citizens and fellow-Greeks in the present
and the future. Yet, these audiences are not the addressees who are
present in the text. When the poets of victory odes use a second
person in order to address an audience for their songs, this second
person nearly always refers to an individual, most often to the victor
himself. So the communicative situation for the performance of the
victory ode is an allegedly intimate one-way conversation from poet to
victor, taking place publicly in the form of a choral song addressing an
audience of the victor’s fellow-citizens and future audiences in different
parts of Greece. To put this in narratological terms: in spite of the
public character of the choral performance, we find one singular overt,
external, and very visible primary narrator (the poet), addressing one
singular primary narratee, who is equally overt, external, and visible
(the victor).
Outside the narrative sections, the relationship between the primary
narrator and the primary narratee, i.e. between poet and victor, is often
thematized in a way that underscores the fiction of intimacy and spon-
taneity in order to hide the contractual relationship from view. Espe-
cially Pindar places a great deal of emphasis on his personal bond with
his patron. An example is Isthmian 8.15b–18, where Pindar emphasises
his own personal link with Aegina by referring to the mythical relation-
ship between Aegina and Thebes. Similarly, Pindar may emphasize his
commitment to the mythical heroes of the victor’s city, as in Isthmian
5.20 and 6.19 ff. He insists on a relation of guest-friendship (xenia) be-
tween himself and his patron on many occasions (cf., e.g. O. 1.11ff.; P.
3.69; 10.64 ff.; N. 1.19 ff.; 7.61; I. 2.48). Occasionally he goes further and
represents his relationship with the victor as one of philia (cf. P. 1.92;
4.1; N. 3.76) or even as an actual family tie (cf. I. 1.1ff. and I. 7.37ff.,
for Theban victors). As has been demonstrated recently by Carey,6 the
vigour of Pindar’s personal commitment is especially clear when it is
contrasted with the way in which Bacchylides addresses similar top-
ics. In contrast with the ‘brief and colourless reference to xenia’ in B.
5.11, the lines describing Pindar’s personal ties with the victor in Pi. P.
10.64–66 are charged with emotion, emphasizing ‘trust, kindness, zeal,
friendship and reciprocity’.7
6 Carey 1995: 94–95.

7 Carey 1995: 94–95, who further discusses Pi. N. 7.61ff. and P. 4.299 as similar
charged and poignant references to the poet’s personal bonds with the victor.
The status of the primary narrative
Since the vast majority of stories found in the victory odes relate
exploits of the great heroes of the mythical past, the status of these
stories is traditional: the narrator retells a story known to his audience,
or, more precisely, the vast majority of stories found in the victory
odes are based on fabulae that are familiar to the audience. It need
not surprise us that purely fictional stories are not favoured by the
poetics of this genre. After all, the poet is the ambassador of truth, or
at least needs to present himself as such (as Pindar and Bacchylides
explicitly do on many occasions). For the task of revealing to their
audiences and to the world how great a man this victor is and how
astonishing his accomplishments becomes utterly hopeless as soon as
there is the shred of doubt about their sincerity. When the poets want
to associate the victor’s qualities with great heroes from the past, they
evidently prefer to tell stories about these heroes that are comfortably
recognizable to the audience rather than made up accounts. Fiction is
counter-productive to the aims of the genre.
Because the poets of victory odes can rely on their audience’s knowl-
edge about the facts of the fabula, they can allow themselves to be
fascinatingly economical in their way of forging the fabula into a story.
Often they rely heavily on the audience’s knowledge. Take, e.g. the way
in which Pindar tells his Aeginetan audience about Peleus and Telamon
murdering their half-brother Phocus:
Of old they prayed that it (Aegina) would be noble in men
and renowned in ships,
when they stood by the altar of the father (Zeus) Hellenius
and together spread out their hands to the sky,
the famous sons of Endaïs and Phocus, the mighty prince,
child of a goddess, whom Psamatheia bore
where the sea-waves are breaking.
Shame prevents me from saying something big,
and not hazarded according to Right:
how they left the famous island,
and what deity
drove the valiant men from Oenona.
I shall halt. (Pindar, Nemean 5.9–16)
The Pindaric narrator can afford to tell a story about fratricide by not
telling it. Instead of telling that Peleus and Telamon murdered Phocus
and that they were banned from Aegina after which they went to Iolcus
and Salamis respectively, he breaks off his narrative in a way that is
in keeping with the fictional mimesis of extempore speech. He merely

alludes to the murder in ‘something big and not hazarded according to
Right’ (14) and to the fact that Peleus and Telamon had to leave the
island (15–16). It will be clear that this narrative manoeuvre depends
entirely on the audience’s familiarity with the events of the fabula.
Sometimes the poets explicitly refer to the traditional status of their
narrative by mentioning its source. When Bacchylides in his fifth ode
begins to tell the story about Heracles descending into the underworld,
he marks it as traditional by adding ‘as they say’ (legousin, B. 5.57).
Similarly Pindar ascribes some of his stories to ‘what they say’ (e.g. in P.
2.22, in the story about the wheel of Ixion; O. 9.49, in the story about
Deucalion and Pyrrha; O. 6.29, in the story about the nymph Pitana).
We are dealing with epinician variants of the ‘anonymous spokesmen’
device, which originates with the historians. Sometimes Pindar ascribes
his stories to proteroi, ‘men of old’, as he does at the end of his story
about the Aeacids in Nemean 6, where he says: ‘Men of old found in
such deeds / as those a highway of song, and I myself follow along,
making it my concern’ (53–54). In Nemean 7.20–23 and Isthmian 4.37–39
Pindar explicitly mentions Homer as the source of stories about Ajax.
In some passages, however, something strange is going on: Pindar
ascribes his story to an older source, but modern scholars have a very
difficult time in tracing his version of the story in other mythological
sources, let alone in sources preceding Pindar. In Olympian 7 he intro-
duces his story about the foundation of Rhodes as follows:
The ancient reports of men
tell that when Zeus and the gods were apportioning the earth,
Rhodos was not yet visible in the sea,
but the island was hidden in the salty depths. (Olympian 7.54–57)
In his commentary on this passage Verdenius remarks: ‘The attempts

made to reconstruct a Rhodou Ktisis which could have been used by
Pindar, have produced mere hypotheses. This part of the story was
probably invented by Pindar himself. … In order to mask the fact that
he partly invented it he applies the same trick as Plato, viz. emphasizing
its remote origin (cf. “having heard an old story from a man who
was not young,” Pl. Tim. 21a)’; similarly Young.8 If Verdenius and
Young are right, the reference to earlier authorities is a rhetorical
means to camouflage personal mythological inventions or adaptations.
8 Verdenius 1987: 74; Young 1968: 87.

A similar situation is found in Nemean 1, where Pindar introduces his

story about the baby Heracles killing snakes by saying that it is an old
story (lines 33–34). Here again Pindar—perhaps together with Phere-
cydes of Athens—is our earliest authority for this story. And here again
scholars have busied themselves in filling up what they believe is a
gap in the transmission of texts.9 However, the passage from Olympian
7 proves the possibility of Pindar pretending that he draws his material
from earlier authorities.10
In one passage Pindar explicitly tells us that he does not follow the
version of his predecessors:
It is proper for a man to speak well of the gods; for less is the blame.
Son of Tantalus, I shall tell your story contrary to my predecessors:
when your father invited the gods to his most orderly
feast and to his beloved Sipylus,
giving them a banquet in return for theirs,
then the Lord of the Splendid Trident seized you,
his mind overcome by desire. (Olympian 1.35–41)
The audience knows exactly what Pindar is talking about. They are
familiar with the story of Tantalus, who served his own son Pelops at
a banquet for the gods, who, upon discovering this, resurrected him
from the cauldron, replaced his shoulder (eaten by the absent-minded
Demeter) with ivory, and punished Tantalus severely in the underworld.
This is the old story, but it is not the story Pindar is going to tell.
According to him, this old story is untenable because it implies divine
cannibalism (52–55). He believes it was invented by one of Tantalus’
envious neighbours (46–51). His predecessors were so uncritical as to
accept it, but he will not. He will tell what really happened. Pelops
did have an ivory shoulder, but he was born with it (26–27). Tantalus’
banquet was very orderly. It was at that occasion that Poseidon fell in
love with him and abducted him to the Olympus (37–45). Tantalus was
later punished in Hades, but this was for a different reason: he had
stolen nectar and ambrosia from the gods (55–64). As a consequence,
Pelops was returned to earth (65–66).
Although Pindar invents an entirely new version of the Pelops story,
rejecting the fabula as the audience knows it and substituting it with a
new account differing from it in all major facts and events, it would be
9 See Braswell 1992: 57, with references.

10 For a similar interpretation of Pindar’s implicit reference to Homer in N. 3.52–53,
see Pfeijffer 1999a: 212–213, 349–352.
misleading to characterize his story as fiction. His claim is that his new
version reveals what really happened. The fabula as the audience knows
it is argued to be implausible and ascribed to an envious neighbour.
Pindar is not inventing a fictional story with familiar characters; he is
rather like a modern historian who argues for a different, and in his
eyes more plausible, reconstruction of an episode from the past. In this
respect Pindar’s criticism of the traditional myth about Pelops is not
very different from his judgment that Homer exaggerated Odysseus’
prowess at Ajax’ cost (N. 7.20–23).
Motivation of the primary narrative
Unlike characters in musicals who suddenly burst out in song and

dance for no apparent reason whatsoever, the primary narrators in vic-
tory odes are sometimes explicit about their motivations for including
a story about the distant past in an ode celebrating the victor’s recent
achievement. Sometimes they do so in general terms, as, when Pin-
dar introduces a narrative about the Aeacids by declaring that it is
his duty to praise them whenever he performs on their native island,
Aegina:
And as for you, Aeacids with your golden chariots,
I declare that I have the clearest duty,
when coming to this island, to sprinkle you with praises.
(Isthmian 6.19–20)
Another way of motivating the narrative is to mark it explicitly as
a story exemplifying a preceding statement, which is also the most
common way for secondary narrators in Homer (→) to motivate their
narratives, e.g.:
A god, a god one
should honour, since that is the best of blessings.
For also the king of the
horse-taming Lydians, …
Croesus, was once saved by Apollo
with his golden lyre. (Bacchylides 3.21–24, 28–29)
The Greek conjunction epei (23, here translated as ‘for’) makes explicit
that the story about Croesus it introduces is meant by the narra-
tor to be an example proving the validity of his claim that to hon-
our a god is always the best policy for a mortal. As we have seen
above, the story does in fact fulfil this purpose, for the end of the
story reveals that Croesus was rescued from his pyre by Apollo, as a
reward for his piety towards the deity. Similarly, Pindar introduces a
story about Heracles in Olympian 9: ‘Men become brave and wise as
divinity determines. For how else could Heracles have had the strength
to fight against Poseidon, Apollo and Hades?’ (O. 9.28–35, my para-
phrase).
Sometimes the primary narrator motivates his narrative by relating
it explicitly to the primary narratee. An interesting example is the
following introduction of the mythical narrative:
For in wrestling following the footsteps of your mother’s brothers,
you do not refute Theognetus at Olympia,
nor Clitomachus’ brave-limbed victory at the Isthmus.
Exalting the clan of the Meidylids you carry the word,
the one that once Oecles’ son stated in riddles when he saw
at seven-gated Thebes the sons holding their ground in battle,
when they had gone the second road
from Argos, the Epigonoi.
Thus he spoke as they were fighting:
‘By nature the noble determination is conspicuous
from the fathers in the sons. I see him clearly,
weilding the coiled snake on his blazing shield, Alcmaeon,
the first in the gates of Cadmus. …’ (Pindar, Pythian 8.35–47)
The victor is addressed in lines 35–38. By winning his victory he
has proved himself worthy of the reputation of his family-members
Theognetus and Clitomachus, who in their days had won significant
victories of their own. And by proving that he possesses the same talents
as his ancestors, he has confirmed the innate excellence of his clan and
thus exalted it. The idea of inherited quality implied by this praise of
the victor is then connected to a mythical narrative about Amphiaraus
and his son Alcmaeon. Amphiaraus fought bravely before Thebes, but
was killed and Thebes was not sacked. Ten years later his son went
along with a second expedition against Thebes, the expedition of the
Epigonoi. This second expedition was successful. Thebes was sacked and
Alcmaeon distinguished himself in battle. We hear about Alcmaeon’s
bravery, staged as an epiphany, in the words of his father Amphiaraus,
who witnesses the climactic moment of his son entering the gates and
comments upon the events. In line 38 the narrator makes explicit that
the story he is about to tell is directly relevant to the victor himself: he
can apply to himself what Oecles’ son Amphiaraus once said, when he
saw the Epigonoi, with his own son Alcmaeon amongst them, sacking the
city of Thebes. The contents of what Amphiaraus said at that occasion
are then immediately given. The connection between Alcmaeon and

the victor turns out to reside in the idea that noble determination is an
inherited quality.
Sometimes the Pindaric narrator motivates his narrative in practical
terms, in a way that ties in with his fictional mimesis of extempore
speech. The clearest example occurs in Pythian 4. After he has told
about Jason, beguiling Medea and with her help accomplishing the
task set by her father, Aeetes, of ploughing with fire-breathing bulls,
he tells that Aeetes reveals to Jason where the golden fleece is kept,
but that he does not expect him to retrieve it, because it is guarded
by a huge serpent. Then the primary narrator intrudes into his own
story and gives the following motivation for being surprisingly brief
about the very climax of the story: ‘But it is too far for me to travel
on the highway, because the hour is pressing and I know / a shortcut’
(P. 4.247–248).
It may not be unimportant to note here that the motivation for
including the narrative provided explicitly by the narrator never en-
tirely concurs with his real motives for telling this story. The thematic
relevance of a narrative included in a victory ode is manifold. The
explicit motivation covers at best one single point of relevance, leav-
ing all the rest to the interpretation of the audience. And the point of
relevance the poets are explicit about is often a minor point, as, e.g. in
those cases where the poet motivates his choice of subject for his story
by emphasizing the geographical link between the heroes he tells about
and the victor. The stories about the Aeacids following the geograph-
ical motivations in Nemean 3.28–32 and Isthmian 6.19–21 are in many
ways directly relevant to the respective victors and the poet’s respec-
tive thematic concerns. And all these points of relevance are in the
stories themselves dealt with more extensively than the geographical
link between the heroes and the victor. Even in the example discussed
from Pindar’s Pythian 8, the explicit motivation is to some extent mis-
leading. For Amphiaraus’ speech is not finished after he has made his
point about the importance of inherited quality, i.e. after the expec-
tation raised on the part of the audience by the explicit motivation
has been fulfilled. Amphiaraus continues for eight more lines (48–55),
prophesying that the leader of the second expedition, Adrastus, will be
the only one to lose his son, who will be the sole casualty. The narrator’s
explicit statements about the relevance of his story are misleading in so
far as lines 48–55, Amphiaraus’ prophecy about the outcome of the bat-
tle, are irrelevant to the explicitly avowed aim of the mythical example.
This part of Amphiaraus’ speech does not bear on the analogy between
the victor and Alcmaeon as exemplifications of the power of inherited
excellence. Thus the motivation the narrator is explicit about is not his
only motive for telling the story.11
There are, however, many examples of narratives that are not explic-
itly motivated. These include all those narratives that are, in keeping
with the fictional mimesis of extempore speech, presented as a mere
afterthought to the preceding section, e.g. by means of the relative pro-
noun. In these cases the relevance of the narrative to the ode as a whole
and the victor is left entirely to the interpretation of the audience. Here
Pindar and Bacchylides are like musical characters who suddenly burst
out in song for no apparent reason, only with one difference: the audi-
ence are taken by surprise, finding themselves all of a sudden in the
middle of a song without having realized when it began.
Secondary narrators and secondary narratees
Both Pindar and Bacchylides make ample use of the possibility of stag-
ing characters who tell a story. As a rule these secondary narrators both
belong to the story-world of the primary narrative and tell about this
same story-world. The typical situation is that the secondary narrator
either reveals what preceded the episode that constitutes the primary
narrative or prophesies its outcome, and the primary and secondary
narratives put together produce one continuous storyline. Since the
secondary narrators tend to tell episodes from the primary narrative
of which they are a part, and since the status of this primary narra-
tive is traditional, the secondary narratives are also traditional. We do
not often find secondary narrators who tell invented stories, the most
notable exception being the anonymous envious neighbour in Pindar’s
Olympian 1, who tells the traditional version of the Pelops myth, which is
rejected by the primary narrator as a lie (lines 47–51). The scope of the
victory ode calls for narrative economy. Whereas the epic poets may
use secondary narratives as a vehicle for introducing narrative material
that is alien to the main storyline, the poets of victory odes tend to use
secondary narrators as instruments to manipulate narrative time: the
11 See further Miller 1993: 31–34; Pfeijffer 1999a: 426–456.

use of secondary narrators providing flashbacks or prophecies allows

them to concentrate a long story in one dramatic moment.
Most secondary narratees are just as internal as the secondary nar-
rators. One character tells what will happen or what has happened to
another character. As a rule the story told is of direct personal rele-
vance to the secondary narratee. One good example among many is
Heracles telling Telamon that he will have a son, that his name will be
Ajax and that he will distinguish himself as a warrior (Pi. I. 6.52–54).
Also the elaborate network of secondary narratives in Pindar’s majes-
tic Pythian 4 follows this same pattern. The story of Pelias usurping the
kingship of Iolcus from Jason’s father, Jason’s secret upbringing by the
centaur Chiron, his return to Iolcus to reclaim his kingdom, Pelias’
demand that he should bring back the golden fleece, Jason’s expedi-
tion with the Argonauts in order to capture the golden fleece and the
subsequent colonization of Libya by Battus is distributed among the
primary narrator and various secondary narrators. The latest episode
of the entire fabula, the colonization of Libya, is told first in the form
of a prophecy by the secondary narrator Medea, who addresses the
Argonauts on Thera (13–56). The prophecy is of direct personal rele-
vance to at least one of them, Euphemus, who will be Battus’ ancestor.
The earliest part of the entire fabula is also told by a secondary narra-
tor. We hear about the unjust usurpation, Jason’s secret education and
his claim to the throne in Jason’s own words, when he first confronts
Pelias on his return to Iolcus (102–119). Jason’s second speech to Pelias
reveals their shared genealogy (138–155). The purpose of the expedition
of the Argonauts becomes clear in Pelias’ response to Jason (156–167).
The personal interest of the respective secondary narratees in the sto-
ries told needs no arguing. Thus the primary narrator focuses his own
story on the dramatic moment of Jason’s return to Iolcus and on the
expedition itself, revealing the previous history and the consequences of
the expedition by means of secondary narrators.
Some of the secondary narratees show strong emotional responses to
the secondary narratives. It is clear that these responses are meant to
steer our perception of the embedded narrative. Especially Bacchylides
uses this type of emotive guidance to great effect. One of the most
impressive examples is Heracles’ response to Meleager who had just
told him in the underworld about his own death:
That was the only time, they say,
that Amphitryon’s fearless son
wetted his eyes, out of pity
for the fate of this miserable man.

And in response to him
he said: ‘The best thing for a man is not to be born …’
(Bacchylides 5.155–160)
Not all the secondary narrators found in the victory odes are tied to
the primary narrative as closely as Jason, Pelias, and Medea in Pindar’s
Pythian 4. If, e.g. we reconsider Amphiaraus’ role in the story about the
two expeditions against Thebes in Pindar’s Pythian 8, he seems at first
sight a very typically Pindaric secondary narrator, both in his function
and in his status. He witnesses the climactic moment of his son entering
the gates of Thebes, tells about the disaster of the first expedition of
which he himself was a part, and prophesies the outcome of the current
expedition (44–55). He is an internal narrator both as a participant
of the former and a witness of the current expedition. However, his
presence on the battlefield is surprising (and in fact not accounted
for), because he died ten years earlier during the first expedition. The
primary narratees have no choice but to accept his presence and to
conclude that he apparently manifests himself as an epiphany. He
remains an outsider to the events that constitute the primary narrative,
does not address anyone explicitly, nor is there any sign that anyone in
the story is even able to hear what he is saying.
Amphiaraus is also atypical in another respect. For while he is con-
nected to the primary narrative very loosely, he is very tightly con-
nected to the primary narrator. After he has concluded his speech in
which he tells about the successes of his son Alcmaeon, the primary
narrator puts himself explicitly on a par with his secondary narrator
by saying: ‘Rejoicing also myself, I throw crowns at Alcmaeon; I sprin-
kle him with song too’ (56–57, discussed above). This is exceptional,
for Pindar and Bacchylides are as a rule much more implicit about the
relation between the primary and secondary narrator. In most passages
it is very difficult indeed to define such a relationship more precisely
than to observe that the secondary narrator relates part of the primary
narrative.
One of the most interesting narrative manoeuvres to be found in
the victory odes of both Pindar and Bacchylides is the blurring of the
boundary between primary and secondary narratives, to the effect that
the primary narrator merges with the secondary. The central narrative
of Pindar’s Nemean 5 is an illustrative example:
For those men [i.e. the Aeacids] also willingly
the loveliest choir of Muses
sang on Mount Pelion. In their midst

Apollo chased his seven-tongued phorminx with his golden quill
and led songs in every kind. First they sang, beginning with Zeus,
of holy Thetis
and Peleus, and that Creteus’ delicate daughter Hippolyta
wanted to shackle him
by a trick, by persuading the lord
who watched over the Magnesians, her husband,
with crafty plans to be her accomplice.
She rigged a lying, made-up tale,
that that man had had a go at their nuptual marriage in Acastus’
own bed. This was the other way around.
For she tried hard to persuade him and begged him
with all her heart. Her steep words stimulated his feelings.
Immediately he repulsed the bride, out of fear of the anger
of the Father, the god of Guests. He took good notice of it
and he approved from heaven, the cloud-maker Zeus,
King of Immortals, that soon
he would achieve that a sea-goddess, one of the Nereids
with golden spindle, would become his wife,
after having persuaded her brother-in-law Poseidon,
who often comes from Aegae to the famous Dorian Isthmus.
There cheerful companies welcome him, the god,
with the sound of pipes,
and they compete with the bold strength of their limbs.
Inborn destiny decides
over everything. You, Euthymenes,
have fallen twice in the arms of Victory in Aegina
and touched elaborate hymns. (Pindar, Nemean 5.22–42)
The story about Peleus and Hippolyta is one of the most skilfully
narrated pieces in Pindar. What is especially admirable is the way in
which the poet has integrated the story in the drift of his preceding
argument by breaking up the chronological order of the narrated events
and using the chronologically last event, the wedding of Peleus and
Thetis, as the ‘frame-story’ of the entire episode. The chorus of the
Muses sang for the Aeacids, represented by Peleus, on Mount Pelion,
on the occasion of the celebration of the wedding of Peleus and Thetis.
‘That’ (hōs) in line 26 introduces the story of Peleus and Hippolyta
as the subject matter of the Muses’ song. Thus the chorus of Muses
becomes the secondary narrator who tells the story of Peleus and
Hippolyta.
The chorus of Muses sings to the accompaniment of Apollo’s phor-
minx. The Muses begin with an address to Zeus; they then proceed
with a song of praise about Peleus and Thetis, the newly wedded cou-
ple. This ‘hymn’ includes the episode that Pindar quotes (26–36) about
Peleus and Hippolyta; this episode does in fact praise Peleus, for it illus-
trates Peleus’ obedience to the laws of Zeus Xeinios in reward for which
Zeus gave him Thetis. ‘First’ (prōtiston men) in line 25 characterizes the
song about Peleus and Thetis, including the episode about Peleus and
Hippolyta, as the first part of the Muses’ performance. A correspond-
ing ‘second’ or ‘next’ (in Greek epeita de or something similar), does
not follow. Pindar does not indicate where the Muses’ song ends. The
closing inverted commas are absent, so to speak. This conveys the sug-
gestion that the remainder of the ode, i.e. lines 41–54 (of which only
41–42 are quoted above), dealing with the victor and his family, is part
of the song of the Muses. By leading the audience to anticipate another
part of the Muses’ hymn, ‘first’ helps to convey the suggestion that the
song of the Muses does not end at line 41 with the shift from the heroic
past to the present occasion. The secondary narrator, i.e. the chorus of
Muses, merges with the primary narrator, i.e. Pindar, represented by
the Aeginetan chorus performing the ode. This strategy is in the first
place suggestive of the quality of Pindar’s ode—a specimen of proud
self-consciousness which is relevant in the light of the superiority of
poetry over other forms of victory immortalization as thematized in the
opening lines, and which contributes to the encomiastic aims of the ode
to the extent that the victor’s immortalization depends on the quality
of the poet’s product (cf. N. 4.6–8, etc.). But the trick of not closing off
the song of the Muses contributes also more directly to the encomiastic
aims of the ode. Pindar exploits the fact of performance. By creating
the illusion that the Muses are still singing when the victories are men-
tioned, the Aeginetan chorus performing the present ode merges with
the chorus of the Muses: it is singing for the victor and his family just as
the chorus of the Muses sang for Peleus. This in itself draws a compli-
mentary analogy between the Aeacids and the victor and his family,
and conveys the suggestion that their victories are a reward for their
moral behaviour just like Thetis was Peleus’ reward for his abiding by
the laws of Zeus.
A comparable strategy occurs in two passages in Bacchylides’ third
ode:
The (immense) crowd shouted.
Ah, thrice-fortunate the man,
who received from Zeus
the honour of ruling over the greatest number of Greeks
and knows how not to hide his towering wealth

under the black cloak of darkness. (Bacchylides 3.9–14)
and:
Lord Apollo
… spoke to Pheres’ son:
Who is mortal must exalt twin
thoughts: that only tomorrow you will see
the light of the sun
and that you will complete a life
of fifty more years of deep wealth.
Delight your heart with deeds dear to the gods. For that
is the highest of gains.
Who thinks understands what I am saying. The high
heavens are unblemished. Water of the sea
does not become foul. Joy is the gold.
It is not right for a man to shed his grey
age and retrieve again the flower of his
youth. The light of great accomplishments does not
disappear with the body of a mortal, no,
the Muse feeds it. Hieron, you have shown
the most beautiful flower of bliss
to the mortals. For whom has success
silence does not bring adornment.
With the beauty of truth
someone will sing the gift
from the sweet-tongued nightingale of Keos.
(Bacchylides 3.76–98 [fin.])
Carey says: ‘Only modern printing methods can distinguish whether

lines 10–12 are a statement by the poet or a cry by the crowd at
Delphi; it is difficult to see how the identity of the speaker could have
been brought out in performance. The result is an indeterminacy in
the authority for the statement. And for the listener who hears the
crowd speaking it is not clear where collective comment ends and
authorial comment recommences. In the same way, when Bacchylides
cites the advice of Apollo later in Ode 3 (78 ff.), it is not clear where
the quotation ends. Indeterminacy of speech-boundary is inevitable in
narrative performance, wherever beginnings and endings are not for-
mally marked. The point here is not the indeterminacy of the point of
transition but the specific effect. The poet’s views merge with those of
the quoted speakers, leaving unclear the point where external authority
ends and the poet’s authority begins.’12 As in the passage discussed from
Pindar’s Nemean 5, it cannot possibly be determined exactly where the
primary narrator takes over from his secondary narrators. It is not
clear and not intended to be clear whether ‘I’ in line 85 refers to
the secondary narrator Apollo or to the primary narrator Bacchylides.
Most editors print the closing inverted commas after line 84, but the
possibility of Apollo continuing his speech to the very end of the
ode cannot be excluded, especially since the tone of 85–92 is just as
oracular as 78–84, and the poet is referred to in the third person in
the final line. As a result, it is not clear whether the praise bestowed
on the victor, Hieron, in lines 92–98 is uttered by the poet or the
god. As in Pindar’s Nemean 5, this trick of merging the primary and
secondary narrators contributes to the encomiastic aims of the ode.
For the suggestion is conveyed that the praise of Hieron is sanctioned
by divine authority. This subtle narrative technique forms a fitting
conclusion to the discussion of the narrators in Pindar and Bacchylides.
12 Carey 1999: 20.

part four
DRAMA
chapter sixteen
AESCHYLUS
J. Barrett
Narrative in drama is always more than narrative. It is also, of course,

the (speech) act of a character on-stage, the utterance of one of the
dramatis personae. Some aspects of these narratives, therefore, are deter-
mined not by the workings of the narratives themselves, but rather by
their dramatic contexts. The task of identifying and demarcating narra-
tive in Aeschylean drama is often difficult due to the variety of speech-
forms on-stage and the relative fluidity with which they give way to one
another. Indeed, the boundaries between these speech-forms are often
blurred. While this situation may demand a degree of provisionality in
identifying narrative sections, it also suggests that Aeschylean narrative
should be seen as one aspect of an interest in the problems and pos-
sibilities of various forms of discourse, an interest that is evident in all
fifth-century tragic texts. Aeschylean drama, that is, attends to ques-
tions such as what narrative is, and what it can do, while deploying its
many kinds of narrative amid a wide range of other speech-types.
One boundary that is on occasion blurred in these texts is that
between dialogue and narrative. In fact this distinction disappears alto-
gether in dialogical narrative. A form of narrative made possible largely
by the dramatic medium itself and exploited in later Greek drama
to great effect, dialogical narrative appears relatively infrequently in
Aeschylus. Some Aeschylean narratives (which I will consider monolog-
ical) may consist of lengthy sections punctuated by dialogue between
narrator and narratee(s), such as occurs at Persians 249–514.1 Others,
however, are emphatically dialogical, such as we find at Supplices 291–
324, where the stichomythia between the choral leader and Pelasgus
produces a (brief) account of Io’s travails. The category of dialogical
narrative in Aeschylus includes explanatory narratives produced subse-
quent to the events reported by a narrator either external (Supp. 291–
1 The text referred to is that of West 1990.

236 part four – chapter sixteen
324; Pers. 703–738) or internal (Pr. 246–262), as well as a predictive nar-

rative produced by a narrator now external, now internal (Pr. 743–781).2
These dialogical narratives are characterized by fast-paced exchange
(often comprised of question and answer), and the pointed participation
of the narratee in the production of the narrative. This latter quality,
in fact, signals perhaps the most significant aspect of these narratives:
they are of great importance to the narratee whose persistent questions
underscore the urgency attached to hearing the story.
I divide the (monological) narratives in Aeschylus into three cate-
gories: those of choruses, individual characters, and messengers. Else-
where in tragedy, particularly in Euripides (→), the narrative prologue
may be seen as a discrete category, but the extant Aeschylean plays
contain only one such prologue (Eumenides) and I therefore group it with
other character-narratives. As a matter of convenience, I include several
passages that might be categorized as catalogues (such as Pers. 302–330
and 759–786) or descriptions (such as those of Septem 375–652).3
Chorus
In choral lyric modes of speech often shift seamlessly into one another.
The narrative section of the parodos in Supplices (1–18), for example, is
initiated by an utterance belonging to the realm of prayer and imme-
diately followed by a return to that realm. This brief ‘narrative’, then,
is framed by and forms part of a prayer. In practical terms, therefore,
identifying narrative in these choral lyrics must often entail some qual-
ification. It has even been said that tragic choral lyric is ‘an intertex-
tual field, alternately evoking and frustrating generic expectations’.4 It
is nonetheless possible to identify passages that are at least principally
narrative and to examine their workings as such.
All Aeschylean choral narratives are sung with none but the chorus
on-stage.5 The isolation of the chorus marks the integrity and continu-
2 Cf. A. 1198–1213; Ch. 514–539.

3 On catalogues in Aeschylus see Michelini 1982: 99–100.
4 Rehm 1996–1997: 47. Cf. Rosenmeyer 1982: 162 and the comment of Kaimio that
Aeschylean narratives are framed so that ‘the chorus cannot be said to be only the
narrator’ (1970: 86).
5 Here I follow Taplin who, however, places Danaus on-stage with the chorus from
the opening lines of Supplices (1977: 193–194). Perhaps the most controversial case is the
parodos of Agamemnon, where Taplin keeps Clytemnestra off-stage until 258.
j. barrett – aeschylus 237
ity of their song, as it endows the song—and the narratives contained

therein—with a degree of autonomy. This integrity, however, comprises
a richly heterogeneous collection of speech types, including a substan-
tial amount of narratorial commentary. The Persian elders’ account of
the troops’ departure for Greece (Pers. 12–139), for example, is punc-
tuated at 93–100 by a moralizing claim about the dangers of divine
deception, as it is at 116 and 122 when ‘the chorus imagine and pre-
emptively utter the kind of cry of lament [oa] which a defeat would
precipitate’.6 Much the same happens at Seven against Thebes 742–791,
where the chorus at 758–771 punctuate their narrative about Oedipus
with a metaphorical description of the present crisis (‘a sea of trouble
drives on a wave … which crashes around the stern of the city’) and an
expression of their fear.
With few exceptions Aeschylean choruses produce narratives that are
subsequent to the events recounted.7 The unusual narrative in the first
stasimon of Seven against Thebes (287–368) deserves notice in this regard.
With the attacking Argive army outside the city gates, the chorus of
Theban women initiate a prayer that turns into a narrative of events
both present and future. They speak of Thebes ‘cast into Hades … its
women young and old taken captive and dragged by their hair like a
horse’ (321–329). The sections of the narrative that concern the future
are not prophetic; they constitute rather an account of what the chorus
imagine and fear, as they themselves indicate (332).8
Kaimio has noted that in Aeschylean choral lyric, use of the first-
person singular with reference to the chorus as narrator marks a link
between different sections of the song.9 In this manner are marked the
beginnings of narrative sections at Seven against Thebes 742, Supplices 538,
and Agamemnon 104, as are the ends of such passages at Agamemnon 248
and 456–460, and Choephori 638. Such transitional moments typically
coincide with a shift in focus between the present and the past events
recounted. Narrative sections begin with a shift from present to past at
6 Hall 1996: ad 116–117.

7 The chorus of Agamemnon produce simultaneous narration at 67–103 and 427–455.
8 For the spectators, who function as primary narratees, the use of Homeric phras-
ing (see Hutchinson 1985: ad 287–368) points to the unusual status of this narrative: as
an account of imagined events, the chorus’ narrative proves to be replete with (tradi-
tional) elements familiar from accounts of the sack of Troy. In this regard one may com-
pare Choephori 585–652, which recounts the ‘myths’ of Althaea, Scylla, and the Lemnian
women. See Garvie 1986: ad 585–651 on the illustrative use of myth in Aeschylus.
9 Kaimio 1970: 86.
Persians 852–856 where the chorus turn to an account of Darius’ reign;

Seven against Thebes 742 and 775; and Supplices 538. Narrative sections
end with a marked shift from the past back to the present at Persians
140 and 904; Seven against Thebes 790; Agamemnon 248 and 456–460; and
Choephori 638.
These sections may also begin with an indication that the narrative
will explain what precedes. The particle gar performs this function at
Persians 12–15, for example, where the chorus explain their anxiety and
gloom: ‘For all the strength of Asia has gone … and no messenger
or horseman has approached the city of Persians.’10 The introduction
of choral narratives in this way is closely allied with one of their
main functions, that of contextualizing the action of the play (for the
spectators), and offering a framework for understanding it.
Finally, narrative sections may be metrically distinguished. At Persians
140 the chorus conclude the narrative and turn to the present moment,
marking this change with a shift from lyric metres back to anapaests.
At Agamemnon 104 a turn from anapaests to lyric dactyls marks the
beginning of a narrative section. This kind of metrical marking is far
from typical, but it does represent one technique available to the poet.11
As a consequence of the chorus’ visible presence on-stage, some
aspects of their identity are readily apparent. For example, the chorus
by definition constitute a group composed of anonymous individuals
who sing with a single voice. In all extant plays of Aeschylus, further-
more, the chorus members are socially marginal in some sense: women
(Th., Supp., Ch.), slaves (Ch.), foreigners (Supp., Ch.), and elderly men
(Pers., A.). And to a large degree their loyalties and interests are evi-
dent. Indeed, interspersed with their narratives are expressions of fear,
prayer, lament, moral judgment, etc., that serve in part to characterize
them.
When the chorus function as (secondary) internal narrators, they
can only know what they have seen or heard: they are subject to the
restrictions of access and place, as are the choruses of Sophocles (→)
and Euripides (→). While Aeschylean choruses often recount events
from their own history (e.g. Pers. 12–139), they only rarely speak of
themselves as agents, and never explicitly mark their narratives as the
product of their own eyewitness status. (In the rare narratives that
10 Cf. Seven against Thebes 321, 742, and 772.

11 Where narrative beginning or end coincides with that of the song that contains it,
it will likewise coincide with metrical shift.
position the chorus as a prominent agent—e.g. Septem 321–36812—the

chorus are by definition witness to the events recounted.) There are,
however, implicit claims to such status, as when the chorus at Agamemnon
248 say that they did not see what followed the preparations for the
sacrifice of Iphigeneia, implying that they did witness what they report
up to this point. The chorus of Persians leave little doubt in 1–11 that
they witnessed what they narrate in 12–139. This general absence of
explicit claims to eyewitness status suggests that these narratives are
offered as the products of knowledge shared by those who do and might
constitute such a chorus (and often also by the theatre audience as well).
As the voices of the chorus singing in unison and the anonymity of the
chorus members imply, these narratives present themselves as what any
member of such a group would, and does, know.13
When the chorus function as external narrators, however, recount-
ing ‘ancient’ events that they could not have witnessed, they become
omniscient narrators, as in Sophocles (→) and Euripides (→). As such
‘ancient’ stories contain much more than any individual(s) could know,
they provide the chorus with a potent narrative device for recounting
events remote in time and of privileged, even paradigmatic, status. The
chorus at Seven against Thebes 742–790, for example, recount the beget-
ting of Oedipus, his self-blinding, and his curse of his sons. Laius, they
report, was ‘overpowered by the foolishness of love’ (750), and they
speak of Oedipus’ state of mind with some precision (artiphrōn, meleos,
algei, 778–780).14 Narratives of this type freely omit much of the fabula
and thus rely on the narratees’ familiarity with the events recounted. It
is, in fact, the status of such fabulae as old and well known that enables
these choruses to narrate with such authority.
Even when they are internal narrators, these choruses employ several
strategies for bypassing their restrictions. They may make use of ex
eventu knowledge, as at Agamemnon 122–123 where the chorus relate that
Calchas turned his gaze to the Atreidae and ‘knew’ that they were the
hare-eating eagles. The same chorus report at 220–221 that from the
moment he decided to sacrifice Iphigeneia, Agamemnon had a change
of mind that embraced extreme acts of daring (to pantotolmon phronein
metegnō). Narratives of events imagined or feared may also enable the
12 Cf. Supp. 1–18.

13 Kaimio 1970: 240–241; Stehle speaks of (non-dramatic) choral song as ‘community
poetry’ (1997: 18 and 26–70, esp. 69); Gould 2001: 326–327.
14 See also Supp. 538–599. Cf. A. 750–756.
chorus to bypass some restrictions. At Seven against Thebes 345–368, for

example, the chorus provide a detailed account of their own city’s
(imagined) fall. In so doing, they succinctly define the nature of the
attacking army as they peer into the mind of an (imagined) attacker
who ‘wants to have a partner’ in his quest for plunder (354). Such
strategies may well serve the narrator’s interests: at Agamemnon 220–
221 the chorus incorporate into the account their own view of what
transpired as a transgression committed by Agamemnon; and in the
Seven against Thebes passage the chorus not only offer a vivid picture
of their own fear, but they also manage to portray the attacking army
as a virtual kinship group characterized by greed and a devotion to
plunder.15
On occasion, however, the chorus may simply be unhampered by
any restrictions. The chorus of Persians, for example, describe at 28
not only the outward appearance of the Persian commanders, but
also their inward resolve (psukhēs eutlēmoni doxēi). Similarly, the chorus of
Agamemnon, recounting how Iphigeneia was ‘bridled’ to keep her silent,
go on to explain that she ‘wanted to address’ those about to kill her
(242–243) ‘because she had often sung for them’ (243–245). While the
chorus might well have deduced from outward signs that Iphigeneia
wanted to speak, the explanation of why she wanted to do so would not
have been similarly observable. Moments such as these serve in part
to endow the narrative with a reach and amplitude that augment its
ability to inform about and to shape (our knowledge of) the events
recounted. Such freedom from restriction, that is, serves the poet’s
twofold interests of representing the events more fully and defining
the chorus themselves by putting on display the scope and quality of
their concerns: if we glean from the narrative something of Iphigeneia’s
perspective, we also understand the chorus to be attentive to her as an
agent endowed with such subjectivity.
Aeschylean choruses are also self-conscious narrators. At Choephori
638, for example, the chorus reflect on the interpretive task performed
by their narrative:16 having recounted the mythic examples of Althaea,
Scylla, and the Lemnian women, the chorus conclude by asking ‘Which
15 See Goward on the ‘double game’ played by the Aeschylean chorus: with their
‘narrative authority … they provide the audience with background information, framing
the play, but at the same time they leave gaps’ (1999: 48).
16 Metanarrative commentary appears also at A.104; Th. 356; and Supp. 580, 590–
591. Cf. proseikasai at A. 163; exikhneusai at A. 368; and pou at A. 711. Cf. also Septem
742 (the chorus claim to tell an ‘ancient’ tale); Supp. 538 (the chorus say that they are
of these [stories] do I not justly collect?’ Thus do they assert the

interpretive propriety of comparing Clytemnestra to these other mythic
women.
The isolation of the chorus on-stage during narratives compels us to
consider the role of the spectators as narratees. While clearly full mem-
bers of the world on-stage, the chorus perform a mediating function
between this realm and that of the theatre audience,17 and although
they do not directly address the theatre audience, it is nonetheless
instructive to consider the degree to which the audience’s function as
primary narratees affects our understanding of the narrative.
When the chorus of Persians at the play’s opening, for example,
refer to themselves as ‘the Persian Faithful’, they offer a self-description
consistent with common Greek belief.18 The (original) theatre audience
will find that the chorus think of themselves just as they expect them to
do.19 Similarly, at 93–94 when the chorus ask rhetorically, ‘What mortal
man can avoid the god’s tricky deceptiveness?’, the spectator may
recognize a familiar theological view.20 Although we must hesitate to
attribute an awareness of this dynamic to the chorus themselves, their
mediating function—and the audience’s role as primary narratees—
open the way to perceiving such irony. Dramatic irony of all forms, of
course, is produced in this and similar ways.
In fact there are more signs of narratees in choral narrative, narra-
tees with whom the spectators may identfy. At Supplices 580–589 the
chorus address the narratees’ scepticism, when they insist that their
story of Io is true. Although the chorus here repeat the claim of ‘the
entire earth’ (583) that Io’s child is truly (alēthōs, 585) the son of Zeus,
this repetition posits sceptical narratees—much as does the original
statement of ‘the entire earth’. What might have made sense as part
of their protestations to Pelasgus that they are Argive, here finds itself
directed at the narratee as an agent armed with some scepticism.
Aeschylean choruses may also call upon knowledge or values shared
with their narratees. They may, for example, identify certain stories as
following an old path in recounting Io’s travails) and 580 (they mark their narrative as
a true logos).
17 On this mediating function see Käppel 1998: 72–75; Calame 1994/95: 140–141;
Nagy 1994/95: 49–51; Henrichs 1994/95: 66–70; Baur 1997: 44–46. Cf. Kaimio 1970:
82 and Pfister 1988: 76–79.
18 Belloni 1994: ad 2–4.
19 Hall 1996: ad 1–2.
20 Broadhead 1960: ad 107–114. Cf. 865–866 with the comments of Hall 1996.
familiar even to the narratees. Thus do the chorus of Choephori refer

to the story of the Lemnian women at 631 with the definite article (to
Lēmnion), suggesting that the story, as well as its import, is well known
to all. Similes provide the chorus with another method of alluding to
shared knowledge and values.21 At Agamemnon 231–234 the chorus sing
that in preparing for Iphigeneia’s sacrifice Agamemnon ordered his ser-
vants to lift her above the altar ‘like a young goat’. Here the narratees’
presumed familiarity with sacrificial procedure is invoked, as is a pre-
sumed understanding of the gravity of Agamemnon’s transgression as
revealed by the comparison.22
The chorus may pose questions that simultaneously serve rhetorical
purposes and also posit the narratees as (silent) interlocutor. At Agamem-
non 681–688, for example, the chorus ask who could have named Helen
so appropriately as they underscore the verbal root for destruction (hel-)
embedded in her name. At Choephori 638 the chorus justify their narra-
tive at the same time as they place the narratees centrally in the nar-
rative (‘Which of these [stories] do I not justly collect?’).23 The chorus
may also pose questions that the narratees might ask: having mentioned
Io’s transformation and suffering, the chorus of Supplices ask at 571–573,
‘Who then charmed suffering, wandering Io, driven on by the gadfly?’
The chorus here go on to answer this question, quite as though it has
been asked by the narratee.
More frequently the chorus anticipate the narratees’ questions, often
marking this anticipation with the connective, explanatory particle gar.
Such a gar-clause introduces at Persians 102 the explanation of why the
Persian army cannot be resisted, and at 126 the explanation of the fear
the chorus have just expressed about the fate of their city. As mentioned
above, a gar-clause may initiate the narrative as a whole, as at Seven
against Thebes 321 where it explains the preceding prayer to the city’s
gods.24
The chorus may also use the presentation-through-negation device
as a way of incorporating the narratees into the narrative. When the
chorus of Persians say at 14–15 that no messenger has arrived to tell
them the fate of the expedition to Greece, they implicitly acknowledge
21 De Jong 1987: 93–94. Similes occur at: Pers. 128–129; Th. 758–761; A. 50–54, 231–
234, 242–243, and 717–738.

22 Fraenkel 1950: ad 232.
23 Similar questions occur at Pers. 93–96; Th. 772–777; Supp. 590–591; Ch. 594–598.
24 Such clauses appear also at Pers. 91, 97; A. 222, 469; Ch. 637.
that the narratee might well expect a messenger to have arrived. Sim-
ilarly, at 864–867 the same chorus recall how many cities Darius con-
quered ‘without crossing the Halys and without leaving home’, either
contradicting or confirming the narratee’s presumed expectation.25 At
Agamemnon 186 the chorus report that Agamemnon, having heard the
seer Calchas’ explanation that Artemis demanded another sacrifice, did
not blame the prophet. Again, the narratees are imagined as expecting
otherwise.26
The use of negation, the practices of posing and anticipating ques-
tions, and the other signs of attention to the narratees, all work to con-
struct a narratee interested and involved in, as well as essential to, the
act of narration itself.
Characters
Character-narratives speak of events in the past (e.g. Pers. 176–214,

A. 1372–1398, Pr. 640–686), the present (Pers. 803–806, A. 320–336),
or the future (e.g. Pers. 813–820, Pr. 700–741, 786–815). We find both
continuous narratives (e.g. Pers. 176–214, Supp. 605–624, A. 1372–1398)
and those punctuated by dialogue (e.g. Pers. 759–822, A. 281–350, Pr.
436–506). Unlike choral narratives, those of characters nearly always
form part of a ‘dialogue’ in that they are addressed to one or more
interlocutors.27 As such, these character-narratives most often respond
to a request (e.g. Supp. 605–624, A. 281–350, Pr. 197–241), which in some
cases is implicit (e.g. Pers. 759–822). When no such request occurs, these
narratives often serve an explanatory function (e.g. Pers. 176–214, A.
1577–1611, Pr. 340–376). Some cases, however, appear to be gratuitous
and serve a persuasive function (e.g. A. 855–894, Pr. 436–506).
25 There is, of course, an implicit contrast with Xerxes’ recent disaster seen as the
consequence, in part, of his crossing the ‘natural’ boundary between Europe and Asia.
Here again, considering the audience as narratee reveals a larger, added meaning:
the spectators will have shared the (Greek) view that the river Halys ‘was the western
boundary of Persia proper’ (Broadhead 1960: ad 865–866). Hall judiciously notes that it
is not clear why the chorus should share such a view and that ‘a differentiation between
Persian and Greek ideals of leadership is almost certainly intended’ (1996: ad 862–866).
26 As Agamemnon’s rebuke of Chryses in Il. 1 might lead audience members to do.
Similar uses of negation occur at Th. 768; A. 228–230, 387, and 396. One may compare
the use of alpha-privative, as at Supplices 561 (athikton) and 580 (apseudei logōi).
27 Of the passages discussed here, only the prologue of Eumenides is spoken with no
others on-stage.
The category of persuasive narratives points to a purpose shared

by all character-narratives: even when responding to a request, these
narratives also serve a rhetorical purpose in their dramatic contexts.
Although it is clear, for example, that Clytemnestra’s account of her
suffering (A. 855–894) serves her own deceptive purposes, it is equally
true that her narrative of the beacons (281–316)—which responds to a
request from the chorus—forms part of her attempt to establish her
own discursive (and political) authority.
These narratives (or narrative sections) are frequently marked as
such at beginning and/or end. At Agamemnon 855–860, for example,
Clytemnestra begins by declaring that she will not hesitate to speak
openly.28 Narrative endings are often marked with a demonstrative
referring to what has just been said.29 Although many of these begin-
nings and endings are marked also by an emphatic temporal shift (Pr.
199, 471, 740), on occasion a temporal shift alone marks the transition
(Supp. 261; Ch. 579; Eu. 30). Both instances of character-narrative in Per-
sians (176–214 and 759–822) are metrically distinguished. The Queen’s
narrative in iambic trimeters is framed by dialogue in trochaic tetrame-
ters.30 The ghost’s narrative in iambic trimeters likewise initiates a shift
away from trochaic tetrameters.
A form of ring-composition not infrequently marks the beginning
and end of a narrative. The Queen in Persians marks her account with
references to the night (176, 200) and Clytemnestra marks hers with
references to Troy (A. 281, 316). Prometheus begins his account of
human suffering with reference to this suffering (442) and ends with
ironic reference to his own (471), as he frames his narrative of providing
tekhnai with references to them (477; 506).31
Characters may act as external or internal (secondary) narrators, or
as a combination of both. Among external narrators, some explain how
they come by their knowledge (the Pythia at Eu. 1–19), while others do
not (Aegisthus at A. 1583–1602). Internal narrators may report events
they have witnessed (Danaus at Supp. 605–624; Orestes at Ch. 269–296),
or events in which they played an active role (Clytemnestra at A. 1372–
28 Elsewhere: Pers. 180; A. 1183, 1373; Ch. 554; Pr. 197, 445, 476, 642, 703, 788, 825,
844.
29 Pers. 200, 786; A. 315, 348, 895, 1239, 1393; Ch. 297; Pr. 469, 816, 842, 873.
30 All character-narratives, and messenger-speeches, in Aeschylus are in iambic trim-
eter.
31 Further examples at Pers. 759–760 and 785–786; A. 1178–1179 and 1195, 1578–1581
and 1610–1611; Ch. 269–270 and 297.

1394; Prometheus at Pr. 197–241). Cassandra (A. 1256–1294) speaks now

as internal, now as external narrator in recounting both her own undo-
ing at Apollo’s hands and the future coming of Orestes as avenger.32
Internal narrators are subject to the restrictions of access and place;
thus, Orestes, reporting the oracle of Apollo, leaves some room for
doubt about the reliability of Apollo’s pronouncements (Ch. 297–298).33
Both Danaus and Orestes recount, largely in indirect statement,
what others have said. They report, however, proclamations about
themselves: Danaus tells of the vote in the Argive assembly and resul-
tant decree that he and his daughters are free to remain in Argos
and that they will be protected; Orestes reports Apollo’s oracle com-
manding him to avenge the murder of his father. The double status of
these narrators—that of eyewitness as well as subject of the reported
speech—shapes these narratives to some degree: although not playing
an active role in what they report, each of these narrators occupies a
central position in the narrative, as the presence of various first-person
forms attests. It is worth noting that Orestes’ narrative encompasses
declarations of Apollo concerning Orestes himself as well as general
statements about what happens to any mortal who fails to fulfil the
duty to avenge a murdered kin. This narrative also turns at 286 from
indirect to direct speech, although he claims throughout to be reporting
Apollo’s command.
Other eyewitness character-narrators also prove to be subject to the
restrictions of access and place, showing no signs of bypassing these
restrictions. Clytemnestra, for example, claims no privileged knowledge
in her deceptive account of her suffering (A. 855–894) and thus offers
a narrative that, though misleading, remains narratologically faithful to
her status. Indeed, this narrative emphasizes her ignorance as constitutive
of her suffering: she heard many rumours without knowing how to
evaluate them (863–876). In recounting her sufferings, Io holds Zeus
alone responsible, while ignoring Hera’s involvement (Pr. 640–686).
Even though Prometheus mentions Hera at 592, that is, Io cannot
place her in the story and thereby displays her limited knowledge as
narrator.34
32 Other external narrators: Darius (Pers. 800–838); Clytemnestra (A. 281–316); Cas-
sandra (A. 1215–1238); Apollo (Eu. 625–639); Athena (Eu. 681–710); Prometheus (Pr.,
700–741, 786–876). Other internal narrators: Clytemnestra (A. 855–894); Pythia (Eu.
34–59); Prometheus (Pr. 351–372, 436–506); Io (Pr. 640–682).
33 Garvie 1986: ad loc.
34 Griffith 1983a: ad 669–682.
Not all character-narrators are subject to such restrictions, however.

Clytemnestra’s sweeping beacon-speech (A. 281–316), which tells of a
chain of fire signals from Troy to Argos, describes not only the qual-
ity of the wood burned by the guards on distant Messapion (295),
but also the diligence and wakefulness of those on Makistos (290–291).
Clytemnestra also remarkably claims that the fire at one point burned
more than was ordered (301). However we read this line, Clytemnes-
tra as (external) narrator here claims a form of omniscience. But this
line also reveals the origin of her narrative: the ‘orders’ to which she
refers, like the ‘laws’ (nomoi) governing the torchbearers mentioned at
312, reveal that her narrative is imagined. The entire system of torches
was established ahead of time, of course, and she produces the nar-
rative on the basis of these arrangements, without acknowledgment.
Such a surreptitious claim to narrative control serves Clytemnestra’s
ulterior motive well: she not only (finally) convinces the chorus that
Troy has fallen (compare 264–277 and 317–319); she also claims priv-
ileged knowledge and a conspicuous form of (masculine) discursive
authority, as the choral leader acknowledges at 351: ‘Lady, like a sound-
minded man you speak wisely.’ This narrative control, then, serves as
an index of her power in the public realm and of her control of the
chorus.
Motivated omniscience appears in character-narratives elsewhere.
The prophetic narratives of the ghost of Darius in Persians, Cassandra in
Agamemnon, and Prometheus in Prometheus Bound fall into this category.35
Prometheus demonstrates his privileged knowledge in identifying Io
(589–592), as she immediately confirms. On two occasions he mentions
that his mother (Themis = Gaia) revealed the future to him (209–
218, 873–874). In spite of his access to some privileged knowledge,
Prometheus also draws upon his own experience in recounting the
Titanomachy. In doing so he relies upon ex eventu knowledge. When
he says, for example, at 207–208 that the Titans rejected his advice,
thinking in obstinate arrogance (karterois phronēmasin ōiont’) that they
would prevail through sheer force, he reports what he was able to
deduce from what they said and did; he does not claim access to
35 As do those of Apollo (Eu. 625–639) and Athena (Eu. 681–710). Darius is called
a ‘god’ at 157 and 643, and ‘divine’ at 651. Reconciling Darius’ prophetic ability with
his ignorance of Xerxes’ disaster, Broadhead remarks that Aeschylus ‘has attributed to
Darius such knowledge as suited his dramatic purpose’ (1960: ad 739ff.). Cf. Hall 1996:
ad 681–851, with references.
their thoughts. One may compare the narrative of Aegisthus at Agamem-

non 1583–1602. An infant at the time of the events he relates (1605–
1606), Aegisthus nonetheless describes in some detail the gruesome
meal offered to Thyestes by Atreus. Although he is presumably recount-
ing a well-known story or relying upon ex eventu knowledge, Aegisthus’
(implicit) claim to such detailed knowledge serves his purpose, even if
the purpose itself is largely thwarted: the account he provides supports
his claim that the killing of Agamemnon is just revenge for an earlier,
even more heinous, act of bloodshed.36
Aeschylean character-narratives contain many reflections on the nar-
rating activity. Here, however, it is again important to keep in mind
that these metanarrative comments—like the narratives themselves—
also perform rhetorical functions on-stage. At Agamemnon 855–860, for
example, Clytemnestra announces that she feels no shame to speak
of her love for her husband; she then says that she will speak of her
difficult life while Agamemnon was at Troy. While affirming her self-
assurance as narrator and underscoring the authority with which she
speaks (‘I shall speak of my own miserable life, not what I have learned
from others’), she also asserts her (masculine) agency: speaking for her-
self and of herself in public, Clytemnestra performs the bold appro-
priation of authority that characterizes her throughout the play. She is
able to fulfil these two very different functions, then, by means of these
metanarrative comments.
Clytemnestra at Agamemnon 315–316 concludes her beacon-speech by
saying that her account constitutes proof of her earlier claim that Troy
had fallen. This statement is in part a guide for the narratee, but it
is also a judgment about her narrative talents. The Pythia of Eumenides
reflects on her account at 45 when she says ‘I shall prove to have spoken
clearly’.37 The Pythia goes on, however, to qualify her description of
the Erinyes, saying that she ‘calls’ (legō, 48) them Gorgons but that
she cannot compare (eikasō, 49) them to images of Gorgons: she offers
eyewitness testimony but makes clear that she is at a loss to say what
she has seen, thus preparing the narratee for an alarming sight.38
36 An instructive analogue appears in the prologue of Eumenides where the Pythia
recounts the ancient history of the temple at Delphi (1–19). Unlike Aegisthus, she
reveals (at line 4) that she is repeating a story she has heard (hōs logos tis). See Som-
merstein 1989: ad loc. on the significance of the indefinite tis.
37 See Sommerstein 1989: ad loc. for this translation.
38 See also A. 1373 (Clytemnestra); A. 1584 (Aegisthus); A. 1183 and 1195 (Cassandra,
whose comments are perhaps the more remarkable in that she is possessed throughout:
Prometheus’ narratives likewise contain many metanarrative com-

ments. At 197–198, for example, he introduces his account of the Tita-
nomachy by saying that it is painful for him to tell the story, but painful
also to remain silent, thus underscoring his emotional involvement in
the (story of) events narrated.39 At 870–876 he employs recusatio, saying
that it would be too long a story to tell how he will be released, and
besides, Io would not benefit from hearing.40 Not long before this, how-
ever, Prometheus tells Io that he will answer any questions since he has
more time on his hands than he would like (818).41
In character-narratives narratees are essential; indeed, they typically
respond to the narratives and thus flesh out, and complicate, the nar-
rator’s (perhaps implicit) assumptions and expectations about the effect
of his words. Thus, for example, if the Queen of Persians equates her
fear with that of her narratees (210–211), the chorus do not entirely
embrace this equation (215–216). Or if Clytemnestra at Agamemnon 348
directs the narratees to recognize that her narrative has been produced
by a woman, the chorus respond by saying that she has spoken just
like a man (351). For such dramatic reasons, then, character-narratives
situate the narratees at the centre of their concern. Such narratives
employ gar-clauses and use negatives to contradict the narratees’ pre-
sumed expectation.42 There are also many cases of more explicit atten-
tion to the narratees’ role in hearing, and even confirming, the narra-
tive.43 Character-narratives further incorporate the narratees by asking
questions. These may be rhetorical (as at Pr. 500–503 or A. 1374–1376)
or apparently sincere (as at A. 1194–1195 or Ch. 297).
In the case of Cassandra’s prophetic narrative, this attention to the
narratees serves in part to thematize the inevitable failure of her proph-
ecies. When the chorus fail to understand her (A. 1112, 1130–1133) and
her metanarrative comments are ‘meta-prophetic’ as well); Ch. 554 (Orestes); Pr. 641–
642 (Io).
39 At Eumenides 34 the Pythia makes a metanarrative comment to similar effect.
40 At 827 he similarly says that he omits a great deal from the story he does tell.
41 Other such comments at: 226–227, 740–741, 788, 801, 824–827, 842–845, 870–876.
The many metanarrative comments in Prometheus Bound, in fact, constitute something of

a theme: with the protagonist chained to a rock, story-telling itself becomes to a large
degree the focus of the drama. See Goward 1999: 78–79.
42 Cf. (gar-clause) Pers. 765, 767, 772, 802, 816, 821; Supp. 262; A. 326; Ch. 278; Eu. 38;
Pr. 224; (negative) Pers. 781, 783, 786, 802, 809–810, 813–814; Supp. 622; A. 290–291; Pr.
206.
43 Hearing: Pers. 211; A. 348, 879; Pr. 443, 476, 505, 641–642, 683, 703–706, 740, 789,
802. Confirming: Pers. 784; A. 1184–1187, 1240–1241; Ch. 574; Pr. 674.
finally confess their helplessness in the face of her enigmatic utterances

(amēkhanō, 1177), Cassandra promises to speak ‘no longer in riddles’
(1183) as she turns from sung lyrics to spoken trimeters. Although
she unambiguously announces the death of Agamemnon at 1246, the
chorus, of course, remain confused. Here, the pointed attention to
the narratees plays a crucial role in characterizing Cassandra both as
narrator and as prophet.
Like choral narratives, character-narratives may also invoke the nar-
ratees’ presumed familiarity with existing stories or ‘myths’.44 Cassan-
dra does so at Agamemnon 1232–1236 when she compares Clytemnestra
to an amphisbaina or ‘some Scylla’, and at 1197 when she identifies the
ills of the house as familiar to the narratees because they are ‘ancient in
story’.45
This pronounced attention to the secondary narratees, suggests that
they serve as guides/foils for the spectators, functioning as primary
narratees.46 There is, however, another way in which these narratives
may guide the audience as narratees, as they exploit the potential
offered by having multiple audiences on-stage, a practice found also
in Sophocles (→), Euripides (→), and Aristophanes (→). In Agamemnon,
for example, Clytemnestra alternately addresses Agamemnon (877–894,
905–907) and the chorus (855–876, 895–905), with Cassandra and the
servants on-stage as well. In each case the narratives can only bear
different meanings for the various audiences on-stage—as the contrast
between Agamemnon’s (reluctant) obedience to Clytemnestra and the
foreboding of the following choral song implies (975–1034). Clytemnes-
tra’s account of the sufferings of ‘a woman apart from her husband’
(861–862) is contextualized for the chorus by their experience of her
during the previous ten years, whereas Agamemnon may hear her
remarks as the generalization she offers, with pointed reference to his
own role in ‘abandoning’ her. At the same time, the theatre audience
witness multiple narratees listening in, so to speak, on what is addressed
to others, and find therein the suggestion that the same narrative may
44 Cf. the dialogical version of this phenomenon at Supplices 291–295 where the
narratee’s familiarity is demonstrated, in part, by reference to what ‘they say’ (phasi,

291) and to the circulation of a common account (phatis pollē, 293).
45 Rather than invoking the narratee’s presumed familiarity with the myth, Io’s
truncated account of Argus’ death (Pr. 677–681) demonstrates her limited knowledge as
narrator (see above). This brief version does, however, rely upon the familiarity of the
audience (as external primary narratee) with the myth (see Griffith 1983: ad 680–681).
46 Goward 1999: 76–79.
be understood in various ways. Thus do the narratees on-stage consti-

tute a complex ‘model’ for the theatre audience in their role as primary
narratees.
Messengers
Although less frequent and less highly formalized in the surviving plays
of Aeschylus than in those of later tragic poets, the messenger-speech
is among the more familiar of conventional elements in tragedy.47 This
form of narrative appears in Persians, Seven against Thebes, and Agamem-
non, the first of which has been called a ‘messenger-speech in its purest
and at the same time most ambitious form’.48 The narrative in Persians
treats the Persian defeat in Greece; that in Seven against Thebes the prepa-
rations of the attacking Argive army; and that in Agamemnon part of the
battle at Troy and the storm that dispersed the Greek fleet. These nar-
ratives are eyewitness accounts of events off-stage, as the messenger of
Persians announces: ‘I was there; I did not hear about it from others,
Persians, and I can recount the awful events’ (266–267). Such a claim
underpins the conventional messenger’s role and constitutes his very
raison d’être.
As in Euripides (→) but not in Sophocles (→), the messenger-speech-
es in Aeschylus are all produced by characters whose role encompasses
little beyond bringing the report. These narratives are preceded by
dialogue in Persians and Agamemnon, while those of Seven against Thebes
begin immediately upon the messenger’s entry. Aside from Seven against
Thebes 39–68, they all appear in scenes that punctuate the narrative(s)
with dialogue, thus producing discrete sections. Indeed these sections
can display significant differences: like only Agamemnon 636–680 among
Aeschylean messenger-speeches, the last of the four sections in Persians
(302–347, 353–432, 447–471, 480–514) presents a narrator who partici-
pates significantly in the events reported, while in the other sections he
is an eyewitness who, although an overt narrator, disappears quickly
into the background of the reported events. This tendency toward
self-effacement at the scene of action proves to be common in the
messenger-speeches of later tragedy, particularly in Euripides (→).49
47 On these formal differences see Keller 1959: 3, 7–8.

48 Rosenmeyer 1982: 198.
49 De Jong 1991a: 5; Barrett 2002: 34–40.
Messenger-speeches may respond to a question, as do each of the

narrative sections in Persians as well as Agamemnon 636–680. They typ-
ically conclude with a summation or judgment: at Agamemnon 680, for
example, the herald assures the chorus that his account is true; the
messenger in Persians concludes at 513–514 similarly asserting the truth
of his report and adding that it is far from a full account. The scout in
Seven against Thebes assures Eteocles that he will never find fault with the
report (651–652).
Aeschylean messengers are self-conscious narrators. Perhaps the least
self-conscious of the three is the figure in Seven against Thebes and yet
he, too, asserts the clarity of his report (40, 67) and the security of his
own knowledge (375, 651–652). The messenger of Persians reasserts his
knowledge at 341, but he also presents himself as actively editing his
narrative. At 329–330, 429–430, and 513–514 he rounds off narrative
sections saying that his account is incomplete. At 343 this messenger
concludes with ‘So the logos has it’ (hōd’ ekhei logos) after reporting
the number of Greek and Persian ships. However we translate, the
messenger here leaves room for some distance between this logos and
himself, acknowledging that his report is a logos.50
But the herald in Agamemnon takes self-consciousness to an extreme.
Two of his three speeches may be considered narrative (551–582, 636–
680), but the first of these fits only poorly into this category. The less
problematic narrative that recounts the shipwreck (636–680), begins
with 14 lines of metanarrative comment. At 551–583 the herald parades
a series of events that he could narrate, but will not. ‘If I were to tell’ (ei
legoimi), he begins at 555. And at 563–567 he produces an anacoluthon
that begins ‘If one were to tell’ (ei legoi tis) and ends ‘Why suffer these
things (again)?’ Here not only does his metanarrative comment displace
the narrative of the Greek army’s sufferings at Troy; it also questions
the value of such a narrative: ‘Why must I speak of those who perished
and count their number, and why must the living grieve at malignant
misfortune?’ (570–571). He then offers a summary to stand in for such
an extended narrative: ‘Having once captured Troy, the Argive army
dedicated the spoils to the gods across Greece, nailing them up at their
shrines as a sign of past glory’ (577–579).
50 Broadhead (1960: ad loc.) warns that logos here does not mean ‘story’, offering
‘score’ or ‘tally’ instead, but the other occurrences of this expression in Aeschylus do
not support him (Th. 225; A. 1661; Ch. 521). Belloni 1994: ad 337–343.
Like other (mortal) narrators in Aeschylus, messengers are in prin-

ciple subject to the restrictions of access and place. The messenger at
Agamemnon 663, for example, speaks of ‘some god’ (theos tis) and at Per-
sians 345 ‘some daimon’ (daimōn tis). At Seven against Thebes 545 the mes-
senger speaks of how Parthenopaeus ‘appears’ to him. The herald in
Agamemnon, furthermore, is unable to report on the whereabouts or con-
dition of Menelaus. As the mention of divine involvement at Agamemnon
663 and Persians 345 make clear, however, messengers also make infer-
ences that enable them to produce more complete narratives: at Per-
sians 372 the messenger reports that Xerxes spoke to the troops with
optimistic thoughts in mind (hup’ euthumou phrenos).51 Similarly, these nar-
rators may take advantage of ex eventu knowledge to enrich their nar-
ratives, as at Persians 361–362 where the messenger says that Xerxes
understood neither the trick of the Greek nor the ill will of the god.
On occasion, however, a messenger may show signs of bypassing
these restrictions altogether. This is particularly true of the messenger
in Persians whose perceptual point of view implicitly embraces both the
expansive vision of the sea battle as a whole (e.g. 419–420) and extreme
close-ups such as his description of Matallus’ beard at 316–317. This
messenger’s self-effacement as an actor in the events narrated supports
this implicit claim to virtual omnipresence: he is virtually absent from
the narrative until line 485 when he begins to report the Persian retreat,
using third-person forms referring to the Persians throughout.52 His
absence here is noteworthy because he offers such a vivid account of
near total destruction: at 464 he reports that all (hapantōn) on Psyttaleia
were killed. His own survival remains unexplained among the carnage
on both land and sea, underpinned as it is by his omnipresence.53
Although messenger-speeches in Aeschylus share much with charac-
ter-narratives, they typically position the narratee less centrally in the
narrative. Whereas Clytemnestra’s interests depart significantly from
those of the chorus in Agamemnon, for example, and thus shape her
narrative and guide its purpose, Aeschylean messengers cannot be said
to have motives of a similar order in producing their narratives. This
is principally a consequence of their having little role in the play aside
from delivering their report. With the messenger’s rhetorical purpose
51 Cf. Th. 392, 491–492, 529–530.

52 He does use the first-person plural personal pronoun at 406 (hēmōn). De Jong
1991a: 5.
53 Barrett 2002: 32–37.
less central to his role, the place of the narratee in the production
of the narrative becomes less pronounced: these narratives work to
privilege the ‘events themselves’ in part by deemphasizing the status
of the narrative as speech directed at the addressee.54 These narratives
do, nonetheless, display some of the same techniques for incorporating
the narratee found elsewhere.
Messenger-speeches anticipate the narratee’s questions with gar-
clauses;55 they also contradict the narratee’s presumed expectation with
a negative.56 They employ similes, thus signalling something of how the
narratee is imagined to make sense of the events recounted. At Persians
424, for example, the messenger compares the slaughtered Persians to a
haul of fish. Similes occur four times in the messenger-speeches of Seven
against Thebes: 53, 381, 393–394, and 498.
Like character-narratives, messenger-speeches invoke the narratee as
a knowing subject whose knowledge may support the narrative project.
‘Know well’, the messenger says at Persians 337, ‘that if it had been
simply a matter of numbers, the Persian fleet would have prevailed.’
He goes on to ask the chorus not to think numbers were decisive (344–
347).57 At Seven against Thebes 651–652 the messenger says to Eteocles
that ‘You will never find fault with this man [the messenger] for his
reports’. Similarly, the herald in Agamemnon concludes by calling upon
the narratee as knowing subject: ‘Having heard this, know that you
have heard the truth’ (680).
Other forms of attention to the narratee include direct address (Pers.
353; Th. 39, 62) and use of second-person forms (Pers. 356; Th. 576, 632).
It is noteworthy that Aeschylean messenger-speeches pose no questions
in their narrative sections, aside from the largely metanarrative Agamem-
non 551–582.58 This is one indication of the narratee’s relatively dimin-
ished presence here.
54 For these reasons, it is of particular value to consider the status of the theatre
audience as narratee in the case of the messenger. For a discussion of the messenger’s
narratives in Persians with respect to the audience, see Goldhill 1988a: 192–193.
55 A few representative examples: Pers. 335, 338; Th. 427; A. 655.
56 As, e.g. at Pers. 373, 417; Th. 378, 491; A. 668.
57 Cf. Pers. 431, A. 681.
58 Questions occur at the end of sections in Seven against Thebes when the messenger
asks Eteocles whom he will send to battle, in the form of a metanarrative comment
preceding the narrative at Agamemnon 646–649, and as the herald turns to the present
moment (671–672).
Conclusion
Aeschylean narratives assume a wide variety of forms, with narrators

adopting a range of positions and attending to the (multiple) narra-
tees in numerous ways. And this rich array of narrative types appears
amid the even more varied congeries of speech-forms that constitute
Aeschylean dramatic texts. This wide range of narrative types, the vari-
ety of narrators and their treatments of the narratee(s), the difficulty of
identifying and demarcating narratives, the fluidity with which narra-
tive may turn into dialogue, and the qualities of these narratives that
compel us to treat them as embedded in spite of the fact that they are
not—all of these point to the unusual status of narrative in this (dra-
matic) context. The many ways in which these narratives are hedged
about with qualification, that is, serve both to complicate our task in
understanding them and to clarify one of their key features: like the
plays of Sophocles and Euripides, the dramas of Aeschylus incorpo-
rate a sustained interest in and experimentation with the workings and
possibilities of narrative. Although we may discern patterns, narrative
practice in Aeschylus—as in Greek tragedy as a whole—is far from
settled. Indeed, the many self-conscious narrators and the pronounced
tendency toward metanarrative comment found in the plays suggests as
much. The richness of these narrative types, then, should be seen not
only as an index of a sophisticated understanding of narrative forms,
but also as a sign of experimentation: in addition to being speech-acts
in a dramatic context, Aeschylean narratives also constitute a sustained
programme of enquiry into the very nature of narrative itself.59
59 Gould 2001: 333–334.

chapter seventeen
SOPHOCLES
I.J.F. de Jong
Introduction. The forms of Sophoclean narrative
Like all Attic tragedians, Sophocles makes ample use of narrative in

his plays: we have messengers, the chorus, and protagonists who, in the
form of continuous narrative, dialogue, or a combination of the two,
relate their own or other people’s experiences, functioning as secondary
or intra-dramatic narrators.1 The dialogical form of narration is new
in comparison to epic and historiography, and it may be useful to
start this chapter by taking a closer look at this development. For
one thing, dialogue is a natural vehicle for narrative in a dramatic
context, and Sophocles, who strives for an unobtrusive presentation of
information, often turns to this form. We find dialogical narrative as
part of expositional prologues, e.g. in Aj., where Ajax’ nightly onslaught
on the flocks is presented to the spectators in two successive dialogues
between Athena and Odysseus (40–50) and Athena and Ajax (91–
117).2 Often narrative precedes the rheseis of messengers or reporter-
protagonists, e.g. in Oedipus Tyrannus 1237–1296, where the report of the
death of Jocasta follows on a dialogue (1223–1236) in which the news
is briefly announced, and Electra 871–919, where in a dialogue Electra
Chrysothemis recounts how she found a lock of Orestes’ hair, and then
proceeds with what she herself calls (892) a full and detailed narrative,
1 There is no single systematic study on Sophoclean narrative, but Goward 1999
and Barrett 2002 discuss many of its aspects and forms in general, while Roberts 1989,
Kraus 1991, and Markantonatos 2002 specifically discuss the narratives of Ph., Tr., and
OC, respectively.
2 Cf. OT 103–131 (death of Laius presented in dialogue between Creon and Oedi-
pus); Tr. 61–93 (Heracles’ latest exploits, recounted in a dialogue between Hyllus and
Deianeira).
256 part four – chapter seventeen
which is in a continuous form.3 The combination of introductory dia-

logue, which presents the main news and whets the appetite for the full
story, and detailed rhesis is a logical one and is also employed by Aeschy-
lus (→) and Euripides. In Sophocles a narrative often evolves back into
a dialogue which, while not narrative itself, does contain a reaction to
the narrative just recounted; an example is Electra 764–803, where cho-
rus, Electra, and Clytemnestra react—in dramatically different ways—
to the news of Orestes’ ‘death’ in the presence of the messenger, the
Old Slave.4
Sometimes Sophocles employs dialogical narration when he wants
to keep up the pace, rather than slowing down to the detail of a rhesis,
e.g. in Antigone 1278–1316 when, after a detailed messenger-speech on
the death of Haemon (1192–1243), the death of Euryclea is reported in
dialogue only.5 On other occasions the dialogical mode of presentation
is exploited for special effects, as when the chorus interrogates Oedipus
about his sad past (OC 510–548), and especially when Oedipus finds
out the truth about himself in a series of interrogations of Jocasta (OT
726–770), the messenger from Corinth (1017–1053), and the shepherd
(1121–1185).
Although Sophoclean drama is full of narrative, there are also mo-
ments when narration is avoided altogether: e.g. Odysseus in Philoctetes
3–11 briefly touches on the fate of Philoctetes, but then cuts himself
short. His avoidance of narration is motivated by the plot (as he himself
says, there is no time for conversation), but the story of Philoctetes is not
one in which he played a positive role, and hence his reluctance seems
to be psychologically motivated as well. Finally, there is the playwright,
who saves up the story, which will be told in full and pathetic detail later
(263–264, 285–316, 676–718). Likewise, Orestes’ cutting short the story
of his own and Electra’s suffering during the past years in Electra 1288–
1292 (and cf. the Old Slave doing the same in 1364–1366) is not only
3 Cf. Aj. 214–284, 719–747; El. 660–679; OT 726–770; Ant. 223–248, 384–406, 988–
997, 1155–1191; Tr. 225–247, 663–671, 734–748; Ph. 317–342; 542–602; OC 324–360,
1579–1585.
4 Cf. Aj. 783–812; OT 834–862, 1286–1296; Ant. 278–331, 441–445, 1244–1256; Tr.
291–334, 375–392; OC 385–420, 1666–1669.

5 Cf. Tr. 180–199 (a messenger informs Deianeira how Lichas reported to the
Trachinian people Heracles’ victory; the story will be told in full by Lichas himself
in 229–290), 1114–1142 (Hyllus tells Heracles about the death of Deianeira, which in
899–946 had been reported in full by the Nurse).
i.j.f. de jong – sophocles 257
acceptable in terms of plot and psychology, but also understandable

from the point of view of the playwright, who had already dealt with
this subject in two earlier narratives (254–309, 585–605).6
The marking and function of narrative
As a rule, when narratives are presented en bloc, the beginning and

end are explicitly marked. Narratives may start off with an emotional
preamble, as they often do in Homer (→), e.g. Trachiniae 1–5, where
Deianeira introduces the story of her life as follows: ‘There is an ancient
saying among men, once revealed to them, that you cannot understand
a man’s life before he is dead, so as to know whether he has a good or
bad one. But I know well, even before going to Hades, that the one I
have is unfortunate and sorrowful.’7 In the majority of cases we find a
metanarrative remark, which announces the act of narrating and often
stresses the reliability of the narrator: ‘I was sent for this purpose and
therefore will tell you [how Orestes died]’ (El. 680), or ‘I was there,
dear mistress, and will tell you, and I shall suppress no word of truth.
For why should I try to soothe you with words which will later brand
me as a liar? Truth is always best’ (Ant. 1191–1195).8 The actual narrative
typically starts off with the particles gar or epei. Sometimes we find epic
devices for starting off a story: asking after the main characters (‘What
mighty antagonists entered the lists … Who set out for the ordeal of the
contest?’: Tr. 503–506; cf. Il. 1.8), or the ‘there is/was X’ motif (‘There
is a cape in Euboea …’: Tr. 537;9 cf., e.g. Od. 9.508–519).
The end of a narrative may be signalled by the use of the present
tense, which describes the situation to which the events of the story have
led (e.g. Aj. 65: ‘and now he [Ajax] is torturing them, bound as they are,
inside his dwelling’),10 or a metanarrative remark (e.g. OC 1665–1666:
6 Cf. OT 685–686 (the chorus declines to tell Iocaste about Oedipus’ altercation
with Creon); OC 361–364 (Ismene declines to tell Oedipus and Antigone about her own
suffering, because there is now more pressing news, viz. the quarrel between Oedipus’
sons), 1148–1149 (Theseus modestly declines to recount his victory over the Thebans;
the playwright’s motive is that this battle had already been evoked by the chorus in
1044–1095).
7 Cf. Aj. 134–140, 430–433, 1266–1271; El. 254–260; Tr. 153–154.
8 Cf. Aj. 284, 748; El. 892; OT 707–710, 771–773, 1237–1240; Ant. 407a, 998; Tr. 472–
474, 672–673, 749, 899; Ph. 603–604a; OC 1291, 1586.

9 Cf. OT 103; Tr. 555, 752, 1159, Ph. 604b.
10 Cf. Aj. 323–327; Tr. 36–48, 943a.
‘And if anyone thinks I speak foolishly, I would not beg for the credence
of those who think I am a fool’),11 or, most often, a conclusion (e.g.
El. 307–309: ‘When things are so, my friends, there can be no good
sense or piety. No, when things are bad, inevitably one’s conduct must
be bad also’).12 The conclusion is usually accompanied by or evolves
into some form of exhortation to action: ‘keep good faith’ (El. 916–919),
‘show what you’re worth’ (Ant. 37–38), ‘I curse you’ (Tr. 807–812), or ‘I
supplicate you’ (OC 1326–1345). This is an important characteristic of
dramatic narrative: it is never told by way of entertainment to while
away time, as stories may be in epic, but it always has a function within
the plot: characters report events because these events call for action
and reaction.
Internal and external narrators
What kind of narrators do we find? Not surprisingly, most dramatic

narrators are internal, characters who are recounting events in which
they themselves played a role. The degree of their involvement may
range from protagonist to affected person to mere witness: Philoctetes
is given ample opportunity to recount his miserable life over the past
ten years (Ph. 263–316) and Deianeira sums up her fearful life (Tr. 1–
4813); an example of a story told by a person affected is the sacrifice of
Iphigeneia as recounted by Electra (El. 558–579), who wields it as an
argument in the agon with her mother;14 typical witness-reports, finally,
are the messenger-speeches, which will be discussed below.
A special form of narration consists of a narrator reporting events
about which he or she has merely heard. In Electra 417–427, for example,
Chrysothemis reports to Electra Clytemnestra’s dream (‘They say that
she was once more in company with your father and mine … That
is the story I heard from someone who was present when she told
11 Cf. El. 761–763; Tr. 289a–290; Ph. 389a, 620a.

12 Cf. El. 916–919; OT 720–725, 814–833, 1280–1285; Ant. 37–38, 1023–1032, 1240–
1243; Tr. 173–177, 484–489, 807–812, 943b–946; Ph. 389b–390, 620b–621, 1343–1347;
OC 382–384.
13 Discussion in Hulton 1969: 51–52; Winnington-Ingram 1980: 75–76; and Heiden
1989: 21–30. Cf. OT 771–833; OC 960–1002.

14 Discussion in Goward 1999: 111–112. Cf. Aj. 284–330; Ant. 21–38, 49–57; Ph. 603–
621; OC 361–384.
her dream to the Sun’).15 By presenting the dream in the form of a

hearsay report (and secondhand hearsay at that) Sophocles adds to
the casualness with which Chrysothemis approaches it. She has not
interpreted it (as Electra will shortly do), indeed has grasped so little of
its meaning that she almost left the stage without telling Electra about
it.
Hearsay narration can be exploited to great effect since—as Sopho-
clean characters themselves are only too aware (cf. El. 885–886 and Tr.
67, 425–426, 747)—hearsay is a less reliable source of information than
autopsy. One example is Oedipus Tyrannus 103–131, where Creon can tell
Oedipus about the murder of Laius only indirectly, repeating what the
one survivor recounted at the time; finding out the exact circumstances
of this event will take up the entire middle part of the play (707–862).
The most protracted example is found in Trachiniae, where Deianeira is
informed of Heracles’ attack of Oechalia in a series of hearsay reports:
(Hyllus:) they say Heracles was a slave to a Lydian woman the past
year, but is now attacking the city of Eurytus (61–93); (messenger:) I
heard Lichas say that Heracles has been victorious (180–199); (Lichas:)
Heracles has been a slave of the Lydian Omphale for a year (he told
me) and because this was due (he said) to Eurytus, he has sacked his
city by way of revenge (229–290; here Lichas’ reference to Heracles
as his source serves to back up the true part of his story and thereby
commend the false part); (messenger, correcting Lichas:) in fact, Lichas
told the Trachinian people that Heracles sacked Oechalia because he
wanted the daughter of Eurytus, Iole (335–374); (Lichas:) I said that I
had heard that the captive woman is Iole the daughter of Eurytus (395–
435) … yes, Heracles sacked Oechalia in order to get Iole (472–489).
The function of this protracted use of hearsay narration is to point up
Deianeira’s isolation: she is last in a convoluted and unreliable chain of
communication.16
Only occasionally do we find external narrators. Here speakers, usu-
ally the chorus, recount events of what for them, too, already is a myth-
ical past in which they themselves played no role: ‘the body of Danae
too endured to give up heaven’s light in a brass-fastened dwelling, and
immured in a tomblike chamber she was held prisoner. And yet she was
highly esteemed by reason of her birth, my child, and she was treasurer
15 Discussion in Goward 1999: 107–108.

16 Gellie 1972: 62; Heiden 1989: passim; Kraus 1991: 83–88; and Goward 1999: 91–
96. Cf. Aj. 134–200; El. 566–579; Ant. 21–36; OC 374–381.
for Zeus’ golden-flowing seed’, says the chorus to the absent Antigone
(Ant. 544–548).17 As in Homer (→), Pindar (→), and Aeschylus (→) these
stories of the past are invoked in order to provide a parallel (whether
by way of similarity or contrast) to the situation at hand on stage; the
way in which the parallel is intended is not always clear (is the cho-
rus consoling Antigone, exhorting her to accept her fate, shying away
from their own responsibility?).18 They display the typical allusive and
elliptical style that highlights certain events (which are necessary for the
‘message’), while at the same time leaving out others. Thus in the case
of the ‘Danae’ story, both prehistory (why Danae was immured) and
aftermath (her escape with her son in a metal chest) are suppressed,
while the nature of her ‘honoured house’ and ‘Zeus’ golden-flowing
seed’ is left for the narratees to fill in. Conversely, the detail in her
story that corresponds most closely to Antigone’s situation, her impris-
onment, is mentioned twice, in terms that recall that heroine’s present
predicament. Their external status turns these narrators into author-
itative speakers: they are not bound by the restrictions of narrators
who are themselves part of the events and thereby lack an overview,
but can narrate omnisciently, at times embracing ‘the whole of human
knowledge’.19 Of course, their omniscience applies only to their compe-
tence as narrators; as interpreters of the play’s action the chorus is as
restricted and as partial as the other participants.
17 Cf. El. 145–152 (Electra recalls the stories of Niobe and Procne), 504–515 (the
chorus relates the chariot race of Pelops, the first of many mishaps in the house of
Agamemnon, to which a new one is now about to be added; for the audience, the
theme of the disastrous race recalls the scheme of Orestes’ ‘death’ in a race, which is
about to be related), 837–847 (chorus and Electra recount the story of Amphiaraus and
Eriphyle), 955–965 (Lycurgus), 966–987 (Cleopatra); Ant. 823–833 (Antigone recalls the
fate of Niobe by way of a parallel to her own sad fate); Tr. 503–530 (the chorus recount
the story of how Heracles and Achelous fought over Deianeira, illustrating the force of
Aphrodite, who is presently responsible for Heracles’ infatuation with Iole; thus a story
from the recent past has already acquired the status of a mythological tale, known to
all); Phil. 676–679 (Ixion). Comparable in function are the brief anonymous paradigms
presented by Menelaus and Teucer in Aj. 1142–1146 and 1150–1156.
18 Burton 1980: 124–132 and Gardiner 1987: 92–93.
19 Goward 1999: 22.
Messengers
One group of narrators is deserving of a separate discussion: the mes-

sengers. As in Aeschylus (→) and Euripides (→), all Sophoclean plays,
with the exception of the Philoctetes, feature at least one messenger-
speech, in which a character, usually anonymous and of low social sta-
tus, reports offstage events, not seldom the bloody climax of the play.20
Sophocles, typically, has well integrated this traditional device into his
plays. Thus his messengers are characters who (i) sometimes play a role
in the rest of the play as well (Lichas and Hyllus in Tr., the Pedagogue
in El.), (ii) have distinct personalities (notably the guard who reports the
two burials in Ant.), and (iii) sometimes act as messengers on more than
one occasion (the guard in Ant., the second messenger in Ant., who first
reports Haemon’s death, then Eurydice’s). The messengers do not tell
their stories of their own accord, as is usually the case in Aeschylus and
Euripides, but are sent by one of the characters in their story to report
what has happened (and hence to set in motion new action). Thus the
messenger in Ajax is sent by Teucer to report Calchas’ prophecy to
Ajax, in an attempt to avert the danger it contains.21 A final device
by which Sophocles integrates and naturalizes this potentially alien ele-
ment of narrative into the dramatic context is by having it not only
preceded by a dialogue (as in Aeschylus and Euripides), but also fol-
lowed by one, in which the messenger is asked for extra information or
joins the discussion about his news (OT 1286–1296; Ant. 1244–1256; Tr.
291–334, 375–392; OC 1666–1669). The preceding dialogue, in which,
after a brief announcement of the main news, one of the characters asks
for detailed information is a fine method of justifying the length of the
narrative that is to follow. The same purpose is served by the dramatic
quality of the narrative, which includes speeches, historic presents, and
many graphic descriptions. Or as one messenger puts it: ‘the briefest
way to tell my news would be to say that Oedipus is dead. But to tell
briefly what has happened nor the words nor the deeds that took place
there allow’ (OC 1559–1560).
Messengers are in principle internal narrators, though their role is
often restricted to that of a witness. Indeed, it is because they have been
eyewitnesses that they can now act as messengers, and hence this fact
20 Aj. 748–782; El. 680–763; OT 1237–1285; Ant. 249–277, 407–440; Tr. 749–812, 899–
946; OC 1586–1666. Goward 1999: 26–32 and Barrett 2002: 76–101.

21 Cf. El. 669–670; Ant. 272–277; Tr. 285–286.
is invariably stressed: ‘So much as this I know, for I was present’ (Aj.
748) or ‘I saw with my own eyes the dire calamity of my father and did
not merely hear about it’ (Tr. 746–747).22 Though often no more than
witnesses, they are emotionally affected witnesses: ‘what we saw next
was terrible’ (OT 1267) or ‘such was this event, terrible to relate, and for
those who saw it, as we did, the worst disaster of all that I have beheld’
(El. 761–763); in the latter case, the conventional emphasis which a
messenger places on his autopsy has additional significance, in that he
is telling a false tale. But the messenger does occasionally play a role
in the events, most notably the guard in Antigone, who reports the two
burials of Polynices (223–331, 384–445).23 It is precisely his involvement
which determines much of the tone of his stories: the first time he is
reluctant to report the burial, and even fearful, conscious as he is of his
own failure as a guard; the second time, he is both exhilarated and sad
to be able to tell who did it and, handing her over, to save his skin.
Overt narrators
Sophoclean narrators are overt narrators, who reveal their presence

as a narrating subject at many points and in many ways. To begin
with, there are the metanarrative remarks which, as we have seen,
often mark the beginning and the end of a narrative, but which also
occur in the course of an account. In the latter case they may have a
structural function, emphasizing a new or crucial phase in the story,
e.g. OT 1169: ‘Ah, I have come to the danger point in telling my story.’24
Next, there are evaluative or emotional qualifications: Electra refers to
the relationship of Clytemnestra and Aegisthus as ‘their final outrage’
(El. 271); the Theban priest calls the Sphinx ‘the cruel singer’ (OT 36)
and the oracles which Oedipus received ‘the shameful predictions of
my cruel oracles’ (OT 796–797); Antigone talks about ‘the unhappy
corpse of Polynices’ (Ant. 26); the chorus refers to the sons of Oedipus as
‘the unhappy two’ (Ant. 144); Deianeira calls herself ‘unhappy’ (Tr. 16),
etc. Then there are comments, e.g., El. 696–697: ‘So far, things stood
thus; but when one of the gods does mischief, not even a mighty man
22 Cf. OT 1238, 1263; Ant. 423, 432, 1192, 1207, 1216; Tr. 742–743, 888–889, 912; OC
1646, 1654.
23 Cf. Hyllus in Tr. 749–812, nurse in Tr. 899–946, messenger in OC 1586–1666.
24 Cf. OT 800; Tr. 678–679.
can escape’; the comment serves to articulate the narrative, marking

its climax.25 Sometimes comments are couched in comparisons, e.g.
Ant. 424–425: ‘Antigone cried out bitterly’, says the messenger, ‘with
a sound like the piercing note of a bird when she sees her empty nest
robbed of her young; just so did she cry out, weeping, when she saw the
corpse laid bare.’26 With this pathetic image, the messenger—thus far
concerned only for himself and hence critical of the burial—now shows
some sympathy for the person who carried it out.
Narratees
Narratives in drama are nearly always recounted to someone: the cho-

rus, but as a rule one or more of the protagonists as well. The pres-
ence of these secondary or intra-dramatic narratees is acknowledged
with a greater or lesser degree of intensity. The mildest form are voca-
tives, usually friendly or ingratiating in nature (‘my dear’, ‘friends’,
‘child’), but occasionally more negative (‘you shameless insolence’, ‘you
wretch’: OC 960, 978). The abundance of vocatives which Philoctetes
employs when recounting to Neoptolemus the story of his solitary years
in Philoctetes 260–316 (cf. 260, 268, 284, 300) underscore the argumenta-
tive function of this narrative: soon he will ask his narratee to take him
with him. Sometimes the narratee is not present on stage, but apostro-
phized: a god (Aj. 172–181, El. 95–102) or the sun (Ant. 100–106), called
on in their capacity of witness. In some cases a more active involve-
ment of the narratees is solicited in the form of ‘there you could have
…’ passages (e.g. ‘if you had been close at hand to see the nature of
her action, you would indeed have pitied her’: Tr. 896–897),27 or ques-
tions (e.g. ‘And then what kind of days do you think I pass when I see
Aegisthus sitting on my father’s throne …?’: El. 266–274).28
Where the narratee is actually a character in the story being recount-
ed, we are dealing with second-person narration, often quite forceful in
tone: the Theban priest, leading up to a request for help from Oedipus,
recalls the king’s former help: ‘For it was you who came to the city of
25 Cf. Aj. 150–161; OT 777b–778. See also nn. 8 and 13 for comments at the opening
and end of narratives.

26 Cf. Aj. 215; El. 98; Ant. 113; Tr. 32–33, 530.
27 Cf. Tr. 365, 692; OC 969, 1587–1589.
28 Cf. Ant. 1194–1195; Tr. 742–743; Ph. 276–278.
Cadmus and released us from the tribute we were paying, the tribute of
the cruel singer …’ (OT 35–36). Conversely, in order to demonstrate the
incompetence of Tiresias, Oedipus recalls the seer’s role in the episode
of the Sphinx: ‘Why, come, tell me, how can you be a true prophet?
Why, when the bitch-sphinx sang her song, did not you speak some
word that could release the citizens?’ (OT 390–392).29
There is also a special form where the narratee who plays a role
in the story is absent and we are again dealing with an apostrophe.
A particularly forceful example is Ajax 134–171+182–191, where the
chorus of Salaminian sailors give their—incredulous—version of Ajax’
nightly massacre of the herds in the ‘you’ form, addressing their lord
Ajax, who is still in his tent. The choice of this form is effective, in
that it underscores the chorus’ dependence on their king (which they
themselves explicitly acknowledge: ‘little men are best supported by the
great … we have not the strength to defend ourselves against them
without you’: 160, 165–166), and leads up to the epode where they
actually call out for him to come out of his tent. Finally, a narrator
who does not use second-person narration may yet abundantly insert
‘you’ forms, as Hyllus does in his report to his mother of the manner
in which she poisoned his father Heracles (749–806): ‘as Heracles was
about to sacrifice, there came Lichas bringing your gift, the robe of
death. He put it on, as you had instructed, … When Heracles asked Lichas
through which scheme he had brought the robe, Lichas told him that
it was your gift alone … Heracles hurled himself to the ground, uttering
many cries and dwelling upon his disastrous marriage with you …’
In this way Hyllus leads up to his forceful conclusion: ‘These are the
plot and the action, mother, of which you are convicted, for which may
avenging Justice and the Erinys punish you!’ (807–809).
In general, the conclusion of a story, containing an exhortation to
action, addresses the narratees: thus Tecmessa ends her report to the
chorus about Ajax’ madness with ‘Come friends, for this is why I came,
go in and help him, if you have any power to do so! For such men
are won over by the words of friends’ (Aj. 328–330). The most active
involvement of the narratees is found in dialogical narratives. Here the
narratee actually speaks, and by his questions prompts or steers the
narrative: e.g. ‘(messenger:) Teucer gave orders that Ajax be kept within
the shelter of the hut and not allowed out alone. (Tecmessa:) And where
29 Cf. Aj. 134–200, 1273–1297; El. 11–14, 585–608; OT 1017–1053; Ph. 1324–1347.
is Teucer, and what is the reason for saying this? (messenger:) He has
lately come back. He believes that this departure [of Ajax] seals the
fate of Ajax. (Tecmessa:) Ah me, from what man did he learn this?
(messenger:) From the prophet who is son of Thestor …’ (Aj. 795–801).30
Narrators can also be their own narratees, as is often the case in
choral narratives, where the members of the chorus address each other
(or no one in particular); naturally, in such cases the primary narra-
tees, the spectators in the theatre, will have a greater sense of being
addressed themselves than when other narratees apart from the chorus
are present (→ Aeschylus and Euripides). When the chorus are narrat-
ing events in which they themselves are involved, their internal narra-
tion employs ‘we’ forms. This is the case in Ant. 100–147: ‘Beam of the
sun …, you who moved off in … flight the man … that was raised up
against our land … and flew to our country …; he paused above our
houses. …; but he went, before his jaws had been glutted with our gore
…’, etc. This ‘we’ narrative, recounting the quarrel between Polynices
and Eteocles in terms of its effect on the city, presents an effective con-
trast with the previous scene, in which the same event had been looked
at from the point of view of the family by the sisters Antigone and
Ismene.31
As always, it is important to take into account the different functions
which narratives may have for the primary and secondary narratees.
Thus the ‘argument’ function of the mythological examples in Antigone
944–987 may be to console or exhort Antigone, and to warn Creon,
while their ‘key’ function may be to suggest to the spectators the
catastrophic consequences of thwarted desire. In the same way, the
report of Orestes’ death in Electra 764–803 has different effects on
Electra (who is shattered), Clytemnestra (who is sad for a brief moment
but otherwise relieved), and the spectators (who know it to be a false
tale and can admire the old man’s narrative talents).
30 For other examples, see the discussion of dialogical narrative in the introductory
section.
31 Gardiner 1987: 84–85.
A Sophoclean speciality: false narrative
Taking his cue from the Homeric lying tales (→ Homer) and Aeschy-
lus (Ch. 674–690), Sophocles loves to insert false narratives. The longest
and least complicated instance is the pseudo messenger-speech voiced
by the Paedagogue in Electra 680–763. It serves to back up the news of
Orestes’ death, which in turn forms part of the larger scheme of Orestes
who, in disguise, enters the palace and kills Clytemnestra and—later—
Aegisthus. His death could, of course, have been reported much more
briefly, along the lines set out by Orestes in 49–50 (‘tell them that
Orestes is dead by an accident, fallen from his moving chariot in the
Pythian games’), but the Pedagogue actually spends more than half of
his tale on an account of Orestes’ successes before his accident and
the first half of the dramatic chariot race, including a detailed list of
the contestants. Of course, this attention to detail serves to increase the
authenticity and hence the authority of his invented tale (which will
indeed be believed without any reservation by his narratees Clytemnes-
tra and Electra). But the picture painted here of an Orestes proving
his mettle in the most heroic of athletic contests (cf. 693–695: ‘He was
proclaimed as an Argive, by name Orestes, son of Agamemnon, who
once gathered the famous armament of Greece’) has a considerable
effect on Clytemnestra, who briefly takes pride in her son and hence
grieves over his death, however desirable it is (766–771), and above all
on Electra, who now has all the more reason to regret the loss of such
a brother (808–822). At the same time, this invented tale embodies the
theme of ‘power brought low’ (the beautiful and apparently invincible
young aristocrat Orestes unexpectedly crashes), and as such it can be
said to anticipate, for the spectators, Clytemnestra’s impending down-
fall.32
Lichas’ report on Heracles’ sack of Oechalia in Trachiniae 248–290 is
not so much false as misleading, focusing as it does on Eurytus’ insults
and Heracles’ revenge, which consists in taking the former’s city, but
leaving out the crucial fact that he had another motive for wanting this
city, viz. Eurytus’ daughter Iole.33
32 Macleod 2001: 107–132. To my mind, her suggestion that the tale actually func-
tions as a warning goes too far: the Paedagogue is clearly not trying to warn her.
33 Analysis in Parlavantza 1969: 28–30; Heiden 1989: 53–64; and Kraus 1991: 84–85.
Two highly intriguing false narratives are those in Philoctetes by Neop-

tolemus (329–390) and the Merchant (542–561). They are a mixture of
true and false, and even their false elements are of relevance to the play
as a whole. Let us take a closer look at Neoptolemus’ tale. Odysseus and
Phoenix came to fetch him from Scyrus (true), telling that he was the
only one who could take Troy (after the prologue Neoptolemus knows
that this is only partly true, since Philoctetes and the bow are also nec-
essary; hence his somewhat bitter comment here: ‘whether it was true
or after all a fiction’: 345). Having arrived in the Greek camp at Troy,
he was greeted as a second Achilles (this may be a detail invented to
increase the persuasiveness of his tale, but nothing speaks against it
being true), but when he asked the Atrides for the armour of his father
they refused, as they had already given it to Odysseus. When Odysseus,
too, refused to hand it over, he decided to go back to Scyrus (false).34
Where Neoptolemus unknowingly foreshadows the course of the play
itself (until the forceful intervention of Heracles): Odysseus and Neop-
tolemus have come to fetch Philoctetes, there will be disputes over his
bow (865–1080, 1218–1302), and at some point Neoptolemus will decide
to go back to Scyrus (1402–1408). Thus for a long time there will be
more truth in Neoptolemus’ narrative than he himself realizes.
Conclusion
Sophocles’ drama features a great deal of narrative and a wide variety

of intra-dramatic narrators. This may be due in part to a technical fac-
tor (his abandonment of the trilogy, which means that what preceded
the play must somehow be presented in the course of the play itself),
but much more to a thematic one (Sophoclean characters are often in
the grip of the past). His predilection for narrative was a challenge to
Sophocles, so much a man of the theatre: how can this undramatic ele-
ment be integrated into his plays? His solution is to chop up narrative
into dialogue or, when it is presented en bloc, to frame it by dialogue;
to make his characters intensely interested in the content of the narra-
tives, to which they react forcefully; to endow even the most functional
34 There are other analyses of Neoptolemus’ tale: some consider it a lie from
beginning to end; others think that the part about his being refused the armour of
his father is true, because when Philoctetes repeats the fact in 1364–1365, Neoptolemus
does not correct him.
narrators of all, the messengers, with real personalities or with a role

elsewhere in the play. Moreover, he varies the status of the narratives:
some are no more than hearsay reports, others are downright lies. In
this way they fascinate, puzzle, and surprise both the characters on
stage and the spectators in the theatre.
chapter eighteen
EURIPIDES
N.J. Lowe
To appreciate the distinctiveness of Euripides’ use of narration, we need

to bear in mind three general principles governing the use of narration
in tragedy. (i) All narration in drama is embedded narration; there is thus
always a signifying relationship between the embedded narrative con-
tent and the framing situation, especially the voice and intentions of the
narrating figure in communication with the intradramatic narratees. (ii)
It is in the nature of tragic performance that all significant action is
in some sense unseen, and mediated through ‘narration’ in the broad-
est sense: the past and future, the offstage world, states of mind, the
designs of gods are all inaccessible to direct representation, and access
comes only through variously problematic categories of verbal report.
(iii) Tragic drama actively explores the boundaries between narration
and neighbouring modes of discourse—between telling and showing,
between narrative and argument in persuasive rhetoric, and especially
between narrative focalization and the modality of other, non-narrative
kinds of world-building through language.
Precisely because these boundaries are so blurred in practice, it is im-
portant for analytic purposes to mark a clear formal distinction be-
tween narrative and more strongly modalized forms of mediated action
in tragedy—such as the proleptically oriented categories of plans, wish-
es, prayers, threats, vows, fears (all of which can, but need not, be pre-
sented in narrative form); analeptic forms such as regrets and accu-
sations; and beliefs, feelings, and desires about the present. For these
purposes, tragic narrative may be circumscribed as dealing with past,
present, or future events (rather than states of mind, judgment, or gno-
mic opinion) presented by their narrator as non-provisional fact (rather
than modalized states of potentiality), whether or not they are actu-
ally true—and normally related to the narratee as news (except for the
special cases of agonistic narrative, where narrative is used to present
competing, contested analepses, and choral paradeigma, where analep-
tic narration expands the causes or significance of present events).
270 part four – chapter eighteen
Within this framework, Euripides’ distinctive uses of narration are

strongly connected to formal elements of structure. Euripidean drama
is patterned around formally positioned slabs of narration, each with
its own characteristic poetics, and the modulation between these gives
the Euripidean play much of its characteristic shape.1 The following
survey addresses the functions of narrators and narratees in the five
main narrative loci: prologues, choral odes, narratives in rhesis within
episodes (including messenger-speeches), agones, and closing narratives.
It necessarily omits such extraordinary sui generis coups as the dying
Alcestis’ hallucination (a narration of counterfactual events in present
time) and the many passing glides in and out of narrative in dialogue,
in the plotting of intrigue, and in non-agonistic rhesis.
Prologues2
Euripides’ pervasive use of the formal prologue was already recognized

as a trademark device by his own contemporaries (Ar. Ra. 946–947;
cf. 1197).3 The form of a monologue on an empty stage is usually traced
back to Eumenides, but there the Pythia does seem to imagine a human
audience of prospective consultants (31–33), though probably offstage.
The two probably post-Euripidean prologues, Iphigenia at Aulis and Rhe-
sus, use an entirely different form, a later development of the Sopho-
clean duologue in which the audience eavesdrop on a two-handed con-
versation: in both cases, a remarkable nocturnal summons and instruc-
tion, deliberated between the two voices in the Iliadic night-time space
between action when armies are silent and plans can be laid.
1 In an important discussion, Goward 1999 identifies as the most distinctive feature
of Euripidean narrative the use of formal narrative in prologue and exodos, and a
corresponding diminution in the role of overt prolepsis (dreams, oracles, curses, etc.)
in the body of the play; this, she argues, is part of a wider strategy of diminishing the
teleological shapeliness of the Aeschylean and Sophoclean tragic universe in favour of a
world-picture dominated by chance. Other consequences, noted by Segal 1992, include
a more self-conscious marking of literary form and contrivance; new kinds of plotting
that require more initial narrative equipment; and a more self-conscious approach to
myth.
2 The key discussion is Erbse 1984, q.v. (6–19) for survey of earlier literature; see also
Imhof 1937: 26–45; Schmidt 1971; Strohm 1977; Hamilton 1978; Segal 1992; Katsouris
1997.
3 Among the many discussions see especially Erbse 1984: 3–4; and Dover 1993: 337–
339.
n.j. lowe – euripides 271
The central contention of Erbse’s detailed analysis is that what Eu-

ripidean expositors tell us is true, and a canon of reliability for the crit-
icism of subsequent untruths. Yet as Hamilton (1974, 1978) and Segal
(1992) have shown, the prologues still leave gaps and ambiguities, and
their narrators are anything but objective, impersonal authorities; a
striking case is Electra in Orestes, who at 16 and 26–27 pointedly elides
parts of the narrative that sit uncomfortably with her public character.
And as we shall see, closer examination also reveals that the so-called
‘detached’ prologues are full of complex, sometimes contradictory, indi-
cations of the presence of narratees within the play which widen rather
than limit their narrative’s engagement with its context.
The range of prologue narrators is extremely broad: in the nineteen
complete plays, we have five gods, a satyr, and a ghost, alongside a
spectrum of mortals ranging from title characters (Andromache, Iphigenia
in Tauris, Helen) to nameless slave (Medea) and peasant (Electra). There are
nevertheless some revealing patterns that reveal the distinctiveness of
Euripides’ approach. Where the prologue speaker is a mortal, Euripides
consistently presents victims or other impotent or marginal figures; the
only male narrators are Electra’s nameless husband and the aged Iolaus
and Amphitryon, while the named female speakers are consistently
captive and/or besieged (often as suppliants). The effect is to enter
the Euripidean play from a perspective of wide access to narrative
content but limited power over its development. In the prologues,
narration is an expression of disempowerment; as speech acts, the
prologues express lament, prayer, the abandonment of hope. Heroines
who do break out from victimhood to violence (Medea, Hecuba, Electra)
are not given prologue narration; viewed initially through other eyes,
they are presented as objects of study rather than ab initio sympathy.
The five divine narrators, and the ghost of Polydorus in Hecuba, are
a particularly complex group, being privy to information outside the
reach of human consciousness. Their narrations are more markedly
proleptic, and sometimes have the illocutionary force of command
or promise. Yet though gods and ghosts know all, they will not (and
cannot) tell all, so that in practice their accounts leave tantalizing
spaces for the spectator or reader to fill; and in contrast to the mortal
narrators, they are invisible to all the play’s human cast,4 and so do not
communicate with the world of the mortal action at all.
4 Hecuba is onstage throughout the dialogue of the gods in Troades (36–38), yet is
oblivious to their presence; the other divine prologists are alone on stage.
This raises a fundamental question: who are the narratees of all these
soliloquies on seemingly empty stages? Aristophanes’ Ranae took it for
granted it must be the Athenian audience, as regularly in comedy. But
there is no explicit audience address of the kind routine in Old and
New Comedy;5 instead, tragedy consistently assumes a Homeric uni-
verse with its permanent audience of unseen spectators, whose tacit
homologies with the audience in the theatre are central to tragedy’s
model of the ironic structure of its world.6 Divine speakers come closer
to acknowledging the presence of a theatrical audience, but even these
never resort to second persons; instead, their use of deictics and other
demonstrative terms (such as Aphrodite’s deixō at Hipp. 9, or Poseidon’s
words at Tro. 36) evokes a subtler pragmatics of implicit expository
address to quasi-overt narratees, of whose presence they are conscious
but whose actual identity and location in the narrative are left unspeci-
fied.
But even mortal prologists do not exposit in a vacuum. Rather, they
draw tacitly or explicitly on a range of available narratees on stage:
self, unseen gods, textually invisible attendants, or simply landscape,
empty air, sunlight. Since all of these are irresponsive, the distinction
between communicative address and non-communicative apostrophe is
not always solid.7 These unresponding audiences may also be blurred,
combined, or juxtaposed: thus the pedagogue in Medea describes the
nurse as talking to herself (51), while she herself claims earth and sky as
her narratees (57). Iphigenia, similarly, describes her audience as aithēr
(43; she at least is unattended), yet 37 makes it clear that she is also
aware of, though not speaking for the ears of, the listening gods. But
while prologue narrators may make free use of such communicative
tropes as deictics, interactional particles, and rhetorical questions,8 they
do not address themselves to a responsive narratee on stage. Orestes is
unconscious during Electra’s narration; Hecuba is oblivious to the gods’
conversation; and even when the expositor does share the stage with
a listener,9 the intended audience is generally more open.10 As a rule,
5 See Segal 1992: 96 on the rhetorical question opening Heracles.

6 See more fully Lowe 1996 (esp. 523–524) and 2000: 180.
7 On apostrophe as second-person narration → Introduction and → Lycophron.
8 Ar. Thesm. 868 makes fun of this trope by introducing a second party to answer
Hel. 56.
9 See the table in Schmidt 1971: 5 with 4 n. 15.
10 Thus Amphitryon speaks of Megara in the third person, with deictic, at 14
and throughout until she speaks to engage him directly in dialogue at 60 for her
the presence of a responsive listener closes off prologue narration: the

arrival of a second character, or of the chorus, forces a suspension of
narration, a targeting of address, and a modulation from past to present
time.
Messenger speeches and related narrative rheseis
In a detailed survey of the engagement of Euripidean messenger-speech

narrators both in their narratives and with their on-stage audiences, de
Jong shows that messengers play a vital role in constructing the ‘open’
dialogic interplay of perspectives and evaluations that characterize the
Euripidean world-picture. Particularly important for present purposes
is her demolition of the myth of the ‘objectivity’ of the messenger as
narrator, through her exposure of pervasive touches of focalization,
references to autopsy, and address to the on-stage narratees.11 Yet as
Barrett demonstrates, there is simultaneously a tendency towards the
messengers’ self-effacement as presences in the stories they tell: by
ambiguous self-positioning in the scene, by referring to spectators in the
third person, by distancing themselves from other spectators mentioned
in their narrative, or by instances of focalization that strain the limits of
autopsy.12
As de Jong points out, the traditional anonymity of the tragic mes-
senger does not mean that such narrators are faceless, disengaged
nonentities. On the contrary, Euripidean messengers are all closely
aligned with one party to the play’s central dialectic. They are anony-
mous not because they are characterless ciphers, but because they are
slaves, and tragic slaves are always unnamed; 24 out of 30 speakers13 are
slaves, while the peasant in Orestes is a feudal dependant of Agamem-
non’s family, and the professional herald Talthybius, though a free man,
supplementary narration. In Heraclidae, Iolaus is accompanied by the mute Heraclids,

but speaks of them in the third person until Copreus’ arrival at 48 prompts an urgent
appeal to the boys to gather close.
11 Barrett 2002: passim.
12 Barrett 2002 (esp. 74–96), who adduces El. 826–829; HF 930–934; and Med. 1156–
1157.
13 Reckoning from the canon in de Jong 1991: 179–180 (q.v. for earlier corpora) and
189–190, but with her definition relaxed for purposes of this discussion to admit also
the six narrative rheseis and one messenger lyric (the Hecuba parodos) excluded from her
canon of messenger-speeches as such.
regularly speaks as though fulfilling the role of a servant of the collec-

tive commanders. Strikingly, not one of the tragic messengers is female
(there is only the therapaina in the prosatyric Alc.)—in part, perhaps
because they need to ferry news from the public world offstage, but
also perhaps for reasons of narrative authority.14
The chorus is always present for messenger narratives, and in nine
instances it is the sole audience; in Supplices it is the principal narra-
tee, though Adrastus is also present; elsewhere the messenger addresses
himself primarily to a solo character, but the chorus still offer their own
response. In such cases the alignment of sympathies is potentially com-
plex, since the choral perspective may be significantly at odds with that
of the principal narratee. Though in a majority of cases the narrator
and all onstage narratees are on the same side, in seven the messen-
ger faces a hostile audience,15 and there are six more where the chorus’
sympathies conflict with those of the messenger’s principal addressee.16
Even the messenger’s own loyalties can be divided: Hippolytus’ groom
is Theseus’ slave as well as Hippolytus’, but champions his young mas-
ter over the old, while Pentheus’ slaves similarly call their master’s judg-
ment into question. And to complicate things still further, the on-stage
narratees may differ from the messenger and/or from one another in
their response to the content of the narration; thus, for example, Medea
is no friend to Jason’s slave, but is delighted by the content of his narra-
tion, at the same time as the chorus are sympathetic to the narrator but
appalled by his tale.
Choral narration
Unsurprisingly in view of the increased narrative element in the pro-

logue, narrative parodoi are rare in Euripides, and limited to cases of
immediate autopsy—as in Hecuba 98–153 (hesitantly counted with the
previous section’s narrative rheseis, since it brings news to an on-stage
narratee), Ion 184–218 (strictly, mimetic rather than narrative), and Iphi-
genia at Aulis 164–302. But narrative stasima are more common, and
14 On the gender, status, and authority of tragic messengers see Barrett 2002: 99–
101.
15 Hec. 518–582 and 1132–1182; Tro. 1123–1155; IT 260–339 and 1327–1419; Or. 1395–
1502; Bacch. 1043–1152.

16 Med. 1136–1230; Hipp. 1173–1254; Suppl. 650–730; Hel. 1526–1618; Bacch. 434–450
and 677–774.
there is also frequent use of shorter narrative inserts as paradigms or

aetiological flashbacks, as well as some lyric recapitulation of events
already witnessed or recounted (as at Or. 327–331). An inventory of nar-
rative odes would probably include Andromache 1009–1046 (the fall of
Troy), Troades 511–567 (the sack of Troy), Electra 432–486 (the shield of
Achilles), Iphigenia in Tauris 1234–1283 (foundation of the Delphic ora-
cle), Phoenissae 1019–1066 (the Sphinx), Orestes 807–843 (Agamemnon’s
murder), and the special case of Iphigenia at Aulis 751–802 (a vivid pro-
lepsis of the coming war for Helen).17
Euripidean stasima are particularly characterized by a movement
from general to specific, far to near, impersonal to personal, and light to
dark; for narrative odes, this generally entails a transition from past to
present and from third-person analeptic narration to more modalized
and proleptic forms of first-person discourse such as prayers, hopes,
and speculations. Thus a common pattern is for stasima dominated
by narrative of past events to modulate into a non-narrative reflection
situated in the present, in which the choral narrators’ present situation
and perspective progressively (or abruptly) takes over from a more
objective narratorial stance. In Alcestis 569–605, the narrative subject
of Apollo’s labour for Admetus leads to reflection on Admetus’ present
hospitality and the chorus’ prayer that it will be rewarded; at Andromache
275–308, a narrative account of the judgment of Paris modulates into
counterfactual speculation about a world in which Cassandra’s advice
had been heeded and Paris done away with at birth.18
The identity and authority of these choral narrators follow well-
recognized patterns. Euripidean choruses are normally female, and the
male choruses in the extant plays19 are all geriatric (Alc., Hcld., HF ), if
17 On all these passages see Hose 1990–1991 (index locorum ad loc.); cf. Panagl 1971,
whose list of ‘dithyrambic’ stasima excludes the Andromache and Orestes odes but includes
the examples listed in the next note (excepting HF 348–441.).
18 Cf. Hec. 629–656 (Paris and the Trojan War, hence the present sufferings of
the chorus of enslaved survivors) and 905–952 (the fall of Troy, hence their prayer
for Helen’s present or future doom); El. 699–746 (the golden lamb of Argos, hence
Clytaemnestra’s crime and its present consequences); HF 348–441 (Heracles’ labours,
hence the present despair of his family and friends); Hel. 1301–1368 (Demeter and
Persephone, hence Helen’s neglect of their cult as the root of her present suffering);
Phoen. 638–689 (the story of Cadmus, hence prayer to Epaphus to save the land of his
descendants now); and IA 1036–1079 (the wedding of Peleus and Thetis contrasted with
the imminent sacrifice of Iphigenia).
19 Evidence for the identity of the choruses in the lost plays is marshalled by Hose
1990–1991: 22–27.
we discount Rhesus’ Trojan warriors and Cyclops’ generically obligatory

satyrs; aside from these last two, there are no young men or male slaves.
All are bystanders or victims, excluded from the processes of power
by age or sex, and in the latter case sometimes also by enslavement
(Hec., Tro., IT, Hel., Ion) or alien status (Phoen.). They generally align
closely with the interests of one party, especially if both the chorus and
a central character are female; this is especially the case in plays of
intrigue and conspiracy (Med. Hipp. Andr., Hec., El., IT, Hel., Ion, Or.),
with only the problematically authentic Iphigenia at Aulis following the
alternative Sophoclean pattern (Antigone, Electra) of a conspiracy hatched
in the prologue from which the chorus are excluded.
All these choral narrators exhibit a standard narrative competence
characterized by (1) unlimited access to earlier myth, whether or not
it is closely continuous with their own experience; (2) limited access at
the time of their first entry to information about the present state of
affairs in the city, based on hearsay rather than autopsy, though not
regarded as problematic on that account; (3) restriction of access there-
after to what is reported onstage, supplemented with highly unauthor-
itative and subjective speculation about offstage and future events. As
generally in tragic choruses, the narratorial persona is a complex fusion
of the communal ‘I’ of choral lyric (the Pindaric ‘I’ of the primary nar-
rator is not heard) with the awareness and focalization of their own
dramatic character.20 Only rarely is choral narration overtly addressed
to an onstage narratee. Rather, the chorus’ narrative stance conforms
to the conventions of lyric narrative, addressing itself to an unspecified
collective narratee: notionally fellow choreutae, but readily extensible, via
the chorus’ role as embodiments of the consciousness of the community,
to embrace the spectators in the theatre (→ Aeschylus and Sophocles).
Agonistic narration
Agonistic narrative is a particularly fluid type, merging readily into

other modes of discourse; the narrative element is strongest in Medea,
Hecuba, Heracles, and Troades, but as Lloyd notes, formal blocks of de-
marcated narration are the exception rather than the rule in Euripi-
20 See especially Gould 1996 (stressing the latter element) with Goldhill 1996 (count-
er-emphasizing the former).

dean agonistic rheseis.21 Instead, the regular pattern is for an argu-

ment to drop in and out of narration ad hoc. As Dubischar’s taxonomy
stresses, the pragmatics of the Euripidean agon involve a clear if flex-
ible structure of roles: there is always (1) a plaintiff; (2) a second party
who (depending on the type of agon) may be either an opponent or a
sympathetic sounding-board; and there is optionally (3) an umpire fig-
ure, who may or may not be the same as the second speaker. (In Andr.
and Supp. the plaintiff and opponent speak their cases to the judge sep-
arately.) The chorus is always present as on-stage (secondary) narratees,
and conventionally offers brief verdicts on the speakers’ cases, but it
is never the principal narratee, whose role defaults instead to the kritēs
where available or otherwise to the rival speaker, and only in Heraclidae
does the choral verdict seem to affect the action.
The engagement of narrators and narratees with the narrative ele-
ment in agones is determined principally by these roles. The narrative is
overtly focalized, and heavily punctuated with analytic argument. The
past exists to validate arguments in the present: to be raked over for
grievances and claims on duty, so that the commonest narrative trope
is a turn from Then to Now. For similar reasons, agonistic narrative is
particularly given to forensic counterfactuals, in either conditional or
interrogative form. Thus the core of Medea’s case against Jason (475–
491) is expressed in a narrative of her services to him and his disser-
vice to her, culminating in a complex counterfactual alternative under
which his behaviour might have been defensible. Jason then replies to
this narrative at 526–568 by systematically re-narrating the same events
with a different motivation, segmenting his retelling into a series of for-
mally itemised forensic responses to the principal accusations embed-
ded in Medea’s narrative catalogue. This agonistic exchange of roles
between narrator and narratee creates two versions of the same story
in which the narrating party is guiltless and all fault lies with the nar-
ratee. But both narrations are embedded in longer and more diverse
forensic rheseis which contextualize the function of narrative as merely
one among many modes of verbal contestation, and the construction
of rival narratorial identities and authority is itself part of this wider
agonistic discourse.
21 Lloyd 1992: 24–25, 42.

Narration apo mēkhanēs and other exodos narratives
Euripides’ predilection for the deus ex machina (ten in the extant plays,
if we admit Rhesus and disqualify Medea) and more generally for pro-
leptic narrative in the exodos sequence, gives his plays a highly distinc-
tive pattern of closure, discussed in detail by Dunn (1997). The deus
proper, marking the sudden arrival (often in the midst of crisis onstage)
of a new, omniscient voice from an essentially outside world, involves
a violent shift of perspective away from the pointedly restricted under-
standing of the characters into a vertiginous glimpse of lives and ages
to come, in which the story of the play becomes continuous with, and
aetiologically foundational to, the history of the audience’s own world.
As narrator, the typical Euripidean deus is omniscient, authoritative,
and beyond the reach of simple canons of moral judgment—whence
the debate, stretching back to Verrall, over how such figures can be
reconciled with the narrative authority of the human body of the
play. But they are also partisan and judgmental, supporting one side
against the other and sometimes taking sides in divine factions (Hipp.,
El.). Euripidean dei speak as though the futures they tell are already
written: ‘Long ago’, says Dionysus at the notorious Bacchae 1349 of the
events both within the play and beyond it, ‘my father Zeus assented
to these things’. Such omniscient narration of things to come runs
close to the modalized forms of proleptic speech-act such as command,
promise, and vow. The closing narrators are more than mere truth-
speakers; their very ability to speak with authority about the future
gives them a force of incontrovertibility which carries the weight of a
determinative order and takes immediate precedence over alternative
courses of willed human action.22
In three cases (Med., Hcld., Hec.) the exodos-narrator is not a god but a
mortal narrator who has been central to the conflict in a strongly bipo-
lar play, and now uses the authority of prolepsis in a final clash with
the surviving antagonist. The source of such figures’ sudden narrative
authority, with its epic roots in the clairvoyance of the dying Patroclus
and Hector, is not always clearly disambiguated in the text. Polymestor
attributes his eerie knowledge of Hecuba’s fate to ‘Dionysus, prophet
to the Thracians’ (Hec. 1267), and perhaps we are not encouraged to
wonder why Polymestor’s own blinding was not included in the god’s
22 Bushnell 1988 (contrasting the much weaker narrative authority of Euripidean
manteis); Barrett 2002: 50.

revelations, while Eurystheus’ foreknowledge is similarly due to ‘an old

oracle of Loxias’ (Hcld. 1028). But though Medea has proven super-
natural powers, she does not spell out the source of knowledge that
allows her so effectively to blur the distinction between what could be
mere statement of intent (1378–1380, 1384–1385), effective prediction of
an institutional commemoration (1381–1383), and outright prophecy of
the bizarrely improbable (1386–1388). This, combined with the strongly
partisan status of such narrators in a polarised, agonistic drama, gives
their final predictions an elusive, troubling force.
The narratees of the Euripidean exodos-narratives are so constructed
as to bring out the fundamental duality between primary and sec-
ondary audiences which is so fundamental to tragedy’s use of the dra-
matic medium to express a view of the world. The very content of the
prolepses underlines the contrast between the on-stage narratees for
whom these events are still in the future and the narratees in the theatre
for whom they are in the past. This is reinforced by the further concen-
tration on proleptic elements that connect to cult foundations or other
narrative traces left in the Athenian audience’s own world. Indeed, the
authority of the exodos-narrators comes primarily from the theatre audi-
ence’s knowledge that their fulfilment has the double certainty of what
has happened and what is still present. In this, as in Euripidean nar-
rative as a whole, the clash of perspectives between different narrators
and different levels of narratee seems designed to maximize, rather than
to disguise, significant contrasts in the way human experience is viewed
and interpreted, and even (as the studies of Segal, Dunn, and Barrett
suggest) to interrogate the forms of representation itself.
Conclusion
Narration in Euripidean drama is closely tied to the heightened mark-

ing of formal elements which characterizes Euripidean dramaturgy as
a whole: not just the formal prologues and ex machina speeches which
bookend the action with static narrative blocks, but the more highly
conventionalized stasima and formal debates within the body of the
play. The choices and handling of narrator are closely associated with
the distinctive Euripidean exploration of issues of power; between pro-
logue and exodos, and often in the prologue itself, the initiation of narra-
tion is overwhelmingly an activity of the disempowered, and narrative
an expression of and a substitute for the access to control of events
that eludes figures marginalized by gender, age, or status. But by the

same token, Euripidean drama is interested in modulations of narrative
sensu stricto into more determinative and empowering kinds of speech-
act: communication, persuasive argument, and above all the plotting
of intrigue, which provides a mechanism by which the very narra-
tor figures who have been marginalized in the play’s initial structures
of power can use their powers of narration, hitherto an expression of
disempowerment, to convert a sequence of events from a proleptically
rehearsed secondary narrative into primary narrative action.
chapter nineteen
ARISTOPHANES
A. Bowie
There has been a good deal of debate, as the Introduction (→) shows, as
to whether drama is a legitimate subject for narratological study.1 Since
embedded narratives are unproblematic candidates for narratological
analysis, it is on these that we shall concentrate.
Narrative
Aristophanic narratives do not involve complex play with levels of

narration, varied narrators etc., but the most salient feature that will
emerge is a general aversion from uninterrupted narrating. The audience
is not often permitted to wallow in the pleasure of a long oration whose
events take them away from the main plot: Aristophanes generally
prefers his narratives to be deferred, fragmented, or obscured. The
result is that there is a great richness in the modes of Aristophanic
narration.
He himself on occasion explicitly suggests the inappropriateness of
such orations to the genre. In Birds, Peisetaerus’ long (and much inter-
rupted) narrative of bird-history and other matters, is prefaced by the
following exchange (463–465):
Pe. Boy, bring a garland! Recline everyone! Someone quickly bring water
for my hands!
Eu. Are we going to have dinner, or what?
Pe. No, but I’ve long wanted to make a big, fatted speech.2
1 This uncertainty may explain the near absence of studies of Aristophanes in
narratological terms; Dunbar 1995 does however have useful remarks on the technique
of various passages of Birds, and see too Gelzer 1976. On comic plots in relation to
other genres, cf. Lowe 2000: 86–88; Silk 2000: 256–300 (‘Causal sequences and other
patterns’).
2 Cf. Thesm. 381–382 where, as the First Woman puts on her garland to speak, the
282 part four – chapter nineteen
Lengthy narration seems to require preparation and a certain amount

of self-deprecation to be tolerable.
At times, the level of interruption is such that it is difficult to decide
when a scene, which conveys a story through jokes, tomfoolery, debate,
etc., can reasonably be called a narrative and when not.3 Since some
of the more problematic and interesting passages tend to grow out of
more obviously narrative ones, I have not hesitated to include such
passages, so long as a narrative can be discerned, running through a
good deal of interruption and other interference.4
This is not to say that Aristophanes does not narrate: there are long
speeches containing narrative, but these are comparatively rare and
usually occur in situations5 or parts of the play, for instance prologues,
debates, and parabases, where such uninterrupted narrating is to be
expected. There are two notable examples in the first two extant plays,
and one in the last.
The longest uninterrupted and formal narrative in Aristophanes is
the Sausage-Seller’s account of his victory in the Boule (Kn. 624–682).
We are prepared for a long speech by the Chorus’ request ‘Tell us how
you fought the contest’ (614) and their remark that ‘I think I’d go a long
way to hear this’ (620–622), and then by the Sausage-Seller’s open-
ing reassurance that ‘It’s worth hearing’ (624). There follows a chrono-
logical account of events, linked by temporal markers, and without
attempts to break up the narration with intrusive jokes, references to
the audience, interruptions by other characters etc. It is a dramatic
narrative monologue, and its end is marked by the Sausage-Seller’s
complete triumph through his monopoly on coriander. This unusual
passage may perhaps be explained by the fact that up to this point
Knights has been a series of confrontational scenes, with quick-fire repar-
tee and very short speeches the norm: apart from the Chorus, no one
Chorus say: ‘Silence, silence: pay attention. She’s cleared her throat like the orators: it’s
going to be a long speech.’
3 This technique is reminiscent of Euripides’ ‘agonistic narration’ (→), but is devel-
oped a lot further.

4 There is an interesting example of highly fragmented narratives, of the expulsion
of the tyrants (274–282, 616–625, 631–634, 665–667) and of the Persian Wars (285, 675,
1247ff.), which crop up at various points in Lysistrata, but in too piece-meal a way to be
discussed in detail here: cf. Moulton 1982: 309–312; Bowie 1993: 178–204.
5 For instance, the debating speeches of the First Woman (Thesm. 383–432) and
of Mnesilochus (466–519) contain a good deal of narrative, as does the Stronger

Argument’s in Clouds (961–1023); the Weaker takes a more eristic and fashionably
elenchic line.
a. bowie – aristophanes 283
has spoken for more than six consecutive lines since the confrontation
began. Together with the parabasis that it follows, this speech thus pro-
vides some 180 lines of respite from the confrontations, which can then
resume in a similar manner. The speech maintains the confrontation,
but in a different form, as an analepsis. It is unusual too in its use of
quoted direct speech.
Dicaeopolis’ speech in self-defence before the Acharnians (Ach. 496–
555) also involves a lengthy narrative (509–557). Here the ‘reluctance’
to narrate comes out, not so much in the speech, as in the build-up to
it: there is a gap of 200 lines between the initial request to be heard at
292, and the actual start of the speech at 496. This deferral is produced
by a wide variety of devices. Dicaeopolis makes several attempts to start
his defence, offers to speak with his head on the block, brings out the
block, threatens the Acharnians’ coal-scuttle, persuades them to listen,
and seems at last about to speak at 366–367. This is a blind, however,
and he then speculates on the difficulties he faces, reminisces about
the past and pays a lengthy visit to Euripides to get a disguise. Thus
Dicaeopolis’ narrative is hedged about by various captationes benevolentiae,
which, along with the strong tragic colour in the Telephus intertext,6
the Euripidean rags and the tragic language, helps to mediate the
introduction of the unusually long speech. After this lengthy build-up,
the actual narrative of the war is somehow reduced in prominence:
200 lines of build-up lead to but 46 of narrative, and of those nearly a
quarter are an almost unrelieved series of 32 genitives, which bravura
passage mark the speech’s end.
Such long narratives are then largely absent until Wealth and Car-
ion’s narrative of events in the shrine of Asclepius. This is introduced
somewhat like a tragic messenger-speech and the scene runs for some
120 lines (649–770), with Carion speaking 105 of them: this is really
unusual in Aristophanic comedy. His speech is broken up into sections
of around twelve lines by one- and two-line remarks by Chremylus’
wife, but these remarks are far less disruptive than those we shall see in
other plays. In Menander (→), we shall find a greater willingness to use
such lengthy narratives.
One might expect straight narrating in prologues, but even here
Aristophanes dilutes narrating with jests and other diversionary tech-
niques. Apart from Acharnians, where a single figure lists his past plea-
6 For the role of the Telephus narrative in the play, cf. Foley 1988.
sures and pains, there is a basic pattern involving two characters,

who play roles of varying importance, and we can see a development
through Aristophanes’ career. In Knights, the prologue begins with two
slaves clearly in great distress, but their dialogue reveals little about the
precise causes. Eventually, Demosthenes asks: ‘Do you want me to tell
the story to the spectators?’ (36). Though his narrative is essentially
a straightforward, apparently chronological account of some 40 lines,
it tells its tale through a series of jokes, as can be seen from 40–45:
‘You see, we’ve got a master who’s rustic in his foul temper, a bean-
chewer and quick to anger, Demos of the Pnyx, a difficult old man,
who’s rather deaf. This chap bought at the last new moon a slave, a
leather-tanning Paphlagonian, a most villainous and slanderous fellow.’
Here, the audience can laugh first at the description of themselves, and
then at the realization that the mysterious ‘Paphlagonian’ mentioned
in line 2 is in fact Cleon, the leather-merchant and politician. Indeed,
Nicias has explicitly asked the audience to show their pleasure if they
like what they see and hear (38), thus actively encouraging laughter, and
so interruption of the narrative. Closure is brought simply by the use of
a hearty obscenity: ‘when the old man beats us, we shit eight times the
usual’ (70).
Wasps builds on this basic form. After initial jesting, Xanthias sud-
denly asks ‘Come on, let me tell the spectators the plot!’ (54). This time
however the narrative is postponed, first of all whilst Xanthias tells the
audience what they must not expect (Megarian jokes, Heracles, Euripi-
des, Cleon; 55–63); then by the information that the figure asleep on the
roof is his master, suffering from a ‘strange malaise’ (71); and finally by a
guessing-game where members of the audience are imagined as making
unavailing attempts to divine the nature of the malaise (73–85). Then
the narrative begins, again with an actorial remark ‘I’ll tell …’ (87).
At nearly 50 lines, it is long by Old-Comedy standards but, even more
than was the case in Knights, it is less a piece of straight story-telling
than a series of jokey vignettes. In the first part, seven jokes occupy
only 23 lines (91–113), and generally the narrative is broken up into
small segments of a maximum of three lines in length. In the second
part (114–135), Bdelycleon’s measures are ordered temporally by mark-
ers (‘first’, ‘then’ etc.), but the temporal relationships are less important
than the rhetorical: ever more extreme measures are required to cope
with Philocleon’s persistent ingenuity.
The prologue of Peace is of greater technical and comic complexity.
That a beetle is being fed we soon grasp, but not why. Food is pre-
pared before our eyes, but its consumption off-stage is narrated first
by a slave coming on-stage and then in simultaneous narration by his
companion looking off-stage into the house: such narration of contem-
poraneous action off-stage is found only here in Aristophanes.7 The
audience’s uncertainty about what is happening is articulated on-stage,
in a brief variant of the guessing-game in Wasps. Slave A then employs
an elaborate metanarrative motif to introduce the plot: ‘I will tell the
plot to the boys and to the young men and to the men and to the great
men and even to the very great men over there’ (50–53).8 The narrative
finally gets under way, but is interrupted after only six lines. On-stage,
the Slave quotes Trygaeus’ earlier words to Zeus, and is interrupted by
very similar words uttered by Trygaeus himself off-stage (56–64):
Sl. For days now he’s been gawping up at heaven like this, slandering
Zeus and shouting, ‘Zeus, what have you got in mind? Put down your
broom: don’t sweep Greece away!’ Ah, ah! Be quiet! I think I can hear
his voice.
Tr. [Off ] Zeus, what on earth are you wanting to do to us people? You’ll
destroy the cities, if you don’t watch yourself !
This technique of on-stage words imitating off-stage action is repeated
ten lines later. On remembering Trygaeus’ command to his ‘Little
Pegasus’ to fly him to heaven, the Slave looks inside again, and his
description of Trygaeus flying up is complemented by Trygaeus’ actual
appearance in the air: narrative and vision, on- and off-stage action
finally come triumphantly together, as the play gets under way with the
theatrical master-stroke of Trygaeus riding his Beetle.
Clouds is undated but almost certainly later than Peace. It moves away
from the two-slave prologue, but a main narrator still interacts with on-
stage characters (a slave or his son); the expository technique is more
sophisticated, if less dramatic than in Peace. Strepsiades introduces the
plot in two equal parts (1–40, 41–78). The first begins with an imitation
of the emotional beginnings to set narratives found in high-style poetry,
such as Homer (→) and Sophocles (→), as, to set the scene, he uses
apostrophes to gods, commands to slaves, monologic description of his
debts and exasperated debate with his son. The second part is more
explicitly narrative and is again divided into two by the brief entrance
7 There is simultaneous narrative of action on stage, in Xanthias’ description of
Philocleon’s dancing before and after the arrival of Carcinus’ sons (Wasps 1482–1515).
8 Comedy uses these motifs generally in a more self-consciously elaborate way than
tragedy (→ Aeschylus).
of the slave: the first part tells simply of the differences between his own
and his wife’s backgrounds; the second is a more elaborate account of
how their son’s name was chosen and of the future plans for him held
by father and mother, and involves quotation of the direct speech of
husband and wife. The pattern of diffuse and emotional reaction to
the current situation, followed by a more expository narrative section,
recalls the previous plays, but is less stylized. The audience are not
explicitly named as the narratees, but gradually become them: the
longer the soliloquy goes on and the more narrational it becomes, the
more it mutates into implicit audience-address. There are no explicit
markers of beginning and end in Strepsiades’ account.
In the prologues of Lysistrata, Thesmophoriazusae, and Frogs, Aristo-
phanes makes more use of the dialogic form of narration, much used by
Sophocles (→) and strikingly developed by Euripides (→). This is in fact
a development of the type discussed so far: there are again two charac-
ters on stage, but the subordinate character plays a much greater role
in the exposition, and there is no movement to a long narrative by the
dominant character.9 The last two plays also use the dialogic form, but
prefaced by a set speech by one character, that in Wealth (1–21) being
the kind of initial straight narrative of the plot up to this point that has
been eschewed elsewhere.
If the early prologues move from diffuse presentation of the situation
to somewhat more formal exposition, elsewhere in the plays Aristo-
phanes uses the opposite technique, as for instance in Peace. Once Peace
has been freed, the Chorus motivate the narration: ‘Tell us’, they say
to Hermes, ‘O god most kind, where she went when she was away
from us so long’, and he replies ‘O wisest farmers, listen to my words
if you want to hear how she was lost. First of all …’ (603–605). This
introduces a long scene (601–705) in which the history of the war is
reviewed. To begin with, Hermes adopts a straight narrational style,
articulated with the temporal markers ‘first …, then …, then …’, and
interrupted briefly by Trygaeus and the Chorus (615–618, 628–631). At
the mention of Cleon (648), things change. Trygaeus begs Hermes to
stop his account, and turns to Peace: ‘But, Lady, tell me why you are
silent’ (657). Anger at the spectators will not let her speak to them, so
Trygaeus suggests she whisper to Hermes. The narrative then becomes
9 The prologue of Birds may have been transitional between the two, but the
uncertainty over speaker-allocation at the start means we cannot be sure; cf. Dunbar
1995: 132–133.
a series of questions relayed by Hermes from Peace about the current

situation in Athens and the likely future, followed by answers from Try-
gaeus. In contrast to Hermes’ earlier speeches, no utterance is now
more than three lines long.
We can see something of a development over time in this technique
of consecutive passages of a decreasingly formal narration. In Ecclesi-
azusae 311–477, a dialogue between Blepyrus and the Man recounts the
preceding night’s strange goings-on, and then Blepyrus and Chremes
discuss the events in the Assembly. In the first scene (311–371), two
fifteen-line speeches by Blepyrus surround 30 lines of dialogue; in the
second (372–477), we find familiar features of disruption, as Chremes’
account is punctuated by Blepyrus’ questions, and there is a mixture
of longer and shorter speeches and much use of antilabē. The fragmen-
tation of the narrative, however, is much less intrusive than in earlier
plays, and the questions are not so much disruptive as designed to
amplify and give variety to it.
Finally, where substantial amounts of information need to be con-
veyed, Aristophanes again avoids lengthy narration, by conjoining sec-
tions that vary the amount of uninterrupted narration and combine
the two techniques just discussed. There is a complex case in Birds.
This play has to create its own mythological past and brave new world
in the future and, as a result, is notable for the amount of narrating
found in it. Peisetaerus’ lengthy account of past and future (465–626)
is divided into two sections. In the first (465–538), he outlines the shifts
in the Persian, Greek, Egyptian, Phoenician and Olympian monarchies
from bird- to men-kings. Though he covers an enormous spread of his-
tory, interruptions by Euelpides (to be discussed below) mean that his
longest uninterrupted utterance is but 21 lines and comes at the very
end (518–538); otherwise he does not speak more than five consecutive
lines before an interruption comes in. The subsequent long proleptic
recital of the establishment and benefits of bird-rule (548–626) begins
with another lengthy speech by Peisetaerus (550–569), broken by but a
single line of Euelpides (553), but subsequently the recital is very much
more broken up by questions and answers than was the earlier section.
Straight narrative, interrupted narrative and densely dialogical narra-
tive are thus all deployed to lend variety and complexity to the scene,
and much information is conveyed in as entertaining a manner as pos-
sible.
Narrators
Apart from in the parabases, where the chorus tell what the poet wishes
them to, all Aristophanic narrators are essentially secondary narrators.
As in tragedy, most narrators are also internal narrators: they are either
involved in the action or relating contemporary events. In the course
of the plays’ action, the narrators tell of events they have had a major
role in, but in parabases the chorus often mention events to which they
were more tangential: the Acharnians tell of the trial of Thucydides,
son of Melesias, (Ach. 703–712), though they were only witnesses. We
can include here those short narratives which the choruses sometimes
sing in lyrics (especially towards the end of the plays), attacking the
morals of individual Athenians; these are in some ways the comic
equivalent of the mythic narratives sometimes found in tragedy, and
their relation to the plot can be similarly oblique (→ Euripides). For
instance, the Knights sing of Lysistratus, Thoumantis, Arignotus and
the disgusting sexual habits of his brother Ariphrades (1264–1289), the
whole given a specious authority by being introduced by a high-flown
Pindaric quotation: ‘what is better when starting a song or bringing
it to a conclusion than to sing the charioteers of swift horses—and
not Lysistratus?’ These passages can also have a proleptic quality, as
when the Acharnians recount Dicaeopolis’ good fortune in terms of the
troublesome Athenians who will not now worry him (836–859), or curse
their former choregus to various fates for leaving them without dinner
(1150–1173).
The only major examples of external narrators occur in Birds, not
surprisingly given its great temporal range. We have already looked at
Peisetaerus’ account of world history, and the Birds themselves narrate
a theogony (685–703). Given that it goes back to the start of time, it
would be natural to consider it an external narration, but the Birds
manage to talk in such a way that they could be taken as internal to the
narrative: they stand as representative of birds throughout history.
Two types of narrative that are prominent in Sophocles (→) also
appear in Aristophanes, but are rare: narrations of things that char-
acters have only heard, and false narratives. It is only in Knights that
we find the former used significantly often, when Aristophanes mocks
contemporary politicians.10 The brief account of Cleon’s restricted abil-
10 There are two brief examples in the Eleusinian gephurismos in Frogs 426–428, 432–
434.
ity as a lyre-player is introduced by ‘his fellow students say …’ (985–

996), and hear-say introduces the micro-narrative of Cleonymus’ glut-
tony (1290–1299), and the narrative of the meeting of outraged Athe-
nian ships when they learn they are to be commanded by Hyperbolus
(1300–1315). One might have taken these as a self-defensive distancing
mechanism, disclaiming responsibility for the rumours, were the whole
play not a satire on Cleon: the defensive stance is thus revealed as a
mere comic pose.
The only significant example of a false narrative is Mnesilochus’
account of his own earlier life as a bride and of women’s supposed
sexual failings (Thesm. 466–519), which lands him in such trouble. There
are no false narratives which are used to shape the plot as in say
Sophocles’ Electra: Dicaeopolis may aim to deceive in his disguise, but
his narrative gives his version of the truth.
The one group who regularly narrate are the chorus, especially in
the parabases. In this part of the play, they often announce their inten-
tion to narrate with phrases such as ‘But, removing our cloaks, let us
get to the anapaests’ (Ach. 627), or ‘But come now you numberless mul-
titudes, make sure the excellent things we are about to say do not fall
uselessly to the ground’ (Wasps 1010–1012):11 the parabasis is thus marked
off as a special part of the play where the chorus can address the spec-
tators as their specific narratees.12 For the most part, the narratives are
fairly straightforward: usually chronological and articulated by tempo-
ral adverbs, they do not show especial narrative sophistication nor the
reluctance to narrate straightforwardly apparent elsewhere. The chorus
act mainly as internal narrators, but they sometimes dip into history,
as in Knights, when they praise their ancestors for their martial spirit
(565–580), or mythology, as in Birds (685–703).13 For the purpose of the
parabases, especially in the early plays, they tend to speak as a general
‘chorus’, rather than in character, though this is not true in plays from
Birds onwards. They use the first person, but, especially in the ‘paraba-
sis proper’, they speak on behalf of the poet in the third person: ‘ever
since he involved himself with comic choruses, our poet has not yet
come forward to the audience to say how clever he is …’ (Ach. 627–628;
11 Cf. Kn. 503–504; Peace 729–733; Birds 685–690; Lys. 615; Thesm. 785; Frogs 686–687.
12 On the parabasis and the relationships set up between poet and audience, cf.
Hubbard 1991.
13 For examples outside the parabasis, cf. Lys. 782–796, 805–820, introduced by ‘I
want to tell you a story …’

they never use second-person narration about the poet). In an interest-

ing variation, there are a few places where they use the first person,
but the ‘speaker’ is in fact the poet. This is the case throughout the
parabasis proper of Clouds (518–562), where they recount Aristophanes’
unhappy experiences with the first version of the play. The same tech-
nique appears with a sudden shift from third to first person in Peace,
where 732–753 are in the third person and 754–774 in the first; compare
Acharnians 656–664, where the metre changes when the person does:
He says he will give you much excellent instruction, to make you happy,
and he won’t flatter you or bribe you or deceive you by villainy or swamp
you with praise; rather he’ll give you the best lessons. So then let Cleon
contrive and plot against me all he wants: I’ll have right and justice on
my side, and you won’t catch me going round the city as a coward and
pervert like him!
This shift emphasizes the chorus’ role as ‘mouth-piece’ of the poet: the
voice of chorus and author fall together.
It is the chorus too, especially in the parabases, who provide the
rare examples of second-person narratives, addressed to the audience.
These tend to appear very intermittently in the more normal third-
person narratives about the poet.14 Though the poet is usually the focus
of the narratives, these parabatic passages are often introduced by ‘we
want to criticize you’ (memphomestha) or the like,15 so they are addressed
to the audience even if it is not usually the grammatical subject of the
verbs. There are two examples outside the parabasis, in Knights, where
first the chorus describe critically Demos’ political behaviour (1111–1121,
1131–1140; he replies with contrasting narratives of his own), and then
later Aristagoras gives Demos an account of what he was like before his
transformation (1335–1357; Demos shamefacedly accepts this). Indeed,
one could say that much of the contest between Sausage-Seller and
Paphlagon is an eristic exercise in second-person narrative, as each
contestant constructs a criminal history for the other.
In line with its reluctance to narrate uninterruptedly, extant comedy
generally does not, unlike tragedy, have narratives by characters whose
sole purpose is to speak a prologue or to bring news as a messenger.
The instance of the latter closest to tragic practice is the Herald in
Birds, who brings news of Athens’ reaction to the bird-kingdom (1277–
1307), but this is more a list of bird-like actions done and of bird-names
14 Cf. e.g. Ach. 637–638; Kn. 531–536; Clouds 532, 577–579, 607–624.
15 Cf. Ach. 676; Clouds 525, 576; Wasps 1016.
assumed than a straight narrative. The other two Messengers in that

play are not allowed to speak uninterrupted (1122–1187). The messen-
ger who brings news of Lamachus’ injury has a short speech with a
tragic flavour (Acharnians 1174–1189), but this fits his role of helping to
reintroduce the tragic intertext of Euripides’ Telephus, earlier applied to
Dicaeopolis but now to Lamachus. More usually, the characters them-
selves bring the news from off-stage. Xanthias’ recounts Philocleon’s
behaviour at the symposium (Wasps 1292–1325), and we have already
noted above the Sausage-Seller’s account of events in the Boule and
Carion’s lengthy account of the rites in Asclepius’ shrine. These charac-
ters are intimately involved in these actions, and thus breach the tragic
convention that ‘Whatever the degree of participation of the messenger,
he is never the protagonist of his own narrative.’16
As to the identity of narrators, we find that there is a general inclu-
siveness: social status and (to some extent) gender do not affect access
to narrating. We have already discussed the key role played by slaves in
prologues; later in the plays, narrating tends to pass to the main char-
acters. It is true that women narrate only in ‘women’ plays, but this
reflects their generally reduced role elsewhere.
Control of narration grants power to its possessor,17 and this is one
way in which comic protagonists are able to achieve their ends. At
the start of Acharnians, the Assembly is happy to listen in silence to
ludicrous narratives by the fraudulent Ambassadors and Theorus, but
suppress through the Scythian Archers Amphitheus’ attempt to use a
narrative of his genealogy to pursue a policy of peace (45–58), and
silence Dicaeopolis (59, 64). It is his narrative of the war (496–556)
that finally gives him power to enjoy his new world. Similarly, it is his
knowledge of history that allows Peisetaerus to impress and persuade
the Birds, and Lysistrata wins over Athenians and Spartans through
her (somewhat historically distorted) brief narratives of past Spartan-
Athenian co-operation (1124–1156).18
16 De Jong 1991a: 8.
17 Cf. e.g. Laird 1999: 1–43.
18 One may compare here Praxagora’s speech which persuades the women to follow
her plan (Eccl. 214–240): it is notable for a passage of ‘refrain narrative’, where an
account of women’s actions is punctuated at the end of each line by ‘as they’ve always
done’ (221–228). This is unusual, but cf. also 773–776, 799–803, 862–864 in the same
play.
Narratees
The main difference between tragedy and comedy in ancient Greece is

that where tragedy virtually never acknowledges the existence of the
primary narratees, the spectators in the theatre, comedy constantly
draws them into its action by direct address or invitation to interact
with the players. There is less ‘dramatic irony’ than in tragedy, where
the disparity between the knowledge of the audience and of the char-
acters is regularly exploited.19 On the other hand, comedy exploits the
relationships between narrators and the two sets of narratees in a vari-
ety of different ways for comic effect and to make the plays’ points.
Knights gives a particular twist to the existence of two sets of narra-
tees by having one of the characters called ‘Demos’. He is absent from
the first part of the play, but eventually arrives on stage, so that the
audience have to negotiate the fact that they are at once narratees of
the play and represented by an actor on stage. This doubling demands
from the audience a sophisticated response to the jokes. Similarly, in
Clouds, when the audience have enjoyed a sense of superiority at Strep-
siades’ inability to cope with philosophy in the first part of the play, they
are reminded in the parabasis (518–533) that this is exactly what hap-
pened when they watched Aristophanes’ ‘cleverest play’, the first Clouds:
they have therefore, this time unawares, once again been both narratees
and the victims of the humour.
Another special kind of narratee specific to comedy is the bōmolokhos,
the technical name for a character who constantly interrupts other
characters with more or less fatuous remarks. Such figures are not just
potentially disruptive of narrative, but also offer ways in which play
may be had with multiple narratees. In Acharnians, when the Ambas-
sador narrates his trip to the Persian court (65–125), Dicaeopolis acts
largely disruptively, but with a purpose. There are in this scene in effect
three narratees: the Assembly, Dicaeopolis and the spectators. Each
receives the Ambassador’s narration in different ways. The Assem-
bly listens in respectful silence. Dicaeopolis reacts sardonically, making
remarks which for the most part appear to be asides to the audience
19 But cf. Acharnians, where Dicaeopolis, in donning the disguise of Telephus, remarks
‘I must appear a beggar today, be who I am but appear not to be. The spectators must
know who I am, but the Chorus must stand there like fools, so that I can mock them
in what I say’ (440–444). The chorus then compliantly treat him as a beggar and not
Dicaeopolis.
since the characters on stage take little notice. Dicaeopolis’ remarks in

effect have two functions: to react to the speeches in the Assembly as
part of the play, and to mould the reactions of the spectators so that
they are not fooled and receive enlightenment about the faults of the
Assembly they attend. The narrators, the Ambassador and then The-
orus, thus interact in different ways with their narratees, only two of
which they are conscious of.
A further variation is found in Birds, which shows that the bōmolokhos
need not always be simply disruptive. In Peisetaerus’ account of world
history, Euelpides acts as a persistent bōmolokhos. The effect is that
there are two on-stage narratees (in addition to the audience), the cho-
rus, who are the main ones whose questions give rise to Peisetaerus’
account, and Euelpides, a subordinate one who does the interrupt-
ing. There are also in fact two narrators, also main and subordinate,
because some of Euelpides’ remarks also involve telling little stories
(492–498, 511–513). The interactions between these three actors are
quite complex. The first section (465–538) is motivated by the cho-
rus’ amazed reaction to the news that they were once kings of the
universe and by their excited questions. Peisetaerus’ narrative is then
interrupted four times by Euelpides with jokes, which he ignores. At
the mid-point, the chorus pose another incredulous question (500), and
Peisetaerus continues. Euelpides then interrupts with three more jokes,
which are again ignored, but his fourth intervention (517) finally draws
a response, and Peisetaerus shifts metre and concludes this part of the
narration. At another request from the chorus, he begins again with
a proleptic account of what they must do in future (550–626), the first
part of which (550–570) is interrupted by but a single exclamation by
Euelpides (553) and concluded in the same way. The subsequent part
(571–626) is again motivated by the chorus’ question, but is broken up
by no fewer than twelve questions and comments from Euelpides, with
which Peisetaerus now engages.20 As Peisetaerus becomes more persua-
sive therefore, he engages more and more with his narratees, and his
influence grows.
It is worth noting too that Euelpides’ remarks, though both regularly
disruptive and ignored, are not hostile, but can actually replicate Peise-
taerus’ arguments in a more comic form.21 For instance, Peisetaerus
20 Some editors give some remarks to Tereus, but this does not affect the point made
above: Dunbar 1995: 316–318.

21 On the agonistic aspect of this scene, cf. Dunbar 1995: 309–310.
illustrates his argument for the power of the cockerel from its name
‘Persian Bird’, and Euelpides interrupts with his story of the loss of
his cloak, for which a cockerel was responsible (483–498): narrator and
bōmolokhos thus function together. Euelpides may interrupt Peisetaerus’
discourse, but his enthusiastic reception of it generates in its warped
way an enthusiasm in the other addressees, on-stage and in the audi-
ence.
The most complex use of multiple narratees occurs in Thesmophori-
azusae, where much of the humour in the ‘Helen’ scene comes from
the existence of the Old Woman, who is guarding Mnesilochus and
plays the bōmolokhos in a special way (850–928). She intrudes herself
into the action as an unwelcomely uncooperative narratee alongside
the audience, when Mnesilochus/Helen enacts the expository prologue
of Euripides’ play (864–865):
Mn. Many souls perished by Scamander’s streams for me.
OW. And you should have too.
Matters are even more complicated when Euripides/Menelaus arrives,
at which point the clash between the reactions of ‘Helen’s intended
narratee ‘Menelaus’ and of the Old Woman gives the scene its humour.
‘Menelaus’ is told he has come to Proteus’ halls (874–880):
Eur. Of which Proteus?
OW. You fool, he’s lying, damn it! Proteas22 died ten years ago …
Eur. And to what land have we plied our ship?
Me. Egypt.
Eur. Alas! How far we have sailed!
OW. You don’t believe the nonsense this villain is talking do you? This
here is the Thesmophorion!
Here the roles of narrator and narratee become confused, as Menelaus
and Mnesilochus play out their tragedy for the Old Woman’s benefit,
but she refuses the role of narratee and in her turn narrates to them
the recent events in contemporary Athens. The Old Woman sees only
deceit in the clash of narratives and insists hers is the true one; the two
men equally insist on theirs—but the audience can enjoy the clash in
all its hilarious confusion.
22 The joke is helped in the Greek by the fact that ‘Proteus’ the mythical king and
‘Proteas’ the imaginary Athenian both have the same form of the genitive.
Thus, if comedy does not display some of the complexity of narration

that can be found elsewhere, it has its own enormous richness in the
interrelationships between narrating, narrators, and narratees.
chapter twenty
MENANDER
R. Nünlist
Menander’s narratives are best explained as secondary narratives be-

cause their narrators are characters who themselves are part of a ‘nar-
rative universe’ (→ Introduction and chapters on other playwrights).
Overall, Menander’s narratives show similarities to those of Euripides
and Aristophanes. With Euripides he shares a certain fondness for using
extended narratives, in particular in the exposition of the plays. Con-
versely, Menander also provides several examples of narratives which
are repeatedly interrupted by the interlocutor, as is common in Aristo-
phanes. Thematically, the majority of Menander’s secondary narra-
tives are closely related to the plot of the play itself: most of them are
instances of external or internal analepsis, occasionally of internal prolep-
sis.1
From a formal point of view, Menander’s narratives are best divided
into ‘monologic’ and ‘dialogic’ narratives. The former category as-
sumes that no secondary narratee is physically present on stage. The
story is narrated, as it were, on an empty stage, or, as an alternative,
the secondary narrator is unaware of another character’s presence, and
this character therefore cannot be the narratee. Dialogic narratives
are spoken to one or several narratees on stage. Here again, two
subtypes can be differentiated, though along somewhat different lines.
Dialogic narratives can be heavily punctuated as in Aristophanes, or
the secondary narrator can be allowed to recount larger sections of his
narrative without interruption.2
1 Given the fragmentary status of the Menandrean corpus, some of the observations
in this chapter are based on conjecture. As a general rule, evidence from Latin adap-
tations of Menander’s plays has not been used because one cannot be sure how closely
they follow their Greek originals. Line numbers refer to the OCT (2nd ed. 1990), quo-
tations are taken from Miller 1987.
2 Needless to say, the difference is one of degree, not of kind.
298 part four – chapter twenty
Monologic narratives
The most prominent of Menander’s monologic narratives are the pro-

logues, which can be found (or presupposed) in all his plays. The pro-
logue either opens the play or is preceded by an opening scene.3 It is
either spoken by a divinity or by one of the human characters.4 In the
former case, the narrator is ominiscient and external.5 Human narra-
tors, on the other hand, are internal and their knowledge is restricted
to what they are entitled to know.6 In the case of plays with postponed
prologue, this leads to a more or less equal distribution of labour. The
information provided by the human character(s) in the opening scene
is supplemented by the divine prologue. The latter can supply informa-
tion to which the human characters are not privy (e.g. the citizen status
of an apparent non-citizen), and the divine speaker can make prolepses
the outcome of which is certain. Consequently, the prologues regularly
observe the chronological sequence ‘external analepsis–internal analep-
sis–internal prolepsis’. In general, the prologue speaker avoids going a
second time in detail over ground already covered. See e.g. the explicit
cross-reference to the opening scene in the prologue to Perikeiromene:
But Someone Else [i.e. the soldier Polemon] was coming up the road,
and he saw what was happening [i.e. Glycera hugging her brother Mos-
chion]. He’s told you the sequel, how he marched off …
(Perik. 157–159)
As an alternative, the divine speaker may expressly correct the informa-
tion given by a human character (e.g. Asp. 110–112: ‘Davus here made
a mistake. Cleostratus … is alive’). The extant evidence suggests that
divine prologue narrators are the rule, human narrators the exception.
The audience of New Comedy is expected to know beforehand the
correct details of the exposition and the imminent happy ending of the
3 Opening prologue: Dysk., Sam.; postponed prologue: Asp., Epitr., Heros, Mis., Perik.,
Sik. On Menander’s prologues in general see Holzberg 1974: 6–113.

4 Divinity: Asp., Dysk., Epitr. (conjectural), Heros, Perik., Sik., Ph. Human character:
Sam.
5 The statement that the divine narrators of the prologues are external needs some
qualification. Mostly, they make clear that they somehow set in motion the action
on stage (e.g. Asp. 148; Dysk. 44; Perik. 164–166). Consequently, they are not entirely
external, but they do not enter the narrative universe and the human characters
remain unaware of their intervention. It therefore seems preferable to treat the divine
narrators as external, particularly in the case of personifications: Chance (Tuche) in Asp.;
Ignorance/Misapprehension (Agnoia) in Perik.
6 The distinction made here applies, of course, to dialogic narratives too.
r. nünlist – menander 299
play. In the case of human prologue speakers, however, the audience

will not have been left completely in the dark due to generic expecta-
tions. As for other generic expectations, divine and human prologue
speakers seem to differ in the following way. Whereas the entrance
and the narrative of the former seem to be taken for granted, the only
extant example of a human prologue speaker justifies his narrative with
two metanarrative comments:
Telling the story [sc. how he raped the neighbour’s daughter Plangon]
will be painful, I reckon, but it will make more sense to you if I explain
in some detail what my father’s like. … Then—I’ll tell you all about us
at one go, I’ve nothing else to do— … (Sam. 5–6, 19–20)7
Monologic narratives are spoken on an empty stage with no secondary
narratee physically present or presupposed by the narrative. In Menan-
der this situation quite often leads to what is, strictly speaking, a breach
of narrative convention: the secondary narrator addresses the primary
narratees (i.e. the audience in the theatre).8 This device, which entails
vocatives and verb forms and pronouns in the second-person plural, is
common in prologues, but can also be found in other monologic nar-
ratives.9 These addresses can be part of the narrative itself, but more
often they mark its boundary at the beginning or the end, e.g. the end
of the prologue to Dyskolos, which includes a metanarrative comment:
There, that’s the outline. Details you’ll see in due course, if you like—and
please do like. (Dysk. 45–46, cf. Sik. 24)
The prologue narratives are comparatively long (approximately 50
lines), and there are a few instances of explicit cross-reference (‘as I
said before’: Asp. 130; Perik. 152).
Monologic narratives other than prologues are quite common in
Menander.10 Mostly a character gives a report of what had happened in
the meantime elsewhere (i.e. internal analepsis), either backstage or off-
stage.11 Unlike the prologue speakers, who are always alone on stage,
7 The same prologue contains yet another metanarrative comment: ‘I hesitate to
tell you the rest of the story’ (Sam. 47). Clearly Moschion is a self-conscious narrator.
8 This breach of narrative convention, though in a different direction, can be docu-
mented from Homer’s time on: the primary narrator addresses a character (→ Homer).
The present type seems to originate with Old Comedy (→ Aristophanes).
9 Prologues: Asp. 113; Dysk. 1, 46; Perik. 127–128, 170–171; Sam. 5; Sik. 24; other
monologic narratives: Dysk. 666; Epitr. 887; Sam. 216, 269. Obviously, the device also
occurs outside monologic narrative; see in general Bain 1977: 185–207.
10 On Menandrean monologues in general see Blundell 1980.
11 Examples include: Asp. 149–153 (a variant insofar as Smikrines narrates what he
other monologic secondary narrators may actually be overheard by

other characters. This sometimes leads to punctuation of their narra-
tives, of which they themselves remain unaware. A particularly strik-
ing example comes from Misoumenos where Getas remains unaware of
Clinias’ remarks for 40 lines until he is finally addressed by him (284–
323).
Apart from these punctuated monologic narratives, there are in-
stances of uninterrupted narrative too. Probably the best known exam-
ple is the monologue by which Demeas opens Act Three of Samia. In
a speech of no less than 77 lines (206–282; the beginning is lost), he
narrates (219–269) what had just happened inside the house when he
had to learn that the father of the baby is actually his adopted son
Moschion. Demeas is mortified (he wrongly believes the mother to be
his concubine Chrysis), but that does not seem to affect his narrative
and histrionic skills in the least. For the central piece of his narrative is
an elaborate instance of ‘speech within speech’. It combines two differ-
ent speakers neither of whom is Demeas himself (two female servants,
one old, one young).12 Demeas’ monologue is, however, exceptional in
terms of length and elaboration, and more representative, perhaps, is
an internally analeptic monologue like Onesimus’ opening of Epitre-
pontes Act Three:
Half a dozen times I’ve embarked on the business of going to my master
and showing him the ring: and half a dozen times I’ve got very close to
it, been right on the brink of it—and then I funk it. (Epitr. 419–422)
Another monologue proves Onesimus to be a highly self-conscious
narrator:
He’s mad, I swear it, quite loopy, really raving, absolutely crazy! My mas-
ter, I mean, Charisius. He’s fallen into black depression, or something
like it. There’s no other explanation. You see, he was crouched inside
the door here for ages, listening to his wife’s father discussing the whole
business with her, apparently, and I can’t tell you, Ladies and Gentlemen,
how he kept changing colour … (Epitr. 878–887)13
did not do in the meantime); Dysk. 259–265, 525–542, 576–586, 670–688; Epitr. 419–422,
879–904, 911–919; Mis. 284–310; Perik. 537–550; Sam. 120–127, 219–269. An example of
external analepsis is Perik. 300–301.
12 On this scene and on speech within speech in general see Nünlist 2002. Despite
the prominence of this device, Menander does not provide an example of tertiary
narrative.
13 The narrative continues for 20 more lines with regular quotations of Charisius’
words.
Dialogic narratives
As explained above, the narrative situation is different here because

the secondary narrator is speaking to one or several narratees who
are present on stage. Consequently, many dialogic narratives are dom-
inated by the interaction between narrator and narratee. For example,
the narrative is often the result of a question or a similar request for
information by the narratee (Asp. 19–20; Dysk. 82–86, 409; Sik. 125–
126, 172). As an alternative, the narrator may introduce his narrative
either with a rhetorical/exclamatory question (Dysk. 576; Sam. 216) or
by means of a metanarrative comment (‘I want to give you a taste of
good news or rather, God willing, of good fortune to come. I want to
be the first to tell you.’ Georg. 43–45; cf. Perik. 780).
The narratee’s opening question indicates his lively interest in the
matter. It is, therefore, not surprising that he then regularly interrupts
the narrative (cf. chapter on Aristophanes). A good example comes
from Dyskolos where Pyrrhias runs onto the stage and reports his dis-
astrous encounter with Knemon. The punctuation of the narrative
is in a way ‘advertised’ beforehand because Pyrrhias begins with the
telling metanarrative comment (96–98): ‘I can hardly talk, I’m so out of
breath. Well, I knocked at the door, and asked to see the owner …’14
No surprise, then, that Chaireas interrupts him more than once.
Pyr.: A miserable old crone answered the door, and from the very spot
where I stand speaking to you now, she pointed him out. He was trailing
around on that hill there, collecting wild pears—or a real load of trouble
for his back.
Ch.: He’s in a proper tizz. [To Pyr.] So, my friend …?
Pyr.: Well, I stepped on to his land and made my way towards him. I was
still quite a way off, but I wanted to show some courtesy and tact, so I
called to him and said, ‘I’ve come to see you, sir, on a business matter. I
want to talk to you about something that’s to your advantage.’ But ‘You
horrible heathen,’ he promptly replied, ‘trespassing on my land! What’s
the idea?’ And he picks up a lump of earth and lets fly with it, right in
my face.
Ch.: The hell he did.
Pyr.: And while I had my eyes shut, muttering ‘Well, God damn you,’ he
14 This is a variation of the topical ‘aporia’ motif. Pyrrhias’ problem is not the wealth
of information he is to choose from, but his (alleged) inability to speak without inter-
ruption.
picks up a stick and sets about me, saying ‘Business is it—what business
is there between you and me? Don’t you know where the public highway
is?’ And he was shouting at the top of his voice.
Ch.: From what you say, the farmer’s a raving lunatic.
Pyr.: To finish my story: I took to my heels, and he ran after me for
the better part of two miles, round the hill first, then down here to this
wood. And he was slinging clods and stones at me, even his pears when
he’d nothing else left. He’s a proper violent piece of work, a real old
heathen. For goodness’ sake, move off!
Ch.: Chicken! (Dysk. 98–123)
In a similarly instructive example from Perikeiromene, the punctuation
and the request for information lead to what almost looks like a cross-
examination:
Pataecus: God, is something still left of my family?
Glycera: Go on, ask me anything you like.
Pat.: Where did you get these things [i.e. the tokens of recognition]? Tell
me.
Gl.: I was found wrapped in the embroidery, when I was a baby. …
Pat.: Tell me, were you alone when you were found?
Gl.: Oh no, my brother and I were exposed together …
(Perik. 779–786)15
This externally analeptic ‘narrative’ of Glycera’s lineage (Perik. 779–812)
is given in almost perfect stichomythia, and in the course of it Pataecus
becomes the narrator and Glycera the enquirer. On a different level the
function of the stichomythia is paratragic, which is, however, undercut
by the presence of Glycera’s brother Moschion who comments on
the ‘narrative’ in the form of asides and thereby punctuates it even
more.
Similarly to beginning the narrative in answer to a request for infor-
mation (see above), the narrator may open his narrative with the par-
ticle gar.16 In a unique case, the narrator opens his narrative with the
‘there is a place x’ motif (Asp. 23).
The narrative itself usually proceeds in chronological order. The
sequence is regularly made clear by temporal markers like ‘after this’,
15 For a similar but much shorter example see Dysk. 409–417.

16 Georg. 46; Dysk. 525, 597, 670; Heros 21; Perik. 318, 537; Sam. 219. The list combines
instances of initial gar in monologic and dialogic narrative. See in general De Jong 1997.
‘then’, etc. A recurring feature is to announce the final step of a narra-

tive with a short metanarrative comment which implies the omission of
part of it (to peras, ‘to cut a long story short’: Georg. 49; Dysk. 117 [quoted
above]; Epitr. 287, 533, 891; similarly Kith. 92). Finally, the narrator may
expressly state that he has reached the end of his story (Asp. 82; Epitr.
292), or, alternatively, an address to the narratee can indirectly indicate
the end of the narrative (Dysk. 123).
Although Menander contains a good number of heavily punctuated
narratives reminiscent of Aristophanes, dialogic narrative does not a
priori exclude longer sections of uninterrupted narrative. By far the
longest example is the messenger-speech from Sikyonioi with no less
than 96 uninterrupted lines (176–271), which is likely to owe its exis-
tence again to paratragic purposes. After an opening dialogue (169–
175) known from tragedy, the messenger explains how he got to know
the events (176–186) and identifies himself as one of the large crowd
who witnessed the scene (188–192) and commented on it with clamour
and sharp remarks (197–199, 202, etc.).17 Apart from initial references to
‘us’18 the messenger recedes more and more into the background and
lets the events ‘speak for themselves’.19 He resurfaces when he reports
how he left the scene (270–271), which ends his speech and his presence
on stage.
The uninterrupted narrative is exceptional but not without paral-
lel. In the arbitration scene that gives the Epitrepontes its title, the shep-
herd Davus is allowed to present his case over a stretch of more than 50
lines with only three short interruptions.20 The explanation is, of course,
generic: forensic speeches must not be interrupted by the opponent.
But with his usual care for well-wrought plots, Menander subtly moti-
vates Syriscus’ silence. His early attempt to interrupt Davus’ speech is
thwarted by the arbitrator, who threatens physical violence in the case
of a repetition (‘If you interrupt, I’ll wallop you with my stick,’ 248–
249). Syriscus is so intimidated by the threat that he only speaks when
17 Unlike his tragic counterparts, he even identifies himself (187–188), but the details
are disputed among scholars (see Belardinelli 1994: ad loc. with bibliography).
18 Sik. 197, 202, 210, 212. It is worth noting that he later refers to the crowd by ‘they’
(244, 245) or simply quotes their remarks (223, 239, 257, 264–266, 269).
19 Minor counter-examples are an evaluative remark (244) and an oath (216).
20 Epitr. 240–292, interrupted in 247–249, 270 (text and distribution of speakers
uncertain), and 274. Davus’ opponent, the charcoal-burner Syriscus, is given a similar
amount of text (almost 60 lines: 294–352), but only a fraction of it is actually narrative.
prompted (270, 274), and at the end he carefully double-checks whether

Davus is indeed done with his speech (293), despite the latter’s closing
formula ‘My case rests’ (292).
Interestingly, there is a similar case in Georgos (54) where Myrrhine
urges the old slave Philinna not to interrupt Davus’ narrative. Con-
versely, Onesimus’ threat to recite a complete rhesis from Euripides’
Auge to Smicrines (Epitr. 1125–1126) may not only be indicative of the
latter’s obtuseness, but may also express general reservations about
uninterrupted speeches. Overall, however, Menander does not seem
to object to longer stretches of uninterrupted speech, and they often
include narrative.
As was mentioned above, most of Menander’s narratives are imme-
diately related to the play itself in that they are either analeptic or pro-
leptic. The only exceptions to this rule are a few mythological exempla
(mostly from tragedy), which are used in order to explain or illustrate
a point of the main plot. All these examples are short and presume
general familiarity with the narrative in question. As a rule, the pur-
pose of the exemplum is humoristic, e.g. when Demeas tries to explain
the pregnancy of Plangon to her father Niceratus.
Dem.: Tell me, Nikeratos, have you never heard actors in tragedies
telling how Zeus once turned into a stream of gold, flowed through a
roof and seduced a girl who’d been locked up?
Nic.: So what?
Dem.: Perhaps we should be prepared for anything? Think! Does any
part of your roof leak?
Nic.: Most of it does. But what’s that got to do with it?
Dem.: Sometimes Zeus is in a shower of gold, sometimes a shower of
rain. Do you understand? This is his doing. How quickly we’ve found the
solution!
Nic.: You’re having me on.
Dem.: Heavens, no! Wouldn’t dream of it. You’re surely just as good as
Danae’s father. If Zeus honoured her, then perhaps your daughter –
Nic.: Oh, dear. Moschion has made a cake of me. (Sam. 589–599)
The general preponderance of analeptic narrative in Menander has the

effect that the main function of secondary narrative is explanatory. But
other functions can be made out too, especially in the case of dialogic
narratives. Pyrrhias’ narrative (quoted above) has a clear persuasive
function (‘let’s go home’). At the same time it nicely characterizes
Cnemon, who has yet to appear on stage, as the grumpy old man of
the play’s title and exposes Pyrrhias’ own qualities as a vivid narrator.
Similarly, Habrotonon’s proleptic narrative (Epitr. 511–535) is to enlist
Onesimus’ support and at the same time characterizes her as a clever
young lady with good intentions.
In conclusion, long uninterrupted stretches of narrative as in Euripi-
des and heavily punctuated narratives as in Aristophanes coexist in
Menander. Whereas the latter tend to be dialogic, the former category
comprises both monologic and dialogic narratives. In terms of content,
Menander’s secondary narratives are closely connected with the main
plot of the comedy, most often in the form of an analepsis, which helps
to explain the action on stage.
chapter twenty-one
LYCOPHRON
N.J. Lowe
Imagine a truly omniscient narrator. Imagine that your knowledge of

the world is unbounded by space and time: that you, alone among
mortals, can see the patterns of cause and effect linking all things
in heaven and earth from the beginning of the world to the end of
history. From where you stand, a powerless figure close to a centre
of power, you can see ripples of consequence spreading out to the
horizons of destiny. You see the story of the world extending through
you and beyond, a river of time passing beneath your feet and out
to a boundless ocean. You know that you can alter nothing, convince
nobody, avert none of what you see; that your own life will be short,
brutal, unfulfilled, and that soon enough every last member of your
family will be slaughtered or enslaved, but that even so the sufferings
of generations to come will dwarf the tragedy of the present. And you
have a power of language beyond any other: power to make new words,
new riddles in which to tell your story. You know that your words
will be remembered, reported, pondered, now and in generations yet
unborn. What do you say? What kind of story will you tell?
The Alexandra of Lycophron attempts an answer to this most daring
of narratological challenges. The result is a text of ferocious difficulty,
uncertain critical repute, and dizzying narrative paradox:1 an unstage-
able dramatic fragment in the form of a play-length messenger-speech,
reporting verbatim a vast proleptic monologue of impenetrable riddles
guaranteed to be disbelieved even if deciphered. Its much-discussed
syncretism of genre marries epic, tragic, and lyric thrēnos with the man-
tic voice of lyric and tragedy to create Greek narrative’s longest and
most ambitious essay in prior narration, with major intertexts including
not only the Aeschylean Cassandra and her important lost counterpart
1 The key treatments of Lycophronian narrative are Fusillo 1984 and West 1984
(and cf. Fusillo–Hurst–Paduano 1991 and West 1983, 1991/2, and 2000); see also
Fountoulakis 1998 and Schade 1999.
308 part four – chapter twenty-one
in Euripides’ Alexandros, but the epic laments of Briseis, Hecuba, Helen,

and especially Andromache (who anticipates Lycophron’s Cassandra in
her distinctive proleptic penchant for mourning what has not yet hap-
pened).
The Alexandra’s central idea is that Cassandra’s prophecy should be
presented not diegetically but dramatically, with no overt primary nar-
rator (‘Lycophron’ does not speak) but within a nested frame of mul-
tiple speakers and audiences: in particular, that there should be narra-
tees within the world of the poem for whom the events narrated are
in the future, unintelligible, and disbelieved, and an audience outside
the poem for whom the proleptic content is past, understood, and con-
firmed. The poem’s main narrative (31–1460) is reportedly the direct
speech of Cassandra herself, as committed to memory by a peculiarly
memorious slave and reported verbatim to Priam in the dramatic frame.
Equally importantly, however, it is a narrative that addresses itself to
a reader beyond this frame, for whom everything that is incredible to
Priam and his slave is history fulfilled, and incomprehensible ravings
become riddles with determinate solutions. But the most remarkable
feature is the Alexandra’s extension of this narrative nesting to four or
more levels, each characterized by a different twist on the relation-
ship between narrator and narratee, and inviting the poem’s readers
to reflect on the symmetry between levels and its implications for larger
issues of narrative authority (as embodied in the fourfold narrators) and
interpretative hermeneutics (as practised by their respective narratees).
level narrator narratee narrative

quaternary Apollo Cassandra inspiration
tertiary Cassandra [none] prophecy
secondary slave Priam messenger speech
primary ‘Lycophron’ Ptolemaic reader the Alexandra
Apollo, Cassandra, and the narratology of inspiration
The status of Cassandra’s narrative is complicated from the start by

its source in mantic inspiration. Unlike her Aeschylean counterpart,
Lycophron’s Cassandra does not herself explicitly credit Apollo as
source of her story; she speaks only of her rejection of his sexual atten-
tion (352–353, 1457) and his retaliatory theft of her pistis (1454–1457).
n.j. lowe – lycophron 309
Nevertheless, the slave introduces her speech as entheon (28), and it is

Apollo himself who will guarantee the fulfilment of her prophecies
(1458), so we are apparently to assume that her knowledge emanates
from the god. Yet the key thing about Apollo as a narrator is that he
does not, strictly speaking, narrate; rather, he instils narrative content
directly into the consciousness of its recipient. In effect, this is narra-
tive unmediated by any form of actual narration, and part of Cassan-
dra’s problem is that the fabula instilled in her—the whole of human
history—has no intrinsic or pre-formed narrative shape; it is up to her
to give it one, and both the intractable vastness of the story and her
own deficiencies as a storyteller result in an attempt that is less than
lucid to its various narratees. It is never, in fact, explicit how far it
is the god, rather than his prophetess, who is principally responsible
for the organization of raw history into a story (the Herodotean war
of Europe and Asia) with a beginning (the rape of Io), a middle (the
destruction of Troy and its aftermath), and an end (the final settlement
in Lycophron’s own times by the enigmatic figures of the Lion, the
Wolf, and the Wrestler2). One striking passage (229–257) is marked by
frequent deictic expressions, present tenses, and first-person references
to seeing the scenes she is narrating—suggesting that the events she
describes here are unfolding before her eyes in real time as the god
prompts, and that her narration is an improvised live commentary on a
hallucinatory montage assembled and edited by Apollo.
Nor is it clear whether the story Cassandra tells is the whole of
the story implanted by the god. Cassandra’s speech stops short at
the point when a sudden flaring of self-consciousness drives her back
into physical and communicative isolation at 1461. This retreat into
silence is apparently not the end of her utterance, which continues as
an interior monologue and may, for all we or the slave know, have
extended her story far beyond its timespan in audible speech:
So much she spoke aloud, and hurrying back in her steps
she went inside her prison. But in her heart
she wailed the last of her Siren song
like a Mimallon of Clarus [i.e. an Apolline prophetess], or a channeller
of [the Sybil] Melancraera
Neso’s daughter, or a Phician monster [Sphinx]
incomprehensibly gibbering her tortuous words. (1462–1467)
2 The principal identifications are conveniently tabulated by Hurst in Fusillo–
Hurst–Paduano 1991: 25.

Finally, it is not even certain how reliable a narrator Apollo is, or how
much of his story its recipient understands. West intriguingly suggests
that ‘Cassandra’s second sight is restricted to what the god cares to
reveal’, and that Apollo has withheld, for example, the fact that Ajax’s
rape will actually be thwarted, leaving Cassandra to assume the oppo-
site.3 If this is so, the reader is already enlisted in the completion of
major interpretative gaps in the narrative that elude even Cassandra—
though there is no extant tradition of a rescue,4 and it must be doubted
whether even Lycophron would wish to stake so much of the inter-
pretability of his text on his readers’ competence to fill a single ellipsis.
Cassandra: the mantic narrator
Whether or not we accept this stronger claim, West is surely right

to see touches of characterization in some elements of Cassandra’s
narration: a history skewed by the perspective of an eponymously
anti-sexual figure ‘temperamentally inclined to see eros as a perpetual
source of trouble’.5 More daringly, West argues that even the causal
pivot of Cassandra’s world history, her attribution of a central causal
role to Ajax’s assault, may itself be a subjective distortion imposed by
Cassandra’s own troubled outlook on history—that the shape of cause
and effect would look quite different if the story of the world had a
different narrator. As West points out, the sorrows of the Trojans cannot
all be attributed to Ajax’ crime as Cassandra claims; the sins of the
Trojans, and especially the crime of Paris, seem played down, though
they are the cause of the war and the city’s destruction, as indicated by
the setting of her prophecy.
Cassandra is of course an overt, internal narrator—present both
as a first-person narrating figure and as a character in the story she
tells. The future she narrates, though large, is centrally her own, and
is organized accordingly. Her story has four phases: the sorrows of
Troy (31–364), the sorrows of the Greeks (365–1282), the conflict of
Europe and Asia (1283–1450), and a short reflective conclusion (1451–
1460). But each in turn culminates in the story of Cassandra herself:
the first in her rape by Ajax (348–364), which then frames the tale of
3 West 1984: 149 n. 105.

4 See Gantz 1993: 652 for sources and a cautious assessment.
5 West 1984: 147.
the Greek sufferings that avenge it (365–386 and 1281–1282); the second
in her death and her posthumous vengeance on Ajax’ homeland by the
thousand-year tribute of the Locrian maidens (1108–1173); the third in
the emergence of the Wrestler from her lineage; and the last in the final
recognition of the truth of her prophecies (1458–1460).
Cassandra’s situation at the moment of narration is very specifically
sketched: the occasion is the dawn departure of Paris’ ships to Greece
(16–22), the place is just outside Cassandra’s sea-cave cell on the hill of
Ate (29, 349–351, 1462, 1469). Yet it is a mystery to whom, if anyone,
Cassandra thinks she is talking. She herself describes her audience as
‘deaf rocks, mute wave, and fearsome ocean’ (1451–1452), an elemen-
tal audience incapable of either hearing or speech, since no human
can understand her utterance (1454); she does not acknowledge, and
perhaps does not register, the human presence of Priam’s slave, whose
invisibility to any but the reader recalls the treatment of silent atten-
dants in drama. But her actual narration is characterized by bursts of
apostrophe—first to the city of Troy (31, 52, 69, 72), then to Paris (90–
146, the longest of several sustained passages of second-person narra-
tion), and sporadically thereafter6—evoking an eerie sense that some of
the time, at least, she is addressing herself to her own vision and a series
of figures seen within it.7
Yet Cassandra is also a highly self-conscious narrator, aware that her
words are true but unbelievable, and that responsibility for this lies
outside her own control and (paradoxically) with the very source of
her own narrative competence. She anticipates that her prophecies will
be remembered in sufficient detail long enough for their truth to be
recognized in time:
But he [Apollo] will make my words true. And folk will understand it to
their sorrow
When there will be no more means to help my fatherland,
And they will praise the Phoebus-possessed swallow [i.e. me].
(1458–1460)
6 To Prylis (221); Cassandra herself (258–264); Hector (280, 300, 1189–1216); Troy’s
enemy daimōn (281); Troilus (307–313); Hecuba (315, 330–335, 1174–1188); Polyxena
(323–329); the mountains of Euboea (373–378); Odysseus (815); Egesta (968–973); Setaea
(1075–1082); ten Locrian landmarks and peoples (1146–1154); Troy (230–231). Full-
scale second-person narration is particularly prominent for close family members; the
striking exception is Priam, any apostrophe to whom would have created a perhaps too
eerie effect when her words were repeated by the slave.
7 Comparable situations in Euripidean (→) prologues.
From the moment of its performance, therefore, Cassandra’s narrative

is conceived as a memorized text rather than a communicative utter-
ance, a speech without a living audience, frozen instead for subsequent
iteration and study by a variety of narratees not physically present at
the time of narration.
Priam’s slave: narrative as drama
Nor can we ever be quite sure whose words we are even reading, thanks
to the notorious problem of determining where Cassandra’s voice ends
and the slave’s begins.8 It is often remarked that, despite the clear mark-
ing of the formal boundaries (30, 1461), the slave’s style is not noticeably
differentiated from Cassandra’s own, though this is something of a sim-
plification;9 the first fifteen lines remain broadly within the lexical and
stylistic range of tragic diction, eschewing the neologisms, compaction
of metaphor, and mythological cryptograms that distinguish the body
of the text. It is only at 16 that we meet with a marked change of
register, as a sudden barrage of proper names introduces the geog-
raphy, genealogy, and mythological background of Troy—all of this,
however, familiar and readily intelligible to the secondary narratee,
if extremely demanding on his primary counterpart(s), or indeed any
latter-day reader unequipped with commentary. The one moment in
the prologue where the slave’s idiom becomes indistinguishable from
Cassandra’s is the astonishing description of the boats hitting the water
at 22–24;10 this level of gobbledegook is briefly mirrored in the epilogue
(1464–1465), which otherwise similarly holds itself within broadly tragic
bounds of clarity. Nevertheless, even such momentary leakages of excess
from Cassandra’s speech to its reporter’s confound any attempt at a
clear separation between narratorial voices and levels—between verba-
tim quotation and paraphrase, and thus between Cassandra’s original
narrative and an edited or misremembered re-narration by the slave.
8 The slave characterizes this particular Cassandran narration as distinctly wilder
than her previous utterances (3–4); it is not clear how these earlier prophecies differed
in intelligibility, but we are clearly meant to think of Aeschylus Agamemnon 1178–1330
(and perhaps also Euripides Troades 353–461), where Cassandra passes from frenzied
lyric to more intelligible but still discredited speech.
9 See Fountoulakis 1998: 293 for a more measured view.
10 Green 1990: 177 translates: ‘the centipede lovely-faced stork-colored daughters of
the Bald Lady struck maiden-slaying Thetis with their blades.’

In contrast to the specificity of Cassandra’s own narrative situation,

the frame of the slave’s report is only very sketchily dramatized, and
its narratee is now specific and sentient, but present in the text only
as a silent auditor. 1467–1471 makes it clear that the report to Priam
follows directly on Cassandra’s withdrawal into her cave; we are thus
still on the day of Paris’ departure for Greece. The opening dialogue
borrows from Platonic quasi-dramatic conversational openings (notably
the Symposium) to suggest a preceding request for information by the
narratee but, as in the Platonic models (→), the narratee remains
textually invisible. It is left for the poem’s own readership to imagine
Priam’s response to the text they are reading over his shoulder, and
to reflect on the similarities with their own response. In some respects
Priam’s competence in decryption actually exceeds that of Lycophron’s
readers, since he is in principle better equipped to decipher arcane
allusions to past events connected with city and family; so at 335–
347, which uses such riddles in referring to Priam’s own death. There
are also some key passages of comparative lucidity (281–306, 348–372,
1087–1089, 1110–1140, 1281–1282, 1451–1460) where Priam would have
to be singularly obtuse not to grasp the narrative outline of what is
being foretold. Yet in the final reckoning the poem’s readers also know
from their vantage of historical hindsight that Priam’s powers of both
understanding and action are circumscribed: that history will not be
changed, and that whatever Priam’s response he will take no action
capable of averting the fulfilment of the prolepsis.
Lycophron, Ptolemy, and the ‘Lycophron question’ 11
This discussion has sidestepped the notorious ‘Lycophron question’,

since even ‘radical unitarians’ (who believe our poem to be a second-
century pseudepigraphon) see the ascription to Lycophron of Chalcis
as more or less pointed—invoking not only an identity for the primary
11 West 1984 influentially distinguishes three fundamental positions: ‘conservative
unitarian’, according to which the poem is essentially the uncontaminated work of

Lycophron of Chalcis; ‘radical unitarian’, dating the poem in its entirety to the second
century or later; and ‘analyst’, seeing the Roman references as interpolations in an
otherwise Lycophronian text. West’s own position might be further distinguished as
‘radical analyst’ (Tim Cornell’s term, from an unpublished paper): she argues strongly
for much more extensive Italocentric interpolation in an original Lycophronian text
than earlier analyst critics have dared to suppose.
narrator, but a historical milieu, a perspective on history, and a specific

primary narratee in Ptolemy II. Whether that image and that narratee
are the actual, historical author and addressee is for present purposes
a question of only secondary importance (though for other aspects of
interpretation it is of course paramount). Nevertheless, though the text
itself says nothing of its author, it has become standard to see the
opening and closing lines as deliberately evoking a real or imagined
performance by Lycophron himself at the Philadelphan court;12 and if
we accept that Ptolemy is the primary narratee of such a performance,13
the comparison between Ptolemy and Priam as contrasting narratees
becomes a large part of the poem’s point. One king is baffled by history,
the other sees it entire; one presides over the dissolution of global order,
the other over its restoration; and one fatally fails to regard the Apolline
authority behind the words of his servant, while the other is fully alert
to the truth of inspiration.
Thus the Alexandra’s nesting of embedded narratives prompts us to
see pointed connections between its levels: in particular, between Cas-
sandra’s attempt to turn Apollo’s mantic gift into narrative and the
poet’s attempt to tell the shape of world history; and between the two
kings’ readings of their place in history, and the reader’s own decipher-
ment of the enigmatizing text.14 Seen as a whole, the structure estab-
lishes a dynamic hierarchy of metapoetic and political value, with nar-
rative authority deferred upwards through the four levels of narrator,
and ultimately from the figure of the poet to the divine figure of his
inspiration, at the same time as hermeneutic authority, the power to
interpret, devolves downwards to the historical reader for whom the
narrative becomes finally intelligible, believable, and true.15
12 So most recently Kosmetatou 2000: 38–39 (but assuming a second-century date
and an Attalid context) and, more cautiously, West 2000: 289 n. 75, following Fusillo
1984: 505–506. The actual performability of the text remains an open question; see
Cameron 1995: 81 and the response by West 2000, as well as Fountoulakis 1998
(valuable despite an unpersuasive conclusion).
13 This is strongly encouraged (contra Fusillo, who imagines a collective court audi-
ence) by the slave’s consistent use of second persons singular; see Fountoulakis 1998:
293.
14 See Fusillo 1984: 500–506.
15 See further Lowe 2004.
part five
ORATORY
chapter twenty-two
ANTIPHON
M. Edwards
Introduction. Narratology and oratory
The application of narratological analysis to rhetorical texts raises var-

ious issues, some familiar to narratologists working on historical texts,
others peculiar to the genre of oratory. Like historical texts, speeches
purport to tell the truth about real events, however the overriding pur-
pose of a speech’s narrator is to mould those events into a story that is
persuasive to his audience, and to that end he will often (it is clear) use
deliberate deception. Frequently the author of a speech and its speaker
are the same person, though in legal oratory a litigant would routinely
call on the services of a professional speech-writer (logographos) to com-
pose the speech for him.1 In these circumstances, therefore, the author
and speaker are not the same person, though the pretence is that the
speaker has composed the speech himself. The speaker will specifically
turn into a primary narrator at the point of the speech where he begins
his narrative of the events in question,2 and this raises the issue of the
function of narrative in oratory.
Aristotle divides speeches into three basic classes (Rh. 1.3): deliber-
ative (i.e. political), forensic (legal), and epideictic (display). Then, in
his discussion of the internal arrangement (taxis) of speeches (Rh. 3.13–
19), he examines the four major divisions of proem (prooimion), narra-
tive (diēgēsis), proof (pistis), and epilogue (epilogos), outlining their main
features within each of the three classes. In oratory, therefore, the
term ‘narrative’ (or ‘narration’) denotes that part of a speech in which
1 I do not accept the theory of composite authorship of speeches by the logographer
in consultation with his client, as propounded by Dover 1968.

2 Note that the primary narrator will always be male, since women had no legal
competence in Athenian courts (where they were represented by their guardian, kurios)
or right to address the assembly and other public gatherings of citizens.
318 part five – chapter twenty-two
past events are narrated. These are regularly the background to a suit
described in the form of a summarized story. The narrative is different
in nature and content from the other parts of the speech; it is on this
narrative that the attention of this chapter will be focused.
Nevertheless, it should be noted that narratives in practice are by
no means always composed in single units: they are frequently bro-
ken down into discrete blocks, with a speech containing two or more
narrative sections separated by sections of argumentation; and in addi-
tion narrative elements will regularly be found in the other parts of a
speech, especially the proofs section. It is also the case that delibera-
tive speeches tend to lack a separate narrative entirely (cf. Arist. Rh.
3.16.11), and we shall concentrate here mainly on the narrative sec-
tions of forensic speeches. In these there is a primary narrator, usu-
ally internal, who is presenting his version of what happened. The pri-
mary narratees in principle are the jurors, who may be either exter-
nal or internal; but in addition a speaker may address his opponent
in apostrophe, who at that point becomes a separate internal narra-
tee.
Antiphon
We shall begin with some general observations on the narrative tech-

nique of Antiphon. All three of Antiphon’s surviving courtroom speech-
es have a discrete narrative section (1, Against the Stepmother 14–20; 5, On
the Murder of Herodes 20–24; 6, On the Chorus-Boy 11–14). There is, how-
ever, a noticeable distinction in strategy between the sole prosecution
speech and the two defence speeches. In the latter, echoing Aristotle’s
recognition that the narrative in a defence speech may be shorter (Rh.
3.16.6), the narratives are brief and selective; only the barest details
are given of the events in question, and in the Herodes even these are
punctuated with argumentation (e.g. §21, ‘And first consider this …’).
Further information is supplied by narrative passages in the proofs sec-
tion (5.29–30; 6.21–24, 34–40), and this mixing of narrative and proofs
is also found in the works of later orators. In his prosecution speech,
on the other hand, Antiphon’s narrative is self-contained and clearly is
vital to the plaintiff’s chances of winning the case, forming over one-
fifth of this short speech as a whole. This strategic difference aside,
Antiphon’s technical practice in introducing and concluding his narra-
tives is consistent over the three speeches. The beginning of the nar-
m. edwards – antiphon 319
rative is marked by a narratorial intervention in the form of a meta-

narrative comment, much in the manner of a Sophoclean messenger
(→), e.g. ‘I shall try to narrate to you the truth of what happened’ (1.13;
cf. 5.19 and 6.10). Narratives in later orators regularly begin with the
particle gar; this is not Antiphon’s practice in the first narrative of the
surviving speeches, though the Stepmother narrative has a crucial second
stage at 1.18, where gar does occur, and the second and third narra-
tives of the Chorus-Boy speech also have the particle. The endings of
the narratives are similarly indicated, either by a concluding remark
(’Such were my arrangements for the chorus’, 6.14), or by a transi-
tional remark (‘Consider now …’, 1.21; cf. 6.41), or by a combination
of both (‘These are the facts; consider now the probabilities arising
from them’, 5.25), but in each instance addressed to the primary narra-
tees.
Finally, as one would expect in a forensic setting where the litigant
regularly presented his own case, the narrators in speeches 5 and 6 are
internal, and their involvement in the actions described is immediately
highlighted: ‘I sailed’ (5.20, with use of the personal pronoun egō) and
‘When I was appointed choregus’ (6.11). The narrator of speech 5 com-
ments overtly on his story ‘and surely I was not deliberately proposing
to send an informer against myself ’ (5.24), and the narrator of speech 6
is overt at 6.34 (‘I must inform you about these things too’). The narra-
tor of speech 1 was in a different position, since the events he describes
had taken place several years earlier when he was still a child, and he
presumably learned of them only from his father when on his death-
bed (1.30). Hence he acts as an external narrator, overt again at 1.18,
where the narratees too are external and overt as the second phase
of the narrative commences with the commonplace formula ‘it would
take too long for me to narrate and for you to listen’ (cf. And. 3.9; Lys.
23.11).
As an example of Antiphon’s practice we will examine the narrative
of the sixth speech,3 where an unnamed Athenian defends himself
before the court of the Palladium against a charge of unintentional
homicide. He had been responsible for the public duty of training a
chorus of boys for the Thargelia festival, one of whom, Diodotus, died
after drinking a potion. In the very brief first narrative (§§11–13 out
of 51) the choregus describes the arrangements he made, including the
3 For commentaries on this speech see ten Berge 1948; Gagarin 1997.
appointment of four deputies to conduct the actual training while he

was involved in litigation. Other parts of the story are told later in
the speech, when it becomes clear that the present prosecution was
part of a continuing legal battle between this wealthy citizen and his
political enemies; but the first narrative concentrates on how he acted
in his capacity as choregus. The narrator is thus internal and overt,
commenting repeatedly on his involvement in the arrangements for the
training of the boys. In the first half of the narrative this is an explicitly
personal involvement: ‘I acted as choregus … I fitted out a training
room … I recruited the chorus’; in the second half he is away on legal
business, but is still involved vicariously: ‘I arranged … I told him …
I appointed … a fourth whose duty it was’. He adopts the persona of
a matter-of-fact and businesslike citizen, who does his duty to the full,
both in prosecuting enemies of the state and in training his allotted
chorus. An effect of meticulous care is achieved by the thoroughness
of the arrangements he describes in the first part of the narrative,
enhanced by narratorial interventions: ‘as best and most fairly as I
could … the most suitable part of my house’. Further to underscore
his fairness the narrator employs the device of presentation through
negation, with more interventions (§11):
Next I gathered together the best chorus I could, neither levying any
fines nor forcing anyone to provide guarantees nor making any enemies,
but everything was done as pleasantly and suitably as possible for both
sides, I giving orders and making requests, they readily and willingly
sending their sons.
A potential weakness in the choregus’ position was that he had not

trained the chorus personally, and to avoid a charge of neglecting his
duty he is at pains in the second half of the narrative to explain his
absence and demonstrate that he had appointed suitable deputies. In
§12 his ‘lack of time’ (askholia) meant that he could not ‘be present and
look after them’, and he ‘happened’ to4 be engaged in legal action of
great importance ‘to the Council and the Athenian people’. The rele-
vance of this seed will become clear in § 35. The impression of the chore-
gus’ thorough and law-abiding persona is maintained by further nar-
ratorial interventions: he prosecutes Aristion and Philinus ‘justly and
properly’, but ensures that the training is ‘conducted with all possible
care’ by a close relative, his son-in-law Phanostratus. He is supported
4 Chance plays a role in all three Antiphontean narratives; cf. 1.16; 5.20, 21.
by two other upstanding citizens, Ameinias, whom the narrator con-

siders ‘an excellent man’ and who was experienced in training cho-
ruses, as was the anonymous second appointee. These three also had
the assistance of a fourth man, Philippus, who was to buy whatever
was needed, and in this way the boys would be trained ‘in the best
way possible’ with nothing lacking due to the choregus’ own indisposi-
tion.
At §§21–24 the choregus gives a second narrative, describing the be-
haviour of his opponent Philocrates on the day of the boy’s funeral,
when Philocrates first laid his charge in the Heliaea court, and the
choregus responded by noting that he was about to prosecute Aristion
and Philinus the following day and challenging his opponents to ques-
tion witnesses. The narrator is again internal and overt: ‘I stood up
before the court and said to the jurors that …’; and his persona of
thoroughness is maintained, now by the careful description of all the
evidence he can adduce to rebut Philocrates’ allegation against him.
Narratorial interventions highlight his sense of righteous indignation:
Philocrates was bringing a ‘slanderous accusation’ (§21, with repetition
of the verb diaballein in §22), and it was the imminence of the choregus’
own prosecutions that explained ‘these allegations’. Philocrates’ refusal
of the commonplace challenge by the choregus to examine slaves under
torture for evidence builds to the climax of this narrative, which takes
the form of another damning narratorial intervention: ‘they well knew
that this examination would not supply them with evidence against me,
but me with evidence against them that their accusation was totally
unjust and unfounded’.
A third stage of the narrative is found at §§34–40, the beginning of
which is marked by a metanarrative narratorial intervention (‘I shall
show you’) and the particle gar. It starts with an analepsis, taking us
back to the day of the boy’s death and the day after, and then to the
day of the funeral described in §§21–22: ‘they (the prosecution) had
been persuaded by my enemies and were preparing to charge me and
to proclaim that I was banned from the prescribed places’ (a phrase
which recurs in §35, twice in 36 and again in 40). Persuasion hints at
bribery, preparation at connivance, and it is the choregus’ political ene-
mies who now take centre stage. Both narrator and narratees are overt,
as the choregus asks two rhetorical questions (who did this and why?) and
adds ‘I must tell you about these things also.’ We have clearly reached
the crucial moment in the narrative, and in §35 the narrator intervenes
to emphasise that the men he was prosecuting had no hope of acquittal
because of the enormity of their offences. Their motive in persuading

Philocrates ‘to register a charge and to proclaim that I was banned
from the prescribed places’ is their ‘safety and riddance of the whole
affair’; and the consequences of the proclamation in this case are pre-
cisely spelled out by the ever-meticulous narrator, with the intervention
that his opponents would ‘not pay the penalty to you for the wrongs
they had done’. A second analepsis coupled with another address of
the narratees helps to ram home the iniquity of their acquittal: this was
not the first time they had ‘devised’ this scheme (the loaded vocabulary
characterizes the opponents as schemers, in contrast to the narrator’s
straightforwardness), they had used it previously against Lysistratus, ‘as
you yourselves have heard’.
In the final section of the third narrative the narrator again addresses
the narratees (‘and you know the sum at which the damages were
assessed’, §38) and makes further interventions as his story builds to its
climax in §§39–40. He is ever mindful of his opponents’ treachery and
describes how, after a reconciliation between the parties was effected,
he met them in a whole series of places, culminating ‘by Zeus and
all the gods’ with the choregus and Philocrates standing together on the
tribune in the council-chamber and talking to each other. The narrative
then ends on a nice note of irony, with the narratorial intervention that
the council was astonished to learn that the choregus was ‘proclaimed to
be banned from the prescribed places’ by the very same people who on
the day before ‘had met me and chatted’.
Students of Antiphon’s speeches, the earliest in the corpus of Attic
oratory, have generally paid scant attention to his narratives, focusing
mainly on his techniques of argumentation, and in particular his exten-
sive use of probability-argument. But even the brief survey above indi-
cates that his narratives offer plenty of scope for narratological analy-
sis. The salient features of his method that will naturally recur in later
writers of the oratorical genre are, first, the use of the internal, overt
narrator, who tells of his involvement in the events that have led to
the trial. The exception here is the speaker of speech 1, but he too
becomes an internal narrator in the proofs section (at §30, ‘just as my
father charged me, even though I was still a boy, during his last ill-
ness’). The narratees are mostly external, though again in the proofs
section the speaker of speech 5 calls on their knowledge of prior events
(5.71, ‘I think the older ones among you will remember this yourselves’;
cf. 6.41). This is a regular tactic, as is the address in apostrophe to the
speaker’s opponent (e.g. 5.11, an example once more outside the nar-
rative proper). Finally, Antiphon already combines the techniques of

the single, extended narrative in speech 1 with discrete, shorter nar-
ratives in speeches 5 and 6, in line with the requirements of each
case.
chapter twenty-three
ANDOCIDES
M. Edwards
Unlike Antiphon, Andocides was not a professional logographer, and

his three surviving works1 were delivered by himself on matters in which
he was directly involved. Since antiquity judgment on his literary abil-
ities has been severe,2 but one area in which he indisputably had tal-
ent was that of telling a good story. The best examples of this are
found in his lengthy defence speech On the Mysteries, delivered in 400
against a charge of impiety.3 Exiled in the wake of the religious scan-
dals of 415, Andocides eventually returned to Athens after the amnesty
of 403, but was prosecuted for attending the Mysteries celebrations of
400 when allegedly debarred by the decree of Isotimides. In his defence
Andocides has to show that he neither was guilty of impiety in 415
nor had confessed his guilt, and so was not liable to the decree; and
he argues that the decree in any case was now invalid. A secondary
charge alleged that he had illegally placed an olive-branch on the altar
of the Eleusinium during the time of the Mysteries celebrations. Ando-
cides is therefore faced with a complex set of circumstances spanning
fifteen years, and his defence is adapted to the needs of the situation.
It is divided into several parts thematically rather than chronologically,
with narrative details provided at regular intervals, displaying a flexibil-
ity that indicates Andocides’ innate talent at narrative. After an indeed
conventional proem (§§1–10), he first addresses the issue of the profana-
tion of the Mysteries (§§11–33), a section which may be sub-divided into
narrative (§§11–18) followed by proofs, with a second short narrative at
§§27–28. The second part of the defence, concerning the mutilation of
the Herms (§§34–70), is constructed on a different pattern, in which the
1 The authenticity of On the Peace is regularly accepted (see Edwards 1995: 107–108),
but was challenged by Harris 2000. The fourth speech in the MSS, Against Alcibiades, is
almost certainly spurious (see Edwards 1995: 131–136; Gribble 1999: 154–158).
2 See Usher 1999: 42.
3 For commentaries on this speech see Makkink 1932; MacDowell 1962; Edwards
1995. On the date see MacDowell 1962: 204–205.

326 part five – chapter twenty-three
narrative of the story is interwoven with the proofs. The reason for this
change of approach may stem from the fact that Andocides had admit-
ted to cognizance of the mutilation, whereas his utter denial of any
involvement in the profanations could not be refuted, hence he begins
with this stronger aspect of his case.4 He next turns to the revision of
the laws (§§71–91), to demonstrate that he was no longer subject to the
decree of Isotimides, and here the narrative elements are very brief,
forming part of the discussion of earlier laws and decrees (cf. §§73, 80–
82, 85–86, 88–89). There follows an attack on the lawless character
of Andocides’ opponents (§§92–136), which contains historical narra-
tive, a narrative of the secondary charge concerning the olive-branch,
and narratives of the behaviour of Andocides’ opponents Callias and
Agyrrhius. The speech concludes with an argument of divine favour
(§§137–139) and an epilogue (§§140–150).
The opening narrative (§§11–18) bears some similarities to the meth-
od of Antiphon. Its beginning is marked in §10 by a metanarrative nar-
ratorial intervention (‘I shall tell you’), and its end is clearly indicated in
§19 with another intervention, ‘you have heard what happened, gentle-
men, and the witnesses have testified to you’ (cf. Ant. 5.25, ‘this is what
happened’). But the use of the particle gar at the start of the narrative is
to be noted,5 and Andocides regularly marks the beginning and ending
of his narratives with an address to the primary narratees, here in §§10
and 19.6 The internal narrator is at first covert; and a striking feature,
not encountered in Antiphon, is the use of oratio recta, as Pythonicus,
the man who first discloses the profanations, becomes a secondary nar-
rator with secondary narratees (the assembly). Andocides had already
used direct speech in On his Return (2.14), and it occurs frequently in
the Mysteries speech.7 The second informer, Teucrus (§15), was a key fig-
ure, since he gave information concerning both the profanations and
the mutilation, and twice the verb connected with his activities is in
the historic present tense rather than the usual imperfect or aorist: he
‘informs’ (epaggelletai) and ‘gives a list’ (apographei) of names. The historic
present is also used in connection with another informer, Lydus (§17),
4 See Edwards 1995: 17–26; contra MacDowell 1962: 167–171, 173–176.

5 Similarly §§111, 130, 133, but not 27, 34, 48, 60, 117, 124.
6 Similarly §§27/29, 46, 54, 60/69, 106/109, 117/123, 124/128, 130, 136, but not 34,
48, 111/113, 133.

7 Already in the proem at §4.
m. edwards – andocides 327
significantly as he lists (apographei) Andocides’ father among the culprits,

and to highlight the reaction of the councillor Speusippus, who ‘has the
culprits sent for trial’ (paradidōsin). The narrator closes his account of the
Mysteries affair by telling how rewards were given to Andromachus and
Teucrus in a continuous narrative (§§27–28), whose opening is marked
by a metanarrative narratorial intervention addressed to the primary
narratees (‘And now, gentlemen, what happened after this?’), and ends
with a similar intervention in §29.
The narrator resumes his story in § 34, with Teucrus’ information
concerning the Herms. Again, the narrative is introduced by a meta-
narrative narratorial intervention (‘I will do as I promised and tell you
the whole story from the beginning’), and the importance of Teucrus
giving the information underlined by the use of historic present: ‘he
informs’ (mēnuei) and ‘gives a list’ (apographei). It now transpires that the
information led to panic in Athens, and in the midst of the troubles
another informer takes centre stage, Diocleides (§37). His importance is
indicated by the use of the present tense eisaggellei (‘brings an impeach-
ment’), and by a narratorial intervention, which emphatically appeals
on the internal status of the primary narratees: ‘Now I want you to
think carefully here, gentlemen; try to remember whether I am telling
the truth, and inform your companions; for it was before you that Dio-
cledes stated his case and you are my witnesses of what occurred.’ His
tale (§§38–43), introduced by gar, is presented as that of a reported nar-
rator; a device familiar from historiography and well at home in foren-
sic oratory. However, it is frequently interrupted by comments of the
primary narrator, for the first time in §39: ‘Now to begin with, gentle-
men, Diocleides gave his story this particular form simply to be in a
position to say of any citizen, according to his choice, that he was or
was not one of the offenders—a monstrous proceeding.’ The same phe-
nomenon can be observed in §40 (‘and told him what I have told you’),
before Andocides’ companion Euphemus is introduced by Diocleides
and given direct speech, as is Andocides’ father Leogoras in §41. The
primary narrator intervenes once more with the comment that Dioclei-
des ‘was in this way trying to destroy’ Leogoras; and the final part of
the reported narrator’s story in §§41–42 is again interrupted by the pri-
mary narrator’s comment ‘he was thereby trying similarly to destroy my
brother-in-law’. The primary narrator indicates the end of Diocleides’
narrative in regular manner by a narratorial intervention (§43, ‘such
was his impeachment, gentlemen’), and this is immediately followed by
the historic present apographei (‘he gives a list’), and a narrative of the
consequences of the information. He carefully details the arrangements

made by the Council for Athens’ security and remarks bitterly that the
author of the ‘troubles’ (kakōn), hailed as Athens’ saviour, ‘was dining’
(edeipnei) free in the Prytaneum—the imperfect tense with which the
narrative concludes conjures up the image of Diocleides serenely din-
ing while all around was panic. The end of this section of narrative is
marked once more by a narratorial intervention addressed to the pri-
mary narratees, in so far as they are internal: ‘first, gentlemen, those
of you who were present recall these things and describe them to the
others’ (§46).
The next narrative tells the emotive story of what happened to
Andocides after his arrest (§§48–53): ‘we were all thrown into the same
prison and it was night and the gates were shut—one man’s mother
came, another’s sister, another’s wife and children—there were cries
and moans of men bewailing and lamenting their present troubles’.
He first quotes his cousin Charmides’ appeal to him for help in direct
speech, and then presents his own thoughts as he wrestled with his
conscience whether or not to inform, first in direct, then in indirect
speech. These thoughts shrewdly contain the evaluations ‘wrongfully’
(adikōs) and ‘rightfully’ (dikaiōs).
In §§60–66 follows the account of how Andocides finally gave his
version of the mutilation of the Herms, with Andocides acting as re-
ported narrator (‘I therefore informed the Council that I knew the
offenders and showed exactly what occurred. The idea, I said, had
been …’). Noticeable here are the repetition of the emotive adverb
adikōs (§60), the intervention of the primary narrator Andocides in the
narrative of his reported narrator Andocides at §62 (‘and that is why
the Herm which you can all see, the one by our family home … was
the only one unmutilated’), and the use of direct speech at §63, as
Meletus and Euphiletus warn Andocides to keep quiet using a phrase
(‘you will keep us as friends as before’) reminiscent of Diocleides’ words
to Euphemus at §40. The investigation of Andocides’ version of events
leads to the denouement of the Diocleides story. Shrewdly exploiting
the internal status of his narratees, he ends this section by a passage
of second-person narration: ‘and when you heard the facts yourselves,
you handed Diocleides over to the court and put him to death. You
released the prisoners …’ (§66). After a summing up of the main facts
of his narratives so far (§§67–68), Andocides ends with a final appeal to
his narratees, in so far as they are internal, an appeal which by now has
acquired the status of a rhyme or incantation (‘Now recollect whether
what I have been saying is true, gentlemen; and if you know the facts,
make them clear to those who do not’: §69).
The next extended narrative in the speech comes at §§ 106–109,
as Andocides surveys past Athenian history to show how grievances
had been put aside in previous times of trouble to the benefit of the
city. The use of external analepses in the form of historical parallels
is very common in the orators and already appears at Ant. 5.67–71,
but has a history of its own stretching back to Homer (→).8 Features
of this narrative such as the introductory metanarrative narratorial
intervention ‘I wish to say a few words about this too’ (cf. Lys. 22.1;
Isoc. 18.4) have been met often enough before; but its patriotic feel
is new, as the now external primary narrator recalls the actions of
his ancestors against the tyrants (§106) and adds a first reference in
the orators to the battle of Marathon, alongside an emphasis on the
nobility (aretē) of earlier Athenians. These themes recur in later speeches
(e.g. Lys. 2.21; Isoc. 4.86–87; Dem. 18.208).
Andocides’ account of the secondary charge against him (concern-
ing the placing of a suppliant’s bough) begins with a short narrative of
more recent events (§§111–112). This time there is no preliminary nar-
ratorial intervention or address to the narratees, though introductory
gar is used. The narrator describes the scene in the Council meeting
as Callias dramatically declares (legei) that a bough had been placed
illegally on the altar during the recent celebration of the Eleusinian
Mysteries. The narrative resumes in §115, with the helpful interven-
tion of Cephalus as a secondary narrator (§116). In order to explain
Callias’ motive for prosecuting Andocides, the narrator goes back in
time and inserts an analepsis (117–123), concerning the daughter of
Epilycus, whom Callias tried to secure for his own son against Ando-
cides. It opens with the regular narratorial intervention (‘I shall nar-
rate why he tried to trap me’), and the narrator suggests his own
decency by a shrewd use of indirect speech (‘I told Leagrus that this
was the time for decent men to demonstrate their respect for family
ties,’ §118). His good intentions are reiterated in another speech to Lea-
grus (§120), before Callias enters a claim (lagkhanei) to the girl, lodges an
information (endeiknusi) and involves (kathistēsin) Andocides in the trial,
and places (tithēsi) the branch on the altar—a remarkable sequence of
present tenses in which, after the first three, the fourth one seems all the
8 On the use of history in the orators see in particular Nouhaud 1982.

more credible. Steadfast to the end, the narrator turns to the narratees
for their assistance in punishing Callias. A third stage of the Callias nar-
rative takes us still further back in time, to the birth of his son (124–127).
Callias marries (gamei, the emphatic first word of this narrative, in the
historic present), but the narrator describes with an ironic humour how
he was soon living with his wife’s mother too, when he was a priest of
Demeter and Kore, the Mother and the Daughter. The mother ‘drove
out’ the daughter, before Callias ‘threw out’ the mother, denying that
her child was his own; and his subsequent altercation with her relatives
is narrated vividly and followed by the narratorial intervention ‘if he
lied he prayed that he and his house would be destroyed—as they will
be’. Callias’ later change of heart is emphasized by present tenses: he
‘brings’ (komizetai) the woman back and ‘introduces’ (eisagei) the boy as
his son; and the narrator describes how Callias swore on the altar to
his son’s legitimacy (§127) in words very similar to the ones used earlier
to describe his denial of paternity (§126). The final part of the Callias
narrative (§130) takes us back to his father, Hipponicus, and the rumour
that he had an evil spirit in his house. The section begins and ends with
an address to the narratees, both times calling on their memories, with
the introductory ‘I wish (boulomai) to remind you’ and an initial gar.
A final section of narrative concerns Agyrrhius’ motive for prosecut-
ing, over a business contract (§§133–135). It is introduced by the meta-
narrative narratorial intervention ‘I shall tell you’ and has an initial gar.
The narrator again adopts the persona of an upright citizen, in contrast
to Agyrrhius, whom he ironically calls a ‘gentleman’ (kalos kagathos), and
his partners—he intervenes at §133 with the remark ‘you know what
they are like’. Their reaction, to get rid of Andocides at all costs, is
relayed for added effect in direct speech, ending with the recurrent
emotive word of the speech’s narratives, adikōs.
In sum, Andocides in the Mysteries speech employs, like Antiphon in
On the Chorus-Boy, an overt internal primary narrator. His primary nar-
ratees are internal, a fact which is frequently, emphatically and max-
imally exploited, the narrator calling on them as witnesses and even
involving them in the form of second-person narration. He adopts
Antiphon’s method of breaking up the narrative into smaller sections,
though here the events of the two main narratives, the profanation of
the Mysteries and mutilation of the Herms, were in fact contempo-
raneous. But a prominent feature of Andocides’ method is the frequent
insertion of interventions by the primary narrator into stories being told
by a reported narrator; and another major advance on his predecessor
is his use of speeches and secondary narrators, both of which greatly

enhance the vividness and effectiveness of the narratives. Andocides
may not have been a professional logographer, but the ability to tell a
persuasive story was one of the tricks of the trade that he possessed in
abundance.
chapter twenty-four
LYSIAS
M. Edwards
In the opinion of Dionysius of Halicarnassus (Lysias 18), Lysias was

‘unquestionably the best’ of the orators at writing narrative. By no
means all of the thirty-five speeches surviving under his name contain
a narrative, but those that do certainly tend to support Dionysius’
judgment, and the narrative frequently plays a central role in the
persuasion of the jurors. A crucial factor is the way in which Lysias
characterizes his clients, who act as internal narrators. The narrative of
the first speech, On the Killing of Eratosthenes, will serve as an example of
Lysias’ method, but we shall begin with some general remarks.
Although there are some speeches in which he mixes elements of
narrative and proof in the manner of Antiphon and Andocides,1 Lysias
generally adopts the basic fourfold pattern of proem, narrative, proofs,
and epilogue. An extensive diēgēsis is a dominant feature of many of his
works,2 and over one-third of speeches 1, 3, 13, 16 and 31 is taken up by
narrative. Various recurrent features are familiar from Antiphon and
Andocides, such as the now familiar metanarrative narratorial inter-
vention that indicates the start of the narrative (e.g. ‘I shall describe’,
1.5, 3.4).3 His speakers regularly add that they will tell the story or the
whole story from the beginning (ex arkhēs, 1.5; 7.3; 12.3; 13.4; 32.3; cf.
And. 1.10, 34), and sometimes that they will do so briefly (12.3; 16.8;
24.4). Almost all of Lysias’ narratives have gar at the beginning, with
the exception of speeches 12 and 32. Speech 12, however, is among a
number of speeches that have a preliminary narrative (prodiēgēsis), and
while there is no gar at 12.4, there is one at the start of the main nar-
rative in 12.6. The beginning and end of most Lysiac narratives are
1 E.g. speeches 7, 16, 31.

2 The circumstances of the case may, of course, militate against an extended narra-
tive: e.g. speech 24 turns on the value of the speaker’s property.
3 Cf. ‘I shall inform’ (19.12), ‘I shall try to inform’ (7.3; 12.3; 25.7; 32.3), ‘I shall begin
to narrate’ (13.4), ‘I wish to describe’ (13.62), ‘I wish to tell’ (22.1).

334 part five – chapter twenty-four
marked, as in Andocides, by addresses to the jurors functioning as pri-

mary narratees, though the end of several narratives is not otherwise
clearly indicated: a significant number have the particle toinun (cf. 7.11;
10.6; 16.18; 22.6; 24.7; 30.6; 32.19), while speech 1 has the adverb houtōs
(‘in this way’, cf. 21.11).4
Speech 1 features an internal narrator, the defendant Euphiletus,
who is accused of entrapment and intentional homicide, after catch-
ing his wife in flagrante with her lover, Eratosthenes, and killing him.
Euphiletus claims the killing was not planned and was sanctioned by
law, but it seems clear that such extreme action was no longer con-
sidered to be appropriate.5 The narrative (§§6–26) occupies twenty-one
out of the fifty sections of this speech, roughly the same length as that
of speech 12, and only the narrative of speech 13 is longer. It is divisi-
ble into three sections, beginning with a preliminary narrative (§§6–8)
describing a perfectly normal Athenian marriage, in which the husband
kept reasonable control of his wife and trusted her the more after the
birth of their first child; next comes the first stage of the main narrative
(§§9–21), in which the narrator tells of events indicating his ignorance of
the affair, until it is revealed to him by an old crone who is the servant
of another of Eratosthenes’ conquests and is confirmed by the confes-
sion of his wife’s maid; then follows the second stage (§§22–26), which
recounts what happened on the fateful night. Throughout the narra-
tor is overt. Thus we have in §7 a prolepsis reminiscent of Homer and
Herodotus, which foreshadows the disaster to come, as the wife is seen
by her future lover at the funeral of Euphiletus’ mother: ‘but when my
mother died, her death became the cause of all my troubles’; and meta-
narrative narratorial remarks, which mark transitions in the story (§9:
‘for I must narrate these things to you as well’; §22: ‘I wish to narrate’).
A noticeable feature of the narrative as a whole is tense variation. A
dramatic vividness is added to Euphiletus’ account by the use of the his-
toric present at certain key points: the birth of the child (§6, gignetai), the
corruption of his wife (§ 8, diaphtheiretai), the deception perpetrated by
the wife when she closes the bedroom door and locks it (§13, prostithēsi,
ephelketai), the approach to Euphiletus of the old crone (§15, proserkhetai);
and in particular the description of the last night, as Eratosthenes enters
4 I take the narrative to end at §26. In Carey’s division (1989: 66) the narrative
extends to §28, and there is an address to the jurors in §29.

5 For commentaries on this speech see Edwards and Usher 1985; Carey 1989;
Edwards 1999.
m. edwards – lysias 335
the house (§23, eiserkhetai), the maid tells Euphiletus (§23, phrazei), who
goes out and calls on his friends (§23, exerkhomai, aphiknoumai), they enter
the house together (§24, eiserkhometha) and Euphiletus knocks Eratos-
thenes down (§25, kataballō).
The narrative plays a key role in the portrayal of Euphiletus’ charac-
ter, as the narrator tells what is, on the surface at least, a simple story
of deception and discovery. The simplicity is, of course, the product of
shrewd narrative art,6 and Lysias thereby builds up a picture of a solid
Athenian citizen, steadfast in his observance of the law, who is at the
same time an unsophisticated man, incapable of the kind of plotting
of which he is accused. He is a somewhat naïve, but caring husband
(§9), firm with his wife when he needs to be and prone to the occa-
sional angry outburst (§ 12). A number of bitter comments on his lack
of suspicion and ignorance (§§ 10, 13, 14, 15) cement this picture.7 But
what is perhaps the most remarkable feature of this narrative is the role
apportioned to the three women, Euphiletus’ wife, her maid, and the
old crone. The wife, though never blamed for the affair, is nevertheless
a clever manipulator (§§10–14), and she is given one of the relatively
rare humorous remarks in oratory as she becomes a secondary narra-
tor: ‘Yes, so that you can have a go at the slave-girl. You tried to grab
her once before when you were drunk. And I laughed.’ (§§12–13). As
well as the wife, the old crone also acts as a secondary narrator, as
she alerts Euphiletus to what is happening. Her tale, in particular the
remark that Eratosthenes made a profession (tekhnē) out of seduction,
and Euphiletus’ confrontation of the slave-girl enable him to paint a
vivid and convincing picture of an extended affair without having to
produce either as a witness, which probably was not allowed in law.8
The slave-girl is, in fact, the only one of the three women who does
not take on the role of a secondary narrator, but she does become a
reported narrator:
And having at once fallen at my knees and obtained a pledge from
me that she would suffer no harm, she accused him firstly that he
approached her after the funeral, then told how she finally became his
messenger and how my wife was in time persuaded and the ways in
which she effected his entrances, and how at the Thesmophoria, while
6 See Carey 1989: 66; Edwards 1999: 67.

7 See Edwards 1999: 58.
8 See Todd 1993: 96. A notable parallel in Lysias is the role of the plaintiff’s mother
in speech 32, who twice acts as a secondary narrator (§§ 12–13, 15–17).
336 part five – chapter twenty-four
I was at the farm, she went off to the temple with his mother, and she
narrated (diēgēsato) precisely everything else that had happened.
This indirect testimony of the women may have made the primary nar-
ratees more ready to believe that Euphiletus’ version of what happened
on the fateful night was not premeditated. These narratees are jurors
sitting in the court of the Delphinium, which heard cases of justifiable
homicide. In contrast to Antiphon’s practice they are addressed reg-
ularly (nine times) in the narrative. However, the interaction between
narrator and narratees is not on this occasion one of deference, despite
the context of a homicide court. Throughout his speech the speaker
treats his addressees as fellow Athenians, who would all have reacted in
the same way as he did, and his tone of righteous indignation is more
that of a prosecutor than a defendant: already in the proem he adopts
the forthright stance of a victim of Eratosthenes’ outrageous behaviour.
This attitude is continued when he acts as narrator. His professed igno-
rance that was noted above is shrewdly emphasized for the narratees
by the common device of presentation through negation: ‘I never sus-
pected’ (§10), ‘thinking nothing of this nor suspecting anything’ (§13),
‘I said nothing about the matter’ (§14); and after the revelation he
reflects on things ‘which had never happened before’ (§17). He also
makes asides to the narratees (‘I must narrate these details to you’, §9,
cf. 11), so that they feel fully in the picture, and remarks such as ‘I was
filled with suspicion’ (twice in §17, sandwiching a recapitulation of all
the signs he had previously missed and standing in stark contrast to the
previous lack of suspicion) add to the persuasiveness of one of Lysias’
finest narratives.
Lysias was the foremost logographer of his day, and while modern
scholars have commented on his relative lack of argumentative skill,
his talent for writing persuasive narrative is undeniable. He employs
devices such as metanarrative narratorial intervention that we have
seen in Antiphon and Andocides, but his particular brilliance lies in his
ability to construct a character for the primary narrator that completely
secured the sympathies of the primary narratees.
chapter twenty-five
ISOCRATES
M. Edwards
The career of Isocrates falls into two distinct phases. Six forensic
speeches survive from the period 403–c. 390 BC, which in their method
bear both striking similarities to and marked differences from the con-
temporary works of Lysias. Only the short speech 21, Against Euthynus,
which may be the earliest of the six, follows the regular Lysiac pattern,
especially if its abrupt ending indicates that the epilogue is lost. Two
other speeches—16, On the Team of Horses and 20, Against Lochites—are
certainly incomplete, and the latter has no narrative. But speech 16,
despite the loss of the proem and narrative proper, contains two narra-
tive sections (§§5–9, 25–38), and the three complete speeches similarly
all have a main narrative, followed by one or more later sections of
narrative (17, Trapeziticus 3–23, 35–37, 42–43; 18, Against Callimachus 2–3
[a preliminary narrative], 5–12, 52–54, 59–61; 19, Aegineticus 5–12, 18–
27).
The beginning of a narrative is, once more, regularly marked in
Isocrates by metanarrative narratorial interventions, all involving a
form of the verb diēgeisthai (‘to narrate’), except at 16.4, the first extant
narrative of this speech whose beginning is lost, where we have ‘I shall
begin to inform you’. Narrators twice claim that they will tell the tale
from the beginning (17.3; 18.4) and one says he will do so ‘as briefly
as possible’ (21.2). Later narrative sections are less regularly introduced,
but noteworthy expressions are ‘I desire to describe to you’ (16.24; note
the verb dielthein) and ‘I shall mention this to you’ (18.58). All the narra-
tive sections have an initial gar, with the single exception of the second
narrative of speech 19, where the particle is absent at §18, though it
does occur in the second and third subsections of this narrative at §§21
and 24. The ends of the first narratives are also regularly and familiarly
marked, as by the comments ‘this is what happened’ (18.12; cf. 16.10;
21.4) and ‘I have told you everything that happened’ (17.24). As with
their opening, the close of later narratives is less clearly defined. Finally,
addresses to the narratees at the beginning and ending of narratives is
338 part five – chapter twenty-five
less common in Isocrates than Andocides and Lysias, with the excep-
tion of 17.3/24, 35/38.
As an example of forensic narrative in Isocrates we may take speech
17. This is his finest legal piece, composed for the unnamed son of
Sopaeus, a subject of king Satyrus of the Cimmerian Bosporus, when
he prosecuted the banker Pasion for fraud. The internal narrator’s
main tactic is to portray Pasion as a callous blackmailer, who took
advantage of his plight after his father had been arrested by Satyrus and
to whom he was forced to entrust his money in order to hide it from
Satyrus’ agents. This he accomplishes by an extended narrative (§§ 3–
23) which is very reminiscent of Lysias in its length and vivid character-
ization and has frequent use of the historic present tense. The primary
narratees may not have been wholly sympathetic towards a wealthy
foreigner who was himself involving Pasion in a fraud (§7), so the nar-
rator highlights Pasion’s unreliable, deceitful and erratic behaviour in
contrast to his own constancy in times of stress.
The narrative may be sub-divided into two parts, the first (§§ 3–
10) covering the period when the narrator and his father Sopaeus
were under suspicion, the second (§§11–23) the events after Sopaeus’
release. In the former he paints an intimate picture of the change in
his fortunes, from the time he left Pontus with money and ships to
his straightened circumstances in Athens and deception by Pasion; he
repeatedly inserts his own thoughts (e.g. §8) and those of Pasion, and
tells how he tried to safeguard his money in the belief that Pasion was
a trustworthy friend. On the news of his father’s release, the narrator
becomes much more confident, and it is Pasion’s turn to be worried
and apparently humble (§ 22), but he is always impudent and full of
ruses (§§19–23).
Underpinning the pathos of the story is the use of highly emo-
tive vocabulary, whereby the narrator becomes overt, such as ‘slander’
(diabolēs, §5), ‘shamelessness’ (anaiskhuntias, §8), ‘the most outrageous
claim of all’ (logon pantōn deinotaton, §12, cf. And. 1.39) and ‘the most
shameful accusations’ (aiskhistas aitias, §13). Additionally, the narrator
attempts to interact with the narratees, who strikingly for an Isocratean
speech are addressed directly eleven times in this narrative, e.g. ‘what
do you think was my state of mind?’ (§10), ‘as you yourselves will learn
in the course of my speech’ (§ 19) and ‘what more need I say to you,
gentlemen of the jury?’ (§ 23). Further, he plays indirectly on the long-
standing friendship between Athens and Pontus by indicating that he
sailed to Athens with two ships loaded with grain (§4), and on the Athe-
m. edwards – isocrates 339
nians’ pride at being the centre of the trading world—he has heard
about ‘this state and the rest of Greece’ (§4), and it is to Athens that he
comes, opening a bank account with Pasion. If the relations between
the two states are used subtly in this narrative, at the end of the speech
the speaker openly reminds the jurors of his family’s services to Athens
(§57). Further sections of narrative are found at §§35–37 (the speaker’s
dealings with Stratocles) and 42–43 (previous legal problems faced by
the speaker in which he had been helped by Pasion; note here how the
narrator becomes overt with the use of the emotive verb sukophantein, ‘to
prosecute maliciously’). In this speech, and speech 18, these later narra-
tives are quite brief, but in speech 19 the second is considerably longer
than the first, as the speaker avoids a long narrative of his complicated
family connections by breaking it up into two sections (§§5–14, 18–29),
so foreshadowing the technique of the inheritance specialist, Isaeus.
So far Isocrates the logographer, whose speeches feature internal
narrators, but it is for his fifteen discourses that he is mainly studied.
These display a variety of content and purpose, ranging from essays
on education and epideictic encomia to protreptic treatises and politi-
cal tracts. They were composed primarily for reading, and the political
discourses in particular provided material for study in Isocrates’ Athe-
nian school, inculcating political and moral ideas in his wealthy pupils.
By the nature of the discourses their narrative sections are broken up,
since interpretation of the facts was more important than extended nar-
ration of them, as Isocrates himself tells us in a narratorial intervention
in Panegyricus 97–98.1 They tend to consist of stories from past history,
included as justifications for the policies being propounded, and hence
feature external narrators. Here we shall concentrate on two discourses:
9, Evagoras and 4, Panegyricus.
Evagoras was written by Isocrates for king Nicocles of Salamis in
Cyprus, to commemorate the life of his father. It is therefore an en-
comium, but one which differs from earlier sophistic encomia in that its
subject is recently deceased and is neither a mythological figure nor
one of the fallen heroes of war celebrated in funeral speeches (epi-
taphioi). Evagoras’ virtue (aretē, a theme of the discourse) and achieve-
ments could be emulated by Isocrates’ readers, and Isocrates was aware
(§8) that he was inaugurating a new literary genre that eventually would
become biography.2 He begins with a conventional element of encomia,
1 Cf. Quintilian 4.2.31–32; Usher 1990: 172.

2 See further on the Evagoras and the origins of biography Bruns 1896; Momigliano
a narrative of the divine and heroic ancestry and the birth of Evagoras
(§§12–21). The overt external narrator introduces this narrative with the
phrase ‘it is fitting for me to describe’ (dielthein appeared in 16.24), and
a gar marks the beginning of the narrative in §13. It culminates in §21
with Evagoras’ birth, which is described in one long periodic sentence
and which again reveals the presence of its narrator, in a praeteritio, ‘I
prefer to pass over the portents, the oracles and the visions appearing
in dreams …’ A metanarrative narratorial intervention then immedi-
ately heralds the second instalment of the story of Evagoras’ life (with
a gar at the start of §22), which again reveals the narrator, this time
in a rhetorical question: ‘the confusion, the fears, the exhortations—
why need I spend time telling them?’ At §33, with Evagoras installed
as ruler of Salamis, the narrative is interrupted, as happens frequently
in Isocrates, and the story of his rule is picked up again in §§47–73,
introduced in §46 by ‘it is easy to learn from his deeds themselves’ and
with a gar marking the start of the narrative. A climax is reached in §64
with Evagoras’ defeat of the Persian king, the narrator commenting ‘the
most amazing thing of all is this’. Narratorial interventions are indeed
numerous (§§48, 51, 57, 61, 69), culminating in §73, where the narra-
tive is brought to a close: ‘I am sure I have overlooked many things
concerning Evagoras.’ One of these details was his violent death—
epideictic biography allowed Isocrates the licence to portray Evagoras’
life as one of unqualified success.
Isocrates is most renowned for his political discourses. The earliest,
Panegyricus, is perhaps the most famous, in which he sets out his plan
for a Panhellenic conquest of the Persians under the joint leadership
of Athens and Sparta.3 Given Sparta’s supremacy in the Greek world
at the time of publication (380), Isocrates is forced to argue Athens’
claims on the basis of justice, and he therefore recounts the mytholog-
ical and historical events that are the material of Athenian epideictic
oratory (§§21–132). The narrator begins with a preliminary narrative of
the antiquity of Athens (§§23–25), marked by the intervention ‘for it is
admitted that …’ (homologeitai men gar; cf. 9.13). The narrative breaks off
at §26, but resumes in §§28–29 with the myth of Demeter and Kore,
a story ‘which deserves to be told again now’ and has gar at the start.
The snippets of narrative continue at §§34–36, after a concluding meta-
narrative narratorial intervention (‘this is what I have to say about …’),
1971; Usher 1999: 312–313.

3 For a commentary on this speech see Usher 1990.
m. edwards – isocrates 341
with the story of Athenian leadership of the Ionian migration. This

time the narrative begins in the second limb of a long periodic sen-
tence, which encompasses the whole story (similarly the story of the
establishment of the laws at §39). After another narratorial summary
(‘let what I have said suffice …’) at §§51ff. the narrator moves on to
Athens’ wars, with the remark ‘I shall pass over recent or insignifi-
cant instances’ of Athens’ aiding suppliants (§54). He in fact starts by
going back into mythological times, with the commonplace story of the
help Athens afforded to the sons of Heracles and to Adrastus (§§54–60,
interrupted at 57). The narrative, then, is Atheno-centric, which indi-
cates clearly that the primary narratees are envisaged as being Atheni-
ans, despite the fact that this tract is ostensibly aimed at a nationwide
audience: Isocrates’ readers among the various Greek states will have
reacted very differently to his version of Athens as the protector of
weaker states and even benefactor of Sparta (§§ 61–63), and as leader
of the Greeks against barbarians (§§66–98; note at §66 the interven-
tion ‘I shall try to describe [dielthein] these things also’, which is fol-
lowed at the start of the narrative by esti gar). The Persian Wars are
powerfully portrayed, naturally with emphasis on the role of the Athe-
nians, who at Marathon, the narrator comments, were ‘a few against
many tens of thousands’ (§86), while king Xerxes is the arrogant fig-
ure familiar from Herodotus (§§89, 90). The part played by the Spar-
tans is diplomatically highlighted, as they rush to Marathon (§§86–87)
and fight to the death at Thermopylae, with courage equal to that dis-
played by the Athenians at Artemisium (§§90–92); but the other Greek
allies are hardly mentioned, apart from the Peloponnesians who were
only looking out for themselves (§93). The Athenocentric narrative cul-
minates in the battle of Salamis (§§93–96), which is depicted as very
much an Athenian victory—the battle of Plataea, won by Pelopon-
nesian infantry, is omitted. As has been pointed out,4 it is the moral
fibre of the Athenians that marks them out more than their physical
capabilities, and so detailed narrative gives way to the general charac-
ter inferences that can be drawn from their actions as a whole. The
narrator, himself explicitly Athenian, even defends the atrocities com-
mitted later against Melos and Scione (§ 102, ‘we deserve praise for
acting harshly in the fewest possible cases’) and asserts Athens’ selfless
moderation in power (§§103–109), in contrast to the oligarchic tyranny
4 By Usher 1999: 300.

that followed the fall of the empire (§§110–121). Further narrative details
pepper the discourse, including Persia’s military weakness (§§140–143),
as Isocrates promotes the policy that was ultimately to be carried out
by Alexander the Great.
It is clear that Isocrates’ early activities as a logographer, even though
he later repudiated them, stood him in good stead for his future career,
not least with regard to his narrative technique. A noticeable trait of
the narrators and narratees in his discourses is their national pride as
Athenians, which colours the way they present, as external narrators,
mythological and historical exempla. In this respect Isocrates’ speeches
are particularly reminiscent of those of the epideictic genre, but Athe-
nian pride also became an increasingly important theme of deliberative
and forensic narrators from the mid-fourth century, as Athens faced the
threat of Philip of Macedon.
chapter twenty-six
DEMOSTHENES
M. Edwards
As one would expect of the orator regarded by most critics ancient and
modern as the master of his art, Demosthenes displays in his speeches
both familiarity with established methods and a readiness to adapt
these for his own purposes. The present discussion will be confined
to the forensic speeches, taking as examples one of the shorter cases
from a private suit (54, Against Conon, which has 44 sections in modern
editions) and one of the longer speeches from a public suit (19, On the
False Embassy, 343 sections).1 We should note, however, that pieces such
as the Embassy and On the Crown speeches, though technically forensic,
were essentially concerned with issues of public policy, and clearly they
were edited for publication as political documents.2
Demosthenes’ forensic speeches tend to open in standard fashion
with a short proem followed by a narrative. A variety of metanarra-
tive narratorial interventions indicates the transition to the narrative,
including expressions such as ‘I shall try to show/tell you’ (e.g. 27.3;
30.5; 36.3) and ‘I must tell/narrate’ (23.8; 34.5). The narrators promise
to tell the whole story, from the beginning and as briefly as possible
(34.5; 36.3; 37.3; 40.5); and they also take pains to interact with the
narratees, with expressions such as ‘it is necessary to remind you, men
of Athens’ (18.17), ‘you will hear’ (32.3), and ‘I ask you to listen to me
favourably, gentlemen of the jury’ (35.5, cf. 55.2). Regularly, the begin-
ning of the narrative is marked by gar, with or without an address to the
narratees (e.g. 18.18; 27.4; 30.6; 40.6; 55.3), though there are exceptions
(e.g. 23.8; 36.4; 37.4). Some Demosthenic speeches may be regarded as
having a single, long narrative in the manner of Lysias, whose end is
indicated by a concluding narratorial intervention such as ‘the dowry,
then (toinun), he acquired in this way’ (27.17, preceded by depositions).
But in the great majority of cases Demosthenes follows the practice of
1 For commentaries on these speeches see Carey and Reid 1985; MacDowell 2000.
2 See, e.g. the remarks of MacDowell 2000: 22 n. 66.
344 part five – chapter twenty-six
his tutor Isaeus, who tended to break up his complex narratives of fam-
ily history into two or more sections, which provide the crucial details
of a case in an apparently straightforward manner.3 This approach is
already evident in Demosthenes’ early private speeches, such as 29,
Against Aphobus III, where a pattern emerges of alternating narrative and
proof sections. Further, although sometimes the end of the first narra-
tive is clearly indicated by a concluding remark such as ‘these are the
facts, gentlemen of the jury’ (30.9; cf. 37.17), on other occasions there
is no clear distinction between the sections of narrative and proof, and
indeed conventional narrative may be abandoned (as in speeches 36
and 38). In the longer speeches in particular, Demosthenes will insert
several sections of narrative, so avoiding monotony. For example, in the
Crown speech he divides the narrative of his own career into three major
stages (18.17–52, 53–109, 160–226) and adds a narrative of Aeschines’
treachery in connection with the war against Amphissa (139–159). The
later sections of narrative are frequently introduced in similar fashion to
the first narrative, with metanarrative narratorial interventions and gar
(e.g. 32.24; 55.23), though there are numerous exceptions in the exten-
sive corpus of this most versatile of orators.
Demosthenes 54, Against Conon, was delivered by Ariston in his pros-
ecution of Conon for serious assault. It recalls Lysias 3, Against Simon,
in its subject matter, but far exceeds the earlier speech in the vividness
of the description of the assault and its consequences. The narrator
in this speech is internal and overt (§3, ‘you will hear’, cf. 9, ‘I would
shrink from saying some of it in your presence’, ‘I shall tell you this’, 11,
‘as you hear’) and makes interventions (3, ‘Conon’s sons camped near
us, as I could wish they had not done’, cf. 4, 7), but admits to gaps
in his knowledge (8, ‘one of them, I don’t know which’). After a brief
proem (§§1–2), whose very first word sets the tone for the rest of the
speech (hubristheis, ‘having suffered gross outrage’), an extended narra-
tive is given of the alleged events (3–12), which characterizes Ariston as
a shy and reserved man, in stark contrast to the drunken and violent
Conon, his sons and their friends. Its start is clearly and familiarly indi-
cated at the end of §2 with the metanarrative narratorial intervention ‘I
shall narrate (diēgēsomai) to you from the beginning (ex arkhēs) how each
incident occurred in the fewest words (dia brakhutatōn) I can’. It is then
broken up into four stages by witness statements, and its end is marked
3 See Usher 1999: 128.

m. edwards – demosthenes 345
by a further set of depositions followed by the summarizing statement

‘that the wounds I received, then (toinun), were not slight or trifling …
I think has been made clear to you on many grounds’ (§ 13). The nar-
rator gives further details later on in the speech to support the char-
acterization of Ariston’s opponents, as a description of their behaviour
at the arbitration hearing is inserted into the proofs at §§26–27: ‘I wish
to tell you now …’ The end of this mini-narrative is marked by ‘and
now (nun) I think most of their defence will concern this point’; and
a brief narrative of the behaviour of Conon and his friends in their
younger days follows at §39: ‘the contempt this man feels for every-
thing sacred I will tell you about … for (gar) I hear, gentlemen of the
jury … surely (dē) Conon, being such a man, is not to be believed on
oath’.
In the first stage of the narrative (§§3–6) the internal narrator de-
scribes the origins of his enmity towards Conon’s sons, which stemmed
from their military service together (note ex arkhēs in §3, which imme-
diately picks up the phrase used to introduce the narrative in §2). He
is overt in the constant contrasts between ‘we’ and ‘they’, beginning
with the reserve of himself and his companions, and their drunken
behaviour: ‘they used to spend the whole of every day after lunch
drinking … but we conducted ourselves in the country just as we were
accustomed to do here’. The repeated use of the imperfect tense indi-
cates that this was no one-off affair (epinon contrasts with diēgomen, ‘we
conducted ourselves’), and imperfect tenses in §4 continue to bring out
the repeated and disgusting abuse of Ariston’s slaves by his enemies,
behaviour which was an ‘outrage’ (hubreōs, the noun here being linked
with ‘brutality’, aselgeias). In contrast, the narrator and his entire mess
took the single action of reporting these events to the general, but his
rebuke of Conon’s sons only served to precipitate actual violence: Aris-
ton was barely saved by the combined intervention of the general, the
taxiarchs and some other soldiers, which indicates the severity of the
attack. In §6 he reiterates both that this was the origin of the enmity
and his own modesty: he had subsequently taken no action, but had
resolved ‘to have nothing to do with men like this’.
The second stage of the narrative describes the events that formed
the basis of the charge (§§7–9). The narrator begins with another fea-
ture of his customary behaviour, an evening stroll in the agora, and how
he encountered (note the historic present parerkhetai) one of Conon’s
sons, drunk and behaving like a drunk. Once more the gathering of
the narrator’s opponents begins with epinon, and Ariston’s meeting with
them is narrated in the present tense (peritugkhanomen, cf. prospiptei). The

story of the fight, like the earlier description of the abuse of the slaves,
is graphically told, with Ariston set upon by three men:
First they stripped me of my cloak, then tripping me up and throwing
me into the mud they jumped on me and beat me with such force that
my lip was split open and my eyes closed; and they left me in such a state
that I could not get up or utter a sound.
The narrator adopts the persona of a retiring young man who will
not repeat the foul language of his assailants, but he does describe
one action which vividly demonstrates their outrageous behaviour (§9,
hybreōs again): ‘he began to crow, mimicking victorious fighting cocks,
and they bade him flap his elbows against his sides like wings’. The
pathos of his own defeat, so total that he is left without his cloak, is
brought out by the wailing of his mother and her maidservants, and by
his having to be carried, bathed, and shown to the doctors.
In the third stage of the narrative (§10) Ariston describes how it
happened (sunebē) that one of his relatives, Euxitheus, and his friend
Meidias chanced on him (peritukhein) as he was being carried home.
This is the fifth use of the verb sumbainō in the narrative after §§3,
4, 6, 8 (and cf. 2, 12), and in addition Ariston was carried home
by chance passers-by (paratukhontōn). This time fortune favours him as
he can stay at Meidias’ house, which was closer to the baths than
his own, and he had more witnesses to his condition. Finally, in the
fourth stage of the narrative (§§11–12), he describes the physical effects
of the assault, and his life is saved by another ‘happening’ (sunebē), a
spontaneous haemorrhage. That his condition was life-threatening is
confirmed by a reported narrator, the doctor (‘the doctor said that
…’, 11, 12), supported by his deposition. The narrator summarizes his
story in §13 by again referring to the ‘outrage and brutality’ of his
opponents (inverting the phrase used in §4). By his graphic narrative
he has certainly ‘made things clear’ to the narratees.
The circumstances of speech 19, On the False Embassy, were very
different. Here Demosthenes himself was prosecuting his great rival,
Aeschines, in 343 for his alleged conduct on the embassy that had
secured the discredited Peace of Philocrates of 346. In this long speech
§§10–66 form the main narrative, but two interesting additional narra-
tive sections are inserted at §§192–198 and 229–231. The metanarrative
narratorial intervention ‘I wish to remind you of what doubtless most
of you remember’ (§9) introduces the main narrative and also, in a
m. edwards – demosthenes 347
manner reminiscent of Andocides in the Mysteries speech, reflects the

internal narrator’s constant attempts to interact closely with the inter-
nal narratees, men of Athens, as he calls on their memories of events
in which they took part (e.g. §§19, 27, 33, 45). The interactional particle
toinun, rather than gar, then marks the start of the narrative in §10. The
narrator is overt, e.g. ‘I wish to tell how’ (29; cf. 17, 25, 27, 35, 44, 65),
as are his narratees, e.g. ‘when one of you hears me describing some
event and it seems to him terrible and incredible’ (25). A special char-
acteristic of this narrator is his constant use of (rhetorical) questions,
often briefly put: §§24, 25, 27, 30 (the simple ‘how could he?’ in refer-
ence to Aeschines), 33, 42–43, 50, 52, 53, 55, 60, 63, 66. His narrative
ends, indeed, with four questions, three of which are short and refer to
Aeschines (compare the ending of the first narrative of the Crown speech
at 18.51–52).
At §§192–198 the narrator relates two anecdotes which highlight
Aeschines’ character in comparison with that of the actor Satyrus and
which are introduced as ‘a trifling story that has nothing to do with
the embassy’. In the first story the narrator relies on hearsay (‘as the
story goes’), though noticeably Satyrus’ moving plea to Philip for the
release of his friend’s captive daughters is quoted in direct speech.
In the second story the narrator is potentially internal, yet absent
at the crucial event and hence relying on ex eventu knowledge (the
Olynthian woman was modest, ‘as the event proved’: §196) and hearsay
(‘as Iatrocles told me on the next day’: §197). What is more, he relies
on the knowledge of the narratees: ‘the story was told in Arcadia
… Diophantus related it before you, as I will compel him to testify,
and it was common talk in Thessaly and everywhere’ (§198). But the
immediate juxtaposition of a story whose provenance is unclear with
one that was common knowledge has the effect of giving more credence
to the former, and so establishes the contrast between the characters of
Satyrus (and even Philip) and Aeschines for the narratees all the more
strongly.
Finally, in §§229–231 the narrator imagines the outcome of an ac-
quittal of Aeschines with an address to the narratees: ‘what in truth will
be the story told about you, if you acquit this man?’ The story centres
on comparisons between the probity of Demosthenes and the wicked-
ness of his fellow ambassadors Philocrates, Phryno, and Aeschines, but
added impact is gained from the punctuation of the narrative with brief
questions: ‘so what?’, ‘what happened after that?’, ‘what do you think?’,
‘what did they think of the prosecutor?’ The addition in §232 of two
rhetorical questions addressed directly to the narratees (‘Who, men of

Athens, seeing this example, will wish to prove himself an honest man?
Who will go on an embassy for nothing …?’) forcefully underlines the
choice that confronted them.
Demosthenes constantly plays in his speeches on the venality of
Aeschines and his own probity, and in turn on their respective treachery
and patriotism.4 One of the key tools he uses in this political battle is
narrative—there is, for example, little or no real evidence offered in
the Embassy speech—and Demosthenes, like Lysias, had an undeniable
talent for presenting his story vividly and persuasively. One feature of
his narratives, as we have just seen, is his extensive use of rhetorical
questions; but a second, even more noticeable, feature that marks him
out strongly from earlier orators is the vehemence of the personal
attacks his narrators make on their opponents. Some examples have
been noted, but he naturally saves his best for Aeschines (18.129):
I am not at a loss for what I should say about you and your family,
I’m at a loss where to begin. Shall I tell how your father Tromes was a
slave in the house of Elpias, who taught in an elementary school near
the Theseum, and how he wore thick shackles and a wooden collar? Or
how your mother practised daytime marriages in the hut near Heros
the bone-setter’s place and how she brought you up, her pretty doll, the
consummate third-part actor?
4 See further on this, e.g., Edwards 1994: 45ff.

chapter twenty-seven
AESCHINES
M. Edwards
During his defence in the Embassy trial, Aeschines laments the fact
that his political life was enmeshed with that of a man whom he
brands ‘a charlatan and a criminal’ (2.153). Unfortunately for him,
it was Demosthenes who ultimately triumphed, both politically and
rhetorically, and Aeschines’ speeches are only now beginning to receive
the attention that they deserve.1
It is in his narratives that Aeschines is at his best, and these form
a significant proportion of all three of the long (196, 184, 260 sections
respectively), technically forensic speeches that have come down to us.
He displays a variety of methods within the three. So, in speeches 1 and
3 the start of the narrative is delayed by extensive discussion of the laws,
in a manner reminiscent of Antiphon’s Herodes speech. Then, while
speech 3 has an extensive narrative of Demosthenes’ political career
(3.58–167), speech 1 has two major sections of narrative on Timarchus’
private and public life, interwoven with proofs, and speech 2 has several
narratives, of the first embassy to Philip (2.12–56), Athenian military
difficulties after the fall of Amphipolis (70–74), the activities of king
Cersobleptes of Thrace (81–93) and the second embassy to Philip (97–
118). In all three further narrative details are interspersed in the proofs.
Most of the main narrative sections are introduced by a metanarrative
narratorial intervention, ‘I shall make my accusation,’ 95, ‘I shall try to
lay it out more clearly in my account’; 2.11, ‘I shall begin’, 96, ‘I shall
give a defence’; 3.57, ‘I shall speak firstly’. In speech 2 Aeschines also
uses ‘I wish to remind you’ (70) and ‘it remains for me to speak’ (81).
The usual promise of speaking briefly or from the beginning is absent
from Aeschines’ introductory remarks, but all the later narratives begin
with gar.
1 Carey 2000; Fisher 2001.

350 part five – chapter twenty-seven
As an example of his narrative practice we shall take speech 1, Against

Timarchus, which was delivered by Aeschines in his successful prosecu-
tion of one of Demosthenes’ supporters in 346/5, to forestall an attack
on himself for treason. The basis of the charge was that Timarchus had
addressed the assembly when debarred from doing so by his immoral-
ity, hence Aeschines immediately stresses his own moderation in his
opening remarks (§§1–3), and the narrative is delayed by a prelimi-
nary discussion of the laws on decency. A transitional section (§§37–
39) reiterates his own decent character, before he narrates Timarchus’
debauched early adulthood in §§40–70. He adopts a mixture of short
and longer narrative sections interspersed with argument,2 as he tells
how Timarchus lived in turn in the houses of Euthydicus (§40), Mis-
golas (§§41–52), Anticles (§53), Pittalacus (§§54–55) and Hegesander
(§§56–70). An immediately striking feature is how overt the narrator
is here, for example, ‘Timarchus did many ridiculous things in those
days, one of them I wish to narrate to you’ (§ 43) and ‘the abuses and
outrages that I have heard were done to the person of Timarchus by
this man were such that, by Olympian Zeus, I could not bring myself
to tell them to you’ (§55). This is a feature of Aeschines’ oratory which
is supplemented by narratorial interventions (§§ 48, 51, 58, 60, 61, 64,
69) and generalizations (§§44, 48, 49, 60). There is regular interaction
with the narratees, in form of references to their own knowledge of
the man or the places involved (§§44, 53, 56, 65, 69–70), and at §49
the narrator prepares them with the only occurrence of oratio recta in
this narrative: ‘Why do I say this in advance? So that when you sud-
denly see him you will not be surprised and have some thought such as:
“Heracles! This man is not much older than Timarchus.”’ The inter-
action serves to mask Aeschines’ total lack of solid evidence, and he is
at pains to create for himself a trustworthy, modest character by calling
on the gods (§§52, 55) and repeatedly avowing that he does not wish to
use the words which describe Timarchus’ lurid behaviour (§§41, 45, 52,
55, 70)—which does not, of course, prevent him from indicating clearly
on numerous occasions what this was (§§40, 42, 52, 53–54, 55, 57, 70).
Two further observations may be made on this narrative. First, like
other orators Aeschines uses the historic present tense to highlight cer-
tain events, here the actions of Timarchus’ various partners (§43, ‘they
find him dining in a lodging house with some foreigners’, 53, ‘after this
2 Usher 1999: 281.

m. edwards – aeschines 351
Anticles takes him up’, 55–56, ‘about this time Hegesander sails back
… this Hegesander arrives’, 57, ‘when he could not persuade Pittalacus
he assails the man in person’) and the story of Pittalacus (§§60–62). The
latter immediately follows the description of the attack on Pittalacus
and his belongings by Hegesander, Timarchus, and some others ‘whose
names I prefer not to mention’ (the narrator perhaps betrays the fact
that he is not omniscient):
First they smashed his equipment and threw it into the street, dice and
dice cups and other gaming items, and they killed the quails and cocks
which the wretched man adored, and finally they tied Pittalacus himself
to a pillar and gave him an inhuman whipping for so long that even the
neighbours heard the uproar.
Secondly, in comparison with Demosthenes, Aeschines is relatively

sparing in his use of questions; but in speech 2 there is a noticeable
concentration of questions in the Cersobleptes narrative, which indeed
ends with a question aimed against Demosthenes (2.81, 86, 87, 88, 92,
93); in speech 1 questions are asked of the narratees before the direct
speech quoted above (§49) and after a call on their knowledge (§65), in
connection with the testimony of Hegesander, which Aeschines knows
he will refuse to swear an oath to (§67), and once again at the very end
of the narrative in §70.
Similar features may be detected in the second major narrative of
this speech concerning Timarchus’ political career (§§95–116). Again
the narrator is overt (e.g. §§ 95, 101, 102, 106, 109, 116), commenting
(§104) and calling on the knowledge of the narratees (§116), and reit-
erating Timarchus’ lewd lifestyle (§§95, especially 105, 107). He is still
sparing in the use of questions, two of them being asked in connection
with Timarchus’ scandalous activities on Andros (§108, cf. 109, 113),
and also in the use of direct speech, which is given to Pamphilus when
he acts as a secondary narrator in §§110–111. The narrator has told
how Timarchus squandered his inheritance—he even sold the farm at
Alopece, which ‘his mother begged and pleaded with him, as I have
heard, to spare and not to sell, but if nothing else to leave this at least
for her to be buried in’ (§99).3 Unsurprisingly, a man who wasted his
own property was also profligate with the state’s, which prompted the
decent Pamphilus to denounce him and his partner Hegesander before
the assembly (§110: ‘Men of Athens, a man and a woman together are
3 The effectiveness of this unsubstantiated remark is noted by Carey 2000: 57 n. 105.

352 part five – chapter twenty-seven
stealing 1,000 drachmas of yours’). It will have been due in no small

measure to his employment of emotive narrative that, despite the flim-
siness of his case, Aeschines secured the condemnation of Timarchus.
Aeschines’ inferiority as an orator to Demosthenes is universally
recognized, but he was in his own right a master of narrative technique.
Indeed, he was far more at home in narration than argumentation, and
extensive narratives are a key feature of his oratory, even when broken
down into separate sections. The fact that he triumphed in two of the
three cases whose speeches survive indicates the persuasiveness of his
narratives, which, as is the case with Andocides, give the appearance of
being products of natural ability. His narrators tell vivid stories, interact
regularly with the narratees and, in line with the fashion of the times,
do not mince their words in their biting attacks on their opponents.
Conclusion
Some of the main points that have emerged in the course of this pre-
liminary survey of the extensive corpus of Attic oratory may now be
summarized. The term ‘narrative’ has a particular connotation in the
oratorical context, being one of the major divisions of a speech. It is
particularly employed in forensic speeches, though mythological and
historical narratives are important elements of epideictic. In all the
orators surveyed the start of the narrative is clearly defined by meta-
narrative narratorial interventions, and various types of concluding
remark indicate its close. The narrative may, in line with the theoret-
ical discussion of Aristotle in the Rhetoric, be a single passage within
the speech, and there is no doubt that Lysias was the master of this
form of composition, using it to characterize the main players in the
case. But it emerges already from the works of the earliest exponents
of the genre that in practice a narrative might well be divided into
two or more sections, which deal with different temporal stages or the-
matic aspects of the story. Narrators in oratory are regularly internal,
even in speeches where the speaker is acting as an advocate (sunēgoros)
for the actual litigant, as is technically the case in Demosthenes’ Crown
speech. These narrators may represent the actual author of the speech,
as in Andocides 1 and Demosthenes 19 (this is especially the case with
deliberative rhetoric), though the great majority of forensic speeches
were composed by a logographer for a client. Since they are trying to
persuade their audience, usually of the mendacious character of their
m. edwards – aeschines 353
opponent, narrators in oratory are overt, lavishly commenting on the

events they recount. Similarly, the narratees are addressed on a regu-
lar basis, and may be asked to use their own recollections of events in
reaching their decisions; but they are more often external narratees in
the forensic context. A particular type of narratee is the speaker’s oppo-
nent, who will be addressed in apostrophe and, in the later orators, with
increasing vehemence. For Aristotle (Rh. 3.13) the necessary parts of a
speech are the statement of the case (prothesis) and proof (pistis): however
the gap between theory and practice was often wide, and Andocides,
Lysias, and Aeschines in particular will have disagreed, instead favour-
ing narrative.
part six
PHILOSOPHY
chapter twenty-eight
PLATO
K.A. Morgan
Introduction
To what extent does an analysis of the Platonic dialogue belong in

a history of ancient Greek narrative? A strict conception of narrative
might result in the inclusion only of embedded narratives such as the
Myth of Er in the Republic.1 More broadly, one might wish to include
those dialogues that self-identify as narratives. Thus the Republic begins
with Socrates’ narrative of his trip to the Piraeus, and the Symposium
with the complex framing narrative of Apollodorus. But what are we to
make of cases like the Meno or the Sophist, where we plunge into direct
dialogue? Are these ‘narratives’?
We learn from Diogenes Laertius (3.50) that ancient classifications
of the dialogues sometimes distinguished between ‘narrative’ and ‘dra-
matic’ dialogues, although Diogenes himself disapproves of this classifi-
cation.2 One might, following this lead, exclude the dramatic dialogues
from consideration (with the exception of secondary narratives con-
tained therein). Narratologists are fond of quoting the Republic (392d2–
394c5) for the distinction between pure narrative and mimesis.3 The
same passage envisages the possibility that work might be ‘mixed,’ that
1 Genette 1980: 29 notes that Spinoza’s Ethics are not narrative, since they do not
tell a story. Yet Plato’s dialogues ground themselves in specific narrative situations; they
tell the story of a particular argument made in a particular context. Cf. Gill 2002:
153–155.
2 Plato was probably not the first to write either narrative or dramatic Sokratikoi
logoi (cf. Clay 1994: 27–28, 42–43; Kahn 1994: 89, 95, 100–101). For modern views that
dramatic dialogue was Plato’s invention and emerged in the Academy, see Thesleff
1982: 61; Tarrant 1996: 136. Plato’s originality as a writer of dramatic dialogues was a
matter of controversy in antiquity also. Aristotle may have believed that Alexamenus
of Teos was the first to write them (but see Haslam 1972 for the ancient system of
classification that may have caused confusion here).
3 Genette 1980: 162–163.
358 part six – chapter twenty-eight
is, combine mimesis and narrative (diegesis). This schema would allow
us to classify Platonic dialogues either as drama (mimesis) or as ‘mixed’
pieces.4 This seems attractive, yet it is important to note that in this
account of diction, Socrates states that everything written by poets is
narrative. Sometimes narrative is pure, sometimes it is effected through
mimesis, and sometimes through both (392d5–6). When a poet repre-
sents the voice of someone other than himself, he ‘composes his nar-
rative through mimesis’. On this account, all poetic production is nar-
rative; distinctions arise only when we ask about the means of effect-
ing the narrative. We might draw a similar conclusion from examining
the opening of the Theaetetus. There, the framing dialogue presents us
with a putative text of a Socratic dialogue whose authority is said to be
Socrates himself. This text will be read aloud, but its author, Euclides,
notes that he has written the text not as a record of the conversa-
tion he had with Socrates (with the ‘he said’s and ‘I replied’s). He has
taken out these ‘narratives’ because they ‘create bother’ and has instead
attempted to recreate the original conversation between the interlocu-
tors. The Theaetetus thus insists both that the body of the dialogue is a
narrative, and presents that narrative as a dramatic conversation. It the-
matizes the question of whether or not the main dialogue is narrative,
and, by implication, extends that questioning to the frame dialogue.5 If
that frame dialogue did not exist, nothing would distinguish the narra-
tive practice of the Theaetetus from that of the Meno. Genette labelled the
suppression of one or more narrative levels pseudo-diegetic.6 This absence
of a framing narrative has a precedent in Hesiod (→), but it is also clear
that Platonic practice in this respect influenced Dio (→) and Lucian
(→).
The only safe conclusion to draw from the passages discussed above
is that all Platonic dialogues are conceived as narratives.7 I propose to
adopt the same approach: even dramatic dialogues are by implication
narratives. The theoretical distinction between mimetic and narrative
forms may have been formulated by Plato, and this formulation did
4 Although Socrates in the Republic declares that the narrative of a good man will
contain only small amounts of mimesis, Plato’s own practice is far from conforming to
this ideal. Cf. Velardi 2000: 126–127; Laird 1999: 51–53.
5 For a more extensive treatment of these issues, see Morgan 2003.
6 Or ‘reduced metadiegetic’ (Genette 1980: 236–237).
7 Clay 1994: 47 blurs the boundaries further (rightly), by pointing out that a narra-
tive such as the Republic becomes dramatic when it is read aloud, since the reader takes
the part of Socrates.
k.a. morgan – plato 359
indeed influence the narrative practice of subsequent writers, but the

merging of mimetic and narrative forms and the narrative experimen-
tation that Hunter sees at work in Theocritus (→) and other Hellenis-
tic poets find a significant precedent in Plato. Indeed, one of the hall-
marks of Platonic writing is the focus on problems of narrative defini-
tion and authority, a preoccupation that Lycophron (→), Plutarch (→),
and Philostratus (→) (among others) share. Plato’s corpus explores the
effects of competing narratives and different narrative voices. This com-
petition, moreover, has philosophical implications: what account are we
to give of the world, and how are we to communicate our account’s
content? A full study of Platonic narrative would account for its rich-
ness and variety while also exploring the philosophical implications
of Plato’s narrative choices.8 The task of this present chapter is more
restricted.
The narrator and the forms of Platonic dialogue
Any discussion of Platonic narrative must begin by noting that Plato

himself, the author of the dialogues, never appears in his own works.
The work of narration is performed by characters who, although many
have regarded them as Platonic mouthpieces, are not explicitly endow-
ed with authority to speak on behalf of ‘Plato’. In the entire corpus we
hear of Plato only to be told that he was not present at the death of
Socrates. There are few places where we feel the looming presence of
the ultimate author. These may be described mostly as instances of ‘Pla-
tonic’ irony,9 as when Critias remarks on how opportune a coincidence
it is that his account of Athens and Atlantis agrees to such an extraordi-
nary extent with Socrates’ ideal city (Tim. 25e), or when Socrates in the
Protagoras (315e) describes a good-looking young boy named Agathon—
the Agathon who was to grow up to be a famous tragedian and who
would have been well known to Plato’s readers. Such interventions,
however, are insignificant compared to the overall absence of any Pla-
tonic voice. The Platonic narrator is never Plato.
The example of the Theaetetus, discussed above, may provide a clue
for interpreting Platonic narrative situations. We might conceive the
8 The narrative structure of the Platonic dialogue is increasingly the focus of schol-
arly attention, as Blondell 2002 and the essays in Casertano 2000 show.
9 Rowe 1987.
texts of the dialogues as Platonic constructions representing a version of

a written text preserved by various disciples and subsequently read out
loud.10 In the Theaetetus the frame presents the traces of this preserva-
tion. On other occasions (as in the Phaedo) we are not given the occasion
of a reading but are made to move back to a verbal narration that pre-
cedes textual codification. Sometimes the narrative takes us even fur-
ther back to Socrates as narrator (e.g. Republic or Protagoras). The final
‘unmediated’ level is that of an unframed dramatic dialogue (such as
Gorgias). If we do not allow the Theaetetus to organize interpretation in
this way, a more diffuse picture of continual experimentation emerges.
My conceptual hierarchy of narratorial mediation or non-mediation
should not, however, be understood to reflect a chronological progres-
sion in Plato’s narrative technique. As Thesleff has pointed out, it is
natural to assume an increase in stylistic complexity over time, but in
Plato’s case this would be a dangerous move.11 The dialogues usually
regarded as Plato’s latest are entirely dramatic, and it is not unrea-
sonable to see the Theaetetus as a turning point after which dramatic
dialogue was Plato’s preferred format.12
In my discussion of narrators I shall distinguish between framed and
non-framed (dramatic) dialogues. In framed dialogues an introductory
section precedes the philosophical discussion and motivates the main
narrative, sketching the setting and sometimes dealing with the history
of the reception of the discussion. The frame may be narrative (that
is, addressed to a silent narratee) or dialogical. In the case of dialogi-
cal frames, one of the interlocutors (like Phaedo) becomes a secondary
narrator. In the case of narrative frames, the narrator may function as
an internal primary narrator (as when Socrates reports past conversa-
tions), or as an external narrator (as when disciples narrate Socratic
conversations that they have heard from others and at which they were
not in attendance). The philosophical discussions in framed dialogues
are thus analeptic. In dialogues without frames, the main conversa-
10 I do not, of course, mean to imply that these texts were actually preserved by
disciples. The history of such ‘texts’ is a Platonic construction.

11 Thesleff 1982: 53.
12 For an attempt to perform a kind of textual archaeology on dialogue frames, see
Thesleff 1982. He argues that the dramatic dialogue was at first confined to circulation
within the Academy (63) with information about the setting being supplied orally
(162), whereas dialogues with narrative frames were intended for wider publication
(63). The direct dialogue intended for publication would then have been a secondary
development (162). Cf Tarrant (1996: 136, 145) for a similar distinction between early
narrative dialogues and later dramatic dialogues.
tion of the dialogue is presented directly. In these cases (the majority)

I shall examine embedded narratives and their narrators. Although, as
we have seen, it is possible to view these dialogues conceptually as the
core of more complex narrative situations, they should nevertheless be
conceived primarily in their own terms, as the unmediated presentation
of philosophical exchange where the drama impinges directly upon the
hearer. Events are presented almost simultaneously (whereas in framed
dialogues the narration is subsequent to the events described).13
The Socratic narrator
Socrates acts as narrator in a dialogue frame five times. Twice, the

frame is dialogical: Socrates acts as internal secondary narrator and
reports the action of a past philosophical discussion to a friend (Euthyde-
mus and Protagoras). In the three instances of narrative frames, Socrates
acts as internal primary narrator, reporting a past discussion to a silent
narratee (Charmides, Lysis, and Republic). The distinguishing feature of
this group of dialogues is that they allow the reader the impression
of unmediated interaction with Socrates, whose narratorial presence is
overt. When the frame narratee is silent, the reader is given the illusion
that (s)he is the narratee, an illusion that would be especially effective
if the dialogue were being read aloud. An obtrusive framing conver-
sation gives the audience a window onto a world of lively intellectual
exchange. In both cases, we see how philosophical conversation and
analysis are greedily sought out and consumed. These Socratic nar-
ratives do not have formal endings, although they do end at a natu-
ral point of closure: the termination of the discussion. In Euthydemus
the frame conversation between Socrates and Crito reasserts itself, with
Socrates urging Crito to the practice of philosophy. Protagoras, Lysis, and
Charmides end with Socrates’ narration of his closing comment (and in
Protagoras and Lysis he narrates his departure). Only the Republic ends,
vividly, with no reminder of the narrative format, as Socrates contin-
ues to address directly his internal narratee Glaucon and exhort him to
virtue.
13 Rutherford 1995: 71 notes that when Socrates is the narrator, the narrative frame
follows the events closely, whereas others’ narration looks back over a more extended
period, thus giving an opportunity for pathos and foreshadowing (Cf. Clay 1994: 44–
Irony is an important feature of the Socratic narrator. We should,

therefore, be on our guard against taking Socratic self-characterization
at face value. Thus in the frame conversation of the Euthydemus, Soc-
rates tells his friend Crito that he was so impressed by the display of
the sophists Euthydemus and Dionysodorus, that he wants to become
their pupil (272b). Although the intellectually plodding Crito appears
to take him seriously, there is small danger of any reader doing so.
Again, in the Republic, Socrates represents himself as terrified by the
onslaught of Thrasymachus: ‘I was panic-struck and was afraid when I
looked at him … I said trembling a little …’ (336d). Or in the Charmides
Socrates says he was overcome with lust at the sight of Charmides’
body (155d–e). Only slowly does he regain confidence and find himself
able to speak (156d). Yet neither in the Republic nor in the Charmides
should we conclude that Socrates was overcome either by fear or by
lust, any more than that he wants to be the pupil of the sophists.
Socrates does not mean what he says. He presents himself the way he
does in order to draw out an interlocutor and sometimes to disarm (not
always successfully) suspicions that he is manipulating the conversation.
Irony is a generic feature of the Socratic narrator.
Socrates uses his narrative to characterize explicitly both himself
and others. Closely related to Socratic irony is the self-deprecating
humorous exaggeration of Lysis 218c, where Socrates reports that he
was rejoicing ‘like a hunter’ at capturing his intellectual prey, when a
strange suspicion entered his head that he was wrong. This causes him
to cry out in distress, in mock tragic fashion. Analeptic narrative thus
enables self-deprecation, a means of generating sympathy for the nar-
rator because of the difficulty of his intellectual quest. To this episode
we may compare Republic 357a, where Socrates (at the beginning of
Book 2) reports that he supposed he was done with the argument,
only to be told that his task was not nearly complete. A similar tech-
nique is employed at Charmides 154b–c, although the deprecation there
is not intellectual. Socrates confesses to his silent narratee, ‘My friend
… with regard to those who are beautiful, I am a white measuring
line [i.e. one without measuring marks]’. The narratee is drawn into a
world, as indeed we all are, where we know ‘what Socrates is like’. The
genre presumes and constructs our familiarity. Plato also has Socrates
employ more indirect self-characterization. Thus his allusion to Homer
45). As Velardi 2000: 128 points out, the narrated dialogues represent the transmission
of Socratic conversation through time.
at Protagoras 315b marks him as a man of culture. Finally, we may note

Socrates’ report of Charmides’ blush (Charm. 158c), which made him,
‘look even more beautiful, for his modesty suited his age. Then he
replied nobly …’ Socrates clearly adheres to broad cultural conven-
tions (youths should be modest), but his remarks on the heightening of
the boy’s beauty and his ‘noble’ reply, characterize him as much as they
do Charmides.
When it comes to his characterization of his interlocutors, Socrates
displays a degree of knowledge that borders on omniscience. Char-
acterization is achieved by describing the tone of an interlocutor’s
speech, or his reaction to speech. Glaucon in the Republic can speak
‘with a laugh’ (451b), or can react with surprise to Socrates’ announce-
ment of the immortality of the soul, ‘looking at me full in the face
in amazement’ (608d). In the same dialogue, Thrasymachus, when
he realizes he is being bested by Socrates, begins to answer ‘reluc-
tantly’ (342c, e). At the close of a long sequence of argumentation
that reduces Thrasymachus to self-contradiction, Socrates comments,
‘Thrasymachus agreed with all my suggestions not as I now easily nar-
rate it, but with difficulty and being dragged along, with an amazing
amount of sweat, since it was summer, and then I saw it, although I had
never seen it before—Thrasymachus blushing’ (350c–d). Socrates’ con-
trol over his narrative here is obtrusive. In order to render the progress
of the argument clearly, he did not reveal to his narratee the faltering
and reluctant answers of the sophist. But once the argument reaches
its conclusion he characterizes Thrasymachus retrospectively. His por-
trayal of the sophist’s reaction is more devastating for being presented
concisely.
But Socrates goes farther than this. He reveals to his narratees even
the thoughts and motives of others. Sometimes he turns his interpre-
tative lens on the dynamics of homoerotic courtship (as we know, love
is the only area where Socrates claims to be an expert: Symp. 177d).
Thus in the Charmides we note the remarkably thick description of the
motivations of Critias, who
had plainly been in distress for a while, and felt jealous of his honour
with respect to Charmides and the rest of those present. Although he had
previously and with difficulty restrained himself, then he could not, for it seems to
me that what I had suspected was absolutely true, that Charmides had
heard this reply about moderation from Critias. Now Charmides, since
he did not wish to maintain the argument himself but wanted him [Critias] to reply,
was stirring him up. He kept on pointing out that he [Critias] has been
refuted, and he [Critias] could not bear it, but he seemed to me to get
angry with him, like a playwright gets angry with an actor who recites
his lines badly. (162c–d)14
Socrates’ current narrative confirms his past suspicions (which he com-

municates to his narratee), and the effect here is to underline his sta-
tus as authoritative narrator. This seeming omniscience extends to the
dynamics of non-erotic exchange with more mature intellectuals. In the
case of Thrasymachus, ‘he clearly wanted to speak so that he could
improve his reputation, since he thought he had an excellent answer,
but he pretended that he was anxious that I should answer’ (338a). The
case of Protagoras is similar: Socrates suspects that he wants to give
a formal display (Prt. 318c). Later on he senses Protagoras’ disappoint-
ment with his own performance and that he did not want to continue
to answer (335a; cf. 348c). Socrates’ conclusions are not unreasonable;
they are based upon his interpretation of the conversational dynam-
ics. Yet, as I have noted, the effect is remarkable. Despite Socrates’
attempts to undermine himself through humour and irony, his narra-
tive technique is powerful and authoritative. He presents and interprets
a wide array of conversational features, reflecting at a narrative level his
more general claim that although he has no knowledge himself, he is
(preternaturally) gifted at exposing the mystifications and hypocrisies of
others. Socrates’ intrusions into the narrative, however, make his own
bias overt. The reader becomes aware that Socrates selects and inter-
prets in line with his own interests and goals. One might ask how fair
Socrates’ interpretations are (given that they mostly show his opponents
in a bad light), and even question the precision of his edited versions of
the conversations.
Disciple narrators
Phaedo, Symposium, Theaetetus, and Parmenides present Socratic conversa-

tion as reported by his disciples. Two (Parmenides and Symposium) open
with narrative frames, while two (Phaedo, and Theaetetus) open with dia-
logical frames. At the beginning of the Parmenides the narrator, Cephal-
us, tells an unnamed interlocutor about the search he and others (‘we’)
engaged in to find someone who could narrate to them Socrates’ con-
14 Cf. Charmides 169c; Lysis 207b.

versation with Parmenides. The search leads to a certain Antiphon,

who has heard the narration from Pythodorus and who is persuaded
to report it. The vertiginous effect of multiple nested narratives is
repeated in the complex opening conversation of the Symposium (Apol-
lodorus repeats to an unnamed friend the contents of a conversation he
had with Glaucon, who had been told the story of the symposium by
someone who had talked to Phoenix. Apollodorus has heard the narra-
tive from Aristodemus, who was the source for Phoenix and who was
present.).15 Here again, Plato’s experiments with narrative were influen-
tial for authors to come, and the complexity of the Symposium was semi-
nal. This dialogue presents, as well as the nested frame narratives men-
tioned above, multiple embedded narratives by the symposiasts, among
which is Socrates’ famous report of what the seer Diotima told him
concerning the mysteries of love. Lycophron (→), Philostratus (→), and
Lucian (→) all seem to have been influenced by the dialogue and by the
role of Diotima.
Theaetetus and Phaedo, by contrast, begin with actual (as opposed to
implied) conversation,16 where one interlocutor extracts the story of
Socrates’ discourse from another who was there (Phaedo) or who heard
it from Socrates (Theaetetus). In all these dialogues, there is considerable
stress on the accuracy of the narrative. Phaedo was there. Apollodorus
(Symposium) has recently reviewed the story and checked details with
Socrates. Antiphon (Parmenides) is said to have heard the story so often
he knows it by heart. Euclides, too, confirmed details with Socrates,
and as we have seen, even wrote his narrative down.
These ‘disciple narratives’ share an important feature with the So-
cratic narratives described in the previous section: they create a world
in which the narration of Socratic conversation is a matter of passion.17
The reader of the dialogues is presumed to share that passion. Just
as we see Socrates waylaid so that people may talk with him (Repub-
lic, Lysis), so the framing conversations discussed here present people
begging to be told what happened at discussions they did not attend.
15 On the complex narrative structure of the Symposium and its significance, see
further Halperin 1992 and Velardi 2000: 112–120.

16 Even the narrative frames have a strong element of conversation, most obviously
the Symposium, which opens with the words, ‘I am not unpracticed in [relating] the
matters you ask about.’
17 Tarrant 1996 stresses how Plato wants the reader to be immersed in the world of
oral tradition in these dialogues (137), and to ‘enter the world of the narration as well as
the world of the narrated story’ (132).
This is explicit when the frame narrator engages in conversation, but

is implied even when the narratee is silent. Indeed, non-presentation of
the narratee makes the point very subtly: the first words of the dialogue
respond to a question (what did Socrates say?) that was asked before
the beginning of the text. The genre assumes we know the question has
been asked and that we ask it ourselves. We enter a world of reception
in which these stories are told again and again, and where narrators
care about getting them right and indicating where memory may not
be exact. Thus at Symposium 199c, the embedded narrator Aristodemus
says that Socrates began to speak ‘more or less like this’. At Protagoras
336e even Socrates draws attention to the difficult project of narration,
‘After Alcibiades, so far as I remember, it was Critias who spoke.’18 At
Euthydemus 275d, Socrates is so concerned about the ability of his narra-
tion to do justice to the original conversation that he invokes the Muses.
We see here both an allusion to the topos of poetic inspiration, and the
standard diffidence of the orator or poet when faced by the greatness of
his subject matter. Yet these topoi are deflated because Socrates’ admi-
ration for the sophistic performance is only assumed.
The frames of all these dialogues, then, explore the authority sup-
posed to stand behind the narrative.19 In some the frame serves only
to identify a more authoritative narrator (sometimes at several removes)
whose memory is then reported in a mix of direct and indirect dis-
course (we might call these figures secondary frame narrators). This
stress on narrative authority has philosophical implications: we are
called upon to examine our criteria for ascribing credibility to any nar-
rative, and also to consider that although the dialogues are presented
as accurate models of philosophical interaction, our main task is to
conduct such discussion ourselves rather than become obsessed with
reportage.
In contrast with the Socratic narrator, disciple narrators are usually
colourless, transparent, and covert. Once we leave the introductory
frame, the influence of the frame narrator ceases to be felt. One result
of this is that the reader has inferential knowledge only of the emotions
of the participants; we have no Socrates to tell us what Protagoras
is thinking. Moreover, in the Theaetetus, and Parmenides, the reader is
never returned to the frame, and the dialogue ends with remarks by
18 Cf. Euth 275b, ‘When I had spoken pretty much these very words’.
19 For further discussion of narrative authority in the Platonic dialogue, see Clay
1992: 117.
the interlocutors in the embedded narrative.20 The distancing qualities

of the frame fade, and the final impression is the immediacy of the
discussion. In Symposium and Phaedo, more obtrusive closure is provided.
The end of the Symposium rapidly reviews Socrates’ actions the next
day after the close of the party and then sends him home to bed. The
Phaedo ends with the moving summary statement by the frame narrator,
Phaedo (‘This was the end of our companion …’). Only in the case of
the Phaedo is the frame narrator present at the events described. This
also explains the greater emotional investment by the narrator in the
narrative here.21 Unlike the other colourless disciple narrators, Phaedo
(himself historically an author of Socratic dialogues) can describe how
he and others were feeling during Socrates’ final day (58e, 88c, 102a,
116a, 117c–e). He is familiar enough with Socrates to be teased by him
(‘he was accustomed to joke about my hair’ 89b), and to be able to
comment on Socrates’ mannerisms (‘he looked straight ahead, as he
often used to’, 86d). These iterative elements add to the impression,
discussed above, of a world of frequent and passionate discussion, which
the reader is invited to join.
On two occasions in the Phaedo the frame dialogue resumes as Eche-
crates, the frame narratee, comments and asks questions (88c–89b,
102a–b). It is hard not to think that his reactions are a guide to reader
reception and response. Echecrates models the narratee’s despondency
when Socrates’ argument for immortality runs into problems, and his
or her curiosity as to Socrates’ reaction to challenge. Later, he interjects
approval at a Socratic proof (‘It seems to me Socrates spoke with
amazing clarity,’ 102a).22 Indeed, a function of some framed Platonic
dialogues seems to be the emphatic presentation of and comment on
the effect of philosophic conversation on its audience.23 We have seen
above how Socrates unmasks the thoughts and motives behind the
reactions of his interlocutors in his narrations. The Phaedo shows us
20 Cf. the ending of the Republic, discussed above. The Parmenides, as Rutherford 1995:
274 remarks, is something of an oddity. Description and comment end at 137c, and by
the close of the dialogue even the compliant interlocutor has faded away.
21 Ebert 2000: 53 draws attention to the unusual prominence of Phaedo as narrator
among the narrators of the other framed dialogues.

22 Phaedo confirms ‘that was what everyone present thought’.
23 For the significance of the emotional effect of Socratic dialectic on interlocutors,
see Blank 1993: 435–437, who draws attention both to the description of the strong
reactions of Socratic respondents and to the important point that Socratic conversation
can have this effect even on someone not present at the original exchange (as is the case
with Echecrates in the Phaedo).
how the passions aroused in the embedded narrations are resumed in

the frame. Other examples of this occur in the Euthydemus, where we
again return to the frame conversation in the course of the narration,
as a device to focus the attention of Crito, the frame narratee (and
us) (‘I said these things, Crito, and I paid particular attention to what
would follow and I watched how they would fasten onto the argument,’
283a). Later, a lengthier return to the frame (290e–293a) results in the
conversation of the frame and the narrative running in parallel as the
frame narrator and narratee rehearse obtrusively the same argument
with similarly unsuccessful results. Again, it is clear that the frame
conversation models the conversation Plato expects his dialogues to
stimulate.
The reactions of narratees are paradigmatic, therefore, as we have
already seen with Homer (→) and will see with Dio (→) and Lucian
(→). The distress, anger, and evasion that Socrates notes in his sophistic
interlocutors have their counterpart in the dedication and involvement
of like-minded conversation partners. True philosophers seek the truth
rather than merely to win arguments. They are gentle, receptive, and
encouraging.24 Whereas contentious sophists are easily annoyed, a Pla-
tonic narrator will often state that a friendly interlocutor answered with
a smile or a laugh.25 The lightness of tone in philosophical discussion
reminds us of the eagerness with which Socrates’ friends and disci-
ples seek to engage him in conversation or hear a narrative of a past
encounter. Socratic dialogue is presented as something it is pleasurable
to listen to and to engage in. One function of the narrative apparatus
is, therefore, to be protreptic.
Secondary narrators and narratives
Embedded narratives may be arranged into several groups. The chief

categories are: Socrates’ accounts of his own past (rare), reports of past
events, reports of past conversations, reports of speeches, mythological
narratives, and hypothetical conversations. These narratives occur both
in framed and non-framed dialogues.
24 For paramuthia and playfulness as the constitutive tone of philosophical discourse,
see Morgan 2000: 164–175.

25 E.g. Phd. 101b, 102d; Resp. 451b.
Socrates’ accounts of his past occur in the Apology and the Phaedo. In
the Apology, Socrates narrates, with commentary, the story of the Del-
phic oracle’s declaration that he was the wisest of men and its (iterative)
aftermath (20d–23b). This is a forensic narratio, but we encounter the
same sort of self-justification in the Phaedo (95e–100a), where Socrates
reports his investigations on causation. It is worth noting, given Soc-
rates’ explicit statements elsewhere in the dialogues on his dislike for
long speeches (Prt. 336b–c), that even these narrations retain elements
of conversation. In the Apology, Socrates reports the conversations he
repeatedly had with his fellow Athenians, and in the Phaedo he punctu-
ates the narrative with orientating questions to his narratee. The most
significant aspect of these embedded narratives, however, and one that
will recur, is the formality with which they are introduced. This fea-
ture is especially striking when we consider that similar formality is
mostly lacking when it comes to the dialogues considered as wholes
(a point to which I shall return). The Phaedo narrative begins with an
offer to tell the story, and then the opening words, ‘Listen, since I
am about to speak’ (96a). In the Apology Socrates states, ‘Listen, then
… I shall tell you the whole truth’ (20d). In both instances, the con-
clusion of the narrative is less marked, as Socrates goes on immedi-
ately to draw the inferences of the narrative for the current conversa-
tion.
Reports of past events are usually brief. Among the more detailed
is Alcibiades’ account of his attempt to seduce Socrates, and Socrates’
courage in battle (Symp. 217b–221b), both part of his eulogy of Socrates.
Yet anecdotes about the past are also used to deflate. Thus Laches
tells the story of how Stesilaus made a fool of himself in battle (Lach.
183c–184a) in order to express his doubts about the value of fight-
ing in armour for display purposes, and Socrates relates briefly the
prosecution of Pericles by the Athenian people (Grg. 515e–516a) to
devalue Pericles’ claims to statesmanship. A final function of such
accounts is to present problem cases for analysis, such as the narra-
tive of the successful villainy of Archelaus (Grg. 471a–c) or Euthyphro’s
account of the death of his labourer (Euthyphr. 172b–d). Narratives of
past events occur, then, mostly for purposes of example and illustra-
tion.
Reports of past conversations are an important part of the philo-
sophical world. In order to approach the truth one must treat the same
subject again and again (Phd. 107a–b), and it may be that recalling what
has been said on previous occasions can help in present efforts or help
focus attention on important points.26 Indeed, the entire corpus of Pla-

tonic dialogues is an example of this narrative function, and the framed
dialogues examined in this chapter make the point even more forcefully.
It comes as no surprise, therefore, that narrative of conversations is
sometimes a feature within the dialogues. The most famous example is
perhaps Socrates’ report of his conversation with Diotima in the Sympo-
sium (201c–212b), but one might also cite Crito’s report to Socrates of his
conversation with an unnamed interlocutor at the end of the Euthyde-
mus (304d–305b). This guides our reception of the conversation between
Socrates and the sophists, as the interlocutor’s disapproval of that con-
versation is a possible reaction for all narratees. Nevertheless, Socrates’
reaction to Crito’s narrative demonstrates how this disapproval has in
Socrates’ opinion no intellectual standing. Instances of reported past
conversation remain fairly limited, however. The reason seems to be
the philosophical premium placed on giving one’s true opinion in any
given philosophical context. One must not merely report the thoughts
of others, but think for oneself. Thus at the beginning of the Meno,
Socrates heads off Meno’s attempt to cite Gorgias as an authority for a
definition of virtue. He pretends to be forgetful of what he himself has
heard Gorgias say, and asks Meno to ‘let him go, since he is absent.
But you yourself Meno, by the gods, what do you say virtue is?’ (71d). It
seems, then, that narrative can be a threat to the philosophic project,
since it encourages mere reportage rather than the active generation of
truth. The status of the Platonic dialogue itself is thus called into ques-
tion: the reader must not be too reliant on what Socrates, or even Plato,
said.
Nevertheless, the questionable philosophical status of narrative does
not prevent lengthy stretches of it within the dialogues. These embed-
ded narratives occur when a narrator reproduces an important set
speech he has heard elsewhere. Examples are Protagoras’ ‘Great
Speech’ reported by Socrates in the Protagoras, Aspasia’s Funeral Ora-
tion reported by Socrates in the Menexenus, the speech of Lysias report-
ed by Phaedrus in the Phaedrus, and the speeches of love delivered by
the symposiasts and reported by Apollodorus (on the authority of Aris-
todemus) in the Symposium. All these narratives are presented in direct
discourse and are clearly marked at introduction and conclusion. Noth-
ing mediates the immediacy of the speeches; we are meant to experi-
26 The Timaeus opens with Socrates’ recapitulation of a Republic-like discussion and
his plea for a narrative account that will bring the perfect city to life.
ence them directly as pieces of affective oratory. Moreover, there is (the-

oretically) no question of any contribution to the speech by the narrator
who reports it.
Things are different when we come to embedded mythological nar-
ratives. Extended philosophical myths, whether narrated by Socrates or
by another philosophical leader (Timaeus or the Eleatic Stranger), sup-
port the narrator’s personal reflection and are marked as the product
of philosophical inference.27 The major myths told by Socrates (in Gor-
gias, Phaedo, Republic, and Phaedrus) share several characteristics. They all
have formal introductions (as, for example, in the Gorgias: ‘Listen then,
as they say, to a very fine story’ 523a). Socrates usually pretends that
the narrative is not his, but that he has heard it from someone else.
Thus the myth of the Gorgias is something that Socrates has heard from
others (524a). The myth of the Phaedo ‘is told in the following way’
(107d); while the myth in Book 10 of the Republic is attributed to Er the
Pamphylian (614b). Even Socrates’ great second speech in the Phaedrus
is ironically attributed to Stesichorus. These attributions seem to be an
attempt to play, once again, with the notion of validation and also to
maintain a prudent reserve with regard to the content of the narra-
tive.28 Even when Socrates cites a source, he still is careful to hedge his
bets (Grg. 527a–b; Phd. 114d). Acceptance of the narrative is conditional;
it can always be replaced by one with a better philosophical basis. For-
mal introductions are matched by formal conclusions, and these con-
clusions, aside from rounding off the narrative are ethically protreptic:
the narratees of the Phaedo, Gorgias, and Republic are presented with the
moral necessity of pursuing justice and philosophy, while the immedi-
ate and less immediate narratees of Socrates’ speech in the Phaedrus
are to pursue love and philosophy. Indeed, all of these Socratic myths
are marked by the obtrusive psychological (and textual) presence of the
narratee. In the Gorgias, Socrates addresses his narratee, Callicles, five
times. In the myth of the Phaedo, the presence of Simmias, the narratee,
makes itself felt particularly in the opening sections (108–110), although
he is also addressed in the conclusion. Glaucon, the narratee of the
Myth of Er, is twice addressed during the body of the narrative (615a,
27 Morgan 2000: 240–244.

28 Cf. Phdr. 275b for confirmation that Socrates’ attributions of his mythological
fantasies to an external source are relatively transparent, even within the world of the
dialogue. The Euthydemus (290e–291a) parades the likelihood that Socrates is attributing
to others arguments that he himself has made.
618b), as well as at the start and finish. Finally, the beautiful boy who
is the imaginary narratee of Socrates’ second speech in the Phaedrus is
addressed explicitly only at the start, finish, and once during the body
of the narrative (252b), but Socrates continually makes his presence felt
by use of the first-person plural (e.g. ‘Let not the argument throw us
into confusion,’ 245b).
These narratives are meant to persuade. Concomitant with this in-
tent is the confusion of clean narrative boundaries and levels. Our
experience of the narrative is always mediated. Socrates injects him-
self into the text with obtrusive statements of belief or logical inference.
We are always made aware that the most seemingly obvious distinc-
tions, such as that between muthos and logos, are contestable and defined
by context.29 We often experience the narrative making a fresh start
and swinging wildly back and forth between different temporal and
spatial realms in vertiginous changes of perspective. Thus in the Gor-
gias we begin with the citation of Homer as a source for the basis of
the story and the division of the world into three kingdoms by Zeus,
Poseidon, and Hades. The story becomes more immediate with the
insertion of direct speech for Zeus, but more remote as Socrates’ heavy
narrative presence resumes. The second part of the myth is presented
as Socrates’ inference on the first part, strongly marked by language
like ‘as it seems to me’, ‘I declare’, and marked also, as was the begin-
ning of the myth, by appeal to the authority of Homer (525d–e). Sim-
ilar changes of perspective occur in the Phaedo, where Socrates uses
the myth to open up a different temporal viewpoint on the signifi-
cance of the discussion. Since the soul is immortal, death is no escape
from wickedness; only education and virtue can help us in the eter-
nal perspective. An introductory paragraph summarizes the content
of the story (as usual, post mortem reward and punishment, 107d–e).
Socrates then disagrees with Aeschylus on the nature of the path to
the underworld, making his inferences from orphic/Pythagorean cer-
emonies (107e–108a). He then makes a fresh start at the beginning of
the story, as the soul descends to the underworld (108a–b). Judgment,
reward, and punishment are related in slightly more detail, but the
meat of the myth starts when Socrates comments that he has heard that
29 Cf., most famously, Gorgias 523a: ‘Listen then, as they say, to a very fine logos,
which you, I think, will consider a muthos, but I consider a logos, since I speak the things
I am about to tell you in the belief that they are true.’ For further discussion on the
implications of this passage see Morgan 2000: 158–159.
the earth is not really what the geographers suppose. He is encouraged

to expand, and then launches into a detailed rhetorical description of
the earth, now presented as the product of inferential reasoning (108c–
110b). Socrates then asks Simmias whether he would like to hear a fine
muthos about what the earth is really like, and then gives more rhetori-
cal description, ending with the rivers of the underworld and thus to the
judgment of the newly dead, now told for the third time (110b–114c).
There does not seem to be a consistent narrative format for these
myths. In the Gorgias, the myth is told in direct discourse, with the
speech of Zeus also quoted in direct discourse. In the Republic, the Myth
of Er is told in indirect discourse, using the accusative and infinitive
construction, apart from the use of direct quotation for the conversation
between two souls that tells the fate of incurable sinners (615c–616a),
and for the information given by a ‘prophet’ who superintends the
choice of lives as the souls of the dead prepare for reincarnation (617d–
e, 619b). In the Phaedo the narrative of the myth oscillates back and
forth between finite and infinitive constructions, while the myth of the
Phaedrus (as we might expect from a product of ‘inspiration’) stays with
finite constructions. It is tempting to associate the predominant use of
infinite construction on the Myth of Er with the strictures of the Republic
on the use of mimetic discourse, although, as I have already remarked,
Plato does not elsewhere seem keen to live up to such requirements for
his narrative.
The myths of the Politicus, Timaeus, and Critias are not narrated
by Socrates, neither are their functions primarily ethical protreptic.
The myth of the Politicus aims to clarify the function of the ruler by
exploring the history of the world, whereas the cosmology and story of
Athens and Atlantis contained in the Timaeus and Critias are a result
of Socrates’ request to set the perfect but static model of the perfect
state (as in the Republic) into narrative motion. Like the ethical myths
described above, however, their status as narrative is strongly marked
by formal beginnings and ends (except in the case of the Critias, where
the narrative of the downfall of Atlantis breaks off curiously just at the
point where we are about to hear direct speech in a council of the
gods). Moreover, these myths share the disorderly narrative motion of
the ethical myths: the tendency to backtrack and make fresh starts and
to forswear attempts at completeness or total accuracy. Leaving to one
side the problematic and incomplete Critias, we may note as examples
Timaeus’ insistence that his story is merely a ‘likely account’ (Ti. 29c–
d), and his anxious explanation that although he has spoken of the
creation of the body prior to the creation of soul, this was not the
order in which they were created by the Demiurge. The reason for
this temporal narrative incoherence is that ‘somehow, because we have
a great share of what is random and happens by chance, we in fact
speak in this sort of way’ (34b–c). Similarly in the Politicus the narrator
(the Eleatic Stranger) struggles with the mass of his material, trying to
focus on what is most germane for the philosophical point he wants to
make (271e, 274b). In the end, however, he is forced to conclude ‘we
raised up a marvellous mass of mythos and were forced to use a greater
portion of it than we had to. And so we have made our demonstration
rather long and we totally failed to perfect the muthos’ (277b).
Despite the insistence of Timaeus and the Eleatic Stranger on the
imperfections of their narratives, we may still ascribe to them a qual-
ified omniscience. Timaeus and the Eleatic Stranger let their narra-
tive choices be guided by what they regard as propriety. Even though
Timaeus stresses that his is only a likely account, he does not hesi-
tate to put himself inside the mind of the creative Demiurge. Since
the narrators in these dialogues believe that a supreme deity is good,
and since goodness is at least partially accessible to the human mind
through philosophical enquiry, the process of creation should conform
to the truths achieved by philosophical enquiry. So the Demiurge must
be good and perfect (because it would be blasphemy to believe other-
wise) and will make the world as perfect as possible (Ti. 28–30). Thus
Timaeus’ conclusions about creation are put forward as both likely and
necessary (e.g. ‘progressing in accordance with the likely and necessary
account’ 53d5), and the Eleatic Stranger declares ‘it is not right for him
[the creator] to move the world now in one direction, now in another.
As a result of these arguments, we must not say that the world always
moves itself …’ (Plt. 269e). Philosophical enquiry puts limits on the nar-
rative possibilities open to the philosophical narrator, but in so far as
he can discern eternal truths, he has a god’s eye view. This is so even
in Socrates’ ethical myths. For all that he disclaims responsibility for
his narratives by displacing responsibility onto his sources or by claim-
ing inspiration, Socrates views the world of the soul in these narratives
from an omniscient perspective that can sometimes (in the Phaedo and
Phaedrus) be outside the world as we know it.
The clear demarcation of the embedded mythological narratives
considered here reminds us of Socrates’ narratives of his past and the
reports of formal set speeches and brings us to a final point. In the
Phaedrus, Socrates famously remarks that a discourse ought to be like
a living creature, with head, feet, and middle fitting together harmo-
niously (264c). The embedded narratives I have been examining in this
section conform to this canon. Yet one is struck by the contrast between
the studied formality of secondary or embedded narratives with the
informality of the beginnings and endings of the dialogues that contain
them. The dialogues often begin with an intellectual event or discus-
sion already in progress (Gorgias, Cratylus, Philebus), or with the meeting
of the frame narrator with a friend (Theaetetus), or with a narration that
is the answer to a question asked outside the text (Republic). Although
the dialogues usually have more defined endings, even here we may
stop in mid-discussion (Philebus), and we are always made aware that,
even if an argument has been concluded, it is part of a larger project
(Phaedo, Theaetetus). I have mentioned above that the openings of the
dialogues presume familiarity with the genre: we are meant to know
who is speaking and to be familiar with the world of the dialogue.
The secondary narratives thus conform to developing canons of rhetor-
ical artifice, while the dialogue as a whole aggressively rejects formal
canons. Why? The answer, in part, must be that the dialogues model
a philosophical process that is meant to be ongoing. However pol-
ished and subtle they are, they must not risk closing off discussion by
being mistaken for treatises that express a defined and already discov-
ered truth. Whereas Socrates’ polished ethical myths are meant (among
other things) to persuade, the dialogues are inspirations and reminders:
models for life, not art.
The last type of embedded narration to be considered here is per-
haps the most characteristically philosophical. It is the imaginative pro-
jection of a line of argument, what one might call a hypothetical narra-
tive. These occur when Socrates, or whoever is leading the discussion,
asks what the answer would be if he were to say something, or alterna-
tively, what would be replied if an imaginary interlocutor were to ask
something. This is an extensive category, although I have space only
to glance at it here. A lengthy example is Socrates’ imagined conver-
sation with the laws of the city in the Crito (50a–54c) in which they
persuade him not to try and escape from prison (‘if the laws and com-
monwealth of the city should approach and stand over me and ask,
“Tell me, Socrates, what do you intend to do?” … what shall we say?’).
Less elaborate examples abound, such as Gorgias 451a: ‘For example, if
someone should ask me … “Socrates, what is the art of arithmetic?” …
I would say …’, or Phaedrus 268a–269c, where Socrates imagines what
would happen if someone who knew the rules of medicine and poetry
but not their application were to pretend expertise in the presence of

a doctor or poet.30 The most fascinating of these exchanges occurs in
the Theaetetus (166 ff.), where Socrates reanimates the dead Protagoras
in order to make a defence of his man–measure doctrine. Not only
does Socrates try to imagine what Protagoras would say, but he talks
as though Protagoras has actually spoken (168c–d) and addresses him
directly in response (170a). Such narratives of potential conversations
are important because they show that the Platonic dialogue seeks to
avoid treating arguments as abstract entities, but always tries to connect
them with a speaker, even if that speaker must be imagined. There is,
as has often been noted, a tendency for the argument in Platonic dia-
logue to become personified and find a voice. Thus the laws, tragedy,
Protagoras, and a host of anonymous interlocutors enter the narrative
when there is need to expand its intellectual scope. Philosophical con-
versation consists of a series of thought experiments (what would the
results be if we were to say ‘x’?). These experiments can be myths,
actual and hypothetical lines of argument, or a combination of the two
(such as the complex of sun, line, and cave in the Republic). Yet they all
become effective only when deployed in conversation in the mouths of
narrators.
What is true for embedded narrations in the dialogues in this in-
stance applies to the dialogues as a whole. Socrates himself, the Eleatic
Stranger, and Timaeus are personified lines of argument engaging
in hypothetical conversations. The elaborate narrative apparatus of
the dialogues brings this point home, but also constructs a world of
reception whereby the reader is guided into the philosophical life.31
30 A selection of further examples: Meno 74b, Philebus 13d, Sophist 243d–245e, 260d–e,
Theaetetus 158e, 163d.

31 This treatment of Platonic narrative had already reached final form when I
became aware of Blondell 2002. Rather than signal laboriously multiple points of
contact between Blondell’s approach and conclusions and my own, I shall state only
that any serious student of the topic will need to consult her book.
chapter twenty-nine
XENOPHON
V. Gray
Xenophon’s Memorabilia, Oeconomicus, Apologia, and Symposium, which all

contain conversations of Socrates, share some important narratological
features. The primary narrators are anonymous, but reveal themselves
in introductory frames, which also state the topic of the conversation.
The narrator claims to be remembering conversations he has heard
himself (Oeconomicus and Symposium), or reports of them from others
(Apologia), or a combination of these two types of memory (Memorabilia).
This makes the works remembered narratives and authorizes them as
the product of eye-witnessing, whether in the first or in the second
degree. The narrator also displays through them the power of memory
that was required of a Socratic pupil and the process that ensures
that the lessons of the conversations remained with him (Mem. 1.2.19–
21; 4.8.11; on this theme → Plato). They could provide private re-
education, as Memorabilia 4.8.11 suggests, or be read aloud and thus
shared with others (→ Plato).
Theoretically Xenophon might have chosen to call his narrator
‘Xenophon’. Instead, as in Anabasis, he introduces the character Xeno-
phon as a dramatized ‘other’, in order to secure an impression of objec-
tivity and hence persuasive force. Indeed, he even allows Socrates to
criticize this character for being ‘foolish’ (Mem. 1.3.8–13). Since Xeno-
phon was probably too young to remember the party described in Sym-
posium (for which the dramatic date is 422 BC), he cannot be identified
as the narrator of this work either. The narrators of Memorabilia and
Oeconomicus might be among those silent bystanders who are mentioned
as witnesses to the conversations (Mem. 1.2.30; 1.6.1; 4.2.1; Oec. 3.1). In
Symposium, he could be one of the named guests, referring to himself in
the third person, or an anonymous guest. The memories are fictional if
they are credited to the author,1 but could still be the genuine memories
of this ‘other’.
1 E.g. Kahn 1996: 29–35.

378 part six – chapter twenty-nine
One common feature of the frames, which can be entirely narrative

or involve dialogue, is their use of formulae. A very full range of
these introduces the sequence of smaller conversations that form the
bulk of Memorabilia 1.4–4.8. A typical instance is: ‘I will first state
what I once heard him say about the godhead in conversation with
Aristodemus …’ (1.4.2). The actual conversation is then started off by
Socrates’ perception of a problem that he tries to solve: ‘Learning that
he did not sacrifice to the gods … “Tell me”, he said, “Aristodemus
…”’.2 There are also other forms of introductory remarks, such as
exhortations to the narratees to join the narrator in his search for the
truth (‘Let us enquire …’: 1.5.1, 7.1), and assertions of the narrator’s
discrimination (‘it seems worthy of report’: 1.6.1). The narrator does not
intervene in the course of the conversations, but he often summarizes
what Socrates achieved (1.4.19, 5.6; 4.2.40, 5.12), thereby rounding off
one conversation before proceeding to the next one.
These formulae are found in later manuals of rhetorical instruction,
which use them to introduce the sayings of the wise as proofs, and
show how to elaborate them into short stories. The longest type of
saying is the apomnēmoneuma, which is the Greek title of Memorabilia.3
Memorabilia indeed elaborates the introductory formulae, as, for exam-
ple, when we hear how Socrates came to view the famous courtesan
Theodote (3.11.1), or how Socrates made Euthydemus receptive to his
instruction over a period of time (4.2.1–8). The first makes a philo-
sophic issue of the beauty of Theodote (it is beyond words, who bene-
fits from it?), while the second characterizes the world of the reception
for Socrates’ teaching method, as in Plato (→). Euthydemus is intro-
duced as an example of how Socrates adapted his approach to those
who were unready to accept instruction (4.1). In a series of short scenes,
Socrates breaks down his pride, speaking to his companions on topics
meant to gain his silent attention, and using irony (4.2.1–8). The main
conversations that follow these elaborations are then introduced in the
formulaic ways: ‘Tell me, Theodote …’ (3.11.4); ‘Tell me, Euthydemus
…’ (4.2.8). Sometimes the remark is developed to prove that Socrates
practised what he preached, e.g. in 4.4.1–4, 5.1–2, 7.1, in accordance
with the programme of the work, which sees imitation of his practice as
an essential part of his education (1.3.1).
2Other examples: 1.2.29–30, 3.8; 2.1.1, 2.1, 3.1.

3Gray 1998: 107–122 has discussed these introductions to the so-called chreiai. See
Hock and O’Neil 1986 for the relevant texts.
The other works also begin with variations of these formulae, and
there is an impression that they form a continuous sequence. Oeconomi-
cus begins: ‘I once heard him [Socrates] speak about household man-
agement as well … “Tell me,” he said …’. Symposium begins: ‘But to
me it seems that not only are the serious pursuits of gentlemen worth
remembering (as is the common opinion) but the playful ones as well.’
He next introduces his actual narrative with the well-known formulaic
frame: ‘On account of what experiences I came to this conviction I
wish to reveal’. Apologia, like Oeconomicus, starts like another one of the
conversations embedded in Memorabilia: ‘It seems to me worth remem-
bering also …’; cf. Memorabilia 1.6.1. The sense of continuity and the
similarity to the frames of the sequence of conversations in Memorabilia
suggest that these other works constitute further conversations in the
sequence and come from the same narrator. Perhaps this expresses the
continuity of the process of remembering and the collected force of the
wisdom that they contain.
The authority of this anonymous internal narrator is expressed in
his claims to have been present at Socrates’ conversations and to have
heard them himself, or at least to have heard about them. He appeals
to his own memory (e.g. Mem. 1.3.1), or simply claims to ‘know’ (Mem.
2.7.1, 2.10.1; 3.3.1; 4.4.5, 4.5.2). Occasionally he introduces anonymous
spokesmen (‘others’), who were witnesses themselves and as such can
confirm what the narrator says (e.g. Mem. 4.3.2). He turns to a named
secondary internal narrator at the end of Memorabilia, where Hermo-
genes tells him about a conversation he had with Socrates, which
revealed his remarkable attitude to death (4.8.4–11). The named wit-
ness gives special authentication to something that the narratees might
find particularly unbelievable. We are in fact told that Socrates’ crit-
ics fail to understand his attitude to death (4.8.1) and that this applied
at first even to Hermogenes. Hermogenes therefore carries author-
ity not only because of his personal experience of Socrates, but also
because he initially shared the ignorance of the critics, but learned the
truth from Socrates himself. Hermogenes appears in this same role in
Apologia, where the narrator refers to the inadequacy of other accounts
and introduces him to tell the true story. The concern for authority
can be found in the frame of a conversation within Memorabilia too,
where Socrates, as narrator of the story of Prodicus, refers to the pos-
sible failure of his memory (2.1.21) but then asserts authority in saying
that he is not using the exact words of Prodicus—which he must know
(2.1.34).
Introductory frames can model a world of reception for Socratic

teaching for the primary narratees, as in Plato (→), characterizing not
only Euthydemus in Memorabilia 4.2 but also Callias in Symposium, who
is presented as one who is passionate to spend time with Socrates, who
prefers the company of the wise to that of military men and office-
seekers, and whose passion is confirmed in his distress at Socrates’ ini-
tial refusal of his invitation to his party. He seems keener to display his
wisdom to Socrates than receive it, but Socrates ironically reveals that
his wisdom is inferior since it has been purchased from the sophists. In
Oeconomicus, Critobulus as secondary narratee shows the same passion
for Socrates’ advice (2.14, 3.1).
Primary narrators often express as their motive for narration the
correction of an erroneous view: that Socrates’ condemnation was
wrong (Memorabilia), that play is unworthy of memory (Symposium), or
that Socrates’ attitude to death was unworthy (Apologia). This makes the
narrators educators in their own right.
The works also have their own distinct features, to which I turn next.
Memorabilia
Memorabilia’s anonymous internal primary narrator embarks on his tale

because he wonders what arguments persuaded the jurors to condemn
a man like Socrates. The first part of the work (1.1–2) refutes the charges
with arguments in defensive courtroom mode, confronting the opposi-
tion (from 1.2.9 onwards) and illustrating some points with dialogues
(1.2.29–46); while the rest (1.3.1–4.8) offers a sequence of dialogues,
framed and sometimes interspersed with narrative, as well as some
plainer reports, such as the account of the limit of Socrates’ instruction
(4.7), which serve as more positive proofs of Socrates’ virtuous instruc-
tion of others.4 The conversations present a comprehensive account of
Socrates’ teaching to a very wide variety of people, mainly on their
relationships—to the gods and to each other. This range of topics
makes the work a compendium of wisdom, unlike any single dialogue
of Plato. The conversations are often in order of the importance of their
topics (beginning with religion: 1.4, and then self-control: 1.5), but they
are also grouped according to topic (conversations about family rela-
4 Gray 1998, especially 123–158, describes the structure of the contents and estab-
lishes that there is a programme in it, which previous scholars have questioned.
tions: 2.2–3, friendships: 2.4–10, relations between political officials and

their constituencies: 3.1–7). The conversations with Euthydemus (4.2–3,
5–6) offer an almost complete process of education in the one interlocu-
tor, raising the original teaching on religion and self-control to a higher
level and culminating in the teaching of dialectic. The last conversation
(4.8) takes place on the eve of Socrates’ death, bringing us back to the
beginning of the work and hence creating closure.
The narrator’s organization of his work (refutation of the charges,
followed by positive proof of Socrates’ virtue in a sequence of conver-
sations) rather interestingly shows him to be one who has learned the
Socratic teaching technique that he portrays in the work.5 The combi-
nation of refutation followed by more positive instruction is the shape
of the character Socrates’ instructional dialogue, e.g. his education of
Euthydemus, which begins with negative refutation (4.2 passim) but pro-
ceeds to positive instruction, which begins at 4.2.40. This characterizes
the narrator as a true disciple of the educational process of his vener-
ated master in his instruction of his narratees and is a unique narra-
tological feature of Memorabilia. In accordance with the clear rhetorical
purpose of the work, the narrator, though anonymous, is overt; he is
throughout present with his emotions and evaluations, e.g. ‘I wonder
then how the Athenians can have been persuaded that Socrates was a
freethinker …’ (1.1.20), and numerous first-person comments (I believe’,
‘I see’, etc.).
The narratees are anonymous like the narrator. Their presence is
implicitly felt, however, when the narrator inserts numerous rhetorical
questions in the first part of his defence, e.g. ‘So, in pronouncing on
opinions of his that were unknown to them, it is not surprising that
the jury erred: but is it not astonishing that they should have ignored
matters of common knowledge?’ (1.1.17).6 They are also the focus of
persuasion when the narrator constructs and refutes the various argu-
ments of anonymous opponents, structuring his work as a virtual dia-
logue with these (from 1.2.9 onwards): ‘The accuser says … but I say
against this …’ (1.2.19); ‘A man might say against this … I do not
oppose that, but …’ (1.2.17); a sustained use of the ‘anonymous inter-
locutor’ device. Perhaps the opposition represents the views of some
narratees, but for others, the victory of the narrator over his opposi-
tion persuades them that he has argued the stronger case. The narrator
5 Gray 1998: 91.

6 Cf. 1.5.1, 2.2, 2.8, 2.11, 2.15, etc.
indeed overwhelms his opposition in the course of the work, first refut-
ing their view that Socrates was harmful (1.2.9–64), then beginning to
prove him helpful (1.3), reintroducing the opposition to argue that he
was only partly helpful (1.4.1), dismissing this with a series of conver-
sations that prove Socrates’ entire helpfulness, letting them put a final
thesis regarding Socrates’ death (4.8.1), and crushing this as well. His
first-person conclusion finally dismisses them, when it invites ‘a man’
who is not pleased with his evaluation to compare Socrates against oth-
ers (4.8.11). This conclusion also uses the ‘continuance’ motif to indicate
to them how influential Socrates has been: ‘All who knew what manner
of man Socrates was and who seek after virtue continue to this day to
miss him beyond all others.’
There are a few embedded stories credited to secondary narrators.
The longest of these is Prodicus’ ‘story’ of Heracles (2.1.21–33), which
the secondary narrator Socrates introduces with a formula similar to
those with which the primary narrator introduces his own conversa-
tions. The purpose of this embedding is to reinforce the teaching on
toil by offering a range of instructional styles on the same topic: poetic
didacticism (Hesiod and Epicharmus) and sophistic epideictic (Prodi-
cus) are embedded within Socrates’ own dialectical conversation. The
question of translation arises when Socrates pretends not to have been
able to capture Prodicus’ own grand language, but this is more likely to
illustrate his knowledge of the original (2.1.34).
Oeconomicus
Oeconomicus presents a multiplicity of narrators,7 in a manner compara-

ble to that found in Plato (→). The primary narrator, anonymous and
internal, without motivation, but in his role of eyewitness, recounts in
dramatized form a discussion between Socrates and Critobulus about
successful estate management. After one sentence (‘I once heard him
discussing the subject of estate management in the following man-
ner’) he disappears never to surface again, except to provide speech-
introductions. This topic (of estate management) is the realm of the
kaloskagathos, who can ‘use his household and his house-holders and rel-
atives, and friends and polis and fellow-citizens for benefit’, like the
7 Pomeroy 1994: 17–18 briefly describes the structure of the work. The embedding
once gave some the impression that the work is not a unity.
kaloikagathoi who associate with Socrates at Memorabilia 1.2.48 and 64.

Socrates teaches in his customary way, defining ‘estate’ in terms of its
function, which is to benefit the possessor (1.1–15). For further instruc-
tion, he introduces reported narrators, recounting what ‘people say’
about the practices of the Persian king (4.4), what Cyrus ‘is said’ to
have told those who received his gifts (4.15–16), what Lysander ‘is said’
to have told a Megarian stranger about Cyrus (4.20–25), and, becom-
ing a secondary narrator himself, reports a long conversation he had
with the supreme estate manager Ischomachus, in which he learned
management (6.12 to the end). This Ischomachus, becoming a tertiary
narrator, embeds his own ‘memories’ (7.43; 8.23) of the conversations
in which he instructed his wife (7.5–10.13) and spoke with the mate of
a ship he once inspected (8.11–16). The secondary narrator Socrates
intervenes only occasionally into the conversations he reports (‘I said’:
10.1, 11.1), while, as was noted earlier, the primary narrator has disap-
peared altogether. Thus, the work ends with Ischomachus addressing
Socrates.
Introductions to the embedded conversations mention the leisure
that is the setting for philosophical discourse, but show that men with
property shun such idleness. (Socrates:) ‘Seeing Ischomachus once …
since he seemed at leisure, I approached him and sat down and said.
“You’re not accustomed to sitting down doing nothing …”’ (7.1); (Ischo-
machus:) ‘I saw this man in his leisure making a full inspection … I
asked him what he was doing and he said …’ (8.15). These introduc-
tions, then, are part of the ‘economic’ lessons imparted.
There is considerable characterization of the secondary and tertiary
narrators and narratees. Socrates as secondary narrator seeks his cus-
tomary definitions, employs irony, describes models for his narratee
Critobulus to imitate and sends him to higher authorities for experi-
ence he does not possess (a habit of his noted in Memorabilia 4.7). In
estate management he is a success himself, since he succeeds in mak-
ing a small profit from his small estate (2.10). The secondary narra-
tee Critobulus is a young man of much greater property, under the
greater obligations that this carries (2.1–9; also 7.3), married, but spend-
ing money on idleness such as boyfriends and dramatic performances
(2.3–8; 3.12), and becoming aware that he is exhausting his means and
eager for instruction in how to become a kaloskagathos (6.12). The ter-
tiary narrator Ischomachus is the ultimate model for Critobulus, and
therefore a fully characterized narrator. He has a great property and is
married like Critobulus, but unlike Critobulus, is not given to idleness,
but to farming. He has a reputation as a kaloskagathos (6.17) and he sup-

plies the instruction that Socrates has foreshadowed—training his wife
to rule the servants and increase his estate (7–10; cf. 3.11–16), develop-
ing his physique and training for war while himself increasing his estate
(11; cf. 4), teaching his workers to increase his estate (12–14), teaching
Socrates how to farm (15–19) and telling him what ruins estates (20–21;
cf. 1.16–23). He is an excellent teacher, instructing his workers in the
art of ruling (13.3–5)—which he considers the peak of achievement (21–
22)—and bringing knowledge to birth in Socrates (19.14–15). Socrates
as his tertiary narratee is eager to learn, which he says is characteristic
of a philosopher (16.9), thus providing a model for his own secondary
narratee, Critobulus.
The possibility that Socrates is ironic arises when he says that he
sought Ischomachus out because he had a reputation as a kaloskagathos
(6.12–17, since reputations can be misleading). Yet there is no irony
in his own praise of farming as secondary narrator (5.1–13; 6.4–10).
Successful farming indeed requires and develops philosophic virtues,
such as self-control (1.16–23), and permits the cultivation of friends
and polis (6.4–11). Ischomachus is ironic about his own reputation (7.3)
and he knows when Socrates is being ironic (11.3–7; 20.22–29), but
he is also serious about his own instruction, and Socrates elicits even
from his remark about profiteering (20.22–29) the serious principle that
‘all men naturally love whatever they think will bring them a profit’.
Oeconomicus can therefore be read as a straightforward lesson in how
Socrates taught his associates to manage their households and other
organizations by extension, such as the polis. Ischomachus says the
principles are applicable to any organization (21.2), and Socrates does
not oppose this; elsewhere, he even endorses it (Mem. 3.4.12).
Other narratees are characterized to a lesser extent. Ischomachus
characterizes his wife through her reactions to his training (7.8, 14, 37
etc.), just as he is seen as beautiful through her eyes (10.6). She has been
identified as the Chrysilla whom Andocides criticized for subsequent
immorality, but to import this into the work and make her portrayal
ironic is narratologically indefensible.8
Secondary and tertiary narratees such as Critobulus and Socrates
steer the reception of the primary narratees in their eagerness to learn.
The truth of the embedded stories is not seriously challenged. Crito-
8 Pomeroy 1994: 261–264 believes that an ironic reading would destroy the point of
the work, but ironists could read it as a parody of the impossibility of such perfection.
bulus asks whether Socrates really believes what ‘they say’ about the
Persian king (4.4), but accepts it as true (4.12) after Socrates proves
it through an ‘enquiry’ into it (4.5–11). He gives the same acceptance
(4.17) to further reported narratives about the king.
The function of the embedded narratives in Oeconomicus is educa-
tional: they are paradigms, which bring the narratees into the presence
of successful managers of great estates (the Persian king, Ischomachus)
and reinforce Socrates’ teaching about the excellence of farming, or the
excellence of his paradigms. They use a range of educational methods:
Socrates teaches through questions, definitions, paradigms and protrep-
tic; Ischomachus’ instructional styles include protreptic (15.10–16.7), and
the Socratic method of questioning and bringing knowledge to birth
(16.10 to 19.14–15). Ischomachus also offers instructional models for all
narratees to adopt in managing their own households: the lectures he
delivers to his wife (7–10), and the praise and blame, rewards and pun-
ishments that he uses on his workers (12–14). His instruction of Socrates
also illustrates Socrates’ customary search for wisdom and his expressed
concern to find expertise he does not have: he has theoretical knowl-
edge, but lacks the experience of running a large estate (this disqualifi-
cation is mentioned at 2.3.11–13). This is why Socrates promises to send
Critobulus to other teachers for various areas of instruction (2.13–18;
3.14–16; 4.4), but in fact he supplies Ischomachus as a person who com-
bines all these.9 This embedded narrative also allows Socrates to cast
himself as pupil and reveal the delight of the philosopher in learning,
which models the reception that is appropriate for Critobulus.
Apologia
The anonymous external primary narrator in Apologia motivates his

own narrative by saying that other writers agree about the ‘high-mind-
edness’ of Socrates over the period of his trial, thus authorizing this fea-
ture, but make it look ‘witless’ by failing to explain his conviction that
death was preferable to life. He wants to correct this situation by intro-
9 For example, Socrates does not fulfil his promise to have Aspasia give advice on
wives (3.14), but this advice is incorporated into Ischomachus’ account. In his own
instruction he remembers more than what has been agreed at 6.6–7 but his expansion
is a natural extrapolation from 4.3, where craftsmen are said to be bad defenders of
their country.
ducing a named internal secondary narrator, Hermogenes, who alone

made Socrates’ high-mindedness appear ‘appropriate’ by explaining its
rationale. This secondary narrator is authoritative, because he was an
eyewitness of the trial, indeed a companion and interlocutor of Socrates
(cf., e.g. 2–3).
The primary narrator offers minimal framing. In 2 he introduces a
report of a conversation that Hermogenes had with Socrates before the
trial (‘He stated that, on seeing Socrates discussing any and every sub-
ject rather than the trial, he had said …’); this conversation runs from
3 to 9 (cf. the similar report in Memorabilia 4.8.4–10). In 10 (‘Hermo-
genes stated that with this resolve Socrates came before his jurors …’)
he introduces a report of the trial itself (11–21), including the embed-
ded responses of Socrates’ accusers (20–21). In 22 he curtails Hermo-
genes’ account of the trial, stating that he knows that much more was
said by Socrates and his counsellors, but that he limits himself to the
proof of his conviction about death. This intervention from the pri-
mary narrator into the account of his secondary narrator demonstrates
his wider knowledge, anticipates any hypothetical objection from the
primary narratees that his account is incomplete, and at the same time
characterizes his desire to discriminate and focus on his theme. From
this point the voices of the primary and secondary narrator merge,
endorsing the same image of Socrates in persuasive harmony. In 23–
31 the primary narrator recounts mainly in his own voice, occasionally
recalling his source Hermogenes (e.g. 27: ‘Hermogenes reports him as
asking …’), how subsequent events proved Socrates’ disregard of death,
recording conversations with his pupils (27–28) and comments on his
prosecutor Anytus (29–30). The narrator even seems to step into the
shoes of Hermogenes when he describes how Socrates’ physical appear-
ance was ‘in agreement with’ his resolve (27), something only Hermo-
genes, who alone was present, could have seen.
There is some direct engagement with the narratees, but not very
much: two instances of the ‘presentation through negation’ device rep-
resent the ignorant narratees’ expectation (31: ‘he was not cheated of
this expectation, but …’; 33: ‘he did not shrink in the face of death, but
he met and saw it through in high spirits’) and an extended instance of
the ‘anonymous interlocutor’ motif at the very end of the work, which
challenges those narratees who do not accept his final conclusion (‘I am
unable not to remember him and, in remembering, not to praise him. If
any of those who desire virtue have been with a man more helpful than
Socrates, I think him a real man worthy of the name most blessed’: 34).
Symposium
Symposium tells the story of a party, from the invitations to the break-up,
and takes as its theme the seriousness that can be found in the playful
activities of kaloikagathoi. Memorabilia 4.1.1 finds the same seriousness
in Socrates’ playful claim to be in love with his young men. The
narrator is anonymous and undramatized and though internal (cf. 1:
hois … paragenomenos: ‘an experience of mine’) after the first framing
chapter invisible in his own story. His initial scene models the world of
philosophy, when it recounts how the company gathered to celebrate
the victory of Autolycus (1.2–6), characterizing Callias, his host and
lover, as an enthusiastic fan of philosophy, and Socrates and his friends
as men who are indeed, as Callias describes them, purified in soul.
Thus they ‘as was natural’ (i.e. for educated men), praised Callias for
the invitation, but at first declined to attend (1.7). The narrator also
frames the speeches that form the bulk of the work with narratives
that contrast different types of play, some worthy of philosophers, some
not. At the beginning Autolycus’ beauty produces the civilizing effects
of love on the company, but this is disturbed by the entry of the
clown Philip with his buffoonish play. Socrates will eventually reject
passive entertainment like this in favour of the more engaged play of
philosophers (3.2). Love is indeed the subject of Socrates’ long speech,
and the kind of play that Socrates will eventually harness to most
serious educational purpose, in order to make Callias a political leader
and lover of the polis (8.36–43). At the end of the work, young actors
‘play’ a dramatic mime of the love of Dionysus and Ariadne, balancing
the homosexual love in the introduction with heterosexual love (9.2–6).
Yet the actors playing the lovers are seriously in love once more, and
under that god’s civilizing influence. The performance ends the work
with the company inspired by thoughts of wedded bliss, while Socrates,
Callias, and others join Autolycus and his father on their walk (9.1, 9.7).
Though invisible as a character, the narrator is overt, in that he
comments on the events of his story. He introduces the various speeches
of the guests that form the inner core of the work with simple tags
(‘he said, they said’), occasionally turning to indirect speech (e.g. 3.14;
4.50). He marks stages of the conversation (e.g., 4.5, 9, 28, 45, 64; 9.1,
7), guiding the narratees to an interpretation (6.10), involving them
as virtual eyewitnesses through the use of the ‘anonymous witness’
device (‘A person who took note of the course of events would have
come at once to the conclusion that beauty is in essence something
regal …’: 1.8), or reminding them of the theme of the work (4.28).
He introduces hōs eikos-comments (1.7, 8), which recall the narrator of
Xenophon Anabasis and Hellenica (→). He also intervenes to give brief
descriptions of the entertainment, sometimes in order to have Socrates
evaluate it (e.g. 2.8, 11–12, 22; 5.9–10; the final entertainment is the
mime). In particular he marks the relation between outer appearance
and inner realities, to make the inner reality clearer. He thus indicates,
as the Socratic narrator in Plato (→), Socrates’ outer seriousness and
inner irony (e.g. 2.17; 3.10; 8.4; cf. 4.19), the outer blush that betrays
Autolycus’ inner modesty, (3.12), and the inner malice that provokes
the Syracusan’s jokes (6.6–10). This has been a theme also of the
introduction, where Autolycus has an inner and outer beauty and
produces the same harmony in the company (their silence, their poses,
their tender looks reflect their inner feelings); the clown uses mock
mournful poses and gestures and voice to convey his inner pain at not
causing laughter in the company; and in the mime too, the ‘players’
resolve their inner and outer appearance, because they truly feel the
serious emotions that they express as stage lovers.10
There is only one secondary narrator, Charmides, when he contrasts
his previous life as a rich man with his current life as a poor one,
adopting a view of democracy in the process that is patently ironic
(4.29–32). Otherwise there are only analytic discussions, but here the
reactions of the rest of the company to the speakers guide the reactions
of the primary narratees, just as the internal audience’s reaction to the
visual spectacles guides their reaction to those spectacles in narrated
form. This is illustrated by the general disapproval of the malice of the
Syracusan entrepreneur, and his polite but firm treatment by Socrates
(6.6–10).
10 Description of physical appearance is relatively rare in classical literature, but
Anabasis 2.6.9 also characterizes the looks and voice of Clearchus as an outer reflection
of his inner character; Apologia 27 marks the conformity of Socrates’ appearance with
his words.
part seven
BIOGRAPHY
chapter thirty
XENOPHON
V. Gray
Cyropaedia praises the achievements of Cyrus the Great as creator of

the Persian empire. In an introduction (1.1) the narrator describes the
process that led him to admire his leadership, then gives an account
of Cyrus’ lineage, natural qualities, and education (1.2–6), his creation
of his empire (3.1–8.1.48), his arrangements for securing it (8.2–6) and
his death (8.7). An epilogue follows, which laments the abandonment
of Cyrus’ practices by Persians in the narrator’s own time (8.8).1 The
intended status of the work is an interesting question. The narrator
could be seen as a historian because he re-writes Herodotus’ Histories
1.107–216, adding an account of the education of Cyrus, giving different
versions of his birth, upbringing, and death, changing his relations with
other main characters, and giving more detail on his battles and his
arrangements for the maintenance of empire.2 However, he spends
so much of his energy on the praise of Cyrus (who is ‘worthy to be
admired’: 1.6) that he looks to be writing an encomium rather than a
history. Praise and blame of different characters is an integral part of
historical works such as Hellenica and Anabasis, but they do not entirely
focus on praise, let alone the praise of one central character. The
narrator’s enquiry into Cyrus’ lineage, natural qualities, and education
(1.5) is the kind of biography we associate with Plutarch, for instance, in
Agesilaus 1–2. Xenophon’s own Agesilaus, which declares that it aims to
praise the central character, also begins with his lineage, and discusses
his natural qualities (3–11) after an account of his military career (1–2).
Cyropaedia uses the same devices as historical texts (e.g. the ‘anonymous
spokesmen’ device), but more exclusively to make the narratees believe
the praise of Cyrus.
1 Due 1989: 31–53 and 108–114 gives the most useful comment on the work for
narratological purposes. Marincola 1997 does not treat Cyropaedia.

2 See for further views about its intended status: Cook 1983: 11–24; Hirsch 1985: 61–
392 part seven – chapter thirty
An external primary narrator
The primary narrator is not dramatized: we hear no name, or any

other personal information. What we can deduce, however, is that
he is an external narrator, who is removed in time from the event
he recounts. This becomes clear from a quite prominent feature of
his narrative: his frequent use of the ‘reference to the narrators’ own
time’ and ‘continuance’ motifs. The instances of the latter motif are
marked by eti kai nun (‘even to this day’) or by mixing the present
tense into descriptions of the past, e.g. in the passage on the best
way to camp and decamp (8.5.2–5). In the epilogue the references are
not to continuance, but to abandonment of the customs that Cyrus
established, and contrasts are drawn between ‘previously’ and ‘now’.
The narrator is indeed quite unusually aware of the relations between
past and present. The time that has passed between the time of Cyrus
and the time of narration is not specified, but can be deduced as about
two centuries, since the epilogue refers to events that occurred as late
as 361 BC (8.8.4).
The ‘reference to the narrator’s own time’ and ‘continuance’ motifs
go back to Herodotus (→) and also occur in Xenophon Anabasis and
Hellenica (→), but the narrator of the Cyropaedia uses them in a distinctive
and interesting way. Due argues that by reminding the narratees of the
contemporary situation they assimilate the work to historical writing,
add liveliness and coherence, and praise the central character, Cyrus. It
could also be argued that the narrator’s distance from the events makes
the praise more objective.3 But more obviously, the device gives another
sort of authority to the narration and the praise. The continuance of
customs into the narrator’s own time prove that they (1) really hap-
pened, since there is contemporary evidence of them, and/or (2) that
they really were excellent, since what lasts and stands the test of time is
good and right. Cyropaedia 1.2.16 makes the first point explicit, when it
indicates that contemporary practice indeed is ‘evidence’ of the ancient
Persian diet and exercise: ‘and even to this day there remain proofs of
their moderate diet and of their exercising off what they consume; for
it is even now a source of shame …’ Cyropaedia 1.2.1 has both effects,
97; Due 1989: 117–135; Tatum 1989: 35; Gera 1993: 13–22. On Xenophon’s contribution
to the development of biography in general, see Momigliano 1971.
3 Marincola 1997: 76, 94, 158–174, 222 examines ‘fear and favour’ among con-
temporaries.
where ‘Cyrus is said and sung by the barbarians even to this day to
have been most beautiful, most kindly, fond of learning and honour-
loving’ (1.2.1). This confirms that not only are there witnesses to his
reputation, but it has not been eclipsed and has stood the test of time
(8.5.28).4
In the case of the epilogue (8.8 passim), the narrator’s position out-
side the narrative allows him to prove the excellence of the customs by
showing that abandonment of them meant decline. The epilogue has
been seen as a later addition to the text, but also as an integral part of
the praise of Cyrus, a conscious narratorial device designed to reinforce
the praise of his practices by demonstrating that the people who once
used them have declined as a result of abandoning them.5 Xenophon’s
Respublica Lacedaemoniorum uses the same device when it states in the
introduction that it seeks to praise the practices instituted by Lycurgus
(1.1–2), and then inserts an ‘epilogue’ (14), which proves the excellence
of the practices by showing that contemporary Spartans have declined
as a result of abandoning them.6 It would be too much of a coincidence
to have a narrator who in both cases inserts after-thoughts that con-
tradict his original thesis (or an interpolator who takes issue with that
thesis).
Sometimes the narrator wants to have his cake and eat it too, since
practices that are said to continue in the main narrative appear to
have been abandoned in the epilogue, but there are usually some
subtle changes. Thus, 1.2.16 says that contemporary Persians continue
to avoid publicly voiding their water and wind, and this is used as
evidence that they once had a meagre intake of food and worked it
off in exercise; 8.8.8 agrees that the avoidance continues, but says that
they no longer work it off.7
4 Cf. 1.4.27; 3.2.24; 4.2.1; 6.1.27–30; 7.1.3, 45, 46, 47; 7.3.15, 5.70; 8.1.23–24, 4.5, 4.16,
4.28, 5.28, 6.5, 14.

5 Due 1989: 16–22 gives a critical summary of the debate and comes down on the
side of the narrative device. Since then Tatum 1989: 220–239 and Gera 1993: 299–300
have given their versions of its significance. Nadon 2001: 139–146 most recently tries
to prove that the seeds of the collapse are already sown in the practices that Cyrus
instituted.
6 As Momigliano 1936 argued; he is followed by Bordes 1982: 198–203 and Carlier
1984: 252–254. Cf. Tigerstedt 1965: 169.

7 Another example is 1.3.2, where it says that the Persians in their homeland
continue to have plainer clothing and food than the Medes, while 8.8.15 says that they
adopted Median dress and food, but continued to practice Persian restraint. What the
Persians did at home in the first passage seems to be different from what they did
It could be argued that the instances of the ‘continuance’ motif in

the main narrative give the narratees the impression that the empire
still remains as it was created by its founder, which would make the
epilogue a shocking surprise, but this underestimates the narratees. The
narrator of Xenophon’s Respublica Lacedaemoniorum could give the same
impression of continuity, but he shows that his narratees have clearly
not closed their minds to the possibility that his practices have lapsed;
for he introduces his account of the decline: ‘If anyone should ask me
whether the practices of Lycurgus seem to remain unshaken even to this
day …’ (14.1). The narratees of Cyropaedia might experience the same
doubt.
The primary narrator and his sources
The narrator describes in his introduction how he reflected on the

general truth that successful leadership was a virtual impossibility, but
was reminded of the exception, Cyrus, and decided to find out more
about him. The resulting narrative therefore is ‘what we have found
out and think we know’ (1.1.6), and his informers turn out to be the
Persians themselves, as appears, e.g. from 1.2.1: ‘what is said and sung
by the barbarians’.8 This gives authority to his account. It departs
from Herodotus’ earlier version of Cyrus’ career, but perhaps exploits
the opportunity left by this author’s mention of other traditions that
he rejected, instead preferring ‘what the Persians say, those who do
not wish to make a legend of the achievements of Cyrus, but speak
the truth’ (Hist. 1.95.1). The remarkable reference to what is sung
establishes the greatness of Cyrus, since poetry is often presented as
a form specially reserved for greatness; cf. Lysias 2.3 or Arrian 1.11.2.
The narrator does not say how he has come by these Persian tradi-
tions, nor does he describe them more precisely, but we do know that
the author, Xenophon, had access to them, since he came to know
Cyrus the Younger, the namesake descendant of his Great ancestor,
by participating in the expedition described in Anabasis and mentioned
abroad. The Persians ‘at home’ retained their original mores (8.5.21–27). Due 1989:
36–37 finds that there is no tension in these instances; Gera 1993: 299–300 and Nadon
2001: 142–144 are more sceptical.
8 Due 1989: 30–31 thinks that these declarations ‘assimilate’ the work to historical
writing.
also in Cyropaedia (8.8.3). Indeed, the narrator of Anabasis describes

Cyrus the Younger’s leadership and achievements in similar terms to
those of his Great ancestor (Anabasis 1 passim).
As an illustration of his use of these sources, the narrator validates
his account by repeated references to ‘(anonymous) spokesmen’ (‘what
was said’), and these most frequently authenticate material that might
otherwise strain the belief of the narratees because of its greatness or
other superlative qualities. The most obvious category of such material
is the superlative qualities of Cyrus, as at 1.2.1 (‘it is said and sung by
the barbarians even to this day that he was most beautiful, most kindly,
fond of learning and honour-loving’).9 Other categories of material
for which anonymous spokesmen are adduced include: the incredible
activities of the gods, such as the omen that favours Cyrus on his first
expedition (1.6.1, cf. 4.2.15), the great physical size of a monument
(7.3.16), Cyrus’ huge armies and subjugation of all the nations as far
as Egypt (8.6.20), so that ‘they say’ that he was, hyperbolically, able to
enjoy perpetual springtime (8.6.22).10
Anonymous spokesmen cluster to validate the extraordinary and
tearful emotion that marked Cyrus’ departure from Media (1.4.25, 26,
27) and culminated in the ‘kiss-story’ (1.4.27–28), for which the narrator
offers not only spokesmen, but also an ‘apology’ (‘if it is necessary
to tell it’). The combination might suggest that he wants to distance
himself from the story for some reason (perhaps as too sentimental), but
Herodotus uses the same phrase to introduce things incredible, such as,
e.g. the amazing construction of the moat and walls of Babylon (Hist.
1.178–179).11 The ‘apology’ seems to be yet another way of validating
the remarkable love that Cyrus inspired in his relatives, which is an
important part of his attraction as a ruler.
The narrator uses even dissonant spokesmen to confirm his account
of the superlative speed of the messengers of Cyrus (8.6.18). In this
instance ‘they say’ that the night relays are succeeded by the day relays
without cease, and ‘some say’ that they cover the ground faster than
cranes, but the narrator adds in his own voice that even if they lie about
the birds, the relays are the speediest on earth.
9 Other examples: 3.2.7 (Cyrus ‘said to be the most warlike’), 4.6.11 (a woman
selected for Cyrus, who was ‘said to be the most beautiful woman in Asia’), 8.5.28
(Cyrus’ wife said to be the most beautiful), and 8.2.13–14 (Cyrus’ generosity).
10 Spokesmen also mark Cyrus’ own more astonishing sentiments: 1.3.4,15; 4.2.13.
11 Cf. Hist. 2.24; Hell. 2.4.10 and 6.3.11.
Sometimes the narrator does not insert references to ‘what is said’

but to ‘what seems to be’ (dokein einai). Some have claimed that in these
cases the narrator is distancing himself and expressing scepticism,12 but
they seem rather to function in the same way as the references to
anonymous spokesmen; they stress the impact on those to whom the
events appeared and confirm that impact rather than denying it. They
particularly validate reputations (which of course are a form of seeming
to be to others), as in 1.5.1 (‘he seemed to be the best’) and cf. Cyrus
himself (‘I am and seem to be the best of those of my years in throwing
the spear …’: 1.3.15).
There are a few places where the narrator seems to have no com-
plete knowledge of the matter at hand and gives two possible reasons
for an action or a custom, in the way of Herodotus (→) and the nar-
rators of Hellenica and Anabasis (→). Cyropaedia 7.2.29 gives two reasons
why Cyrus kept Croesus at his side: for security reasons or because he
recognized that he could be a source of benefit to him.13 However, the
rarity of this lack of resolution, and its restriction to motivation rather
than action, makes it difficult to accept that it just reflects the narrator’s
inability to decide between his sources. The device seems rather to give
greater significance to exceptional actions by amplifying their causes.
An overt narrator
Although the primary narrator does not tell us anything about his per-
sonality, he is an overt narrator, who often indulges in first-person com-
ments, which are all aimed at pointing out Cyrus’ excellence to the
narratees. Clusters of such devices mark his most important achieve-
ments, such as the great battle against the Assyrians (7.1), and the tech-
niques he developed to secure his imperial position (8.2). These first-
person interventions vary between the use of the singular and plural.
The narrator uses the first-person plural in the introduction, when set-
ting out the process of reflection that led him to his conclusion that
Cyrus was an exceptional ruler; the first-person singular in the epi-
12 Nadon 2001: 168 reads ‘seeming’ in Cyr. 1.6.20–21 as ‘mere appearance’, as
opposed to the sinister reality.

13 Other examples: 8.3.14 (two reasons why his people prostrated themselves at the
sight of Cyrus in procession); 4.3.2 (two motives for why the Medes bring their women
on campaign even to this day).
logue, when claiming knowledge (8.8.2), and challenging to those who

disbelieve his picture of decline (8.8.27). Perhaps the plural makes the
narrator appear more dignified, in keeping with the grandeur and dig-
nity of his subject.14 It may also express the natural concord of narrator
and narratees, as in Plutarch (→), though this raises the question why
it is not used all the time. Tatum’s observation, that ‘(the epilogue) is
filled with anger and exasperation’ and Gera’s reference to its ‘sharp
sarcastic tone’, seem to arise from the use of the first-person singular,
which can convey this kind of impression.15 The first-person singular
certainly marks superlatives that reinforce the praise of Cyrus, as in
3.3.59 (‘So Cyrus’ army was filled with enthusiasm, ambition, strength,
courage, exhortation, self-control, obedience; and this, I think, is the
most formidable thing an enemy has to face.’).16 Frequently, the narra-
tor qualifies first-person comments as what he ‘seems to know’ (1.1.6
‘we have found out and seem to/think that we know’; 8.1.40 ‘we seem
to/think that we have observed’); these could express uncertainty, but
are regularly used in philosophical works to accentuate the process of
reflection and thus confirm its findings.
Another use of the first person is to mark the disposition of the con-
tent and introduce new topics: in the epilogue (8.8.8, 16) and in the
account of Cyrus’ education (1.2.9, 15, 16; a mix of singular and plu-
ral).17 The narrator does not use this kind of first-person reference much
in his main account of the creation of the empire, perhaps because it
would distract from the sense of active expansion and conquest. The
marking of new sections is more important when the narrator proceeds
from topic to topic rather than through a connected narrative of mil-
itary action. First-person references come back therefore to mark the
more descriptive arrangements for empire (8.1.17, 48; 2.7; 3.1; 5.2—
mainly plural except 8.2.7). Here they also validate the material, by
stressing that it is the product of his enquiry (8.1.40: ‘we think further-
more that we have learned’; 2.2; 2.10: ‘moreover we have discovered’;
4.5; 6.16, 17).
Cyropaedia also uses categories of narratorial interventions, which are
found in Xenophon’s Hellenica and Anabasis (→), most often to praise
14 Hell. in contrast uses the singular in its first-person evaluations and disposition of
material, e.g. 4.8.1; 5.1.4.

15 Tatum 1989: 216 and Gera 1993: 299.
16 Other examples: 8.2.6, 12.
17 Due 1989: 33 concludes that they make the work resemble historical writing and
‘stress points of special interest and importance’.

Cyrus. He comments that it was ‘no wonder’ that Cyrus gave great gifts
because he had such resources, but ‘more wondrous’ that he showed
concern for his friends (8.2.13), for which cf. Anabasis 1.9.24 and Hellenica
7.5.19. He also appeals to ‘many other occasions’ in order to support
the case in question (7.1.30), for which cf. Hellenica 5.4.1. The comment
‘as was natural’ (hōs eikos) and its equivalents mark a natural reaction
in order to contrast it with a subsequent unnatural but admirable
reaction, as in 4.2.32 (the Medes and Hyrcanians let themselves go
in victory ‘as was natural’, but Cyrus did not let himself go and took
action to preserve the victory).18 The phrase is also used of natural
occurrences and family feeling, e.g. the death of the parents of Cyrus
in old age (8.7.1).19
As can be expected in a narrative about non-Greeks, the narrator
sometimes intervenes to explain foreign customs, like the narrator in
Xenophon’s Anabasis (→). These are usually quite short interventions,
and they are not designed just to inform the narratees for whom these
customs are unfamiliar,20 but to enhance the praise of Cyrus or the
broader understanding of the narrative. Thus, the explanation of how
the Assyrians fortify their camps in order to be able to choose the
moment of attack and rely on horses which they have to hobble at
night, which makes them difficult to prepare for action in a hurry
(3.3.26–27),21 is essential to the understanding of the subsequent action,
in which Cyrus is faced with the problem of attacking an impregnable
fortification. The comment on earlier techniques of chariot warfare
proves the uniqueness of the changes that Cyrus made in chariot-
fighting (6.1.27–30).22
Sometimes the narrator reports Cyrus’ thought about best practices
in the past tense and then confirms this in the present tense in his
own voice, for example in 2.1.29 on the effects of hard work on men.
Due says that these interventions have the same function as first-person
comments and the ‘continuance’ motif,23 but they also show how closely
the narrator endorses the practices of Cyrus.
18 Other examples: 3.1.7; 4.2.9 (where it is unnatural that the Medes joined Cyrus—
since they belonged to Cyaxares); 6.2.12. (3.1.41).

19 Other examples: 3.1.41 (bedding of a recently married couple); 4.1.7 (pleasure at a
victory); 4.2.27 (place of cavalry on flanks); 5.4.29 (a rich ally brings Cyrus money).
20 Due 1989: 109–114.
21 An. 3.4.35 explains this custom without reference to fortifications.
22 Other examples: 1.2.6, 3.10; 8.2.4–6.
23 Due 1989: 109–114.
The primary narratees
The primary narratees are occasionally visible. The narrator does not
address them in the second person, but indirectly addresses them in the
form of the ‘anonymous interlocutor’ device, who might raise objec-
tions to the praise of Cyrus. Of course, he makes such references in
order to overcome the narratees’ imagined objections, but they also
suggest a world in which the narratees are actively engaged in the nar-
rator’s investigation, as in Plutarch (→). An example is 8.2.11 (‘if any-
one thinks that the king selected only one man to be his “eye”, he is
wrong’), where the narrator contradicts the assumed opinion of his nar-
ratees (this might be based on Herod. Hist. 1.114, which mentions only
a single ‘eye’), in order to illustrate the generosity that is such a feature
of Cyrus’ rule (cf. 8.2.7–12). Elsewhere the narrator asks his narratees to
remember the pleasure of basic foods to the hungry man if he objects
to the plain diet of the Persian boys (1.2.11 ‘if anyone thinks that they do
not enjoy their food … let him remember’).
The narratees are also invited to visualize or imagine particularly
sensational events, in the form of the—since Homer (→) well-known—
‘indefinite second-person’ device: e.g., ‘then one might have seen that
the equals were trained as they should be; for they obeyed at once and
at once sent orders to the others’ (3.3.70).24 A highly effective appeal
to the imagination of the narratees occurs in 4.2.28. The narrator
has briefly summarized a total of fourteen frenetic reactions by the
enemy to the sudden appearance of the forces of Cyrus, then he ends:
‘One must also think that they did many other things of many kinds,
except for anyone fighting back; because they died without a fight’.
The narrator involves the narratees in the effect of his priamel: they did
everything but fight back.
Secondary narrators and narratees
From time to time characters in their speeches tell stories to other

characters, particularly Cyrus. These include stories about new recruits
that are told in the course of the banquet in the tent of Cyrus (2.2),
Araspas’ story of his first meeting with the beautiful Panthea (5.1), and
24 Another example: 7.1.38.

the personal histories by Gobryas (4.6) and Croesus (7.2). There are
good reasons why these are put in the mouths of secondary narrators,
such as the sympathy Gobryas and Croesus secure by telling their own
stories, which are sad illustrations of the human condition. Most of all,
though, secondary narrations characterize the fine qualities of Cyrus
through his response and add to his praise. His response is immediately
sympathetic toward Gobryas, even though the main narrator notes that
he is aware that Gobryas’ story could be false (it turns out to be true
of course), and toward Croesus, even though he is an erstwhile enemy.
Araspas’ account of his meeting with Panthea could have been part of
the main narrative, which has already described how she was specially
selected as a prize for Cyrus (4.6.11), but the secondary narrative allows
us to see Panthea through the eyes of Araspas, who will eventually fall
in love with her, and allows Cyrus to make him a paradigm of the
dangerous distractions of eros, which he is determined to avoid.
The stories about the new recruits whom Cyrus had drafted into
his army could also have been part of the main narrative, since the
narrator has already described their training, but they are told instead
by secondary narrators, who raise a great deal of laughter among their
dining companions regarding the inexperience and incompetence of
the new men. The primary narrator has indicated that during such
dinners Cyrus ‘took pains to make the conversation as entertaining as
possible while still inciting to good’ (2.1.1). His reaction to their stories
again illustrates this character. The first story is about a recruit who
never was able to secure a good portion of food at dinner, the second
about the inability of another recruit to comprehend even the most
basic military instructions. The stories show the secondary narrators
to be not entirely unkind to the recruits, but having a quiet laugh at
their expense. However, Cyrus reinterprets their significance in a far
more kindly fashion (2.2.10), admiring in the first case how easily you
can make a friend of new men by giving them even a small portion,
and in the second admiring the obedience of the new recruits, even if
they did misunderstand the orders. This fulfils his characterization as
one who turned the conversation toward the good. When another man
then complains that the stories were just lies, made up to amuse, which
he does not consider worth much, Cyrus fulfils the second function by
letting them pass for the sake of amusement (2.2.12).
Conclusion
Cyrus’ apparent perfection has provoked sinister interpretations of Cy-

ropaedia, which find sub-texts of blame under the surface of praise.
This is contradicted by the author’s other works, which all display the
same positivity about paradigms of leadership.25 Moreover, our survey
of the primary narrator and his activities has shown his continuous and
persistent inclination to praise the subject of his narrative, without any
reservation or a trace of irony.
25 See the very brief survey in Due 1989: 185–206.

chapter thirty-one
PLUTARCH
C.B.R. Pelling
Narrative is a slippery thing, and Plutarch ensured that his readers

knew it. Let us begin with a few passages of embedded narrative,
and see how he portrays storytelling in action. Such passages are rare:
Plutarch normally prefers to tell stories on his own authority. But where
they come they often point out the uncertain relation of narrative to
the events described.
Take Proculus Iulius’ narration of Romulus’ apotheosis at Romu-
lus 28.1–3, especially interesting because it is an embedded narrative
(Proculus’) of a further narrative, Romulus’ account of his apotheosis:
so Proculus is a secondary narrator, Romulus a tertiary. The primary
narrative had given several possible explanations of the disappearance
in the previous chapter. Had the senators torn him apart and concealed
the fragments in their clothing? Or had he genuinely been snatched up
to the gods? Now Proculus comes before the assembled Romans, and
tells how Romulus appeared to him and explained that this had been
the gods’ will, to allow him to return to Heaven: now he will protect
Rome as the god Quirinus. The story was believed ‘because of the
character of the man who told it and the oath which he swore’ (so
narratorial authority does make a difference, Rom. 28.3); but an impor-
tant element of uncertainty remains. Plutarch as primary narrator is
decisive enough to conclude that there was indeed ‘something super-
natural’ at play—but in the way that everyone accepted Proculus’ story,
28.3, which is not the same as saying that this divine version was true.
He goes on to make his own scepticism clear, first relating the parallel
‘mythical tales’ (28.4, 7) of Aristeas and Cleomedes, then arguing that
it is ‘stupid’ (28.7) to think that bodies, unlike souls, can be taken to
Heaven. Something of Romulus’ apotheosis may remain, a matter of
spirit rather than corporeality;1 but that is not the way Proculus and
1 Pelling 1999: 441–442 = 2002: 185.

404 part seven – chapter thirty-one
Romulus described it, or at least not the way that people at the time
apparently took their story.
Plutarch’s version of the meeting of Solon and Croesus is a further,
very elaborate case (Sol. 27–28).2 Misreading is in the air. It is set up by
Solon’s own initial misreading of the court, where he cannot tell which
of the sumptuously dressed figures is Croesus himself (27.3). When
questioned, Solon gives embedded narratives of Tellus, then of Cleobis
and Biton: the stories are reported in indirect speech and given briefly
and enigmatically, presumably because Herodotus’ original is taken as
familiar. As in Herodotus, Croesus does not get the point, ‘and so
Solon left: he had given Croesus pain, but left him no wiser’ (27.9). But
Herodotus’ Croesus does become wiser later, and can pass on Solon’s
lesson (or at least an interpretation which only mildly trivializes, Hist.
1.86.5) to the conquering Cyrus. Plutarch’s Croesus tries to do the same,
and there is embedded narrative here too as Croesus tells of Solon’s
advice (28.4–5). Croesus concludes that ‘it was a greater evil to lose
this wealth than a good to gain it’ (28.4): that, for him, was what Solon
must have ‘foreseen’ when he urged him to ‘look to the end’ (28.5). The
primary narratees would be unlikely to have read Solon’s wisdom quite
like that. They would recall Solon’s exchanges with Thales (Sol. 6–7), a
case where an embedded narrative was straightforwardly false, in that
case Thales’ carefully wrought story of the death of Solon’s son. The
text had there pointed the folly of concluding that it is a mistake to have
anything good at all, ‘wealth, or glory, or wisdom’ (7.1), simply because
one might one day lose it—very much the opposite of the moral that
Croesus now draws. In this case, Cyrus is ‘wiser than Croesus’ (28.6),
and takes Solon’s lesson to heart: but it is not clear that Cyrus, this
great man of insight and achievement, reads that lesson as simply or
subjectively as Croesus has done.
So the wise adviser Solon knows that telling stories is a good way of
conveying wisdom; but it also emerges that stories are not easy to read,
and their point can be missed—as it is missed by Croesus, certainly at
the beginning and possibly even at the end.
There are implications here for Plutarch’s narrative too, but they
are subtle ones. It would be wrong to suggest that his own master-
narrative is infected by similar uncertainties, at least most of the time:
2 For a close comparison of the scene with Herodotus see Frazier 1992: 4499–4506:
she particularly stresses Plutarch’s psychological focus on Croesus’ reactions and the
contrast with Solon.
c.b.r. pelling – plutarch 405
there are many devices for establishing Plutarch’s narrative authority.

But there remains a potential uncertainty about narrative: uncertainty
about fact, about interpretation, about moral implication. That is one
reason for the frequent nests of scholarly citation. Not merely do they
establish the narrator’s learning, they also point to the number of
variants attested by other reputable authorities. If the narrator wins the
narratees’ confidence, it is against a background of potential slippage,
the knowledge that other narratives might be possible and that the
narrator himself may not always be confident that this is the right story
to tell.
That suggests a rather sophisticated brand of complicity between
narrator and narratees, and in this chapter I shall explore the ways
in which this complicity is established and developed. That is a vast
topic, for Plutarch’s narratorial interventions can take many forms; he
can convey a response and make it infectious—approval or disapproval,
or simply engagement and excitement—in many ways, and we can-
not look at them all here. Here I will examine only his uses of an
explicit ‘me’ or ‘you’ in the Lives. I begin with his first-person state-
ments.
These are often there to explain the origin of a story.3 While ‘I’
was travelling over the battlefield of Bedriacum the consular Mestrius
Florus told me of the piles of corpses (Oth. 14.2); ‘I’ discover that
Nicias’ shield is still on display in Syracuse (Nic. 28.6, cf. Ages. 19.10–
11); Sextus Sulla of Carthage has given ‘us’ a particular explanation of
the Roman wedding-cry ‘Thalassio’ (Rom. 15.3); Philotas of Amphissa
told Plutarch’s grandfather Lamprias an anecdote of the feasting of
Antony and Cleopatra, then Lamprias passed it on to ‘us’ (Ant. 28,
cf. 68). There are times when such ‘we’s clearly extend to his narratees
as well, or at least some of them. ‘Our’ fathers still tell a story of
Lucullus at Chaeronea (Cim. 1.8); inscriptions of ‘Lucius Cornelius Sulla
Epaphroditus’ are still found ‘among us’ (Sull. 34.4); honours are paid to
Themistocles’ descendants to ‘our own day’, including one who himself
bore the name of Themistocles and was ‘our’ friend in the school of
Ammonius (Them. 32.6)—this last an instance of how a first-person
plural can blur between the inclusive ‘we’, embracing the narratees,
3 For this sort of ‘I’ cf. Russell 1993: 428. It is of course very common in the histo-
rians, from Herodotus (→: cf. also Dewald 1987 and forthcoming) through Thucydides
(→), Xenophon (→), Arrian (→), Appian (→), Cassius Dio (→); also in biography from
Xenophon (→) to Philostratus (→) and beyond. Cf. Marincola 1997: 80–83.
and the authorial ‘we’ which equals ‘I the narrator’. Similar instances
of the ‘continuance’ motif are found elsewhere too.4
Such passages add to the narrator’s ‘authority’ by citing evidence;
they also convey a world where the past has vitality, where ‘we’ still
care, where stories are still told and memorials are on display. The
same goes for those passages, whether or not they include ‘we’s or
‘I’s, which stress continuing controversy, with arguments still being
made: was Aristides really poor (Arist. 1, including a ‘to our day’)?
Should we follow the traditional version of the Megarian decree or
what ‘the Megarians say’, using Aristophanes’ Acharnians to turn the
blame on to Aspasia and Pericles (Per. 30.4)?5 The past is still alive in
other ways too: the Athenians’ magnanimity towards Aristides’ family
was followed by later cases, and ‘even in our own day the city still
produces many examples (deigmata) of generosity and kindness, and
is justly admired and emulated (zēloutai) for it’ (Arist. 27.6–7). Part of
the narrator’s own purpose in the Lives is to provide such ‘examples’
himself; that of Aristides has been followed by many, and the present
city, true to its past, still gives examples for the future and is ‘emulated’
for it. There is a continuing process of inspiration and imitation here,
one in which Plutarch’s own writings play a part. The ‘we’s and ‘our’s
invite narratees as well as narrator to join in this milieu of moral and
intellectual immersion in the past.
Proems and epilogues are particularly important in the narrator’s
characterization of self, of narratees, and of the dynamic between the
two. In proems we often find a strong self-characterization, or char-
acterization of the reading or writing process:6 a display of critical
learning (Arist.-Cat. Ma.; Lyc.-Num.), or moral debate (Demetr.-Ant.; Per.-
Fab.; Agis-Cleom.; CG-TG), or a setting of a hero’s life in a wider ethi-
cal or historical perspective (Cim.-Luc.; Phoc.-Cat. Mi.). These herald the
4 E.g. Sol. 21.7 and 25.1; Lyc. 31.4; Rom. 13.6 and about a dozen other instances
in that Life; Popl. 10.7, 11.6, 15.3, 24(1).3; Arist. 1.3; Them. 22.3; Cim. 19.5; Alc. 21.3;
Alex. 69.8; Phoc. 18.8, 22.2; Fab. 1.8; Flam. 16.5–7; Sull. 21.8. Cf. Frazier 1996: 38. It
is also found in e.g. Herodotus (→), Polybius (→), and Xenophon—interestingly, more
suggestively in Cyropaedia (→) than in Hellenica and Anabasis (→).
5 Frazier 1988: 301–302, like Dover 1966, assumes that these ‘Megarians’ are written
sources; in Pelling 2000b: 272 n. 60 I give reasons for assuming that Plutarch is here
conveying, and very likely constructing, what Megarians would still be saying.
6 Cf. on beginnings Stadter 1988: 292; on ends Pelling 1997a: 231–236 = 2002: 367–
370. See also Russell 1993: 431 on similar projections of a learned persona in the Moralia,
sometimes extending to making a little fun of himself (e.g. Table Talk 675a or 731a–b):
not the case, I think, in these cases in the Lives.
sorts of reflection which are expected of the narratee during the rest
of the narrative too, once the narrator’s personality has receded into
the background (not that it ever disappears); then similar points recur
with particular frequency in the comparative epilogues. Naturally, then,
first-person statements are common in proems and epilogues.7 Many
of those are undeveloped—‘it seems to me’, ‘I praise’, ‘I blame’, ‘I
infer’8—though even these have their point in setting the tone for the
sorts of response which the narrative invites.
There are a few second-person statements too: let us start with the
formal dedicatees. These can be important in setting a work’s tone.
Outside the Parallels, Aratus is dedicated to Aratus’ descendant Poly-
crates of Sicyon, giving ‘examples drawn from their own household’ to
his sons to encourage them to emulation (Arat. 1). The theme of ‘sons’
and Aratus’ descendants ‘to our own day’ recurs symmetrically at the
end (54.7–8). That does not mean that Polycrates’ family are the only, or
even the target, narratees: the very reading of the work tells every new
reader that it extends to a larger audience. But the moralism of Ara-
tus is more explicit than that of the Parallels, with a particular stress on
education;9 and that fits a more straightforward protreptic work aimed
at the young. Polycrates and his sons give a signal of the type of nar-
ratee expected, even if an extreme example of that type; they indicate
narratees to which real readers may assimilate themselves, flattered and
intrigued to think of themselves as moral classmates of the man’s real-
life descendants.
The Parallel Lives give a more refined version of this. Their dedicatee
is Q. Sosius Senecio, twice consul, perhaps himself of Greek origin, and
also the dedicatee of Table Talk and Progress in Virtue.10 The series may
have been initially dedicated to him during his first consulship in AD
99.11 If so, the proem to the lost opening pair Epaminondas and Scipio
would probably have made the appropriateness explicit: this is a lover
7 Just as they are in Xenophon’s Cyropaedia (→).

8 E.g. Dem. 1.1, Demetr. 1.1, Luc. 44(1).8; Cat. Ma. 32(5).3; Num. 24(2).10; Ages.-Pomp.
83(3).1–2 (‘it seems to me’, ‘I think’); Marc. 33(3).1–2; Crass. 36(3).2; Sulla 41(3).7 (‘I praise’
or ‘do not praise’); Sull. 39(1).5; Cat. Ma. 32(5).3 (‘I blame’ or ‘do not blame’); Sull. 41(3).7
(‘I infer’). Outside proems, e.g. Luc. 36.6; Phoc. 4.1; Alex. 8.1; CG. 2.1; Per. 39.2; Sol. 27.1,
Marc. 21.3, 28.6: notice how many of these are close to the beginnings or ends of Lives,
as the narratorial persona gradually recedes or re-emerges.
9 Pelling 1988: 266–267 = 2002: 291.
10 Jones 1970: 103 and 1971: 55; but the eastern origin is doubted by Halfmann 1979:
211 and Swain 1996: 426–427.

11 Thus Jones 1966: 70 = Scardigli 1995: 114.
of the Greeks and yet a great Roman, a military man with a taste for
the past and for culture, a symbol of the interplay of different worlds
and pursuits which the Lives will explore.12
Sosius’ name recurs in various ‘re-addresses’ at Theseus 1.1, Demos-
thenes 1.1, and Dio 1.1. Those placings are not random. Theseus 1 marks
out Theseus-Romulus as the point where the series reaches its extreme
boundary in the past. The mind-set of the critical but sympathetic nar-
ratee is also in focus, as we shall see, and a narratee of ‘ideal’ sophistica-
tion is here constructed with special care. Demosthenes-Cicero will present
two figures who combine culture and a life of action, and investigate
the tensions which that can bring. Sosius, as a contemporary example
of the cultured man of affairs, adds a valuable further perspective. Dio-
Brutus will investigate the Platonic picture of the ‘philosopher in poli-
tics’, especially the Academic philosopher: and ‘… it is right for neither
Romans nor Greeks to complain about the Academy, for they both gain
equally from this book which contains the Lives of Brutus and of Dion’
(Dio 1.1). That suggests a world of cultural fusion, where both Romans
and Greeks learn from philosophy and are interested in its effect on
political action. Sosius sums up that world too.
Sosius, however, is hardly the typical narratee. The Lives often ex-
plain basic Roman terms and institutions—the meaning of hoc age, for
instance, or deliciae, or even magnus (Cor. 25.3–4; Ant. 59.8; Crass. 7.1); or
how the tribunate worked (Ant. 8.5; Cam. 5.1; Fab. 9.2).13 At other times
too they seem to imply Greek narratees, for instance in their comments
on the lack of Roman aesthetic taste (Popl. 15.4), uncharacteristically
abrasive if aimed only at a Roman narratee but wistfully nostalgic if
aimed at a Greek.14 Still, we need not narrow the real-life audience
down, even there: Roman readers might feel flattered to be expected to
share Greek tastes. Real-life readers doubtless extended over a wide
12 Wardman 1974: 39: ‘Sosius is … the reader who already exemplifies by his life and
achievement the kind of activity to which the Lives exhort us’.

13 Ziegler–Gärtner 1980: 200–203 give a list of Latin terms which Plutarch explains.
For other instances where the audience seems Greek, cf. Wardman 1974: 39–40; Duff
1999, index s.v. ‘audience, constructed as Greek’, esp. 302 on the Parallel Lives. Stadter
2000: 494 n. 4 objects that ‘Plutarch frequently explains Greek terms and institutions’
too, ‘especially those of Sparta and Athens. His practice is more a feature of his literary
technique than an indication of a restricted audience’. I agree that the practice does
not give a firm guide to the real audience, but it does give a guide to the constructed
audience, the narratees: where Greek institutions or terms are explained they tend to
be distinctly more arcane, the sorts of thing where even a Greek might flounder.
14 For similar cases of a Greek viewpoint on Roman issues, cf. Swain 1996: 139–145.
range, from the most distinguished of Plutarch’s Roman friends to

impressionable young pupils at Chaeronea. Sosius may be valuable
as intimating one end of that range, rather as in Philostratus’ (→)
dedication of his Lives of the Sophists to Gordian (VS proem), or in his
claim that he re-edited his Life of Apollonius at the behest of Julia Domna
(1.3).15 Readers can reflect on the implications of being included along
with Sosius in that range: the work is suitable for him as well, and that
has its own implications on the value and applicability of what they will
read.16
The proem to Demosthenes tells a tale about Plutarch as well as about
Sosius. One can be virtuous anywhere: it would be odd if small towns,
Ioulis or Aegina, had produced great actors or poets but could not
generate people of goodness and justice (1.2–3). Plutarch himself has
made his home in the small town of Chaeronea (2.2). Still, there is
value for a writer too in living in a great city, where one has not merely
a lavish supply of books but also hears the stories which people still tell
about the past (2.1). When he was in Rome and Italy, he had not had
time to refine his Latin because he was too busy with political affairs
and with those who came to hear him speak on philosophy. Then,
when he did read Latin sources, he had found it a great advantage
that he was already familiar with the substance (2.2–3). He is not
equipped to give a stylistic contrast of Demosthenes and Cicero as
orators—people ‘who have more leisure and whose age is more suited
to ambitions of that sort’ (2.4) might do that—but he can at least
compare them as politicians and men of action (3.1).17
15 For similar dedications of narrative works cf. Marincola 1997: 52–57, pointing out
that they are less frequent in what he calls ‘Great historiography’ than in related,
smaller-scale genres—autobiographies, memoirs, monographs, works with a strong
panegyric element. Those genres are also more ‘personal’ than historiography in that
the narrator too often emerges as more of a character, either as more ‘self-conscious’
about the writing process (as in Plutarch, or, say, in Sallust) or as a figure in the
narrative itself. The two points go together, with both narrator and narratees being
in sharper focus.
16 Swain 1996: 144–145 argues ‘that Plutarch probably looked on … Senecio … as a
man who needed encouragement towards attaining the peace of mind that comes from
Greek philosophy’. Swain bases this particularly on the dedication to Sosius of Progress
in Virtue. I should put this less in terms of Plutarch’s view of the man and more in terms
of the rhetoric of the dedication, the suggestion that even a Sosius might be improved:
but the basic point is similar.
17 Cf. Cat. Ma. 7.3, on the comparison of Cato’s style with Lysias’: ‘This is a matter
for those with a greater feeling for Latin style to decide, but we will include a few of his
bons mots, for we think that human character appears more clearly from what people say
Scholars frequently quote this passage for what it tells us of Plutarch’s

Latin;18 they less often ask what its function is in this pair,19 or how it
characterizes narrator or narratee. Once again we have the intimation
of milieu: in big cities people are still exchanging anecdotes about the
past, in little ones too people are examining their moral health. As for
the narrator’s own self-characterization, (a) small towns can produce
great artists, and (b) Chaeronea is a small town. His distaste for self-
praise—a subject on which we have his moral essay, and on which he
dwells in this pair (Cic. 24, 51[2])20—prevents him from drawing the
conclusion from the two premises, but the self-applicability is not far
to seek. Yet he is more than just a writer, and his aspirations have not
been only to be the good man in private life which the first chapter of
the Life has sketched. He has also been a man of affairs: those ‘times in
Rome and Italy’ make a point, together with those distractions which
prevented him from perfecting his feeling for Latin style. The distrac-
tions consisted in ‘political affairs’—presumably diplomatic missions,21
though again he is diffidently vague—as well as ‘those who came to
listen to me on philosophy’. This is a doer as well as a man of letters
and ideas, a narrator who is well equipped to understand the inter-
play of culture and politics which he will explore in Demosthenes and
Cicero. And what of his narratees? As we saw, this is an appropriate
place to introduce Sosius Senecio; but the wider audience is hinted at
as well. The ‘we’s of this proem are sometimes clearly ‘Plutarch’, ‘we
the narrator’: ‘we’ began to read Latin late in life and visited Italy (2.2–
3), ‘we’ are writing this fifth pair of Parallel Lives (3.1). But some ‘we’s
are vaguer: ‘if we fall short of thinking or living as we should, we shall
ascribe this not to the smallness of our country but to ourselves’ (1.4).
There is an intimation here of a value-scheme which narrator and nar-
than (as some think) from how they look.’ That intimates the narrator’s distance from
the physiognomists as well as from the stylistic critics.
18 As I did myself in Pelling 1979: 75 = 2002: 2 = Scardigli 1995: 267.
19 An exception now is Mossman 1999, who dwells particularly on the contrast of
substance and style and its resonance in the later narratives. Rosenmeyer 1992: 221
does address the question, but reaches the opposite conclusion: ‘The arguments of
the first two chapters … are largely unrelated to what follows’. Russell 1993: 428 has
some good remarks on the self-characterization here: ‘this is both apology and self-
recommendation …’
20 On the problems of narratorial self-praise cf. Marincola 1997: 175–182, and, for
Plutarch in particular, Russell 1993; Pelling 2002: 249. The issue is also addressed by
Polybius (→).
21 Jones 1971: 20–21.
ratee share, just as there will be in the epilogue to the pair—there, for
instance, the valuing of wide culture (Cic. 50[1]), the sympathy for Pla-
tonic views on philosopher-kings (52[3].4), the strong views on political
venality (52[3].5–6). All have the tone of dispensing approval and dis-
approval among a community of morally serious people who think and
feel in similar ways.
Then there is the parade of eschewing stylistic comparison (2.4).
That theme too returns in the epilogue (Cic. 50[1].1), though he goes
on there to do something very close to it anyway—perhaps itself self-
characterization, suggesting that even though he rates substance above
style he can make stylistic points as well. In each case, though, the
tone suggests that the narratees are likely to feel the same way. That
reference to those different types of reader ‘who have more leisure and
whose age is more suited to ambitions of that sort’ (2.4) is not especially
warm, nor does it imply that such a project would be triggered by their
own reading of Plutarch. Then in Demosthenes 3 Plutarch is dismissive of
the stylistic criticism of Caecilius, and that too is not likely to produce
any identification of most narratees with this potential stylistic critic.
Or rather, perhaps, we should distinguish between two different sorts
of constructed narratee. There are those whom the narrator welcomes
and accepts, those whom he is writing for: his ‘target’ narratee, perhaps.
Such a narratee is expected to share his assumptions, in this case
a privileging of substance above style. But there is a second sort of
constructed narratee as well, those who he knows will read his work but
may not be so sympathetic, those who may put quite different questions
to the material. They are not neglected, but not welcomed with such
inclusiveness or warmth.
The inclusive techniques, though, are the more usual ones, and
they can be more far-reaching. Those first-person plurals are here
important. It is indeed often unclear exactly how that category of ‘us’
is envisaged: ‘we Greeks’, ‘we cultured beings’, ‘we people of humane
sensibility’, ‘we who are interested in the past’?22 Does it include real
22 Cf. e.g. Dem. 22.5, the actors playing kings and tyrants ‘whom we see in the
theatres crying and laughing not as they themselves wish, but as the plot demands’;
Per. 8.9 (quoting Stesimbrotus), ‘we do not see the gods either, but we infer that they
exist from the honours they receive and the goods which they give us’. Per. 39.2, ‘… just
as we think it right that the gods, as responsible for good things but not for bad, should
rule over and control all reality, not in the way that the poets terrify us …’; Cor. 32.6,
Homer attributes everyday responses ‘to us’ but more irregular ones to the gods; Arist.
6.5, ‘our’ nature does not allow immortality. In the proems, e.g. Per. 1.4–5, ‘we often
readers in subsequent generations as well as those ‘in our day’, i.e.

Plutarch’s own?23 But in any case it is evidently a category that includes
narratee as well as narrator. Elsewhere too, as in the Demosthenes proem,
a ‘we’ may begin by seeming to be Plutarch himself, but drifts into
being a genuine plural ‘we’ which equals ‘you and I’, narratee and
narrator: ‘now that we have delivered our first narrative, we have
to go on to contemplate experiences and sufferings of a similar size
in the Roman pair, comparing the life of Tiberius and of Gaius …’
(Gracch. 1.1). As in that Gracchi passage, it easily reaches the stage where
the whole project of the Lives is envisaged as a joint investigation
of narrator and narratee: when ‘we compare’ two people (e.g. Phoc.
3.6, Pomp. 81[1].1; Popl. 1.1; Ag.-Cleom. 2.7), or ‘bring on first’ one of
them (Dio 2.7), or ‘contemplate’ the pair’s qualities (Pomp. 84[4].11;
Ant. 88[1].1), that ‘we’ is not restricted to Plutarch himself. There are
also blurred intermediate cases, sometimes very uncharged, where it
is unclear whether narratee is included or not: ‘if we were to say that
those writers were lying (though there are a fair number of them …)’,
Cic. 52(3).6; ‘we do not have anything parallel in Pompey’s career’,
Pomp. 82(2).2. These are not very different from some cases without
an explicit ‘we’, such as ‘but this, I suppose, will seem to support
Lycurgus’ case’ (Num. 26[4].14), or Nicias’ moneymaking ‘will seem
more respectable’ than that of Crassus (Crass. 34[1].1). ‘Seem’ to whom?
‘To me’, or ‘to us’? The blurring is important in insinuating that of course
narrator and narratee are people who think along similar lines.
Such ‘we’s create an impression of happy unanimity between narra-
tor and narratee. There are fewer cases where instead of an inclusive
‘we’ there is a disjunction of ‘I’ and ‘you’, though there too the text usu-
ally suggests basic concord, or at least the likelihood of concord. One
of those ‘you’s comes in the proem to Aemilius-Timoleon, which gives
an interesting twist to the relation of narrator and narratee. Normally
Plutarch presents himself as the model for his narratees, almost an
‘ideal’ narratee and moral respondent to the stories he tells. This time
the movement goes the other way. He began his biographical project
for others, he tells us, but continued it for his own sake, using history
despise the craftsman but admire the work,’ etc., and 2.3; epilogues, cf. e.g. Ant. 90(3).4,
‘as we see in paintings’. Russell 1993: 427 observes that even an ‘I’ can often amount to
‘I, as a typical rational being … ’ The same issue of a blurred ‘we’ arises in Polybius
(→).
23 For this complication cf. de Jong 1987: 36.
as a mirror for making up his own life on the model of those of the
past (Aem. 1.1): so, instead of the narrator’s response cueing that of the
narratee, the process here works the other way round. Soon there are
‘we’s that seem inclusive: ‘it is as if we were entertaining each of them
in turn, welcoming them in the history and examining “how great he
was and what sort of man”, and taking the most important and finest
things we might see in their actions …’ (1.2). Then the ‘we’ becomes
less certain: ‘we use our historical reading and our familiarity with its
writing to mould our own life, welcoming always the recollection of
the best and most glorious figures into our souls …’ (Aem. 1.5): is that
‘we’ narrator alone, or narratees too (as the plural ‘souls’ particularly
suggests)? They too by now have ‘familiarity’ with his writings. Then
the first person becomes more clearly the narrator, but that goes with
a blurring of the narratee: ‘From such examples we have now taken
for you the life of Timoleon of Corinth and of Aemilius Paullus …’
(1.6). Is that ‘you’ just ‘Sosius Senecio’? Or any reader? In any case, the
two lives will generate a ‘debate, whether it was good fortune or good
judgment which brought them their greatest successes’ (1.8), and it is a
debate in which both narratee and narrator will participate.
So the debates are shared ones: there are times, too, when the
text gestures towards the possibility that narrator and narratees might
disagree. In the proem to Agis-Cleomenes, the text gives a summary
interpretation of the Gracchi (2.7–8), and goes on ‘you will judge this
for yourself from the narrative’ (2.9).24 It is up to ‘you’ to ‘judge for
yourself ’, and there is again an intimation that other narratives might
be possible. It is at least conceivable that narratees might construct
an alternative interpretation for themselves. But it is also not very
probable: that summary interpretation had been given in confident
indicatives, ‘they … did not realize that they were entering on a course
where it was not possible to withdraw’. That mild encouragement to
an independent verdict is then echoed in the pair’s conclusion: ‘you
24 Cf. Solon 19.4–5, discussing whether there was an Areopagus before Solon. The
text weighs various learned arguments and inclines towards the view that there was,
though allowing that the crucial evidence could be taken another way: ‘well, then,
consider this for yourself ’. The addressee is taken as engaged and discriminating, one
who might conceivably disagree with ‘Plutarch’ but one who will accept that this is the
way to approach the problem. The atmosphere of debate there continues into the next
few chapters, with vigorous discussion of the rights and wrongs of several laws: notice
the ‘someone might say …’ at 20.8; and the continuation of the principles into ‘our
own laws’ at 21.7 and the preservation of the cylinders to ‘our own time’ at 25.1.
see for yourself the difference on the basis of the narrative. If it is

necessary to set it out in detail, I say that Tiberius came first of them
all in virtue, that young Agis made fewest mistakes, and that in action
and daring Gaius was not far behind Cleomenes’ (Gracch. 45[5].7). It
would not be in keeping with Plutarch’s narratorial persona to assume
that so elaborate a judgment would be taken over in each particular by
every narratee: hence the affectation of diffidence. Other verdicts show
similar tentativeness: ‘perhaps it is time to consider whether we shall
not be far off the truth, if we declare that Sulla got more things right
but Lysander fewer things wrong, and give the one man the prize for
self-control and restraint, the other for generalship and courage’ (Sull.
43[5].6); ‘consider whether, if we give the crown to the Greek for military
skill and leadership and to the Roman for justice and generosity, we
shall not seem to be doing too badly’ (Flam. 24[3].5).25 Yet, despite the
diffidence, more important is the underlying assumption that this is the
sort of judgment that narratees might eventually make, that they are at
least playing the same comparative game. It is still that joint project of
comparison which we saw earlier.
And in all those cases—Agis-Cleomenes-Gracchi, Lysander-Sulla, Philopoe-
men-Flamininus—there is a further twist, for all make particular use of
the idea of competition:26 the ‘crown’ for Philopoemen or Flamininus,
the ‘first place’ for Lysander or Sulla, Tiberius ‘coming first’ and Gaius
‘not far behind’ Cleomenes. All those pairs have also made use of the
idea of competition in their narratives. It was competitiveness which led
to the rifts between Lysander and Agesilaus, then between Sulla and
Marius; it was contention for glory which led Agis, Cleomenes, and
the Gracchi astray; it was ambition for glory which drove Philopoemen
and Flamininus too, and they eventually recognized it as a competition
between themselves.27 In all these pairs the men are contestants, and
the agonistic contests of the narrative prepare for the final synkritic
competition. That is only resolved in the final words, and the judges
are narrator and narratees.
25 Cf. Duff 1999: 203–204 (Lys.-Sull.), 268–269 (Ag.-Cl.-Gracchi), and more generally
286 (though these are not all necessarily ‘courtroom metaphors’, as he says at 286
n. 45). For some related points about the complicity of narrator and narratees in the
epilogues see also Pelling 2002: 361.
26 Pelling 1997b: 329–331.
27 Esp. Lys. 2.3–4, 23.3 and 7; Sull. 4.6, 5.10, 13.1, 39(1).7; cf. Stadter 1992; Duff 1999:
179–180; Ag.-Cl. 2.5 (‘contesting’, hamillōmenoi); Phil. 15.1–3; Flam. 13.1–4; and Pelling
1997b: 91, 220 n. 93.
The same lack of real discord is seen in those passages when an

epilogue imagines an objection: ‘here someone might say that …’ (the
‘anonymous interlocutor’ motif).28 Those objections are sometimes re-
butted, sometimes accepted, or at least accepted in a modified form.
The implication is certainly that the narratee has been pondering the
line of argument critically, and is capable of making independent steps
in the argument; he or she is not wholly a follower. But the implication
is still that both are conducting the investigation according to similar
rules—perhaps, indeed, that the narratee has been led by Plutarch’s
own example, in this and in earlier epilogues, to understand how this
comparative exercise ought to be conducted. Sometimes it is the imag-
ined interlocutor who makes the telling point: ‘yet here someone will
draw a distinction between them’ (Alc. 41(2).8, cf. e.g. Gracch. 42(2).4).
The effect is not very different from the rhetorical questions which
often punctuate epilogues (often in close conjunction with an imagi-
nary ‘someone-might-say’ objection): ‘or is this the first point which
tells the other way?’ (Brut. 56[3].6, cf. e.g. Num.25[3].4–5, 26[4].7; Crass.
37[4].3). Are these soliloquizing reflections of the narrator, or are they
questions put to the narratees? By now the distinction does not matter:
the assumption is that both are engaged, weighing issues and putting
the same sorts of question. The dialogue can become more elaborate
still: a reflective question about attitudes to wealth, ‘or is this the first
point which could tell either way?’ (Cat. Ma. 31[4].1), leads on to ‘I
should like to put the point to Cato himself …’ (31[4].5). By now it
is a three-way moral debate, with narrator, narratee, and subject all
engaged.
Those engaged and sympathetic narratees—following the narrator
most of the way, sharing his tastes and assumption, with an indepen-
dence which remains within limits—may also be sensed when Plutarch
apologizes for a digression: they are independent enough to need an
apology, but are expected to be indulgent. ‘We think that this mate-
rial is not unsuited to our biography, nor that it will seem unhelpful
28 E.g. Sol. 20.8; Marc. 32(2).2; Rom. 32(3).1, 2, 3; Numa 23(1).10, 26(4).1; Brut. 57(4).5;
Ant. 90(3).2 (an imagined objection which ‘one could not make’); Tim. 40(1).3; Popl.
27(4).4. Such a tis-intrusion may not always be an objection, of course: ‘if one examined
their battles’, Flam. 22(1).3; ‘one might particularly think Lucullus fortunate in the time
of his death’, Luc. 44(1).1; ‘one should not wholly excuse Lucullus for this’, 45(2).5; also
e.g. Brut. 56(3).5; Mar. 1.4; Ant. 91(4).5; Cic. 54(5).1; Crass. 34(1).1, 38(5).1; Fab. 30(3). 6; Alc.
44(5).2; Cat. Ma. 29(2).5; Ages. 15.3; Pomp. 84(4).4; Gracch. 42(2).4.
to readers who are not in a hurry and not too busy’ (Tim. 15.11).29
Such narratees may also be felt in Plutarch’s moralism: there is rarely
a sense of telling them anything they might be reluctant to accept; he
rather gives the impression of providing thought-provoking test-cases
within an acknowledged framework of moral values.30 Learning as well
as ethical taste is taken for granted, and such narratees will not be
bewildered by comparisons with other historical events and charac-
ters.31 Sometimes those are great—‘Agesilaus, Lysander, Nicias, and
Alcibiades … Salamis, Plataea, Thermopylae, and Cimon’s successes
at the Eurymedon and Cyprus’ (Flam. 11.5–6); sometimes more mixed,
as with the ‘Fabii and Scipios and Metelli, … or Sulla, Marius, and
both Luculli’ (Caes. 15.2), where the lesser Lucullus brother might not
be in the front of everyone’s mind. Literary culture is also assumed,
enough to welcome the quotations and allusions which lace his nar-
rative;32 enough, even, to catch allusions which the narrator does not
label, confident that the narratee will be able to fill in the gap—‘in
that city of Sophocles’ (Ant. 24.3, referring to OT 4–5), or ‘Greece that
had “endured so very much”’ (Ant. 62.1, quoting Euripides’ HF 1250,
and the Herculean suggestions are important); and many others.33 The
same goes for allusions to myths.34 At Theseus 28.3 the text has just
mentioned Theseus’ marriage to Phaedra: ‘as for the misfortunes which
concerned her and his son, there is no disagreement between the his-
torians and the tragic poets, and so we must assume that it was as
they have all made out’. The narrator clearly relies on the narratees
29 Cf. also e.g. Per. 39, building to the Life’s elevated ending with an excursus on the
moral goodness of the gods, suggesting that Pericles is indeed ‘Olympian’: ‘but these
things will perhaps seem appropriate to a different type of enquiry’. The narratee has
a feeling of appropriateness to context, but will also not mind too much (otherwise
the emotional rhythm of the closure would be wrecked), and may not mind at all
(‘perhaps’). Rom. 12.6 is similar but more elaborate. The text has just mentioned an
attempt to fix Rome’s foundation date by reverse astrology, reading back from its future
greatness: ‘these things, perhaps, will attract by their strange and far-fetched character
rather than alienating those who come across them because of their air of myth’. But
there the possibility of a more cross-grained reaction (‘alienating’) is more explicit:
the formulation ‘those who come across them’ may fit this possibility of a grumpier
response.
30 Or so I argued in Pelling 1995 = 2002 ch. 10. On the moral texture of the Lives see
esp. Duff 1999.

31 On these cf. Duff 1999: 251–252.
32 They are usefully collected by Helmbold and O’Neil 1959.
33 E.g. the Homeric cases now collected by Alexiou 2000.
34 And also for an appreciation of some finer points of philosophy and mathematics:
Wardman 1974: 41–42.

to know what is meant: an important point is built on the Hippolytus

story, again rather allusively, in the epilogue (Rom. 32[3].1–2).
There are moments, however, when less concordant narratees are
envisaged, people whose approach was so at odds that they would be
looking for, or even assuming they had found, one of those alterna-
tive and very different narratives. In Demosthenes-Cicero we noticed that
a different sort of person is acknowledged, someone who might con-
duct the stylistic discrimination that Plutarch and his regular narratees
would avoid. We also noted that these were not treated with the same
inclusiveness: we may have to class them as ‘cross-grained narratees’,
in that the text acknowledges their potential existence, but they are
not proper narratees, not people entering into the spirit of the project,
not the readers whom the writing is for. Dorrit Cohn has drawn a dis-
tinction between ‘consonant’ and ‘dissonant’ narrators;35 we may have
to make a similar distinction among narratees. For elsewhere too such
dissonant, cross-grained narratees are treated in a similarly unwelcom-
ing way; they constitute a foil for the more appropriate response which
more sympathetic narratees will develop.36
It is interesting to see how they are described. At the beginning of
Nicias, ‘it is time to request and call upon those who come across (tous
entugkhanontas) these writings not to assume that I have suffered the same
affliction as Timaeus’ (that is, the ambition to outdo Thucydides: Nic.
1.1). These are ‘those who come across these writings’, a phrase which
in post-classical Greek is often hard to distinguish from ‘my readers’37
35 Cohn 1978: 26–33, discussing the degree of ‘consonance’ a third-person narrator
shows with the psychology of a central character. In this case the ‘consonance’ or
‘dissonance’ will be not with the psychology of any agent within the narrative, but
with the self-presentation of the narrator himself.
36 Xenophon too, both in his more historical works (→) and in Cyropaedia (→), ac-
knowledges narratees who may be reluctant to follow his lead; so perhaps does Polybius
(→). Plutarch develops the notion more elaborately.
37 Schmid (1887) I.41–42, 300, and IV.651 collects post-classical instances of such uses
of entugkhanein and assumes that throughout they mean simply ‘sich befassen mit etwas’,
‘studieren’: cf. LSJ s.v. iii. That is clearly sometimes the case, e.g. at Dio Prus. 18.9,
where it is used of careful reading of historians. But such a rendering elides nuances in
some cases: in a careful discussion, Chantraine 1950: 122–126 finds several cases where,
for instance, a more private entugkhanein is contrasted with a more public reading aloud
(anagignōskein). Our nuance is a rather different one, but a certain casualness of such
chance reading ‘encounters’ (as classically at Pl. Smp. 177b) might also be detected at
e.g. D.H. Dem. 43 (‘I shall take examples, not ones that were carefully chosen but ones I
came across in the Philippics’); Lucian VH 1.4; Strabo 1.16.12; or Philostr. VA 6.27 (and
probably 1.3).
but seems to have a different nuance in Plutarch: as in Demosthenes,

they are not ‘proper’ narratees. The same phrase is used in the proem
to Demetrius-Antony. The narrator there gives his reasons for including
characters whose lives were less creditable, ‘not (for Heaven’s sake!) to
give variation to my writing so as to give pleasure or diversion to those
who come across (tous entugkhanontas again, Demetr. 1.5) it’. The better
approach is to realize that ‘we will become more enthusiastic in our
contemplation and in our living of better lives, if we pay attention too
to those who are bad and are censured’ (1.6). That is what ‘we’ do—
and that ‘we’ gives a more regular embrace of the narratee as someone
who reacts as the narrator himself does.
Or consider the famous passage which begins Alexander. ‘We shall ask
our readers mē sukophantein if we do not include everything or go into
every detail of famous events, but abbreviate most of them. For it is not
histories we are writing, but lives …’ (Alex. 1.1–2). There mē sukophantein
is usually translated as ‘not to complain’ (Perrin, Waterfield, Hamilton,
Duff) or ‘not to regard this as a fault’ (Scott-Kilvert). There is more to it
than that. The word always carries a notion of something disingenuous
or disreputable: ‘criticize in a pettifogging way’, ‘quibble’ (LSJ I.2) is
better, or ‘de ne pas nous chercher chicane’ (Chambry). Whether or
not the complex suggestions of classical Athenian ‘sycophancy’ are felt,
there is always a hint that the objector is not being sufficiently generous,
or that he is not saying what is really in the mind.38 Such complaints are
some way from the engaged and sympathetic ‘someone might say …’
objections in the epilogues.
Theseus-Romulus provides a more elaborate example where the nar-
rator toys with a degree of narratee independence. The proem indeed
38 Perhaps non-coincidentally, the word recurs twice at the end of the narrative at
Alex. 74.4–5, where the issue is whether those accusing Antipater are doing so falsely.
At Num. 9.3 the lawgiver does not sukophantein in the case of a genuine impediment
in conducting sacrifices, that is ‘does not make unreasonable objections’. At Cat. Mi.
11.4 some critics esukophantoun at the expense of the funeral of Cato’s brother, failing
uncharitably to realise the depth of his capacity for emotion. At Pomp. 2.10 Pompey
esukophanteito as neglecting public affairs because of his wives. The narrative will show
there is some truth in this, but for the moment the critics are stigmatized as ungenerous:
Pompey is ‘careful and guarded’ about his love life, but ‘nonetheless was blamed by his
enemies’. Naturally, Plutarch also uses the word in contexts of classical democracy:
Sol. 24.2; Arist. 26.2; Per. 37.4; Alc. 13.6, 19.7, 34.7; Tim. 37.1; Phoc. 12.3 etc. It is never
friendly or neutral. See more generally on the word’s range Harvey 1990, singling out
the suggestions of monetary motivation, false charges, sophistical quibbling, slanderous
attack, taking people to court, and raking up old scores; ‘sophistical quibbling’ is the
nearest to the present use.
asks ‘our listeners to be indulgent and to accept ancient tales in an

acquiescent mood’ (Thes. 1.5): that assumes the same sort of read-
erly independence as before, as if the indulgence cannot be taken for
granted. The narratee would normally be critical of such unreliable
material, and in ways which are not merely triggered by passages of
explicit discussion in the text: thus Plutarch cannot simply avoid such
criticism by refusing to question veracity himself. To ask for such dis-
crimination among different types of material is to demand, and to
assume, considerable sophistication in a narratee: in that sense there is
flattery here. But it is also a sophistication which is close to Plutarch’s
own, or at least to the sophistication which he temporarily affects for
this pair. As in the epilogues, even the independent narratee is assumed
to be conducting games which are not too distant from those played by
Plutarch himself: for even if that narratee does decide that the mythical
has not been made to look like history, he or she will be doing so by
applying criteria similar to those which Plutarch has acknowledged he
would apply elsewhere.
The same goes for the epilogue, where the narratees may still not be
sure how serious, and how convincing, the whole exercise of ‘making
myth look like history’ has been.39 Are these then narratees who are
constructed as thinking differently from the narrator, who have been so
perplexed by the clever moves and ironies that they are finally at a loss
to work out what sort of text they have been reading? Not at all: for
the proem itself approached this singular project with some diffidence,
affecting uncertainty whether it would come off. Now, at the end, we
again have the effect of rumination, with narrator as uncertain about
narrative status as narratee. Narrative is still slippery, especially in this
pair, and both parties are assumed to know it. Even in uncertainty,
even when the narrator has highlighted the possibility that the narratee
may not be able to go with him the whole way, narrator and narratees
are not so very different, and share the same sort of patience with the
material and subtlety in the way they toy with it.
The same is true in those cases when the narrator reveals a moral
response to his material which goes beyond those views that the narra-
tee would naturally share. ‘For myself, I would not even sell a working
ox because it was too old, never mind an elderly man’ (Cat. Ma. 5.6).
‘I would not myself agree with Demaratus of Corinth, who said that a
39 Or so I argued in Pelling 1999 = 2002 ch. 7.

great pleasure had been denied those Greeks who had not seen Alexan-
der sitting on Darius’ throne’ (Ages. 15.4). ‘The wisest judges put par-
ticular weight on Tigellinus’ impious and unspeakable cavortings with
prostitutes … This, those wise persons thought, was the worst punish-
ment of all, outweighing a multitude of deaths’ (Otho 2).40 The narra-
tees might not go that far; some would not gibe at turning an honest
drachma from an aged ox; some might even prefer the odd whorish
cavorting to even a single death.41 But at least we are expected to find
the narratorial persona attractive rather than repellent, someone with
whom we can engage and even identify, at least most of the way.
‘We are expected …’, ‘we can engage …’: those are phrases which
the modern scholar uses unselfconsciously, and which have many paral-
lels with the sort of inclusiveness for which I argued above.42 The same
goes for the rhetorical questions: ‘are these then narratees who …?’:
am I asking my own readers, or myself ? The implications are similar
too, of a barely conscious attempt to insinuate the notion that reader
and author are at one in a joint investigation. It is not that scholarly
discourse has stayed the same: the manner is different from that, say,
of nineteenth-century scholarship. It is rather that the ‘Plutarch’ which
40 This essay has been confined to the Lives, but I cannot resist one example here
from the Moralia. At Epicurus makes a pleasant life impossible 1093c the text has been dis-
cussing the absorbing power of imaginative literature, and concludes ‘who would take
pleasure in sleeping with the most beautiful of women rather than staying awake with
what Xenophon wrote about Panthea, or Aristobulus about Timocleia, or Theopom-
pus about Thebe?’ The answer was never going to be ‘no one’.
41 Pelling 1995: 206 = 2002: 238. Contrast Booth 1983: 157: ‘From the author’s
viewpoint, a successful reading of his book must eliminate all distance between the
essential norms of his implied author and the norms of the postulated reader’. Not ‘all
distance’, if the argument here is correct: the remaining distance should not be large,
but it may exist. The important point is that any disjunction of views should not be
genuinely alienating, and those of the narrator, or in Booth’s terms ‘implied author’,
should be found attractive even if not irresistible.
42 Compare the response of a modern philosopher to friends who had questioned
his use of the ‘ubiquitous “we”’ (e.g. in phrases like ‘our ethical ideas’ or ‘what we
think’). ‘It refers to people in a certain cultural situation, but who is in that situation?
Obviously, it cannot mean everybody in the world, or everybody in the West. I hope it
does not mean only people who already think as I do. The best I can say is that “we”
operates not through a previously fixed designation, but through invitation. (The same
is true, I believe, of “we” in much philosophy, and particularly in ethics.) It is not a
matter of “I” telling “you” what I and others think, but of my asking you to consider
to what extent you and I think some things and perhaps need to think others’ (Williams
1993: 171 n. 7). It is hard to better this description of the ‘invitational “we”’, and it fits
closely on to what I have been suggesting here for Plutarch’s Lives.
the Lives construct, with that combination of learned disquisition with

vitality, engagement, and genial characterization of self and narratee,43
has much in common with the more informal style of much current
scholarship, or at least with the scholarly persona which many of ‘us’ try
to project.
43 And a heavy degree of self-referentiality too: this paragraph is sufficient testimony
to that.
chapter thirty-two
PHILOSTRATUS
T.J.G. Whitmarsh
The Philostratean corpus1 is large and diverse (incorporating Love letters,

two collections of Descriptions of paintings, dialogues and essays). His two
longest and most sophisticated narrative texts are In honour of Apollonius
of Tyana (henceforth VA) and the Lives of the sophists (VS), both of which
are—in the broadest sense—biographical works.
In honour of Apollonius of Tyana
I begin with the earlier text, a magisterial and voluminous hagiog-

raphy of the first-century sage and holy man. ‘Hagiography’ is not
a neatly compartmentalized form, though, and critics have empha-
sized the text’s generic affinity to the Greek novels.2 What it shares
in particular with the ‘sophistic’ novels (those of Achilles, Longus, and
Heliodorus) is attentiveness to the authorization of the primary narra-
tive, layered narratives, and ‘overt’ narration at every level. For some,
this is evidence of its playfully ‘sophistic’ approach to its subject mat-
ter,3 but formal sophistication does not necessarily imply a lack of com-
mitment to content. As we shall see, the narrative strategy is largely
consonant with the text’s express purpose (1.3) of magnifying its hero,
although at the same time its many delicious ambiguities do indeed
play self-consciously to the learnedness of the primary narratees.
1 On problems of attribution, see esp. Lannoy 1997. In honour of Apollonius of Tyana
and the Lives of the sophists are by the same author, as guaranteed by the cross-reference
at VS 570.
2 For the VA as quasi-novel, see Bowie 1978: 1663–1667; Swain 1991: 150; Bowie
1994; Anderson 1996; see also Reardon 1971: 189–190; Hägg 1983: 115–117; Reardon
1991: 147–148; Billault 1991. Other scholars, however, consider it to be, at least partially,
historically accurate, e.g. Jackson 1984; Dzielska 1986; Puskás 1991. For crucial objec-
tions to the use of Eastern literary traditions to support the VA, see Anderson 1986: 173
n. 106; Koskenniemi 1991: 10 n. 34.
3 Anderson 1986.
424 part seven – chapter thirty-two
The primary narrator does not name himself, but nor does he gain-
say any assumed identification with the rhetorical superstar Philostratus
(he is a rhetorician and a member of the ‘circle’ of Julia Domna: 1.3).
Even so, narrator cannot be conflated simply with author. The marked
style of the writing (simple but authoritative) is a device chosen to con-
struct a persona for the occasion. Literary style is an issue throughout:
Apollonius uses laconic brevity (7.35), and ‘cultivated a literary style’
that was ‘not dithyrambic or tumid or swollen with poetical words, nor
again was it tongue-tied and hyperatticized’ (1.17; for his style, see also
3.36, 41; 8.6). These descriptions of the subject of the narrative also
function as a stylistic programme for the primary narrator. The narra-
torial persona is thus assimilated to the subject, a phenomenon that we
shall encounter again.
The primary narrator narrates overtly and (as one would expect
at 150 years’ distance) wholly externally.4 Limitations in the narrator’s
knowledge are acknowledged, with ‘I think’ (oimai) a common paren-
thetic interjection;5 elsewhere, he affirms that he ‘knows’ (2.2; 7.1). Such
markers of focalization (particularly ‘I know’ [oida] and ‘it is my opin-
ion’ [moi dokei]) are, as it happens, said to be features of Apollonius’
‘oracular’ style (1.17); this further underscores the Apollonian aspect of
the narrator’s style. In addition, these markers focus the primary narra-
tees’ attention upon the problems of constructing a factual account of a
historical figure, particularly a religious leader around whom so many
layers of myth have accrued. In the Herodotean manner (→), the narra-
tor claims personally to have seen artefactual evidence (3.41, where he
has seen one out of the two books attributed by tradition to Apollonius;
contrast 8.20, where he has not seen the book himself, relying on oth-
ers’ testimony instead); he has travelled to Spain to confirm Apollonius’
theory of Atlantic tides (5.2); he has personally seen a satyr (6.27).
The narrator and his sources
But these ‘autoptic’ elements, overall, represent either vestigial traces or

circumstantial evidence: the primary testimonia to Apollonius’ life come
in literary sources, and it is to these that we now turn. The narrator
4 See further Billault 1993 on the self-presentation of Philostratus in this text.

5 1.21, 1.37, 2.9, 3.14, 5.43, 6.24, 7.3, 8.1, 8.9.
t.j.g. whitmarsh – philostratus 425
is highly self-conscious in his use of sources. The guise of the source-

critical historian is assumed from the outset, in a passage that it is worth
citing in full:
It seems to me that I should not condone or acquiesce in the general
ignorance, but write a true account of (exakribōsai) the man, detailing the
exact times at which he said or did this or that, as also the habits and
temper of wisdom by which he succeeded in being considered a super-
natural and divine being. And I have compiled my information partly
from the many cities where he was loved, and partly from the temples
whose long-neglected and decayed rites he restored, and partly from the
accounts left of him by others and partly from his own letters … But the
more precise details (akribestera) are as follows. There was a man, Damis,
by no means stupid, who formerly dwelt in the ancient city of Nineveh.
He resorted to Apollonius in order to study wisdom, and having shared,
by his own account, his wanderings abroad, wrote a study of them. And
he records his opinions and discourses and all his prophecies. And a cer-
tain kinsman of Damis drew the attention of the empress Julia to the doc-
uments containing these memoirs hitherto unknown. Now I belonged to
the circle of the empress, for she was a devoted admirer of all rhetorical
exercises; and she commanded me to rewrite (metagrapsai) these essays,
paying attention to their style and diction; for the man of Nineveh had
told his story clearly enough, yet somewhat awkwardly. I also read the
book of Maximus of Aegae, which comprised the life of Apollonius in
Aegae; and furthermore a will was composed by Apollonius, from which
one can learn how rapturous and inspired a sage he really was. For we
must not pay attention to Moeragenes, who composed four books about
Apollonius, and yet was ignorant of many of the circumstances of his life.
That then I have combined these scattered sources together and taken
trouble over my composition, I have said; but let my work redound to
the honour of the man who is the subject of my composition, and also
be of use to those who love learning. For assuredly they will here learn
things of which as yet they are ignorant. (1.2–3)
The pose here is partly Herodotean / Thucydidean: the narrator has
travelled and ‘compiled’ sources from authoritative witnesses and docu-
ments, bringing together assiduously the ‘scattered sources’ to produce
a text that is akribēs (‘accurate’, a key word for Thucydides, from the
latter’s famous claim at 1.22.1 onwards). In four cases, he alludes to
information he has ‘discovered’ (1.25; 6.40; 8.5, 8), presumably during
his travels. Information (whether generally ethnographic or specifically
dealing with the Apollonius narrative) gathered from tradition is fre-
quently marked, with an unspecified ‘they say that’ or ‘it is said that’
(the ‘anonymous spokesmen’ device).6 The Herodotean voice resonates

throughout the text, particularly in the description of events as ‘wor-
thy of record’ or ‘worthy of remembering’.7 From another perspective,
however, the narrator is defiantly un-Herodotean: in addition to the
role of autoptic traveller, he also plays that of learned archivist, consult-
ing the books already written on the subject. The narratorial claim to
akribeia is substantiated by reference not only to experience but also to
learning. This latter very much bespeaks a post-classical context domi-
nated by didactic institutions (schools, seminars, libraries).8
Maximus and Moeragenes have attracted a certain amount of atten-
tion from scholars wishing to reconstruct their beliefs, but in narrative
terms their importance is limited (each is mentioned only once more).9
They serve primarily as foils to the narrator’s own project, examples
(supposedly) of distortion whether through want of akribeia or mal-
ice.10 By far the most significant source is Damis. The Damis mem-
oirs serve as a Beglaubigungsapparat,11 a technique designed to enhance
the authority and plausibility of the narrative. This is most prominently
exemplified at 4.25, where the narrator presents ‘the best-known story
of Apollonius’, but supplants the common version, handed down ‘in
a vague and general manner’, with his own truer version which he
anchors explicitly in Damis’ memoirs. But the Beglaubigungsapparat is
6 2.20, 2.21, 3.1, 3.2, 3.4, 3.8, 4.23, 5.1, 5.3, 5.4, 5.39.
7 1.9; 2.4, 16, 28; 5.1, 7, 9; 6.13; 7.3, 28, 35.
8 See further Elsner 1994.
9 See Raynor 1984 on Moeragenes; Graf 1984–1985 on Maximus; and more gen-
erally on the sources for the VA, Flinterman 1995: 67–88. The title of Moeragenes’
work, ‘Memorabilia of Apollonius of Tyana’, is recorded at Origen, Against Celsus 6.41.
The second mentions are: ‘These and many similar incidents are given by Maximus of
Aegeae in his treatise, a writer whose reputation for oratory gained him the position of
imperial secretary’ (1.12); ‘… four books concerning divination by the stars, a work that
Moeragenes has mentioned’ (3.41).
10 The narrator also prominently disputes with other writers on ethnographic issues:
2.9 (the Alexander historians; cf. also 2.18), 2.13 (Juba on elephants), 3.6–8 (various
poets on dragons), 7.35 (unspecified malicious accounts). Similarly, he disagrees with
received opinion (unspecified) at 3.2, 4.25, 5.39. Secondary narrators also use sources:
Apollonius reports Juba on elephants (2.16); the account of the shadow-footed men
(attributed by the primary narrator to Scylax) is refuted by Iarchas (3.47). Other writers
are alluded to without disagreement at 2.12 (unspecified), 2.17 (Nearchus and Pythago-
ras), 3.53 (Nearchus and Orthagoras), 4.28 (unnamed local historians). Quellenforschung
confirms the importance of literary antecedents in the narration of Apollonius’ trav-
els: Rommel 1923: 1–59 argues that the ethnographic digressions can be traced to the
author’s reading in conventional sources or his own invention.
11 Bowie 1978: 1663–1667. For more on the ‘Damis question’, see Edwards 1991.
also an established device in the tradition of literary fiction, which fact

inevitably raises unsettling questions about the truth-status of the text.12
The uncertainty over the text’s generic status—history or novel?—feeds
the (playful) uncertainty over the degree of authority possessed by its
primary narrator.
The use of Damis as a source is proclaimed on numerous occasions,13
as is that of Apollonius’ writings.14 The problems raised by the limita-
tions of the sources are confronted overtly and ingeniously. At 3.27, a
conversation to which Damis was not party, is nevertheless derived from
him, on the grounds that ‘he heard [it] from Apollonius and wrote
it down in his book’—this technique of narrative authorization (sup-
plying a plausible explanation for the narrator’s later grasp of events
that would have been unknown at this stage) is already found in the
Odyssey.15 Again, Damis does not record how Apollonius died (8.29),
and so the narrator is constrained to offer contrary and conflicting nar-
ratives (‘some say … others say’, 8.30). This is the only time in the text
where the narrative presents an unresolved amphibolē,16 but the source
limitations are turned to an advantage. The haziness surrounding the
sage’s death raises doubts about whether he died at all, thus allowing
the insinuation that he was in fact immortal (8.31).17
In many other instances, Apollonius and Damis are conflated as
reported narrators, the primary narrator simply stating that ‘they say’
(phasi).18 All these instances come in the midst of the travel-narratives,
which employ in these cases the ‘experiencing’ focalization of the re-
ported narrators. In several cases, however, there is a lack of clear
12 For the literary Beglaubigungsapparat, see Speyer 1971: 78–79; for the Damis memoirs
as self-conscious advertisements of novelistic genre, see Bowie 1994: 194.
13 1.24, 1.32, 1.34, 2.10, 2.28, 3.15, 3.27, 3.36, 3.41, 5.9, 6.22, 7.15, 7.21, 7.34, 7.42.
14 1.32, 2.20, 3.14, 3.15, 5.39.
15 Hom. Od. 12.389–390; cf. e.g. Ach. Tat. 8.15.
16 Except 2.3 on the Prometheus narrative, which is however subsequently dismissed
as a myth.
17 A further case of source limitation turned to profit: at 3.13, neither Apollonius
nor Damis is said to know how many approaches there were to the Brahmans’ hill,
‘for the cloud around it did not allow them to be seen’. Apollonius, who possesses the
advantages of second sight and prophecy, is usually an entirely authoritative source; the
narratees conclude that only among the Brahmans, where he plays the role of acolyte
rather than teacher, does his omniscience fail him.
18 2.1, 2.4, 2.12, 2.19 (a tertiary narrative: ‘they also say that they learned from the
Indians …’), 2.20, 2.24, 2.25, 2.42, 2.43, 3.5, 3.6, 3.9, 3.13, 3.53, 3.54, 3.56, 3.58, 5.5,
5.6, 5.14, 6.26. Reported narrators also occur in Herodotus (→), Thucydides (→) and
Xenophon (→).
distinction between phasi = ‘Apollonius and Damis say’ and phasi =

‘tradition reports that’. Thus at 3.5, there are two uses of phasi that
clearly refer to the reported narrators (‘they say that they saw …’,
‘they say that they came across …’), but sandwiched between them
is a use that seems to refer to a tradition: ‘they say that this soil is
the best in India, and constitutes the greatest of the territorial divisions
…’ How would Apollonius and Damis know this, and what would be
their interest in reporting it? Similarly ambiguous is 3.9, where a phasi
referring to Apollonius and Damis follows directly on the statement
that the Indians ‘are said’ (legontai) to understand animals by eating the
heart and liver of dragons. There is no reason why Apollonius and
Damis should not be narrating this interesting extension of dragon-lore,
but the use of the passive distances it from any attribution to a specific
source.
The primary narrator’s aim of ‘rewriting’ (metagraphein) the mem-
oirs of Damis raises interesting narratological issues. Many of these
are questions of focalization: how much of the text is Damis, the naïve
faithful, star-struck, provincial who experienced these miracles with his
own eyes? And how much is Philostratus, the metropolitan sophisti-
cate and expert on rhetorical style, who, however, has no personal
experience of the sage? These matters are largely beyond the scope
of this volume, but there are also questions to do with narratorial self-
presentation, and these must be addressed here.
The principal issue is how to refashion source-material—particularly
the memoirs of Damis, which constitute the fabula for large parts of the
text—into a meaningful narrative.19 The narrator expresses concerns
for the ‘length’ (mēkos) of the text, seeking to control the proliferation
of material by judicious selectivity (‘Let us give up this prolixity [mēkos]
and proceed …’, 4.34; ‘my account does not permit prolixity [mēkos]’,
7.2). He omits material that narratees can consult in other sources
(Apollonius’ Epistles, 5.39; the sayings of Musonius, 5.19), ‘mythical’
accounts (muthōdē, 5.1: an allusion to Thucydides’ pointed statement
of omission at 1.22.4), digressive material (‘no more on this subject’,
6.1; cf. 7.39), and the obvious (‘I need not describe the exaltation of
Egypt, nor how the people … applauded him for this action’, 5.24).
He includes things ‘worthy of memory’ and ‘the more serious’ (ta
spoudaiotera / spoudaiotata) events (4.22; 6.35); there are indeed occasions
19 Narratorial selectivity is also an issue for Herodotus (→).

when a part of the fabula is so significant that ‘my account does not
allow me to pass over’ it (2.18), or it is ‘unworthy of omission’ (2.28).
Yet selection can never be a neutral narratorial act. The operative
principles of selectivity throw light upon the self-positioning of the
narrator, and also upon the way he constructs his sources.
The most important consideration relates to the question of how the
primary narrator excerpts Damis. In book seven, he declines to include
all of the events that occurred in prison, contrasting his own selectivity
with Damis’ supposed practice of recording everything that happened:
There followed other episodes in prison, some of them insidiously con-
trived, others of mere chance and not of sufficient importance to merit
my notice (Damis, I believe, has recorded them in his anxiety to omit
nothing). The following, however, are relevant to my account. (7.28)
Damis is constructed by the primary narrator as an uncritical recorder

of everything. This construction is a rhetorical one (no narrative can
present everything, and Damis’ own selectivity is averred elsewhere),20
and of course serves the narrator’s agenda: a naïve source is by impli-
cation an honest one, and the latter is implicitly constructed as using
truthful sources while adopting a more knowing and sophisticated posi-
tion himself (a strategy that has many parallels in modern scholarly
discourse). The tendentiousness of this pose can be pointed up by look-
ing to those passages where the primary narrator’s practice varies. In a
passage in book one, he once again confronts the issue of omission, but
takes a different tack:
For the sake of accuracy (akribologia), and in order to leave out nothing
of the things Damis wrote, I should have liked to have related all of
the incidents that occurred on their journey through these barbarous
regions; but my subject hurries me on to greater and more remarkable
episodes. (1.20)
At first sight, this contradicts the rhetoric of the previously quoted pas-
sage: omission is now held to compromise ‘accuracy’. Akribologia, how-
ever, is not quite the same as akribeia, having a negative connotation
of nit-picking pedantry.21 To narrate all of Damis’ fabula would (it is
20 ‘Many were the discussions that, according to Damis, the sage held in Athens;
but he did not write down all of them, only the more indispensable ones that han-
dled [spoudastheisas] great subjects’ (4.19). The use of the spoud- root implicitly asso-
ciates Damis’ own selectivity with the primary narrator’s emphasis upon recording the
spoudaiotera/spoudaiotata.
21 LSJ s.v. 2.
implied) betray an unsophisticated inability to control material criti-

cally. From this perspective, this passage intermeshes with the other,
constructing Damis’ memoirs as naïve but ontologically unproblematic,
a foil for the flamboyant but veridical primary narrator. The two pas-
sages can be seen as strategic attempts to deproblematize the process of
metagraphein, of ‘rewriting’ source material.
Yet it is not all one-way traffic: the narrator is also at the mercy of
his sources. At 3.45, he asserts that he must not omit a discussion that
took place ‘since it has been recorded by Damis’, a claim that contra-
dicts the principle of selectivity advanced elsewhere. (Similarly at 3.6,
an anecdote ‘must’ be presented, the narrator tells us, because Apol-
lonius records it.) Issues of accuracy also arise when different sources
report different versions.22 At 3.15 Apollonius’ elliptical account in an
unspecified medium of Brahmanic dwelling (they live ‘on the earth but
not on it, fortified yet without fortification, possessing nothing yet hav-
ing the riches of all men’) is said to have been presented ‘too cleverly’
(sophōteron): Damis’ more descriptive version is used by the primary nar-
rator to decode Apollonius’ allusions. This presentation of two different
focalizations within the text allows the narrator simultaneously to mys-
tify Apollonius and the Brahmans, and to preserve his self-arrogated
narratorial stance of lucid akribeia. At the same time, however, it points
up his dependence upon his sources, with all their limitations and opac-
ities. Although Apollonius’ version is stigmatized as ‘too clever’, Damis’
is implicitly too mundane, failing to communicate the theurgic power
of the sages. Indeed, we have argued above that the narrator artfully
makes use of the gaps in Damis: over the text as a whole, the episodes
not witnessed by Damis himself (the interview with Iarchas and the
Indian king at 3.27–33, and particularly the ‘death’ of Apollonius, 8.30–
31) are arguably marked as wondrous, ‘mystic’ moments.
Narratees
Although the narrator claims to be writing for the empress Julia Domna
(1.4), the predominant role he assumes in relation to his narratees is
pedagogical. At a number of points, he guides interpretation with sub-
22 Sources do not have to contradict each other: Damis is said to have recorded the
conversations of Apollonius and Vardanes, while ‘Apollonius himself composed a letter

junctives in the first-person plural (‘let us …’),23 thus imagining a com-

munity of readers steered by the narrator. Elsewhere, he adopts a more
imperative tone: ‘one must think that …’ (khrē gignōskein … 3.1); ‘it is
clear from what I have said that …’ (5.12). The narration targets ‘those
with no experience (apeirois) of Apollonius’ (6.35), a statement that is at
one level self-evident (anyone of Philostratus’ time with direct experi-
ence of Apollonius would have to be at least 120 years old), but serves
primarily to cast the narrator in the role of informed empiricist, a
man with peira (experience) he wishes to impart to students. At 7.39,
he explains his digression on the question of sorcery as motivated by a
desire to protect ‘young men’ from its lures. The narrator here assumes
the guise of the responsible citizen in loco parentis. Although there are
occasions where he proclaims the risk of a negative response from his
primary narratees (because he goes beyond what the Alexander histori-
ans say, 2.7; when he implies Plato’s foolishness, 7.3; because he includes
a speech of Apollonius that does not conform to contemporary orator-
ical fashion, 8.6), but these cases are best interpreted as devices that
procure consent (by stigmatizing negative reaction) and simultaneously
dramatize the narrator’s fearless resistance to received opinion. Overall,
the narrator presents himself as an enlightened, if unconventional, edu-
cator of young acolytes; and this is clearly another point of convergence
between the narrator’s representation of himself and of Apollonius.
Embedded narration
Because of the primary narrator’s insistent focus upon sources, it is per-

haps best to conclude that the entire text is the product simultaneously
of primary (by ‘Philostratus’) and reported narration (by the sources).
There are, in addition, tertiary narratives. On several occasions, Apol-
lonius uses narrative exempla (cf. paradeigmata, 1.34) as philosophical para-
bles; that is, as moralizing reflections upon the situation in which the
characters find themselves.24 Parables, however, cannot guarantee their
containing them, and has sketched out in his epistles much else of what he said in
conversation’ (1.32).
23 ‘Let us consider …’ (2.2); ‘Let us proceed …’ (4.34); ‘Let us consider …’ (5.12);
‘Let us not disbelieve …’ (6.27); cf. ‘We might deduce this from the following evidence
…’ (8.2).
24 1.34, 2.14, 2.37, 5.14. The primary narrator also makes interesting use of this
own correct reading: the process of mapping the tertiary narrative

parable onto the primary/secondary narrative situation is not secure,
since the ‘meaning’ of a narrative is infinitely (re)interpretable. Thus at
7.30, Apollonius and Damis trade stories and interpretations of stories
about lions, as each attempts to establish the best position to take rela-
tive to Domitian. Apollonius begins by narrating how Phraotes advised
lion-tamers to approach the beasts without striking them or fawning
upon them (so he will approach the emperor with cautious authority).
Damis responds with a narrative from Aesop about the lion who pre-
tended to be sick, so that he could eat animals that visited him, until
a fox realized what happened (so he counsels avoiding the emperor,
though he may look docile from a distance). Apollonius responds that
the fox would have proven himself wiser if he had entered the cave but
not been caught. This is the end of the exchange, but the primary nar-
ratee may well imagine a counter-narrative to re-establish the fact that
tyrants are best treated with caution.
Elsewhere, embedded narrative is treated with great sophistication.
There are four levels of narration in the account of Achilles (4.16:
primary narrator reports that Damis reports that Apollonius reports
the narrative presented by Achilles’ spirit), and in Iarchas’ account of
the Hellanodicae (3.30: primary narrator reports that Damis reports
that Apollonius reports that Iarchas reports).25 At the tertiary level,
narrators are often those in positions of particular symbolic importance,
especially esteemed philosophers. The only major tertiary narrator who
is not a philosopher (i.e. Apollonius, the Indian sage Iarchas, or the
Egyptian sage Thespesion)26 is the Indian king Phraotes (2.29–33). In
the latter case, however, philosophy is precisely the matter at issue.
Apollonius asks how he was educated, given the presumption that there
are no philosophical teachers in India (2.29); Phraotes responds with an
autobiography that shows how some few, including himself, have been
selected for philosophical training (2.30). When this tertiary narrative
device at 7.21: the event recorded by Damis is ‘both like and unlike’ the story of
Aristides the Athenian.
25 This narrative originally derives (so Iarchas says) from the Egyptians: in a sense,
then, there are five levels here.

26 In addition to Apollonius’ parables, discussed above, see 4.16 (on his visit to
Achilles’ tomb). For Iarchas’ narratives, see 3.16 (narrates Apollonius’ life), 3.20 (how
the Ethiopians were expelled), 3.23–24 (Apollonius’ prometempsychotic life), 3.25 (story
of Tantalus), 3.30 (his grandfather, and the customs of the Hellanodicae). For Thespe-
sion, see 6.22 (Palamedes, Socrates, Aristides).
wanders from the subject of philosophy, Apollonius ‘interrupts’ and

steers it back to the subject of the Brahmans (2.33).27 This tertiary
narrative is a site of negotiation (with a slight, bristling, intercultural
tension) for philosophical authority. The close interrelation between
narration and authority in the VA is enriched by the thematic emphasis
upon the role of interrogatives. In a programmatic passage in book
1, Apollonius states that ‘I asked questions when I was a lad; it is
not now my business to ask questions, but to teach others what I
have discovered’ (1.17). Asking questions is the converse of narration:
it jeopardizes rather than reinforces philosophical status.
Issues of credibility
The events of the VA are frequently presented as occasions for thauma

(‘wonder’) and ekplēxis (‘awe’).28 At one level, this merely reflects Philo-
stratus’ Herodotean (→) legacy, but there is more at work here than just
intertextual homage. Throughout the narrative, wonder and awe mark
an abrogation of power in the face of a superior being: to manifest
amazement reveals the beholder as philosophically inferior. In a key
passage, the primary narrator states that ‘wisdom renders awe-struck
that which meets with it, but itself is awestruck by nothing’ (7.22). Thus
the various wondrous beasts on the journey to the East occasion thauma
only from Apollonius’ companions, and specifically not from him (1.38;
2.11). If the primary narratee wants to learn how to read this text, s/he
will have to learn not to be amazed at its contents.
But thauma suggests, as well as wide-eyed wonder, a sceptical disbelief
in the face of the incredible. To what extent does the primary narrative
probe its own plausibility? This issue is particularly explored during the
trips to the East, which is after all the traditional location of implausible
narratives.29 The East is constructed in the VA as a place beyond
the limits of Greco-Roman experience. As he travels East, Apollonius
crosses a succession of boundaries symbolically marking the journey
into the unknown.30 Book 3 sharpens the focus on the question of
27 The model for this interruption mid-narrative may be Hieroson at D. Chr. 36.24.
28 On the central role of wonder in biographies of holy men, see Cox 1983: 60–61.
On the theme of thaumata in the VA, see also Reitzenstein 1906: 39–54; Padilla n.d.;
Elsner 1997: 23–24, 28–29.
29 Romm 1992: 82–120.
30 Even when he contemplates the travel, Apollonius speaks of it as ‘travelling over
plausibility, beginning with a description of the wondrousness (thauma,

3.1) of the river Hyphasis: extraordinary trees, fish, worms and wild
asses, pepper trees and dragons (3.1–9). Two of these marvels come
from the pages of the notoriously untrustworthy Ctesias (Indika 27; 25):
informed narratees may be predisposed to distrust the story of the
oleaginous worm (3.1) and the unicorn with its magic horn (3.2).31 But
the narrator does not mention any literary authorities, relating instead
that ‘they’ (unspecified) ‘say’ that these phenomena exist (3.1, 3.2). We
have already discussed above the ambiguity of the primary narrator’s
use of ‘they say’ (‘they’ can refer either to tradition or to Apollonius and
Damis), and it is arguable that he exploits this ambiguity strategically
to mislead his narratees here. At first reading, we might take ‘they’
to be the reported narrators, Apollonius and Damis, relating their
experiences. But in 3.2, Damis is said to ask if Apollonius believes the
story told by the Indians about the unicorn’s horn (anyone who drinks
from it gains immortality), to which Apollonius replies with a sceptical
evasiveness (he will believe it if he finds that the king is immortal). By
subverting the plausibility of the magic properties of the unicorn’s horn,
and hence attacking the credulousness on the part of any narratees who
have assumed the ‘truth’ of these phenomena (which they might have
supposed to be guaranteed by the autopsy of the reported narrators),
the primary narrator conducts an object lesson in critical reading.
Yet the line between the plausible and the implausible is not so easily
drawn. Towards the end of the Indian sojourn (3.45–49), Iarchas dis-
cusses a series of wonders, arguing that while some are untrue (shadow-
feet and long-heads: 3.47), some exist (magnets, pygmies, gold-digging
griffins, the phoenix), and so (as he instructs his narratees, the tertiary
narratees, in the case of the magnet) ‘you must not disbelieve … but
you may wonder (thaumazein)’ (3.46). This sequence is introduced by a
comment on the part of the primary narrator that ‘there would be some
borders’ (1.18). The ‘borders’ of Babylon are marked by a frontier control (1.21). Border
crossing is also a central tool of textual organization: the beginning of a book repeatedly
marks the crossing of a boundary. At the end of the first book, Apollonius resolves to
leave Babylon; while at the beginning of the second book, Philostratus refers to the
Caucasus as the ‘beginning’ of the Taurus (2.1–2). At the end of the second book,
Apollonius reaches a column inscribed ‘Alexander got this far’, which Philostratus
supposes to have been erected either by Alexander to mark the ‘limit’ of his empire,
or by the Indians out of pride that he ‘got no further’ (2.43). The words ‘got no further’
close book 2, so that Alexander’s column also marks the end of a book.
31 Romm 1992: 117. For Ctesias’ unreliability, see Luc. VH 1.3.
profit in neither believing nor disbelieving all of this’ (3.45). Whereas

the passage that opens book 3 teaches primary narratees to be sceptical
of thaumata, this passage teaches them to use that scepticism judiciously.
The model for this type of reading, whereby scepticism must be tem-
porarily (but knowingly) suspended, is found in Herodotus (→),32 but
in the context of the VA, it serves a new role. The exploration in the
Indian episode of the credibility of thaumata serves as a programmatic
education in reading the miracle-working figure of Apollonius. If his
actions strain credibility, we must nonetheless tread a line between crit-
ical acuity and acquiescence, because a divine man necessarily does, by
definition, exceed the usual limits of human ability.
The VA is an opulent text, in narratological terms: clever, knowing,
self-conscious, manipulative. Yet this richness is not simply the work of
a sophistic amator ingenii sui. As we have seen throughout, Philostratus’
narrative sophistication serves a larger point, the exploration of the
divine character of Apollonius. It is precisely because the narrative sites
itself in the midst of the complex battle to interpret the phenomenon of
holy men that it deploys such an array of narrative techniques.
The Lives of the sophists
Set against the richness and depth of the VA, the VS comes across as
plainer fare: there are no instances of extravagantly layered narration,
there is nothing corresponding to the VA’s elegant treatment of sources,
and less ‘play’ with the truth-status of the text. Yet this relative lack
of complexity is itself a stylized literary pose: in general, the primary
narrator of the VS presents himself as a magisterial, didactic figure,
whose project is largely uncomplicated by doubts.33
The preface is addressed to the dedicatee, a ‘consul’ named Gor-
dian (VS 479), probably the future emperor Gordian I.34 This figure is
described as a man of culture (a descendant of Herodes Atticus, and a
‘leader of the Muses’, 480), and the narrator promises to ‘lighten the
weight of cares on your mind, like Helen’s cup with its Egyptian drugs’
32 Our narrator’s phrase may even echo Hdt. 4.96.1 (‘I do not disbelieve or overly
believe in this’), where the Herodotean narrator presents himself as a model for detach-
ed, sceptical interpretation.
33 I am grateful to Thomas Schmitz for allowing me to see unpublished work on the
narrator of the VS.

34 Flinterman 1995: 26–27.
(480). Yet the narratee of the main text, although (unlike the commu-
nity addressed by the VA) a single figure,35 is not this specific political
high-flier, but a more generally conceived student, whose reception of
the biographical narratives is firmly steered by the primary narrator.36
In the course of the text, narration is conceived of as ‘revelation’;37 the
primary narratee is ‘one who wants to know a lot’ about his subjects
(480). To this extent, the VS dramatizes its role as straightforward trans-
mission of knowledge from the knowing to the ignorant. The frame,
however, highlights the theatrical quality of this drama: the narratee of
the preface (not just ‘Gordian’, of course, but also general student) is
jolted into a different role in the main text. The effect of this is to cre-
ate a double consciousness: the narratee can simultaneously acquiesce
to his role as student, guided by the masterful direction of the primary
narrator, and look on, observing the master–student relationship from
the outside. The apparently ‘simple’ pedagogic stance of the narrator,
then, emerges as an artful narrative device.
The biographical narratives are ordered in simple fashion, life by life
in chronological sequence,38 with most simply introduced by the name
of the sophist. The pedagogical narrator marks his presence in the text
overtly,39 selecting paradigms on the basis of their instructional value,
whether moral or stylistic. Exemplary instances from the lives or texts
of the sophists are repeatedly introduced to substantiate general rules
or pointed lessons. The primary narrator draws his narratee’s attention
to noteworthy and memorable cases. ‘I wish to reveal how this came
about, for it is good and worth remembering’ (VS 536); ‘This is another
amazing thing about this man Lucius …’ (VS 557). ‘Philostratus’ does
not simply report what his sources tell him, but recasts his fabulae
35 At least, the one passage in the text where the narratees are directly addressed
employs the singular soi (515). See also below on injunctions to the narratees.
36 ‘Let us not consider …’ (487, 547); ‘one must consider …’ (480); ‘let us proceed
to …’ (510); ‘one must not marvel about this …’ (517); ‘this is what one must know
about …’ (545); ‘let us not fail to remember …’ (544); ‘let Varus be considered worthy
of narrative …’ (576).
37 The root dēl- (‘reveal’/‘revelation’) appears at 498, 515, 520, 523, 535, 536, 567,
574–575. This is a rather different, but no less authoritarian, usage to that of Aristides
in the Sacred Tales (→).
38 The text begins with an enclosed, parenthetic section dealing with philosophical
sophists in chronological order (484–492); it then loops back in time to deal chronologi-
cally with the sophists stricto sensu.
39 ‘I do not consider it right to call this exile …’ (488); ‘I see that the man …’ (503);
‘I shall discuss Scopelian …’ (514); ‘Let me not omit this …’ (524); ‘I shall not omit …’
(527); ‘my narrative summons me to …’ (566); ‘My narrative leads me to …’ (605).
into a story with pedagogical purpose. In selecting and editing in this

way, the narrator casts himself in the empowered role of gatekeeper
of sophistic culture. The narratee is thoroughly dependent upon his
editorial qualities.
Indeed, the narrator not only carefully polices the examples chosen,
but even seeks to control his narratee’s interpretations, guiding her or
him through the inferential processes that will allow them to deduce
the point of the example. ‘The quality of his arguments can be sur-
mised from the following. For (gar) …’ ‘The good fortune that attended
his embassies we may also surmise from the following. For (gar) …’ (VS
521). Gar (‘for’) is the most important particle in this text, investing the
narrator with the power of explication. ‘Let us regard this process as
chewing over his matter, not eating it, for (gar) …’ (VS 583). Yet the
interpretation that is taught is not always straightforward or unequivo-
cal. One of the narrator’s favourite devices is to confound the narratees
with an aporetic or contradictory utterance, before explaining (with a
gar clause) why a simple explanation will not suffice. ‘As for Antiphon of
Rhamnus, I do not know whether one should call him a good or a bad
man. For (gar) on the one hand (men) he may be called a good man, for
the following reasons … but on the other hand (de) there are evidently
good reasons for regarding him as a bad man, and they are the fol-
lowing …’ (498). Such cases of moral or stylistic complexity, expressed
through the prima facie uncertainty of the narrator, recur throughout
the VS;40 their effect is to underscore (through the subsequent solu-
tion of the puzzle) the narrator’s status as a pedagogical expert, one
who understands the problems involved in evaluating the richness and
diversity of human life.
If the narrator is styled as an expert, however, he also presents nar-
rative ‘truth’ as a complex quality that must be striven for amidst mul-
tifarious competing, and often conflicting, biographical variants. ‘Some
(hoi men) say that … but others (hoi de) that …’ and ‘There are those
(enioi) who think …’ are formulae that recur repeatedly. Sometimes they
are unresolved, and the narratee is left unguided,41 but commonly the
narrator intervenes with a flourished ‘but I consider …’.42 The nar-
rator’s opinion concerning individual sophists is constructed as particu-
40 486, 536, 592, 597, 604. Comparable is the device of posing a question, then
immediately answering it (523).

41 498, 521, 524, 570.
42 502, 530, 543, 554–555, 604.
larly insightful, in contradistinction to those of others. The best instance

of this is the case of Scopelian, as whose saviour from slander the nar-
rator represents himself. The account begins as follows:
I shall now speak of the sophist Scopelian, but first I shall deal with
those who try to calumniate him. For they say that he is unworthy of
the sophistic circle and call him dithyrambic, intemperate in his style
and thick-witted. Those who say this about him are sluggish quibblers,
not inspired with extempore eloquence. For man is by nature a creature
prone to envy. For instance, the short disparage the tall, the ugly the
good-looking, the slow and lame the light-footed swift runner, cowards
the brave, the unmusical the musical, those who are unathletic the ath-
letic. So one must not be surprised if certain persons who are themselves
tongue-tied, who have set on their tongues ‘the ox of silence’, who could
not of themselves conceive any great thought or sympathise with another
who conceived it, should sneer at and revile one whose eloquence was
the readiest, the boldest, and the most elevated of any Greek of his time.
But since they have failed to understand the man, I shall reveal what he
was and how illustrious was his family. (514)
Here, the narrator sets himself at odds with the received tradition on
Scopelian, here presented as a form of ‘malicious’ history,43 and does
so in a manner that states his own right to pronounce authoritatively.
Using a technique we have already observed, he passes from the specific
instance (the slanderers of Scopelian) to a general conclusion about the
‘nature’ of man (the explanatory section introduced by ‘for’). In this
case, he then returns to specifics, exemplifying this natural law with
a series of instances (the transition to exemplification now marked by
‘for instance’). This pedagogical self-representation seeks to legitimize
the narrator as a revisionist, the proponent of a true (‘I shall reveal’),
authoritative (‘one must not be surprised …’) account of Scopelian not
vitiated by jealous carping. As in the VA, the narrator casts himself as
the defender of true wisdom against its assailants.
The agonistic temper runs throughout the Lives of the sophists. Those
who ascribe the Araspes to Dionysius of Miletus, are ‘uncultured’ (524);
those who accused Herodes Atticus of hitting Antoninus Pius were
‘ignorant’ (554); those who called him the ‘stuffed orator’ were ‘petty
and trivial’ (565; cf. also 547; 602). Of particular interest is the court-
room episode in the account of Herodes, where the sophist is accused
of murdering his wife, Regilla (VS 555–556). The narrator tells us that
43 For ‘malicious’ history, see Plutarch On the malice of Herodotus, and further Gray
1990. Cf. 531, where the narrator’s interpretation is opposed to that of the masses.
this is a charge trumped up by her brother, Braduas, and proceeds to

present (in indirect speech) Herodes’ grounds for defence, one of which
is his extraordinary grief at the death of his wife. Next, he adds further
reasons of his own to believe that Herodes’ grief was genuine; imper-
ceptibly we have shifted from a covertly narrated account of events in
court to the primary narrator’s overt assessment. The agonistics of the
courtroom have subtly metamorphosed into the agonistics of biography,
with the narrator as lawyer for the defence.44
In general, however, relative to the VA, the VS downplays the attri-
bution of narratives to sources.45 In addition to the instances discussed,
there are references to the works of Aeschines (483), Herodes Atticus
(537, 538), and Polemo (539); also to the will of Antoninus Pius (534)
and to unspecified traditions (‘x is said to have …’, 494, 527). Outside
of these exiguous examples, the text avoids reported narration entirely.
Unless, that is, we count sophistic epideixis as narrative, for there are of
course numerous descriptions of sophistic performance and its recep-
tion.46 Even though the narrator usually assesses a sophist’s literary
ability by passing his own judgments on his extant written works, the
value of an individual figure is frequently articulated in terms of the
audience’s response to his work. The most common form of response
is ‘enchantment’ (thelgein and cognates) or ‘amazement’.47 Occasionally,
negative responses are recorded (laughter at the tubby figure of Leon,
485; the tense exchanges between Philagrus at the students of Herodes
at 579–580).
The general lack of reported narration, however, need not be viewed
as a deficiency: we can treat the emphasis upon primary narration,
instead, as a rhetorical device. The narrator of the VS is understated,
‘simple’, but pedagogically authoritative. To this extent, he bears com-
parison with the VA; and, as in that text, he is artfully revealed to be a
role assumed strategically to bolster the generic self-presentations and
the literary ambitions of the text.
44 For the narrator as legal apologist, cf. also 532 (Polemo ‘is charged with …’), 595–
596 (‘they accuse him of … let me put the case for the defence …’), 600 (‘let him be
acquitted of this …’).
45 On the question of Philostratus’ sources in compiling the VS, see most recently
Swain 1991.
46 See further Korenjak 2000 on the role of the audience in sophistic performance.
47 Thelgein: 491, 496, 520, 593; agamai: 589, 611. See also thaumazein at 529.
chapter thirty-three
AELIUS ARISTIDES
T.J.G. Whitmarsh
The Sacred Tales (orations 47–52) are an idiosyncratic and intriguing col-
lection of ‘autobiographical’ narrations.1 Although they envisage a pub-
lic performance (the primary narratees are a plural ‘you’), they present
an intensely personal narrative, focusing upon a communion with the
god Asclepius. They thus occupy a junctural position between rhetor-
ical exposition and private introspection. The primary, ‘public’ narra-
tion (‘Aristides’ to his audience) embeds a second, ‘private’ order of
narration (Asclepius to ‘Aristides’), presented through the opaque and
fragmentary medium of dreams. These secondary narratives operate
simultaneously at two levels, the apparent and the protreptic. At the
first level (marked by the recurrent use of the verb dokein, ‘seem’/‘think’)
stand the various oneiric phenomena that manifest themselves cryp-
tically to ‘Aristides’; at the second stands the god’s deciphered mean-
ing, what he is in reality commanding. On some occasions, ‘Aristides’
mediates between the two levels (or an interpreter has to be sought:
e.g. 50.16–17); on others (e.g. 50.1, where an apparition speaks to him
directly), the two converge.
Given this delicate equipoise between inner devotion and rhetorical
address, the use of the verb dēloun in the only explicit acknowledgment
of any primary narratees is significant: ‘But now I wish to reveal (dēlōsai)
to you the matter of my belly’ (48.2; cf. 51.67). The narration is (con-
structed as) a ‘revelation’, a bringing of private matters into the public
glare. This word also has theurgic, epiphanic connotations (it is also
used of the god’s revelations to ‘Aristides’ of what he should do, 48.75):
the primary narrator appropriates some of the power of the secondary
narrator. A further function of the term is to insist that the principal
function of the primary narratees is merely to witness: the narration
is (to be) an act of devotion to the god (cf. 48.1, where the narrator
1 Behr 1968; Pearcy 1988, with many interesting reflections upon narrative in the
Sacred Tales; Bompaire 1989; 1993. Translations are modified from Behr 1981.
442 part seven – chapter thirty-three
proclaims his desire to commemorate the god’s acts), rather than a

rhetorical performance designed to engage the primary narratees. This
narrative technique is sustained throughout the Sacred Tales: although
there are periodic markers of overt narration,2 primary narratees are
only very rarely (and obliquely) alluded to.3 This disengagement from
narratees is of course precisely a narrative effect, which serves to con-
struct them as detached observers of a miracle. The narrative repre-
sents itself in general as simply ‘showing’, or (as we might say, using the
language of evangelism) testifying to the god’s power, irrespective of and
unconcerned with primary narratees. But a full account of the Sacred
Tales must seek to get underneath this self-representation, to expose the
covert strategies whereby Aristides engages his narratees.
The autobiographical narrator’s ‘revelations’ are not, therefore, to be
taken at face value as straightforward confessionals.4 Indeed, so far from
being an innocent naif, ‘Aristides’ is extremely self-conscious about the
act of narration.5 In a crucial passage at the beginning of the first Sacred
Tale, he compares himself to Helen in book 4 of the Odyssey, who states
that she could not narrate ‘all the sufferings of enduring Odysseus’
(Hom. Od. 4.241 = Aristides 47.1); instead, she selects and ‘narrates’
(diēgētai) a single deed to emblematize the total. He proceeds as follows:
Likewise I myself would not tell all the achievements of the Savior, which
I have enjoyed to this day. Nor at this point shall I add that Homeric
phrase, ‘not even if I had ten tongues, ten mouths’ [Il. 2.489]. For that is
a small number. Not even if I should surpass all human strength, speech,
and wisdom could I ever do justice to them. I have never been persuaded
by any of my friends, whoever have asked or encouraged me to speak or
write about these things, and so I avoided the impossible. For it seemed
to me the same as if after swimming through the whole sea under water,
2 ‘I want to reveal …’, 47.4; ‘let us recall …’, 48.1; ‘it would perhaps now be
appropriate to discuss …’, 48.45; ‘It has been told how …’, 49.44; ‘I shall return to
the place where a little earlier …’, 50.71.
3 ‘Perhaps someone might ask to hear of …’, 48.60; ‘As for what happened next,
anyone who wants to believe, let him believe; but anyone who does not, be off with
him!’, 49.40 (discussed below). The second and fourth orations make use of the first-
person plural subjunctive (‘Let us record …’, 48.1; ‘let us turn …’, 48.71), but it
becomes clear that the ‘we’ in question refers solely to the narrator: ‘let us give the
logos mentioned at the start …’ (48.37); ‘let us say …’ (50.38). The imperative ‘Come
now’ (phere dē) that percolates the texts (see below) is arguably a command to a primary
narratee, perhaps best interpreted as an assumption of Socratic authority on the part of
the narrator (cf. e.g. Pl. Rep. 348c; Grg. 455a; Cra. 385b etc.).
4 For the general point, see Sturrock 1993: 9.
5 See on this particularly Pearcy 1988.
t.j.g. whitmarsh – aelius aristides 443
I should be compelled to produce records of the total number of waves

that I encountered, and how I found the sea at each of them, and what
it was that saved me. For each of our days, as well as our nights, has
a story (sungraphē), if there were someone willing to record (apographein)
the events or narrate (diēgeisthai) the providence of the god, wherein he
revealed some things openly in his presence and others by the sending of
dreams, as far as it was possible to obtain sleep (but this was rare, due to
the tempests of my body). In view of this, I decided to submit to the god,
truly as to a doctor, and to do in silence whatever he wishes. (47.1–4)
The narrator self-consciously tries out the voice of the epic poet who
appeals to the Muses for help in the face of the vast hordes of Achaeans
he must catalogue (Hom. Il. 2.489). His own narrative task, however, is
constructed as the greater: Homer might need more than ten tongues
and ten mouths, but from Aristides’ perspective ten ‘is a small number’.
The definitive expansiveness of the epic voice is usurped, incorporated
and roundly trounced. Aristides’ quasi-epicism is also inflected with the
‘aporia’ motif (‘not even if I should surpass all human strength, speech,
and wisdom could I ever do justice to them’). This topos is, of course,
already found in Homer (→), but by Aristides’ time it is closely asso-
ciated with rhetorical encomium.6 The quasi-epic temper announced
here permeates the Sacred Tales at two levels: not only are they spotted
with citations of and allusions to early epic,7 but also the almost com-
plete excision of any reference to narratees (discussed above) recaptures
the magnificent grandeur of Homeric narration. Again as in epic and
encomium, the statement of narrative aporia is ironic and subsequently
undercut. After this overwrought display of self-deprecation, the nar-
rator recants his stated decision to submit ‘in silence’ and proceeds to
a lengthy description of the god’s works. The statement of inability to
narrate exposes itself as a rhetorical topos, designed both to magnify the
subject and to indicate the scale of the orator’s task (and hence to laud
his narrative powers).
For all that this device is deeply established in the encomiastic and
hymnic tradition, it does have a specific role to play in this narrative
programme. Every moment of our life, he tells us in the passage cited
above, whether waking or dreamed, has a ‘story’; but it is immensely
difficult to turn this into a text. The problem lies with finding a nar-
6 Cf. Menander Rhetor 368.10–11 Russell-Wilson; X. Ages. 1.1; Isoc. Evag. 48; for its
status as cliché, see Lib. Or. 59.5.

7 Direct quotations at 48.39, 48.72, 51.44; allusions at 48.42, 48.72 (Hesiod), 50.34,
51.12, 51.27.
rator with the will to record the events or ‘narrate’ the providence of
the god. Aristides’ reference to the ‘story’ (sungraphē) of his dreams is
comparable to the narratological concept of a fabula.8 There is, there-
fore, a notionally infinite number of possible autobiographical ‘stories’
that could be written out of the fabula of experience; although in prac-
tice not all the stories could be written, since that would presuppose a
superhuman will on the part of the narrator. The problem that the nar-
rator confronts self-consciously concerns the principles of ordering and
selection necessitated by narrative composition.
This problem, however, is not limited to sorting through a vast num-
ber of events facing him: he must also represent that vastness, or tes-
tify to the magnitude of the god’s power. The distinction here between
the ‘recording’ (apographē) of events and the ‘narration’ (diēgēsis) of the
god’s works has been linked with the two stages in the composition
of the text, the first a rudimentary diary of his dreams (cf. 48.3–4;
49.30 for these apographai) and the second the worked-up narrative we
are presently experiencing.9 There is, however, an alternative (and, I
think, preferable) explanation. Rather than distinguishing between two
chronologically and materially distinct phases of the compositional pro-
cess, the narrator is differentiating between two representational regis-
ters within the same composition: firstly, a banausic effort, editing the
infinite fabula into a finite (hence selective) narrative; and secondly, per-
haps more importantly, producing an artful narration worthy of Ascle-
pius. It may be that these two registers do in practice correspond to
two stages of composition, but the contrast specifically effected here
is not between compositional forms but between objects of narration:
apographein represents ‘the events’, diēgēsis ‘the providence of the god’.
It is this second element that constitutes the abiding concern of
these texts. The ineffability of his subject matter is a recurrent theme
in the Sacred Tales. In the second, he asks himself ‘Where should I
start?’ (48.11), again a use of ‘aporia’ motif. The narration proper is also
circumscribed by doubts. ‘What happened next it is beyond the powers
of a mortal to narrate (diēgēsasthai); nevertheless, I must try’ (48.8). ‘Who
could display what happened as a consequence of this?’ (48.22). ‘You
could not tell in language (eipois legōn)’ what happened in the Achaean
straits (48.67; cf. 47.59; 50.80). Sometimes narratorial doubt stems from
a supposed abundance of divine works to record (50.70). Elsewhere,
8 Bal 1997: 5, 175–219.

9 Pearcy 1988: 381–383.
the narrator self-consciously addresses the need for selectivity, the need,
that is, to compress the fabula so as to convert it into narrative: ‘it is
beyond or like the address to Alcinous, but I shall try somehow to speak
briefly’ (48.60); ‘… to speak briefly and vaguely …’ (48.70; cf. e.g. 50.85;
89; 104). Like the aporia motif, this topos is rooted in both rhetorical
convention and pre-rhetorical tradition.10 Narrative compression is not
merely a question of pragmatics, of how to shoehorn an immense
number of events into a finite literary space. Or, rather, when narrators
claim to be compressing, they are making claims about the unspeakable
amplitude of the subject matter. In this context, it is the god’s works
that strain the very limits of mortal narrative. In the second tale, for
example, the narrator reports that the god explicitly ordered him to
compile only ‘summaries’ (kephalaia) of his works (48.4; cf. 49.5, 13).
At a later point, he writes that ‘it would be more chilling and vivid
(enargesteron) for me to narrate (diēgeisthai) unadorned the very visions
I beheld’, but ‘necessity’ (i.e. the god’s will) constrains him to report
only ‘in summary’ (kephalaia: 48.29). ‘Vividness’ (enargeia) is the quality
of rhetorical description that permits narratees to perceive events as
though with their own senses, the closest that language can come to
effacing its own status as second-order representation.11 At the same
time, however, the overt reference to the act of narration reinstates
precisely that second order of narration. The Sacred Tales offer (or
construct themselves as offering) glimpses of a terrifyingly powerful,
numinous world only dimly represented by ‘mere’ language; but they
are only glimpses.
The Sacred Tales are certainly constructed as ‘astounding’. The nar-
rator refers to events as thaumata (‘marvels’, 47.64; 48.55) or describes
them as thaumastos (‘marvellous’: 50.7, 63, 80; 51.38), following in the
footsteps of Herodotus (→). The ‘anonymous witness’ device is used to
introduce spectators of the miracles that the god performs (48.82; 51.41).
Occasionally, even stronger language is used: his narrative on one occa-
sion encompasses something ‘more chilling’ (phrikōdesteron, 49.48–49;
cf. 48.29). As has been emphasized, this numinous quality is repre-
sented through the violence done to (what is constructed as) ‘ordinary’
narrative. The repeatedly stated inability of mere language to represent
the power of the god becomes a means of representing the power of the
god.
10 See e.g. Dem. 60.6; Hyper. 4.2 Blass; Philostr. VA 4.34; 7.2.
11 See Zanker 1981 on this term.
The language of thauma, however, has further connotations in a lit-

erary context. Thaumata also strain the belief of the listener or observer,
and the term is often used to hint at the existence of an uncertain bar-
rier between truth and fiction (→ Philostratus). There are indices in
the Sacred Tales of a certain self-conscious play with these categories. In
particular, the association between ‘Aristides’ and Odysseus (qua inter-
nal narrators) invites an ironical reading on the part of sophisticated,
informed readers. Odysseus was not only the canonical traveller and
sufferer, but also the paradigmatic liar. At one point, the narrator states
that his own experiences are similar to, or even beyond, Odysseus’
‘address to Alcinous’ (Or. 48.60); and the phrase used here (Alkinou
apologon) had since Plato become proverbial for unsubstantiated self-
vaunting (Pl. Rep. 614b).12 Later, he refers to an episode in the same
tale as a ‘kind of Odyssey’ (48.65). He is alluding primarily to the ele-
ment of shipwreck in the narrative, but it is impossible to exclude a
more self-conscious, and self-subversive, reading. At another point in
the same tale, he receives a visitation from Athena, who tells him that
the Odyssey is not a collection of ‘myths’, but that she would help him
as she helped Odysseus and Telemachus (42). This assertion is finely
balanced: is the Odyssey, like the Sacred Tales, in fact true despite its prima
facie implausibility? Or does the equation of the two texts sow the seeds
of doubts in the mind of discerning narratees?
The Sacred Tales pose a self-conscious challenge to sceptical narratees,
reading against the grain. ‘As for what happened next, anyone who
wants to believe, let him believe; but anyone who does not, be off with
him!’ (49.40). This is a playfully coercive narrative contract between
narrator and narratees: the precondition for an appreciation of these
marvellous tales is the willingness to suspend disbelief; but the very
foregrounding of the requirement to credit the incredible also draws
knowing attention to the strenuous demands made of the narratees. It
is, in fact, impossible for the latter to attain the degree of familiarity,
intimacy, and credulity that Aristides requires, precisely because he
cleaves so closely to his role as narrative gatekeeper. His readers are
constantly aware that the dream-world he constructs as anterior to the
Sacred Tales is inaccessible in all its plenitude and richness, precisely
because he underlines the mediatory effort involved in his translation
of the dream-text into legible narrative.
12 Aristides himself uses the phrase with this meaning at Or. 36.88.
Aristides’ Sacred Tales, then, mark a new stage in this history of bio-
graphical narrative, not because of the personal voice tout court, but
because of the profoundly self-conscious dialogue between inner com-
munion and narrative expression. Despite the extremely limited en-
gagement with narratees, despite the repeated statements of aporia,
despite the apparent rejection of formal markers of narrative (notably
sequence), they do narrate; but what they narrate is constructed as non-
narratable, beyond the limits of narrative form. In this respect, Aris-
tides manifests a generic kinship not so much with other biographers,
Xenophon, Plutarch, and Philostratus, as with such late philosophers
as Seneca and Marcus Aurelius, whose difficult, crabbed writing is pre-
sented as part of the very process of authorial self-scrutiny and self-
correction.13
13 For the interiorizing turn of self-representation in the high empire, see esp. Fou-
cault [1984] 1990: 39–68; Perkins 1995.

part eight
BETWEEN PHILOSOPHY AND RHETORIC

chapter thirty-four
DIO CHRYSOSTOM
T.J.G. Whitmarsh
Most of the orations of Dio Chrysostom pursue what has been called
(broadly) a ‘moralizing’ agenda, in line with the author’s self-projection
as a hardy, practical philosopher. His use of narrative to serve this end
(finding an obvious precedent in Plato→) has been the focus of a cer-
tain amount of recent work; but, as commentators have stressed, there
is also a strong current of irony, and indeed indulgent pleasure, running
through his works, sometimes running contrary to the narrowly moral-
izing trajectory.1 Dio pays sustained, and self-conscious, attention to the
role of narration within his works, which emerges as a complex, devious
and even morally ambiguous phenomenon.
Dio’s orations are designed for public performance, and the primary
narratorial voice almost always reflects upon the speaker:2 either ‘Dio’
himself presents autobiographical experiences as an internal narrator,3
or we are dealing with a narratorial alter ego.4 Most importantly, the
narratives almost always at some level reflect analogically upon the
pedagogical relationship between ‘Dio’ and his primary narratees: it
is, as I shall call it, ‘metapedagogic’.
There are two principal literary contexts for Dionic metapedagogy.
Firstly, a significant number of his texts are dialogues. Dialogic utter-
ances are not necessarily ‘narrative’ in the conventional understand-
1 ‘… one typical trait … [is] a degree of reticence, and the sense of narrative
resourcefulness is held in check by an overriding moral responsibility’ (Anderson 2000:

143); see also Saïd 2000, who sees a less hierarchical relationship between the two
elements.
2 The exceptions are arguably Orations 28 and 29 on Melancomas, at least if they
are to be read side by side: the primary narrator of 28, who states that he has never
seen Melancomas (28.5), cannot be the narrator of 29, who claims to have been a close
friend of his (29.1).
3 Especially Orations 1, 7, 13, 36.
4 Principally in orations 6 and 8–10, the ‘Diogenes orations’; but also in orations
53–55, on Homer and Socrates respectively, and in 56–57, which focus upon the role of
Nestor.
452 part eight – chapter thirty-four
ing of the term, but it is helpful to consider them in this context,

partly because there are important overlaps with Dio’s practice in more
straightforwardly narrative contexts, and partly because (at the level of
literary form) dialogue is so often interwoven with narrative: although
some texts are ‘straight’ dialogue,5 some frame the dialogue with an
indication on the part of the primary narrator of where he heard it,6
after the fashion of Plato’s Republic (→).7 Still others use dialogue to
frame a central speech: the intriguing piece Charidemus (oration 30) uses
a dialogic frame to report a deathbed oration by its subject (a struc-
ture derived from Plato’s Phaedo);8 while oration 74 begins as a dialogue
(74.1), but the next 28 chapters adopt the form of an oration (with the
frame unclosed at the conclusion).
In several of these dialogues, the dramatic scene is metapedagogi-
cal: the interlocutor plays the role of acquiescent narratee (55, 56, 67,
70), his function being to represent conventional opinion (doxa), off-
setting the brilliant but deviant pedagogical narrator. In the sixtieth
oration, the narrator early on secures the narratee’s assent that they
must speak ‘contrary to the doxa of the many’ (60.2). The responses
of narratees may express shock at the narrator’s challenges to received
opinion (‘What! Do you consider it the mark of insanity in a man to
wish to be very highly prized and to amass great wealth?’ 77/78.9), or
‘amazement’ (thauma) at his outlandish utterances (55.9; cf. in a non-
dialogic context 80.1, where the narratees are said to express thauma at
the solitary wandering of ‘Dio’). The course of the dialogue can trace
the ‘conversion’ of the narratee, most notably in oration 56, where
the narratee’s conclusion contains the observation that ‘I am at last
beginning to understand the drift of your argument’.9 Of course, the
assent of a ‘metapedagogic’ narratee does not guarantee that of a ‘ped-
agogic’ reader: one function of such all-too-acquiescent interlocutors
may be, paradoxically, to inspire disagreement (and hence to develop
the reader’s critical autonomy). Elsewhere the narratee may be more
feisty (especially in oration 60), though none aggressively challenges the
5 Orations 14, 21, 23, 25, 26, 55, 56, 59, 60, 61, 67, 70, 77/78.
6 Orations 15, 28, 36.9–15, 58.
7 Cf. ‘After coming up from the harbour …’ (28.1) / ‘I went down yesterday to the
Piraeus …’ (Pl. Rep. 327a).

8 See Trapp 2000: 223–224; and esp. Moles 2000: 200–202 on the intertwined
narrative voices.
9 For ‘conversion dialogues’, or logoi protreptikoi, see Schäublin 1985.
t.j.g. whitmarsh – dio chrysostom 453
pedagogical authority of ‘Dio’, as Thrasymachus does to Socrates in

Plato’s Republic, or Callicles and Polus in the Gorgias.
The principal vehicles for Dionic metapedagogy, however, are the
exemplary stories, or ‘parables’ (as I shall call them), that percolate his
oeuvre.10 I shall begin by focusing upon the first four orations, which
contain some of the most extravagant and intriguing examples. These
so-called Kingship orations are assumed by most scholars to have been
addressed to the emperor Trajan,11 although it should be noted that the
narratees of the second and fourth orations are covert (whereas the first
and third are explicitly addressed to the emperor). Dio artfully varies
his parabolic practice in these texts: the first oration is framed by two
narratives, one an apparently traditional encounter between Timotheus
and Alexander (marked as such by phasi, ‘they say’: cf. ‘it is said that’,
2.1; ‘they say that’, 4.1), the other a quasi-autobiographical presentation
of a supposed event in Dio’s life; the second and fourth orations consist
of dialogues reported by a narrator, who asserts his presence in an
opening frame; the third opens with a narrative about Socrates and
the Persian, which is then revealed as a negative paradigm (3.2). All
these narratives constitute attempts to use analogy to steer the primary
narratee’s (‘Trajan’s) reception of the primary narrator’s (‘Dio’s) speech
by providing paradigms of kingship, evaluated in accordance with the
receptivity of the king in question towards philosophical learning.12
Moreover, each (apart from the negative paradigm of the third ora-
tion) dramatizes the response, be it good or bad, of a secondary or
tertiary ‘narratee’ to a ‘narrator’: the parabolic narratives construct an
interpretative frame for the response of primary narrator to narratee.
I use apostrophes for the secondary ‘narrator’ and ‘narratee’ because
the situations alluded to are not always straightforwardly narrative. Not
every presentation in the dialogues is narrative, on a strict definition.13
10 I concentrate upon the richest examples, although Dio uses this device with great
frequency: 1.1–3; 1.50–84; 2; 3.1–2; 4; 5; 7.1–80; 16.10; 17.13–18; 20.19–23; 21.4, 6; 43.4–
6; 57; 58; 60.9–10; 62; 66.6. See also Saïd 2000: 171–174; → Lucian’s parables. I shall not
discuss here the various orations to the cities (orations 31–35) or those on civic matters
(orations 38–51): although these are frequently in one sense narrative (defending one’s
conduct, for example, necessarily involves telling a story), they are more amenable to
the techniques of rhetorical analysis than narratology.
11 For bibliography and discussion of the problem of audiences, see Whitmarsh
2001a: 325–327.
12 Moles 1990.
13 → Introduction and → Plato for the argument that dialogue can be considered a
form of narrative.
In the parable that opens the first oration, can we count the flautist,
Timotheus, as a narrator, and his audience, Alexander, as a narratee?
In the narrated dialogues of the second and fourth orations, do the
moral points exchanged count as narrative?
In what follows, I have adopted generous definitions, because the
metapedagogic strategy does depend fundamentally upon the distri-
bution of narrator/narratee roles between the figures in the dialogue.
In the fourth oration, for example, Alexander (serving on this occa-
sion as a secondary narratee) responds passionately: ‘He flushed and
grew angry’ (4.18); ‘in fear’ (4.26); ‘he became upset and aggrieved’
(4.49). Clearly, this can be taken as a negative paradigm for an imperial
response to moral improvement (and indeed the description of Alexan-
der’s arrogant character that opens the oration has already prepared
the way for this). This chimes with what we might suppose to be the
primary frame for the delivery of this oration, i.e. Dio to the emperor.
In the second oration, however, the responses of Alexander are not
described, whereas those of Philip are: Philip ‘laughs at’ (2.13, 17) and
‘teases’ (2.19) Alexander; he also ‘betrays awe’ (2.7), ‘something close
to anger’ (2.16), and ‘delight’ (2.79) at him. The ambivalent responses
of Alexander’s ‘narratee’ suggest perhaps that the target audience of
the second Kingship oration is not Trajan but a Greek audience, who
are being encouraged to consider their responses to Roman imperial
power.14
In the complex parable that concludes the first oration, however, it is
certainly a narrative (in the strict sense) that we are dealing with. ‘Dio’,
functioning as internal narrator, promises a ‘sacred and salutary story
(logos) in the guise of a myth (muthos)’ (1.49): in a sequence modelled
on Socrates’ encounter with Diotima in Plato’s Symposium, he claims to
have met with an Elean or Arcadian prophetess (1.50–84), who pre-
sented to him a narrative. The larger part of this secondary narrative
consists of a version of the famous story of Heracles’ choice, familiar
from Xenophon’s Memorabilia (2.1.21–34): Hermes guided Heracles to
two mountains, one representing kingship and the other tyranny, and
asked him to choose between the two (1.69–77). Hermes’ advice makes
him a tertiary narrator and Heracles a tertiary narratee. In this narra-
tive, Hermes ‘figures’ the pedagogic role of the primary narrator, and
Heracles that of the primary narratee, who is thus steered towards true
14 Whitmarsh 2001a: 204.

kingship (and simultaneously to a favourable reception of the present

oration).15 In this particularly ingenious text, then, the metapedagogic
fiction analogizing the primary narrator–narratee relationship resides
at the third level of narration, the secondary narration serving as a
flamboyant, platonizing intermediary between the two.
Prior to this narrative, however, comes another, particularly sophisti-
cated and knowing, engagement with the primary narratee. The proph-
etess prognosticates a future occasion when her (secondary) narratee
‘Dio’ will report the tale to ‘the ruler of many lands and peoples …
even if there will be those who will ridicule you as a rambling (planēs)
wind-bag’ (1.56). This future-time narration maps onto the supposed
narrative scenario in the present (‘Dio’ addresses the emperor). Even in
this context, where there is no figuration, i.e. where the secondary nar-
rator/narratee maps exactly onto the primary narrator/narratee, there
is a protreptic directive: the negative response of deriding ‘Dio’ as a
windbag is stigmatized and (ideally) defused. (The passage is also a joke,
however: Dio the planēs, the ‘wandering philosopher’, often ironizes his
own ‘wandering’ narrative style.)16
This playfully involved imbrication of primary and embedded nar-
rative situations exemplifies nicely Dio’s manipulative narrative skill. It
is arguable, however, that matters are even more complex than this.
The fifty-seventh oration, Nestor, appears to be a preamble for a per-
formance of one of the Kingship orations before a Greek audience: at
any rate, it presents the narrative of Nestor’s speeches to Agamem-
non as a parable for the ‘words I spoke before the emperor’ (57.11).17
Whether all or any of the Kingship orations themselves were originally
performed before Trajan we cannot tell for sure; the Nestor, however,
means that we may have to refine our account of the narrative scenario
of at least one of them. If preambled by this text, the Kingship oration in
question—that is, the sum of words addressed to the emperor—would
itself become a secondary narrative, the imperial addressee (if we are
dealing with the first or the third orations, which explicitly address the
emperor) would become a secondary narratee, the parabolic narratives
tertiary narratives, and what we have called the tertiary narratives of
the first oration (assuming that is the oration in question) quartary nar-
ratives.
15 Moles 1990.
16 Whitmarsh 2001a: 160 n. 108.
17 Whitmarsh 2001a: 327.
That this narratological complexity in Dio’s parabolic narratives

is artfully thought out should be clear enough from the examples
above, but there are also passages where his practice is explicitly dis-
cussed. The sixtieth oration is a dialogue between (apparently) Dio
and an unnamed interlocutor, interpreting the narrative of Nessus and
Deianeira. The dialogue form was discussed above; for now I want to
concentrate upon the closing speech of the interlocutor, who is permit-
ted, almost exceptionally among Dio’s dialogues, a remarkably vigorous
challenge to Dio’s interpretation:
By heaven, it seems to me not a bad or unconvincing [interpretation],
either. And somehow or other I have the feeling that the method of some
philosophers in dealing with their arguments resembles in a way that
of the makers of figurines. For these craftsmen produce a mould, and
whatever clay they put into this they form into the shape of the mould;
and some of the philosophers before now have proved like that, with the
result that whatever myth or story they take in hand, by tearing it to
pieces and moulding it to suit their fancy they render it beneficial and
suited to philosophy. (60.9)
The interlocutor’s final speech, from which this passage is drawn, closes
the dialogue without reply: the reader is left with an unresolved, and
provocatively critical, assessment of Dio’s parabolic practice. This is
interesting in terms of the dynamics of the dialogue form: the combat-
ive rejoinder of the narratee (already pre-empted by his earlier expres-
sion of caution that ‘We may destroy the myth’, 60.30) ironizes the
speaker’s (‘Dio’s’) moral authority, and steers primary narratees towards
a more active engagement with his parabolic practice. It is notable also
that this interlocutor is granted a certain narratorial authority by his
magisterial use of simile himself: the comparison to a maker of figurines
constructs him as a pedagogically competent figure.
A more explicit commentary on the function of parabolic narratives
comes in the fifth, Libyan, oration. The larger part of this (5.5–27) is
taken up with the narration of a supposedly traditional myth (‘it is
said that …’, 5.5), but the opening frame (5.1–5) explains the role of
this narration. The primary narrator ‘Dio’ begins by observing that
a ‘myth’ (muthos) at first sight does not provide promising material, but
that ‘subjects that are guided in the proper direction and act as parables
for (paraballomena) true reality’ provide no small amount of usefulness
(khreia). The implicit connection here between muthos and pleasure, on
the one hand, and logos (‘rational account’) and utility, on the other,
is deeply embedded in Greco-Roman thought, going back via Stoic
theory at least to Thucydides (1.22.4). In the following passage (5.2–

3), he compares his practice to agricultural grafting: he is attaching a
‘useful, edifying logos’ onto ‘useless myths’ (5.3). Framed by this explicit
moralizing preamble, the myth of the sexy, siren-like serpent women
of Libya who tempt men to their death, a brood later wiped out by
Heracles, is easily decoded as an allegory of the conquest of sexual lust
by philosophical hardiness (an explanation that is in any case made
explicit at 5.23–24).
This programme is at first blush straightforward: the primary narra-
tees are being directed towards moral improvement. The exposition of
the myth, however, is more complex. The narrator’s promise to ‘grat-
ify’ (epikharizesthai) the younger members of his audience with a little
bit more mythologizing (5.24–27) ties up worryingly with the general
emphasis upon the dangers of male lust within the narrative proper: the
last story tells of two Greek ‘young men’ travelling in Libya, led onto a
gruesome death by the sight of what they take as a sexy local har-
lot.18 On this interpretation, the promise of exotic sexual pleasure to the
young men within the narrative encodes an allegory for the promise of
exotic narrative pleasure to the primary narratees, both false promises:
pleasure is substituted in the first case by painful death, in the second by
painful but salutary moral lesson. This interpretation cannot be a com-
plete one, however, because it insists that narrative pleasure must nec-
essarily be overmastered by moral vigour; whereas nothing can guaran-
tee that primary narratees will not take pleasure in this pornographic
snuff-movie of a parable. Unless the reader is truly Heraclean—and for
Stoics, no man is truly wise—the beast within will survive the heroic
moralizing.
The question of the pleasure of narratees is one of abiding inter-
est to Dio, especially given his self-consciousness about the potentially
uncongenial nature of his material. In several places, secondary narra-
tees are used as a device to stigmatize negative responses to his mor-
alizing. In the dialogue Agamemnon or on kingship (Oration 56), the Dionic
figure begins by asking whether hearing about as ancient a figure as
Agamemnon is likely to ‘annoy’ his interlocutor; the latter responds
that he would not be aggrieved even if he were to hear of still older
18 Cf. the two cases of an uncontrollable lust among male onlookers: 5.14, 26. The
serpent women are presented in notably erotic terms (‘bosom and breasts’: 5.12, 14, 25);
the myth has been justifiably compared (cf. now Anderson 2000: 155–156) to that of the
alluring but deadly Vine-women and Ass-legs in Lucian’s True stories (1.8; 2.46).
figures, ‘providing I am likely to be improved’ (56.1). This commend-

able narratee rightfully puts instruction first. Analogously, the contuma-
cious interlocutor of the sixtieth dialogue (On Chryseis), discussed above,
is advised by the Dionic figure: ‘Do not be grouchy (duskolos) when you
follow the argument’, advice to which his interlocutor explicitly accedes
(60.2).
One important passage in connection with narrative pleasure comes
in the eighth oration, Diogenes or on virtue. Orations 6 and 8–10 form a
group of narratives presented by an external narrator focusing upon
Diogenes the Cynic, each reporting a substantial speech uttered by
Diogenes to unnamed individuals or masses in the cities. As so often in
Dio, the relationship between secondary narrator and narratee reflects
metapedagogically upon that between their primary counterparts. The
external primary narrator’s voice is almost entirely covert, the occa-
sional parenthetic oimai (‘I imagine’) constituting an isolated and hardly
obtrusive exception (8.36; 9.1): in general, the persona of Diogenes is
allowed to shade imperceptibly into that of the narrator, his near-
homonym. The secondary narratees of the Diogenes orations (i.e. those
addressed by Diogenes) stand for doxa (common opinion), the secondary
narrator Diogenes (and implicitly the primary narrator ‘Dio’) for the
radical voice of philosophical enlightenment. In general, the responses
of secondary narratees are predictable enough, dramatizing the eccen-
tric but powerful role of the philosopher: ‘some admired him … to
others he seemed crazy’ (9.8). But the eighth oration concludes in an
interesting and unusual manner:
While Diogenes thus spoke, many stood around and listened to his words
with great pleasure. Then, possibly being put in mind of the act of
Heracles, he stopped speaking, squatted on the ground and performed
an indecency. Immediately, the masses scorned him, and the sophists
started up their din again, like frogs in a pond when they do not see the
water-snake. (8.36)
In this case, the secondary narratees respond with pleasure … and the
secondary ‘narrator’—though it is hardly a narrative act he performs
here—reacts by defaecating on the ground. Or, rather, by performing
an adoxon, translated ‘indecency’ but also suggesting an affront to doxa
or common opinion. The Augean stable that Diogenes attempts to
clear out is the mistaken assumption, fostered by the sophists, that
pleasure is the proper aim of language. As with the conclusion of the
Libyan oration, this represents a coded threat to the primary narratees,
warning them not to miss the water-snake concealed in his genial
narrative. At the same time, this closing narrative itself operates at

a pleasurable, or perhaps seriocomic, level: it is an amusing twist to
find defaecation used to establish a philosophical position after the
sermon has been misinterpreted by the secondary narratees. This is
a cunning joke on all who would wish to make sense of this oration:
what sort of meaning does this shit have? Is it just a joke? Is there any
metapedagogical value in it? What would be gained from treating the
observers of this shit as secondary narratees? An ingenious double bind,
then: pleasurable responses to narrative are stigmatized in a passage
itself saturated with ludic narrative pleasures.
Some of Dio’s parables are more complex. In the thirty-sixth (Olbian)
oration, ‘Dio’ functioning as an internal primary narrator recounts to
(according to the manuscripts) a Prusan audience about a journey of
his to Olbia; embedded therein is a narrative that he claims to have
presented as a secondary narrator to the Olbians. This consists largely
of a tertiary narrative, a (Stoicizing) cosmic myth supposedly told by
the Persian Magi (39–61), which completes the oration. The secondary
narratees (the Olbians) are characterized as culturally ambiguous (they
practise pederasty, 8; they like Homer, 9; they have beards and Home-
rically long hair, 17; but they wear trousers and Scythian garb, 7; and
‘no longer speak Greek clearly’, 9).19 The metapedagogical role of the
secondary narratees is not self-evident: do they preserve a true, vital
Hellenism, unlike that artificially ossified by Roman conquest (cf. 17 for
anti-Roman sentiment among the Olbians)? Or are they debased, bar-
barized Greeks, childlike in their lack of sophistication (cf. 10–11, where
the secondary narrator ‘Dio’ recommends to them the simple, gnomic
poet Phocylides)? The challenge, for the primary narratees, is to estab-
lish what is the relevance of the secondary narrative to them, and how
to respond to the cues of the secondary narratees (since this will impli-
cate their own cultural definition); that is, how the ‘argument’ function
of the narrative relates to the ‘key’ function. This sense of uncertainty is
characteristic of Dio’s pedagogical style,
The famous eleventh (Trojan) oration is also in a sense a metaped-
agogic parable: the primary narrator ‘Dio’ claims to instruct his pri-
mary narratees (11.1), and to remove the doxa (‘common opinion’) from
them (11.3). In line with his complaint elsewhere that only the intelligent
few understand philosophy (cf. 5.1; 11.145; 60.2), he predicts a frosty
19 See further Moles 1995; Trapp 1995.

response: ‘I know you will all think that this is false, except the wise’
(11.124). Yet for all that its narratorial voice is parasitic upon that of the
moral parables, this oration is a jeu d’esprit. It purports to address nar-
ratees in contemporary Troy,20 offering them the supposedly comfort-
ing suggestion that Troy was never captured, the Homeric texts being
implausible fictions. The Beglaubigungsapparat that underpins this claim is
a narrative, presented by ‘Dio’ functioning as internal narrator, about a
trip of his to Egypt (11.37–124): a ‘very old priest in Onuphis’ (11.37), he
reports, told him the true story about Troy (a device that looks know-
ingly to Hdt. Hist. 2.118–119). Homeric correction is, of course, rife in
the literature of the empire, and Dio’s contribution should be viewed in
the context of the journals of Dictys and Dares, as well as Philostratus’
Heroic tale.21 A further consideration is the possible congeniality of Dio’s
revisionist account to Roman readers, the supposed descendants of Tro-
jan Aeneas.22 But ultimately, the oration acts not as a simple vehicle for
Romanizing ideology, but as a ludic challenge to its primary narratees.
The text begins with the observation that ‘I am almost certain that
while all people are hard to teach, they are easy to deceive’ (11.1): a bril-
liantly playful ambiguity (will the following words teach or deceive?),
which is only partially resolved by the narrator’s subsequent insistence
that he has the true account. This oration exploits the metapedagogic
paradigm principally to pleasurable, ironical effect.
His most brilliant and celebrated moral parable comes in the sev-
enth, Euboean, oration.23 This is another autobiographical tale: ‘I shall
now relate events I saw myself, not things I heard from another’ (7.1).
The action is set on the island of Euboea, in ‘practically the middle
of Greece’ (7.1): this marked location indicates to the readers that the
narrative is to be paradigmatic of Hellenic values, while the surprising
choice of Euboea (rather than, say, Athens or Delphi) as the near-centre
figures the reversal of perspectives that ‘Dio’ will enforce upon his pri-
mary narratees. After a shipwreck, ‘Dio’ narrates, he was cast ashore
in the Euboean wilderness, where he was given hospitality by a hunts-
20 Seeck 1990; Anderson 2000: 152–153; Saïd 2000: 176–186.

21 Esp. Merkle 1994: 194.
22 That different versions of the myth will appeal to narratees of different cultures is
a theme of this oration: ‘You [Trojans] should be grateful and hear me gladly, for I have
been zealous in defence of your ancestors’ (5); ‘there was some advantage in [believing
the Homeric account] for the Greeks of those days [the time of the Persian Wars]’ (147).
For the ‘Roman’ theory, see Saïd 2000: 178–179, with references.
23 Highet 1973; Russell 1992; Swain 1994; Moles 1995; Trapp 1995.
man. As they walk to his home, he becomes a (secondary) narratee:

the huntsman tells him of his farming practices, and reports a visit to
the town when he was charged with tax evasion. Within this secondary
narrative, tertiary narration takes place: the huntsman and his prose-
cutor deliver speeches to the assembly (24–53); the huntsman is saved
from the assembly’s wrath by a certain Sotades, who narrates how he
once benefited from the huntsman’s hospitality (54–59). When the sec-
ondary narrative closes, ‘Dio’ reports that they reached the homestead,
where he was well if simply entertained; presently, it transpired that the
huntsman’s daughter was to marry her boyfriend, the son of another
cohabiting huntsman and the first huntsman’s sister. The metapeda-
gogical value of this narrative is then decoded, as it segues into a long,
moralizing section about the superiority of rural over urban dwelling
(7.81–152). The different parts of the oration are artfully interlinked.
In the main autobiographical narrative, ‘Dio’ links his observations to
aspects described in the secondary narrative;24 while the narrative in
places artfully foreshadows the moralizing section.25
As we have seen, Dio recurrently presents his narratives in terms of
pleasure and utility. This particular example is framed by references to
these principles. At the start, the narrator states that as an old man he
recalls such events in his life ‘not without pleasure’ (ouk ahēdōs, 7.1); but
at the conclusion of the narrative section, he states:
Now I have not narrated this entire story idly, or—as some might think—
because I am a chatterbox, but to present a paradigm (paradeigma) of the
manner of life I originally adopted, and of the lifestyle of the poor …
(7.81)
A parable that promised its primary narratees pleasure is revealed

simultaneously to be paradigmatic of a larger moral argument about
the superiority of rustic simplicity over urban decadence. Interestingly,
in the present context, both the pleasure and the moral instruction
(which are inextricably interlinked) are engendered by clashes of per-
spective between narrators and narratees. Dio’s primary narratees are
implicitly constructed as sophisticated urbanites, who are to be both
24 See 64, where he jokes that the huntsman concealed ‘the fairest of your pos-
sessions’ (i.e. his vegetable patch) when he was being prosecuted; 68, where he asks
whether this is the daughter who gave Sotades the cloak (cf. 58).
25 Cf. 9: ‘[I found that] poverty is in reality a sacred and inviolable thing’; 65: ‘I
could not help deeming these people fortunate and thinking that of all men that I knew,
they lived the happiest lives’. See also Swain 1994: 169.
amused by the extent of the rustics’ naïveté and educated by their sim-
ple virtue. The narrator shuttles between the two perspectives, mediat-
ing between (and simultaneously embodying) the knowing intellect of
the urbanites and the moral probity of the rustics.
On occasion, the pleasant humour is explicitly marked in the amused
reaction of narratees at moments of miscomprehension between rustics
and urbanites. In the huntsman’s secondary narrative, his speech in the
city (a tertiary narrative) is greeted with intermittent laughter (7.23, 24,
29, 30).26 These occasions are not straightforward cues to the primary
narratees, however: the aggressive and unsympathetic reaction of the
city folk (tertiary narratees)27 is offset against the more compassionate
response of ‘Dio’ (the secondary narratee). Yet the events are amusing,
and all the more so for that they are presented by a narrator (the hunts-
man) who remains apparently as ignorant as he was when the events
took place (an issue to which we shall return). This is particularly evi-
dent when he narrates how he saw, ‘square buildings on the walls’ and
‘ships peacefully moored as though in a lake’ (7.22):28 his naïve perspec-
tive upon two familiar phenomena of urban life, towers and harbours,
is enclosed within and framed by the knowing perspective of the pri-
mary narrator/secondary narratee (‘Dio’) and primary narratees.
In the course of the narrative, the role of the internal narrator
‘Dio’ shifts from that of translator of an unfamiliar world—that is, a
townsman interpreting the country for townsmen—to that of apologist
for rustic values (and thus the concluding part of the narrative serves as
a transition to the fiery moralism of the second part of the oration).
At the beginning of the narrative, the character ‘Dio’ is cast as a
powerless figure abandoned on the shore of an unfamiliar world (ever
since the Odyssey, the shipwreck on the beach has been a familiar
narrative topos). This powerlessness is cognitive as well as physical:
‘Dio’ the narrator recalls how he needed to make inferences from his
surroundings. The presence of a deer lying on the beach and dogs
barking on the cliff indicated to him that the deer had been forced
over the cliff; the clothing of the man he subsequently met told him
that he was a huntsman (7.4). As at the beginning of Heliodorus’
26 The laughter at 64 and 68, on the other hand, is benign and non-aggressive.
27 Characterized as a fickle ‘mob’ (okhlos, 7.23, 24, 29; plēthos, 30), whose reactions are
easily manipulated by disreputable orators.
28 The manuscripts transmit the glosses ‘towers’ and ‘in the harbour’ in these two
phrases, but editors rightly delete them as marginal notes that have been incorporated
into the text. See Russell 1992: 117.
Aethiopica, the primary narratees’ knowledge of the scenario is built

up inferentially: the internal narrator ‘Dio’ narrates according to his
experiencing focalization, suppressing any ex eventu knowledge, and the
narratees are dependent upon his interpretative powers.
By the conclusion of the narrative, ‘Dio’ is still apparently dependent
upon the huntsman for an interpretation of the world: he wrongly
suggests (with a misplaced laugh) that the rustics concealed their garden
from the city folk (7.64). He assumes that the rustics will need the help
of the rich in the village, when in fact the situation is more the converse
(7.69); and he needs to ask a number of details about the arrangements
for the wedding (7.70–72). There is, however, an important distinction
to be drawn here between ‘Dio’ the agent in the story and ‘Dio’
the narrator: while the former may have been rooted in the town,
the latter is (constructed as) an experienced traveller with a diverse
range of cultural and phenomenal experience. One of the processes
staged in this text is Dio’s own transition from the blinkered city-man
he was before his exile to the enlightened moralist and exponent of
simple values he is at the time of narration. ‘Dio’ the agent is a naïf,
needing instruction; ‘Dio’ the narrator, however, is a knowing figure, an
instructor himself, closer to the huntsman. It is possible even that the
huntsman (the secondary narrator) might be considered as an alter ego
of ‘Dio’ (the primary narrator), a technique that looks back to Diotima
in Plato’s Symposium, or indeed the prophetess in Dio’s first oration. It is
also worth noting that none of the rustics is named: this too compounds
the sense that they are fictitious paradigms of rusticity contrived to
serve Dio’s narrative purposes (and so the claim to autoptic experience
that opens the oration would have to be read as an ironical play
with a well-worn narrative formula, rather than a simple statement of
fact). The Euboean oration relies upon the elegantly constructed distance
between the perspective of the rustics and that of the townsfolk, but Dio
himself shuttles artfully between the two, both engendering narrative
irony and arrogating to himself moral authority.
Storytelling appears throughout Dio’s corpus. Usually, as we have
seen, it serves a metapedagogical function, analogizing the pedagogical
relationship between primary narrator (‘Dio’) and his narratees. This
metapedagogy, however, is not straightforward or self-evident; in fact,
its instructive value may even be said to lie in its inculcation of critical
skills that might allow primary narratees to negotiate the gulf between
their situation and that of their second-, third- or fourth-order counter-
parts. Moreover, the role of metapedagogical presentations is compli-
cated by the narrator’s stated awareness of the pleasurable seductions

of narrative. The ‘useful’ effects of narrative never fully prevail over
the ‘pleasurable’. Rather than consider this philodiegetic impulse as a
‘sophistic’ dilution of ‘philosophical’ ideals (the dominant view since
Philostratus, who includes Dio among the ‘philosophers in the guise of
sophists’: Lives of the sophists 479, 492), it is preferable to conclude that
Dio’s narratives engage in a self-conscious, sophisticated, and (for sure)
playful manner with the received theories of narration. His primary
narratees, that is to say, are not simply to be instructed; or, perhaps
better, they are not to be instructed simply. Narratees should think hard
about narrative, and the role it plays in pedagogic communication.
chapter thirty-five
LUCIAN
T.J.G. Whitmarsh
Lucian’s corpus is huge, heterogeneous, and complex: his narratives

are invariably overlaid by fiction and role-play. One unifying strand,
however, is the author’s consistent involvement with his narrators and
narratees.1 In the major narrative texts, as conventionally defined, the
primary narrator is always internal;2 in the dialogues, a figure often
appears at some point representing the persona of the primary narrator,
‘Lucian’ (there are some exceptions, which will be discussed below).3
Lucian’s narrators repeatedly foreground and problematize their own
identities. Narration, for Lucian, is a self-disclosing, but also a self-
concealing, act.
Let us begin by surveying the narrative forms we shall be considering
in this chapter. Lucian’s most celebrated narrative text is the True
stories, a quasi-Herodotean account of the narrator’s fantastic travels.4
The narrator is identified as ‘Lucian’ (elsewhere the author’s name
only rarely intrudes into his works)5 in an inscription towards the end
(2.28), but this identification is subverted by the acknowledgment in the
prologue that he made it all up: ‘I had no true story to narrate, since
1 For the purposes of this section, Lucian’s corpus is taken to exclude the works of
doubtful authorship transmitted under his name: On the dancers (as opposed to On the
dance), Philopatris, Charidemus, Nero, Timarion, Halcyon, Swift-footed, Lucius or the Ass, and the
epigrams.
2 If we are excluding Lucius or the Ass (on grounds of inauthenticity). The Demonax
(a quasi-biography of the philosopher) is at one level the presentation by an external

narrator of a series of deeds and sayings of the philospher, but at times the narrator
becomes internal, when telling about his long experience as a student under him (cf. 1).
3 The exceptions are the sophistic declamations, where the narrative situation is
fictitious: Phalaris I (ambassador of the tyrant to the Delphians), Phalaris II (Delphian to

his fellow-citizens), The tyrannicide, and Disowned (imaginary defendant to jurors).
4 See Rütten 1997; Georgiadou and Larmour 1998; Fusillo 1999 and von Möllen-
dorff 2000.
5 Except in titles and paratextual apparatus, the name ‘Lucian’ appears only here
and at Alex. 55; Peregr. 1; [ps.-Luc.] Epigr. 1 Macleod. See further Dubel 1994; Whit-
marsh 2001a: 253; Goldhill 2002: 60–82.
466 part eight – chapter thirty-five
nothing worth mentioning had ever happened to me; and consequently

I myself turned to lying; but I am more honest about it than others
are, for I will say one thing that is true, and that is that I am a liar’
(1.4). This is a celebrated appropriation of Socratic nihilism, the ironic
assertion that wisdom consists in knowing that one knows nothing.6 In
narratological terms, however, it also constructs a delicious paradox.
The story is presented by an internal narrator, but one who deliberately
rejects any claim to autopsy or personal experience (although autopsy
is claimed in the text: e.g. 1.22—a Herodotean touch→). The internal
narrator, then, discloses himself as (in a sense) external to the events
described. The entirety will be ‘meta-literary pastiche’,7 narration of
‘things that I have neither seen nor experience nor heard tell of from
anybody else: things, what is more, that do not in fact exist and could
not ever exist at all’ (1.4). The narrator claims to be motivated not, like
Herodotus, by the desire to record true events, but by ‘vanity’, being
‘eager to transmit something to posterity’ (1.4). The primary narratees
of the True stories are identified as ‘those who take literature seriously’
(1.1), an ironic vaunting of spoudē (‘earnestness’) in a self-professedly
comic text (but also an index of the sophistication of the pastiche).
The text also employs embedded narratives, with ‘Lucian’ and his
crew invariably as secondary narratees: Endymion’s account of the
battle between the sun and moon, presented in a mixture of direct and
indirect speech (1.11–12); the autobiographies of the old man Scintharus
in the whale (1.34) and Homer (2.20); the prophecy of Rhadamanthus
(2.27); and the autobiographical letter of Odysseus to Calypso, the
intended recipient, which is also discretely read by ‘Lucian’ (2.35–
36). As one would expect from a text that has already jettisoned any
pretence to realism, these narratives are also visibly angled towards
primary narratees for comic effect: most notably in Odysseus’ letter,
where the received tradition, sanctioned by ‘serious’ literature, of like-
minded mutuality between Odysseus and Penelope is subverted by
Odysseus’ expressed desire to run away to Calypso.
Next, there are three narrative texts on religion, which are also
presented by internal narrators: On the Syrian goddess, Alexander or the false
prophet and Peregrinus.8 The first is another quasi-Herodotean narrative
6 Pl. Apol. 21d; see Rütten 1997: 30–31; Georgiadou and Larmour 1998: 57–58.
7 Fusillo 1999: 351–356.
8 I exclude the ‘diatribes’, On sacrifices, Astrology, and On funerals, which are not in any
strong sense narrative.

t.j.g. whitmarsh – lucian 467
(the Herodoteanism this time signalled by the adoption of the Ionic

dialect), an account of the cult of Atargatis at Syrian Hierapolis.9 The
narrator, who does not self-identify as ‘Lucian’—but does style himself
‘an Assyrian’ (Syrian goddess 1), i.e. a Syrian, like Lucian—claims to have
derived his narrative partly from autopsy and partly (in the case of
past events) from priests. In addition, there are (in the Herodotean style
→) numerous non-specific reported narrators (‘they say’) substantiating
his narrative throughout the text. The other two texts are satires on
would-be religious leaders, and they share certain formal features. In
both, the narrator self-identifies as Lucian (Alex. 55; Peregr. 1), both are
addressed to a named narratee (Celsus in the Alexander, Cronius in the
Peregrinus), but in both (as in the True stories) the narrator’s stance is
playfully undercut. In the Peregrinus, additionally, secondary narration is
employed. The narrator reports to his primary narratee (Cronius) his
journey to the Olympic games, where he heard a certain Theagenes
(a secondary narrator) rant in support of Peregrinus. The internal
primary narrator here becomes a secondary narratee, and records his
contemptuous response (5–6). Then a lengthy narrative is presented by
an unidentified speaker (7–30). The secondary narratees include as well
as the primary narrator the undifferentiated ‘bystanders’, who cheer for
all the Cynics to be burned (31). Finally, a third secondary narrator—
Proteus himself—gives a speech, but the primary narrator’s ability to
report this is hindered by his inability to hear, i.e. by his ineffectiveness
as a secondary narratee (32–33).
A third category are the prolaliai (or ‘introductory speeches’, which
would probably have prefaced a longer performance): Dionysus, Hera-
cles, Electrum, Dipsads, Herodotus, Zeuxis, Harmonides, and The Scythian.10 In
each of these, we may assume a suppressed primary narrator, ‘Lucian’,
although there is no disclosure of the name. All of these texts include
narratives, and the motivation for the narration is made explicit: the
primary narratees—that is, the narratees of the primary narrator ‘Lu-
cian’—are directed to view certain points of comparison between the
situation in the narrative and the primary narrator’s own situation
(Dion. 8; Heracl. 7; Electr. 6; Dips. 9; Zeux. 2–3; 12; Harm. 3; Scyth. 9).
These narratives serving as ‘parables’ to reinforce the captatio benevolen-
9 See Elsner 2001; Lightfoot 2003: 184–208.

10 On these, see Branham 1985, recapped at Branham 1989: 38–46; also Nesselrath
1990.
tiae for the performances to follow.11 Although this device is characteris-

tically Lucianic (albeit parallelled in the works of Dio Chrysostom →),
the manner in which the parable is introduced, i.e. whether at the start
or the end, whether at length or briefly, whether constructed around an
exegetical description of an artwork or not, varies widely.12 Usually the
primary narrator is external to these parabolic narratives, but Heracles
and Electrum are ‘autobiographical’ and hence the narrator is internal.
The final and largest class of Lucianic narratives are the dialogues.
It is questionable at the outset whether a simple distinction between
narrators and narratees is helpful in the case of dialogues. It is arguable,
and this is the assumption upon which I shall work, that dialogues
are narratives, in that they constitute reports, the primary narrator of
which is ‘Lucian’ (who is however entirely reticent), of conversations
between interlocutors; these interlocutors become secondary narrators
and narratees when narrative (rather than argument) is at stake.13 Each
interlocutor potentially plays the role of both narrator and narratee,
even simultaneously (a narration by one character can represent a
response to that of another).
In the case of Lucian’s dialogues, however, all interlocutors are not in
every case equally valorized: the focal point of the text is often one indi-
vidual, privileged interlocutor. This figure is frequently given a name
that evokes, without replicating, the primary narrator’s (i.e. ‘Lucian’s’)
identity: ‘Lycinus’ (in Greek ‘Lukinos’, while ‘Lucian’ is ‘Loukianos’),
‘The Syrian’, ‘Momus’ (the personification of ‘blame’, i.e. literary invec-
tive), ‘Parrhesiades’ (‘son of free-speaker’), ‘Cyniscus’ (‘Cynical’), ‘Me-
nippus’ (see e.g. Twice accused 33 for Menippus as Lucian’s literary
model), ‘Tychiades’ (‘son of fortune’, a traditional way of describing
bastards and hybrids: see e.g. Soph. OT 1080). In these texts, which
I shall call the alter-ego dialogues, the role of narrator is most often—
though not exclusively—given to the alter-ego figure.
11 A similar ‘parabolic’ function of narratives can be identified in On the hall 1 and
You are a Prometheus in words 5, where however the externally narrated parable apologizes
on behalf of the work in which it is included (rather than an impending speech). The
dream also contains a parable (the relevance of which is decoded at 17–18), an internally
narrated symbolic dream (modelled on Prodicus’ Choice of Heracles at Xen. Mem. 2.1.21–
34), in which Sculpture (the usual translation, though ‘Artisanship’ might be better for
tekhnē) and Education vie for the youthful narrator’s attention.
12 Nesselrath 1990: 114–115.
13 → Introduction, → Plato, and → Dio Chrysostom.
Among the alter-ego dialogues we may distinguish further between

‘mimetic’ and ‘narrative’ dialogues. In the first category, a dramatic
situation is enacted (as in Twice accused, where ‘the Syrian’ is tried in
the course of the dialogue; or Hermotimus, where Lycinus refutes the
would-be philosopher), and in the second the alter-ego figure presents
a narrative, which is responded to by a narratee. The first category
includes the Solecist, Zeus refuted, the Fisherman, Twice accused, the Para-
site, Portraits, Defence of ‘Portraits’, Lexiphanes, Hesiod, Hermotimus, the Ship,
the Cynic. On the terms presented above, these dialogues are narra-
tives, presented by a suppressed primary narrator, reporting conversa-
tions that are represented as having taken place. Given that there are
no markers of primary narrative in the ‘mimetic’ dialogues (the pri-
mary narrator has no voice, and a priori the primary narratees can-
not be acknowledged), however, it is the second category (consisting
of Nigrinus, the Symposium, Icaromenippus, the Lover of lies, Menippus, and
the Eunuch) that is of principal interest in this context. These ‘nar-
rative’ dialogues are often fiendishly complex: behind the dialogue
between the alter ego (functioning as secondary narrator) and his nar-
ratee lies another narrative level, presented by the silent primary nar-
rator, ‘Lucian’. It is no coincidence that the Symposium belongs to this
category, for the dominant literary model is of course Plato’s Sympo-
sium (→).14 The complex, recessive structure of that text is arguably
matched in one Lucianic text, the Nigrinus: a framing letter from the
primary narrator (identified in the MSS as we have them as ‘Lucian’) to
a primary narratee (Nigrinus) presents an embedded dialogue between
two characters, one of whom tells about the effects of Nigrinus’ teach-
ing.15
The distinction between mimetic and narrative is not entirely clear-
cut, though. A number of ‘mimetic’ dialogues contain embedded nar-
ratives (Fisherman 29–37; Twice accused 26–29, 33, 34; at In defence of ‘Por-
traits’ 1–12, the interlocutor Polystratus narrates the speech of Pantheia,
i.e. his experience as a narratee). Conversely, in the ‘narrative’ dialogue
the Lover of lies, Tychiades narrates an event that is itself in effect a
dialogue (a debate on credulity between Tychiades and other interlocu-
tors), except that the secondary narrator introduces speeches (e.g. 36:
14 The Euthydemus and Theaetetus, which are also narrative dialogues, apparently
interest Lucian less.

15 Whitmarsh 2001a: 267; 269 on Plato Symposium and Lucian Nigrinus.
‘At this juncture, Arignotus the Pythagorean came in …’). Even this
feature, however, is in a sense ‘authentically’ dialogic: at any rate, it has
a precedent in Plato’s Republic.
Not all of Lucian’s dialogues incorporate an alter-ego figure. Those
that do not are uniformly mimetic: the Consonants at law, for exam-
ple, which represents squabbles between various phonemes that have
been ousted thanks to the drive towards linguistic Atticism.16 As we
have already seen, however, narrative elements can be found embed-
ded in mimetic dialogues; most notably, for this group, in the Dream or
Cock, where Micyllus narrates his dream (9–11) and the cock his prome-
tempsychotic life (24–25).17
Two particularly interesting cases of ‘mimetic’ dialogue that do not
feature alter-ego figures are the dialogues on cultural relativism, Toxaris
and Anacharsis. Here, the theme of the dialogue is played out in the
equal weighting given to the two participants, one Greek and one non-
Greek.18 Toxaris is of especial interest, because the dialogue revolves
around the presentation of stories motivated by the desire to proclaim
the superiority of friendship in one culture or the other. The roles
of narrator and narratee are thus exchanged in accordance with a
formal sequence. At the same time, the culturally determined position
of the external primary narratee (with whom the Greek reader will
identify) is brought into play, since only the Scythian Toxaris includes
ethnographic information upon his culture’s practices (‘It is not the
Scythian habit …’, 35; ‘I wish to tell you how we make our friends’,
37; ‘our custom is …’, 48).
Ultimately, however, too firm a distinction should not be drawn
between the dialogic and non-dialogic works of Lucian: there is a sub-
stantial degree of crossover between the roles of narrators and narra-
tees in both. We have already mentioned the extreme example of the
Lover of lies, where the (secondary) narrator presents a narrative that
constitutes a ‘virtual dialogue’. In two other ‘narrative’ dialogues, a
16 Cf. the Downward journey (although the figures Micyllus and Cyniscus, who appear
briefly, have alter-ego characteristics), Zeus rants, the Dream or Cock, Prometheus, Timon,
Charon, the Sale of lives, the Judgement of the goddesses, Anacharsis, the Runaways, Toxaris,
Saturnalia, the Dialogues of the dead, the Dialogues of the sea-gods, the Dialogues of the gods, the
Dialogues of the courtesans, Gout—if this last, a paratragic drama, is to be counted as a
dialogue.
17 Other narratives can be found at Zeus rants 15–18 (featuring an internal narrator),
Prometheus 7–19 (internal narrator), and Runaways 12–21 (external narrator).

18 Branham 1989: 104.
figure appears who might arguably represent the alter-ego persona of a

dialogue: Nigrinus, in the presentation of his speech in the eponymous
text (already, as we have noted above, doubly embedded, by a letter
and a dialogue) and the unidentified secondary narrator of the Peregri-
nus.19
How do the secondary narrator–narratee relationships steer the re-
ception of the text by primary narratees? To answer this question we
must look first to the ‘narrative’ dialogues, to see how the secondary
narratees respond to the alter-ego secondary narrator. The simplest use
of secondary narratees is to prompt for information (e.g. Menippus 8:
‘What was his object in that?’) or to draw attention to specific points
of importance (e.g. Symposium 38, where, cued by the narrator, the
narratee agrees to ‘remember’ a detail). A more complex function
is to express awe or pleasure, serving to construe the narrative as
an object of desire.20 The secondary narratee in Icaromenippus asks for
the ‘end’ (telos) of the story, complaining of his suspense (3; cf. 11;
16; Men. 2). The most extravagant example comes in the Nigrinus,
where a speech by the philosopher Nigrinus is presented in a two-
way dialogue, which is itself presented in a letter from ‘Lucian’ to
‘Nigrinus’. The unnamed secondary narrator within the dialogue states
his own response (as in effect a tertiary narratee) to Nigrinus’ speech:
spellbound, speechless giddiness (35). This response is then transmitted
back down the narratorial chain to the secondary narratee in the
dialogue like rabies (in the text’s arresting simile, 38; cf. Lover of lies
40).21 In these cases, the proposed value of the narrative is marked by
the pleasurable responses of the (tertiary and secondary) narratees. The
Symposium playfully inverts this topos. The (secondary) narrator Lycinus
piously refuses to tell the shameful story; but his interlocutor, instead of
begging him to do so (as at Menippus 2), accuses him of putting on an
act, claiming that ‘I know that you are much more keen to talk than I
to listen’ (4)!
Turning now to the ‘mimetic’ dialogues, here narratees can steer the
primary narratee’s responses, too. In Twice accused (32) and the Fisher-
man (38), the legal acquittal by the narratees of the Syrian / Parrhe-
siades vaunts the proper social function of literary satire. Hermotimus
19 For this point in relation to Nigrinus, see Whitmarsh 2001a: 270–271; in relation to
Peregrinus, Harmon 1936: 8–9 n. 2.

20 See also Jup. trag. 3–4; Downward journey 16–17.
21 See Whitmarsh 2001a: 277–278.
concludes with the tearful enlightenment of the narratee (and erstwhile

philosopher, 83), again marking the victory of satirical scepticism (this
time over dogmatism).
An important aspect of this steering is the role of laughter. When
narratees laugh, they join with ‘Lucian’ or his alter ego, united against
a common target: the text thus effects a reincorporation of commu-
nity through satirical mockery. This community-defining function of
satire is made explicit in the ‘mimetic’ dialogue the Fisherman, where
the philosophers begin by joining ranks against Parrhesiades (the alter-
ego figure), but conclude by joining with him in condemning the false
philosophers. In On sacrifices, the narrator opens by predicting that his
readers will ‘laugh at the stupidity’ of sacrifices (1), and concludes that
such practices ‘need a Heraclitus or a Democritus, one to laugh at peo-
ple’s ignorance, the other to lament their folly’ (15). In Peregrinus, like-
wise, the primary narrator (‘Lucian’) states that ‘I imagine that you’—
that is, his primary narratee, Cronius—‘are having a hearty laugh …’
(2; cf. 37, 45).22 The laughter of his narratee is matched by that of the
unnamed figure within the narrative ‘Lucian’ tells, who laughs both qua
secondary narratee (i.e. in response to the overblown secondary nar-
rative of the Cynic Theagenes, reported by ‘Lucian’: 7) and qua sec-
ondary narrator (at the conclusion of his own narrative: 31). ‘Lucian’
also reports his own laughter (qua secondary narratee) in response to
the overblown devotion of the Cynics (33). The laughing narratees also
figure in the dialogues, both as an interlocutor in a ‘narrative’ dialogue
(Eunuch 6) and as an embedded narratee in a ‘mimetic’ dialogue (Down-
ward journey 16–17). By interlinking all such figures through the motif
of laughter, the boundaries of a satirical community are reinforced.
Friendship, for Lucian, is dependent upon the identification of a com-
mon enemy.
One of the fundamental constituents of the satirical society is philia,
‘friendship’. The narrator frequently addresses his narratee(s) as ‘my
friend(s)’ (e.g. Alex. 1; Zeux. 1; Dips. 9; cf. Fisherman 38–39, the acquitted
Parrhesiades is now counted as a friend by the formerly hostile philoso-
phers): this form of address manifests the subtle bond of complicity
shared by all subscribers to satirical laughter. In other contexts, the nar-
ratee can be an enemy: An apology for a slip in greeting and Against an une-
22 Arguably, matters are even more complex: ‘Lucian’ and ‘Cronius’ are in fact
secondary narrator and narratee, the Peregrinus as a whole being the implicit report
(by ‘Lucian’ to his readers) of that narrative situation.
ducated book-buyer are addressed to a hostile figure, whether (in the case
of the first) one who has initiated the aggression or (with the second)
one who has merited the narrator’s mockery. The satirical dynamics,
however, are not fundamentally different in these cases: the entire work
is implicitly a narrative by (the primary narrator) ‘Lucian’ to primary
narratees (with whom the real readers may identify) of his (qua sec-
ondary narrator) clever put-downs to his explicit addressee (secondary
narratee). The primary narrator is inviting his unaddressed primary
narratees to join him in friendship against an inimical target (secondary
narratee). The existence of such primary narratees is in fact acknowl-
edged in An apology for a slip in greeting, which concludes with a fear
that ‘some’ may think that he deliberately made the slip in order to
have a pretext to write the apology (19). He proceeds to express to his
narratee—now ironically styled ‘my best friend’ (philtate) Asclepius—his
hope that ‘all’ may receive this work as the beginning of an epideixis
(‘showpiece’), not as a defence-speech. In this indirect appeal to other
narratees, it is revealed that the explicit narratee is not (of course) the
only destined recipient of the text.23
In other contexts, a difference in perspective can emerge between
narrator and narratee, despite their friendship. In the ‘parabolic’ narra-
tives, the narrator characteristically distances his own views of his work
from those attributed to his narratees. When he has presented the story
of his allegorical dream, the narrator of On the dream reports that one of
his narratees has interrupted his speech with the comment, ‘Heracles!
What a long and legal-sounding speech!’; another has compared it to a
winter dream, calling it ‘an idle tale’ full of ‘pointless yarns’ (17).24 The
principle function of these responses is to attempt to steer the recep-
tion of the text, by defusing criticism. Although these interjections are
subsequently corrected by the narrator (‘No, my friend’, 17), they are
nevertheless allowed to resonate, providing an ironic, playful alternative
reading of the narrative. In Zeuxis, likewise, a ‘wrong’ reading is simulta-
23 Also evident from e.g. Alex. 61; and the Apology, which is addressed to a different
narratee (Sabinus) than the text for which it apologizes (On salaried posts). Apology 15 is
of interest, though, because there the narrator states that Sabinus is the only narratee
whose opinion he values (‘As for the rest, even if they all condemn me unanimously,
I shall be content to quote “Hippoclides doesn’t care”’). This passage should not,
however, be read ‘straight’, but as a ‘performance’ of a close friendship between
narrator and narratee, and thus as a paradigm of satiric incorporation.
24 Cf. the narratee implied by the narrator’s rhetorical question at Hippias 1: ‘Why
have I said all this? It was not out of an ill-timed desire to air my knowledge of history
…’
neously acknowledged and defused. The narrator begins by telling how

after a performance to ‘you’, he was waylaid by ‘many’ of the audience,
who praised his speech for its innovative qualities: ‘Heracles, what nov-
elty! What marvellous paradoxes! How inventive he is! The freshness of
his thought is beyond compare!’ (1). This response, however, displeased
the narrator, who is (he states) proud of his fidelity to tradition as well as
his innovation (2), and the subsequent, parabolic narratives are explic-
itly pointed to exemplify this point. At the conclusion, we return from
the narrative to the frame, with the narratorial comment that he is
confident of their responses because ‘you are real artists and examine
each detail with craftsmen’s eyes’ (12). This is the same audience he was
earlier chiding for their limited grasp of his ambitions: the narrator is
simultaneously allowing for the possibility of ‘misreading’ and protrep-
tically directing his primary narratees towards the ‘proper’ interpreta-
tion.
These ‘wrong’ responses also serve a wider purpose in siting the nar-
rator as a marginal, deviant figure, as against the ‘normal’ expectations
of society. In a number of texts, narratees play the ‘straight’ role to off-
set the ostentatiously eccentric satirical narratorial persona. This can be
a reasonably simple prompt on the part of the narratee for explana-
tion (common in the narrative dialogues).25 In the dialogue Menippus,
the narratee’s question as to whether those who were powerful when
alive have more honour among the dead is answered by a curt ‘Non-
sense!’ (17): the narratee’s conventional assumption is displaced by the
privileged, Cynic view. In the Symposium, the device is used in a slightly
different way: the narratee interrupts, presuming that the address deliv-
ered by Hetoemocles was a praise of the bride or wedding-song, to
which Lycinus replies: ‘Of course, we ourselves expected something of
the sort, but it was far from that …’ (21). Here it is not that the narra-
tor has a truer perspective (because he previously shared the traditional
expectation), but that narrator (at the time of the narrative) and narra-
tee are alike educated in the vanity of philosophical pretension.
In such cases, the narrator is the winner, and the narratee con-
structed as foolishly naïve. But in other contexts, the narrator is him-
self ironized. This is most obviously the case in the True stories, where
(as we saw above) the advertisement at the start of narratorial lying
undercuts any credibility that the subsequent narrative might have had.
25 Cf. also the ‘virtual’ narratees of Hipp. 2; Dion.5; Alex. 21; You are a Prometheus 1.
It is also conspicuous in the case of the Teacher of rhetoric, where the

narratorial persona adopted is that of a rhetorician advising his youthful
charge to take short-cuts, and not to bother with the technicalities of
learning. Other cases are subtler, however. The narratee of Icaromenippus
expresses, as we noted above, pleasure in listening; but he also com-
plains at the implausibility of the narrator’s trip to the heavens (12).
Responses of ‘wonder’ (Men. 1; Ship 9) are necessarily ambiguous: a sign
that the narratee is impressed, for sure—but perhaps also incredulous?26
Subtler still is the Alexander, where the close relationship between
narrator (identified as ‘Lucian’) and narratee (identified as ‘Celsus’)
threatens to expose itself as a self-serving bond of complicity.27 The
two are presented as sharing fundamental values: they both are to feel
‘shame’ at the story of Alexander (2); they both are to look down upon
‘those accursed Paphlagonians’ (11), ‘thick and uneducated people’ (17).
Shared values are further hinted at by statements such as ‘You see
what sort of school the man that I am describing comes from!’ (5).
‘Lucian’ and his narratee would, of course, never display the gullibility
of Alexander’s provincial victims: ‘As a matter of fact, this trick, to a
man like you, and if it is not out of place to say so, like myself also, was
obvious and easy to see through, but to those drivelling idiots …’ (20).
The complicity of narrator and narratee, however, is arguably satirised.
Not only is Epicureanism (to which both are said to devote themselves:
61) elsewhere mocked by Lucian, but also the progress of Alexander’s
cult from the margins to the imperial court in the centre sends up the
anti-provincial snobbery appealed to by the narrator.
It is instructive to observe, in conclusion, that two of the texts in
which the narrator’s voice is most ingeniously ironized, namely the
True stories and Alexander, are the two texts in which the narrator self-
identifies as ‘Lucian’. Narrators are never to be taken at face value in
Lucian’s writing, especially when they lay claim to transparency. Lying
narrators feature heavily in his corpus, whether traditional poets (Men.
2; Jup. trag. 39–40; Sat. 5–6; Hesiod passim) or contemporary frauds (cf.
esp. Lover of lies passim), but the echo chamber of satirical voices prof-
fered in response is no more veridical. Lucian the self-confessed cul-
tural outsider, the self-constructed literary scandal, refuses the terms
(plausibility, realism, autopsy) of traditional narrative—while happily
mobilizing critiques against other traditional or contemporary nar-
26 → Herodotus and Philostratus.

27 My views on the Alexander have been influenced by Julie Lewis.
ratives on precisely these terms.28 At one level, this is an aspect of

Lucian’s satirical, Socratic, negative epistemology: true knowledge lies
in destroying the illusion of knowledge.29 At another level, however, the
debunking of narratorial authority is an act of cultural self-positioning,
a self-conscious projection of the author’s knowingly oblique perspec-
tive on Greek culture. Identity (whether cultural or narratorial) is for
this author always an issue, and also a problem, a resource, and a game.
28 Esp. True stories 1.3–4; also How to write history 29 on autopsy.

29 Rütten 1997: 30–31.
part nine
THE NOVEL
chapter thirty-six
CHARITON
J. Morgan
Callirhoe, probably the earliest fully extant Greek novel, is narrated by

an omniscient external primary narrator, who is the most obtrusive
of his kind in the extant genre. His frequent appearances in the text
enable us to characterize him, and his covert narratee, with some
precision. He addresses his narratee directly, in the second person, on
only two occasions: at 5.8.2 he says, of a particularly dramatic scene,
‘you would have thought you were in a theatre’,1 and at 8.6.11 he
implicates his narratee in the appreciation of the heroine’s beauty: ‘you
truly would have thought you were looking at Aphrodite herself as she
rose from the sea’. Nevertheless, the features noted in this chapter imply
a more or less continual interaction.
The narrator introduces himself in a short prologue:
My name is Chariton, of Aphrodisias,2 and I am clerk to the attorney
Athenagoras. I am going to tell the story (diēgēsomai) of a love affair that
took place in Syracuse. (1.1.1)
However, on several occasions he uses the present tense of institutions
of the Persian Empire, as in the following example:
Persia can mobilize its forces very easily. The system has been in force
since the time of Cyrus, the first king of Persia. It is established which
nations have to supply cavalry for a war, and how many; which are to
supply infantry, and how many; who is to supply archers; how many
chariots each people is to supply (both ordinary and scythed); where
elephants are to come from, and how many; and from whom money
is to come, in what currency, and how much. Everybody participates in
these preparations, and they take no more time than one man takes to
get ready. (6.8.6 –7)
1 Quotations are taken from the translation by B.P. Reardon, in Reardon 1989, with
a few minor alterations.

2 It has been suggested that ‘Chariton of Aphrodisias’ is a pseudonym (which would
cleanly separate the narrator ‘Chariton’ from the author ‘x’), but although it is almost
480 part nine – chapter thirty-six
This locates his voice in the period before the conquests of Alexan-
der, about four hundred years before the novel was actually written,
and about two hundred years before the city he claims as his home
was founded. The narrator is thus fictitiously configured as more or less
contemporary with the events he relates, in the manner of the Athe-
nian Xenophon (→).3 His persona as a contemporary historian also per-
haps accounts for his eschewal of cheap literary effects of surprise and
suspense: he tends to keep his narratee fully informed of what is hap-
pening. So, to take an example which will enable comparison with the
other novelists, when the heroine is presumed dead and buried in her
family’s tomb, the narrator has already told us that she is only uncon-
scious (1.4.12–15.1). Similarly, events leading to the reappearance of the
hero and the final recognition and reunion of the protagonists are also
fully conveyed to the narratee. The one major exception to this ‘histo-
riographical’ manner is the heroine’s pregnancy, which is not revealed
until she herself becomes aware of it.
The fullness of the information he provides extends beyond what
a normal ‘historian’ could have offered. The narrator has access to
events on the divine plane, and can tell us of the agency and motives of
Eros in getting the story started, and of Aphrodite in bringing it to its
conclusion:
Eros intended to make a match of his own devising … Eros likes to win
and enjoys succeeding against the odds. He looked for his opportunity
and found it as follows. (1.1.3)
Aphrodite thought this too harsh; she was growing less angry with him.
At first she had been incensed by his misplaced jealousy: she had given
him the fairest of gifts … and he had repaid her kindness with arrogance.
But now … Aphrodite took pity on him; having harassed by land and sea
the handsome couple she had originally brought together, she decided
now to reunite them. (8.1.2–3)4
too appropriate for a romantic novelist to be true, both the proper names in this
opening sentence are epigraphically attested at Aphrodisias.
3 Also 5.1.3, 5.2.2, 5.4.5, 5.9.1, and possibly 4.6.1 if the MS reading dokei is retained.
The last example talks of one of the minor characters in the story as still alive, fixing
the narrator close to the dramatic date in the fourth century BC. In addition he makes
some generalizing comments on the nature of barbarian despotism which are more
appropriate to the narrator’s than to the author’s date (5.2.6, 6.5.10).
4 Similarly 2.2.8, 2.4.5, 2.8.3, 3.3.8, 4.5.3, 6.8.1 (Fortune), 3.2.17 (the evil spirit),
3.3.10, 3.4.7 (Providence), 3.4.10 (some avenging spirit), 3.9.4, 4.7.5, 6.4.5, 6.7.1 (Eros),
8.3.6 (the god). In isolation some of these could read as metaphor, but others are so
j. morgan – chariton 481
He can also take us inside the minds and hearts of his characters and
tell us, as objective fact, what they thought or felt. For example, at 1.12
we are told that the pirate Theron acted out of rapacity not humanity;
that he did not judge it prudent to look openly for a buyer for Cal-
lirhoe; that he was afraid to approach someone eminent; then that he
could stand the delay no longer; there follow an interior monologue
and a dream. Quite often the narrator will draw attention to the truth
behind appearances:
The majority advised the opposite course, partly on the grounds that
Callirhoe’s father had done the royal household no little service, and also
because this was not a separate case that he was bringing to his court,
but virtually part of the case already before him. They did not want to
admit the real reason—that they could not tear themselves from the sight
of Callirhoe’s beauty. (5.8.7)5
An important by-product of this omniscience is a recurrent and explicit
emphasis on dramatic irony: the narrator and the narratee always know
more than the characters, as when Callirhoe builds a cenotaph for her
husband which
was like her own tomb in Syracuse in all respects—shape, size, costli-
ness—and like hers it was built for someone who was still alive!
(4.1.6)6
One major function of the highly visible narrator is to articulate the
structure of the story. Following his introductory appearance, he makes
major re-entries to sign a switch between narrative threads, when he
leaves the heroine at her second wedding in Miletus and backtracks
in time to pick up events in Syracuse following her abduction from
the tomb (3.2.17, ‘how he did so I shall tell shortly; first I want to
relate what happened in Syracuse during the same time’); at the exact
half-way point of the text, when he makes a lengthy summary of the
story so far, and moves forward into the second half of the novel (5.1.2,
‘This has all been set out in the story so far. Now I shall describe what
happened next’); importantly at the beginning of the last book, where
he makes another recapitulation and looks forward to the resolution of
the plot (8.1.1ff., discussed below); and at the very end to round the
precise and personal (even when the reference is to a personified abstraction) that they
can only be taken as a literal report of the activities of an anthropomorphic divinity.
5 Similarly 2.5.12, 5.9.7, 6.9.4 (expanded with a pseudo-interpretive deduction, ‘In
fact I rather think …).

6 Compare, for example, 3.4.18, 4.1.1, 4.2.1, 8.1.2.
whole thing off (8.8.16). Less prominent narratorial interventions mark

the operation of Chariton’s interlace technique.7
He is also active in guiding the narratee’s response. Sometimes this
is done quite overtly, as in the section at the beginning of the last book,
which gives us important pointers as to his, and his narratee’s, system
of values. After saying that ‘Fortune was minded to do something as
cruel as it was paradoxical’, he points up the irony inherent in events:
‘Chaereas was to have Callirhoe in his possession and fail to recognize
her; while taking others’ wives on board his ships he was to leave his
own behind.’ Finally he comments on the pleasure that will be derived
from the story’s happy ending:
And I think that this last chapter will prove very agreeable to its readers:
it cleanses away the grim events of the earlier ones. There will be no
more piracy or slavery or lawsuits or fighting or suicide or wars or
conquests; now there will be lawful love and sanctioned marriage. So I
shall tell how the goddess brought the truth to light and revealed the
unrecognized pair to each other. (8.1.4–5)8
The ultimate point of telling and reading this story then is for pleasure;
and its most pleasurable part is concerned with happy love, though
there is at least an implicit recognition that the happy ending owes its
effect to the preceding tensions.
This is not the only place where the narrator comments on the
nature of his own story. The paradoxicality highlighted in this passage
is similarly drawn to the narratee’s attention when Fortune throws Cal-
lirhoe’s pregnancy into the equation (2.8.3–4, ‘an unexpected, indeed
incredible state of affairs. How she did it is worth hearing’), when
Theron’s ship is discovered with Callirhoe’s grave goods (3.4.1), and
when the protagonists are reunited (8.1.9). The novelty of the situa-
tion is emphasized when Chaereas prosecutes himself in court (1.5.4),
or when Persian spoils arrive at Syracuse in peacetime (8.6.12). At cli-
maxes the narrator is not content to let the drama of the situation speak
for itself; rather he makes sure that the narratee appreciates the full
effect. A prime example of such self-advertisement, featuring the ‘apo-
ria’ motif, is when Chaereas makes his unexpected appearance at the
law court in Babylon:
Who could fitly describe that scene in court? What dramatist ever staged
such an astonishing story? You would have thought you were in a the-
7 As at 3.7.1, 7.4.11, 7.6.1, 8.6.1; see Hägg 1971: 139–154.

8 The idea of cleansing is a faint echo of Aristotle on tragedy; cf. Rijksbaron 1984.
atre full of emotions beyond number: all were there at once—weeping

and rejoicing, astonishment and pity, disbelief and prayers. How happy
all were for Chaereas! How glad for Mithridates! For Dionysius, how sor-
rowful! … I think the King himself, at that moment, would have liked to
be Chaereas. (5.8.2–3)
Another such example is when the protagonists, safely returned home,
are theatrically displayed to the populace of Syracuse:
Thunder never so stunned the ears nor lightning the eyes of those who
beheld them, nor did anyone who had found a treasure of gold ever cry
out as did that crowd then, when beyond all expectation they saw an
indescribable sight. (8.6.8)
A recurrent trope in such passages is a rhetorical question, addressed
to the narratee, usually making the point that it is impossible for any
writer to do justice to the material. This protocol is established very
early when a Syracusan assembly raises the question of the protago-
nists’ marriage: ‘Who could describe that assembly? It was dominated
by Eros’ (1.1.12).9
These rhetorical questions have no expressed addressee, but must
be read as part of the narrator’s communication with his narratee. We
turn now to other aspects of this communication. If the narrator occa-
sionally aims a question at his narratee, there are many places where
the questioning goes in the other direction, as it were. Although the
primary narratee is perforce without a voice, the narrator’s habit of
inserting explanatory parentheses can only be for the narratee’s benefit,
a pre-emptive response to an implicit request for clarification. So, for
example, Theron is surprised at the size and luxury of Dionysius’ coun-
try residence; whereupon the narrator inserts a syntactically indepen-
dent aside (‘it was in fact [gar] equipped to receive the king of Persia’,
1.13.1) before continuing with the narrative and the structure of the sen-
tence he has just interrupted, almost as if he had seen his narratee raise
a questioning eyebrow.10
The passage from the beginning of Book 8 has already demonstrated
the narrator and narratee’s shared taste for romantic and exciting sto-
ries. Similar shared values are also apparent in the many generaliz-
ing and normative sententiae with which the narrator litters his text.
The function of these is twofold. First, they naturalize the story by
grounding it in a description of the real world acceptable to both nar-
9 Similar rhetorical questions at 1.6.2, 4.1.11, 8.1.14, 8.4.1.

10 Other examples are at 1.14.1, 2.1.5, 3.1.2, 5.3.5, 5.8.1, 7.1.3.
rator and narratee; second, they invoke that shared perception of real-
ity as explanation or motivation, as indicated by an appropriate con-
nective participle. To take two examples: after Chaereas accuses his
wife of enjoying a riotous party in his absence, the narrator contin-
ues:
But lovers are easily reconciled; they gladly accept any justification from
each other; and so (oun) Chaereas changed his tone and began to talk
winningly to her, and his wife welcomed his change of heart. (1.3.7)
Or when Dionysius renounces the use of force against Callirhoe, the
narrator adds:
But for all that he did not give up hope of winning Callirhoe over, for
(gar) Love is naturally optimistic. (2.6.4)11
A similar effect is achieved when Eros is described as ‘a cruel tyrant’
(4.2.3); when the narrator comments on typical Greek curiosity (4.5.4),
contrasts Greek spirit with barbarian servility (6.4.10) or talks of the
‘innate superstition that barbarians feel towards the royal title’ (7.6.6).
In every case the effect depends on an appeal to a community of
experience that secures the reader’s consent to the fiction (turns a real-
life reader, in other words, into the narratee which the text constructs).
A similar effect is achieved at a rather more prosaic level (dependent
on general knowledge rather than shared moral values or literary taste)
when the narrator digresses at some length and in the present tense on
the national character and geographical location of Tyre (7.2.7–8).
The shared value system is operative also in the many evaluative
judgments that the narrator makes as his story proceeds. He is not
an objective reporter, but his judgments define the narratee to whom
they are acceptable as much as the narrator who passes them. So
the heroine is introduced as a ‘wonderful girl’ (1.1.1), and the villain
Theron as a ‘scoundrel’. As the trial at Babylon begins to decide
which of Callirhoe’s two husbands will get to keep her, the narrator
comments:
But the prize was not a wreath of wild olive or fruit or pine, but supreme
beauty, for which the gods themselves might fitly have contended.
(6.2.2)
11 Similar sententiae occur at 1.4.2, 3.3.16 (connected by oun); at 3.2.6, 3.4.13, 3.9.3,
6.4.3, 6.5.1, 6.5.5, 7.1.4, 8.6.5 (connected by gar); at 5.2.6, 5.8.4 (connected by hōste); at
8.5.14 (connected by houtō).
And the acquittal of Theron would have been ‘the worst outrage
possible’ (3.4.10). The constantly judgmental tone of the narrative is a
feature that distinguishes Chariton from the other Greek novelists.12
Another way in which the narrator appeals to shared experience is
through comparisons, whose ostensible purpose is to help the narratee
grasp something unfamiliar by assimilating it to something he knows.
Many of these are neutral enough: two of them denote a specifically
Hellenocentric view of the world: when the interest of the court case
is likened, in a rhetorical question, to the Olympic Games and the
Eleusinian Mysteries (5.4.4), and again when the arrival of the con-
testants in the courtroom is compared to ‘the competitors at Olympia
arriving in the stadium escorted by a procession’ (6.2.1). An interest-
ing subset of comparisons is specifically literary or artistic. At the very
beginning of the story, Callirhoe’s beauty is adjudged superior to that
of a Nereid or mountain-nymph, and Chaereas is compared to ‘Achilles
and Nireus and Hippolytus and Alcibiades as sculptors and painters
portray them’ (1.1.2–3). Polycharmus is introduced by being compared
to Homer’s Patroclus (1.5.2). Dionysius’ awareness of the inconstancy
of Love is glossed with a comment from the narrator about the depic-
tion of Love by poets and sculptors (4.7.6). The set-up in the Persian
courtroom is compared to a Homeric divine council, with a quotation
from Iliad 4.1 (5.4.6). Callirhoe’s appearance in the courtroom is com-
pared to Helen’s appearance among the Trojan elders in Iliad 3 (5.5.8).
This assumption of a shared level of literary cultivation also underlies
one of the most eccentric features of Chariton’s narration: on many
occasions he quotes Homer, not, as it were, in illustrative parentheses,
but to carry the main narrative forward; for a brief moment the narra-
tor’s microphone passes to the poet.13 Every one of these quotations is
in effect a familiarizing comparison for the benefit of an ostentatiously
bookish narratee.
12 Further examples at 2.5.2, 3.4.12, 3.6.5, 5.3.9, 5.5.1, 5.9.8, 6.5.8, 7.2.5.
13 1.1.14, 1.4.6, 2.9.6, 3.4.4, 3.6.3, 4.7.5, 5.2.4, 6.1.8, 6.2.4, 7.4.3, 7.4.6, 8.1.17, 8.5.2;
at 4.7.7 a similar effect occurs with a quotation from Menander’s Misoumenus. On the
quotations from Homer, see Müller 1976.
Embedded narratives
There are numerous embedded narratives in Callirhoe, few of them

of any great length or technical interest. Without exception they are
internal analepses, and only one of them introduces information that is
not already known to the external primary narratee, when a messenger
brings news of the death of the Egyptian king and the approach of the
Persians (8.2.3). Their main function is to convey information between
the characters, and to make it clear who knows what; this is necessary
because the narrator makes sure that the primary narratee knows
everything. Occasionally the primary narrator will draw attention to a
detail deliberately omitted from the embedded narrative, again in order
to clarify the actors’ areas of ignorance, as when Callirhoe tells her
story to Dionysius but omits any reference to Chaereas (2.5.10) or when
Theron confesses but conceals the name of the person to whom he
has sold Callirhoe. A secondary, narratorial, function is to recapitulate
events for the benefit of the primary narratee, in a way analogous to
the narrator’s own retrospectives at the beginning of Book 5 and Book
8.14 Most of the embedded narratives are in the form of direct speech,
but occasionally indirect speech is used or the narrator indicates that
a narrative took place without going into details, presumably in order
to avoid tedium for the primary narratee.15 Occasionally narrative is
included in a speech whose actorial motivation is something other than
the transmission of information, such as a lamentation or a persuasive
or forensic oration.16 And sometimes the primary narratee knows that
the embedded narrative is false or partial. Thus the villain Theron is
characterized by being allotted two lying narratives, clearly signed as
such, at 3.3.17 (echoing Odysseus’ Cretan tales), and 3.4.7; at 3.10.2
some country people tell Callirhoe of the destruction of Chaereas’ ship
in a way that leads her to believe him dead, although the narrator has
already told how he escaped; and at 6.1.2–3 two factions in Babylon
each present a slanted narrative of events, according to whether their
sympathies lie with Chaereas or Dionysius.
14 On this aspect see Hägg 1971: 253–263.

15 Indirect speech at 3.9.9–11 (which starts in direct speech), 8.1.14; abbreviated
reports at 7.2.4, 8.1.17, 8.5.9.
16 1.14.7–8, 5.1.4 narrative within a lamentation; 3.2.8–9, 5.6.5–8, 6.7.8 narrative
within forensic or deliberative speech; 4.4.7 narrative within a letter.

In most cases the response of the secondary narratee is indicated.

This reflects the character and position of the actor within the fiction:
so at 2.1.5 Dionysius is excited to hear of Callirhoe’s beauty, but disap-
pointed that she is a slave, and at 2.5.11 he reacts to Callirhoe’s nar-
rative with tears of thwarted desire. However, these narrated responses
are also signs to the primary narratee, at least in the sense that the
story invites from him an unashamedly emotional response (even if the
emotions do not coincide with those of the secondary narratee).
The final embedded narrative of the novel requires separate com-
mentary. The story has reached its resolution, but before the text can
attain closure the Syracusan populace has to hear the whole story. At
this stage their reaction has no function within the story. However,
the narrator presents his narratee with the longest and most elabo-
rate embedded narrative of the entire work (8.7.9–11). At first Chaereas
wants to start at the end to avoid the traumatic episodes of the begin-
ning, but the audience insists on hearing the whole thing in order. As he
is still reluctant, his father-in-law Hermocrates narrates the first part of
the story (an interesting case of second-person narration), and Chaereas
takes over at the point where he sailed to Miletus. He is in constant
interaction with his audience, telling them at one point not to be afraid,
because Callirhoe did not really become a slave. At the point where his
ship is destroyed, the audience breaks out into lamentations, and then
again prevents him omitting a painful part of the story. At the very end
there are ‘cries of well-wishing from everyone’. This complex embed-
ded narrative is in effect a resumé of the novel in its entirety, and in
a sense the fictitious provenance of the story that the primary narrator
tells. The response of the in-frame audience, therefore, can be read as
endorsement of the primary narratee’s response to the novel, and of the
narrator’s strategy of giving full information in its proper order.
chapter thirty-seven
XENOPHON OF EPHESUS
J. Morgan
The Ephesian Story is narrated by an anonymous external narrator who

is far less perceptible than Chariton’s.1 On no occasion does the narra-
tor speak of himself in the first person or to his narratee in the second,
and there is virtually no overt self-referential commentary on the qual-
ity of the narration, with just two scenes being described, in the same
phrase, as ‘a pitiful spectacle’ (1.14.2; 2.6.3). Since he refers at the very
end of the novel to the protagonists’ lives after the end of the story
(5.15.3), the narrator is not configured as a contemporary of the events
he narrates; but there is no indication by how much the story antedates
the act of narration, nor in what circumstances the latter should be
imagined as occurring.
Like Chariton’s, this narrator knows the story in its entirety, and
communicates it fully to his narratee. The heroine Anthia, like Cal-
lirhoe, falls into a death-like sleep and is entombed alive: but the nar-
rator has already explained that the poison with which she believed
herself to be committing suicide was in fact a sleeping-potion given
her under false pretences by a kindly physician (3.5.11). He also dis-
poses of full information about events on the divine plane, telling us
that the jealousy of Eros begins the plot (1.2.1), that Eros is enraged
by Habrocomes’ resistance (1.4.4), that Fate had not forgotten (1.10.2),
and that the god of the Nile took pity on Habrocomes’ prayer (4.2.6).
1 We must immediately enter the caveat that the exact status of the extant text is
disputed. It has been argued that we have only an epitome of a once more extensive
text (Bürger 1892). Although this thesis is untenable in the form in which it was orig-
inally argued (see the assaults on it by Hägg 1966 and O’Sullivan 1995), almost every
modern reader has the sense that the narrative is cripplingly bare and undeveloped.
O’Sullivan’s hypothesis of residual oral technique suggests that our text may be just one
realization of a fluid texte vivant, or a skeletal summary on which oral performance could
be improvised. Either way, a text whose function is to record the story in the simplest
way possible is precisely the kind of text that is likely to minimize the visibility of a
potentially more interesting narrator.
490 part nine – chapter thirty-seven
Oddly the gods’ role in the dénouement is merely implicit. The narra-
tor has access to the thoughts and emotions of his characters,2 though
he does not always explain why they do what they do. In contrast to
Chariton’s, however, this narrator is often content to let the charac-
ters speak for themselves: roughly two-thirds of the text can be clas-
sified as showing rather than telling.3 Similarly the narrator can offer
opinions about events and characters: for instance, Manto is beauti-
ful, but ‘not nearly as beautiful as Anthia’ (2.3.1), Cyno is ‘hideous to
look at and much worse to listen to’ (3.12.3), and Anchialus ‘pays the
price for his wicked passion’ (4.5.6).4 But, to a surprising extent, the
narrator as often maintains a laconic objectivity, offloading the judg-
ments on to his characters. To take a single example, it is the charac-
ter Habrocomes, not the narrator, who describes Cyno as a murderess
(3.12.5).
Communication between the primary narrator and his primary nar-
ratee coheres with this pattern. There is, for instance, a short digres-
sion about the temple of Apis at Memphis, which serves to locate the
story in relation to the narratee’s knowledge of the real world. The
narrator offers a couple of sententiae stressing the difference between
Greeks and barbarians (2.2.4; 3.11.4), and a number of explanatory
parentheses, particularly when a new character is introduced (2.9.1,
2.14.1, 3.5.9, 5.2.2, 5.4.5, 5.5.2, 5.9.7, 5.12.1). These occur particularly at
points of transition between the narrative lines and are a by-product of
Xenophon’s extravagantly primitive interlace technique: the pretence is
maintained that a strand is resumed not at the point where it was left,
but at a time exactly coinciding with that reached by the strand to be
dropped. There are thus many fictitious gaps in each strand, which the
explanatory asides appear to fill. One might say, in fact, that the most
visible function of this barely visible narrator is precisely to control the
rapid transitions between the novel’s two storylines.
In one respect, however, Xenophon marks a clear difference and
arguably an advance over Chariton. This is in the matter of embed-
ded narratives. Sometimes these, like Chariton’s, concern events within
2 E.g. at 1.9.6, 1.14.7, 1.15.4, 2.4.1.

3 The figure is from Scarcella [1979] 1993: 175–177.
4 Quotations come from the translation by G. Anderson in Reardon 1989. A
complete list of the narrator’s observations can be found in Scarcella [1979] 1993: 172–
174.
j. morgan – xenophon of ephesus 491
the novel and clarify who knows what and when. So, at 3.3.4 the bandit
Hippothous (a structurally important figure who acts as girder between
the two story-lines) tells his new friend Habrocomes about his earlier
encounter with Anthia. There is no new information for the primary
narratee here, but it is a vital part of the story that Habrocomes should
now learn what the primary narratee has known for some time; the
news motivates his journey to Cilicia. Similarly at 3.9.4 an old woman
called Chrysion (whose only reason for existence is to transmit this
information) tells Hippothous’ men the tragic story of the death of a
woman whom Hippothous recognizes as Anthia, and the theft of her
body from its tomb by pirates. The narrator has already made it clear
that Anthia is still alive, but the partial information provides motivation
for Habrocomes’ voyage to Alexandria. It is also worth noting that this
embedded character-narrative avoids the kind of omniscience that the
primary narrator takes for granted. Thus Chrysion shows uncertainty
over the motive for Anthia’s suicide by giving alternative explanations
(‘either because she was mad or because she was in love with someone
else’). Chrysion’s narrative precipitates an extreme emotional reaction
and lament from Habrocomes, which the primary narrator and nar-
ratee are able to read, from their positions of superior knowledge, as
dramatic irony.
Three embedded narratives cover events outside the story and are of
no organic relevance to it. One of these is Anthia’s tale of a childhood
encounter with a ghost, which she uses to account for the epilepsy she
has feigned to frighten off the clients of the brothel into which she has
been sold (5.7.7–9); this is clearly marked as a fiction, but nonetheless
stands in analogic relation to the main narrative, the horrific assault
of the ghost corresponding to the sexual assaults intended by the cus-
tomers of the brothel. The other two embedded narratives, both exter-
nal analepses, function more clearly as didactic analogies. The first
is Hippothous’ account of his love for and loss of the beautiful boy
Hyperanthes (3.2.1–15). This is set up as being equivalent to the story of
the hero Habrocomes, for which it is exchanged (3.1.5). Like the main
story it concerns passionate love disrupted by the aggression of a rival,
resulting in separation and travel; as in the main story, the first meet-
ing occurs at a festival, and the lovers are more or less equal in age.
However, the homosexuality of the embedded narrative forms a clear
contrast to the love of Habrocomes and Anthia; and Hippothous’ story
embodies a quite different paradigm of sexual relations, substituting for
the equality of the main story a division of roles into active and submis-
492 part nine – chapter thirty-seven
sive partners on the classic pederastic model.5 Its argument function is

very limited; it elicits Habrocomes’ narrative of his adventures (given in
extreme summary at 3.3.1), which in turn prompts Hippothous to recall
his encounter with Anthia; Habrocomes shows no signs of understand-
ing its importance as a counter-paradigm to his own experiences, but
for the primary narratee it shapes the understanding of the main story
in as profound a way as this novel ever achieves. Again we can note
that the secondary narrator avoids the omniscience of the primary one:
at 3.2.4 Hippothous talks of the envy of ‘some evil spirit’, rather than
attributing events to the agency of a specific deity in the manner of the
primary narrator.6
The second external analepsis is the narration by the old fisherman
Aegialeus of his love for and elopement with the young girl who became
his wife, and whose embalmed corpse he keeps in his bedroom (5.1.4–
11). Aegialeus has no function in the novel other than to tell this story.
Again the motifs of the embedded narrative reflect those of the main
story: inamorations, rivalry, separation, and reunion. By casting the
theme of separation and reunion in its most extreme form (the separa-
tion of death redeemed by the continued presence of the mummy), this
narrative succeeds in producing a profound learning experience in the
protagonist, who realizes ‘that true love knows no age limits’ (5.1.12).7
Although they differ widely in the visibility and the characteristics
of their primary narrator and in the uses to which they put embedded
narratives, the novels of Chariton and Xenophon stand together, in
this as in other respects, as relatively straightforward examples of their
genre. As we move to the three sophistic novels of Achilles Tatius,
Longus and Heliodorus, we shall encounter far more complex narrators
and narrative situations.
5 On this see Konstan 1994: 26ff.

6 Virtually the same phrase recurs in Aegialeus’ narrative at 5.1.6.
7 On this see Morgan 1996: 174–175.
chapter thirty-eight
ACHILLES TATIUS
J. Morgan
Leucippe and Clitophon opens with a narrative by an unidentified internal

primary narrator,1 describing a painting of Europa at Sidon and report-
ing the narrator’s meeting with Clitophon, whose narrative of his own
adventures forms the bulk of the novel. The only information forth-
coming about the primary narrator is that he had just experienced a
severe storm at sea; he is, of course, a Greek speaker, and his ekphrasis
displays a pretty taste in rhetoric. No clues are given as to the occasion
or setting of the framing narration, except that, by obvious implication,
it must be imagined as occurring within the same lifetime as the event
narrated. The primary narratee is more or less invisible: an explicative
aside about the identity of the Phoenician goddess ‘whom the Sidonians
call Astarte’ (1.1.2)2 implies a Hellenocentric ignorance of such things;
a single indefinite second-person verb within the ekphrasis (1.1.13, ‘you
might have said that the picture was even moving’) draws the narratee
into the rhetoric and implies a community of interest and taste with the
narrator, implicit also in the mere assumption that the primary narratee
will be interested in Clitophon’s story. The primary narrator becomes
the external secondary narratee of the novel’s romantic substance. He
characterizes himself as ‘under the influence of Eros’ (1.2.1 erōtikos), and
actively elicits Clitophon’s narration by taking him to a suitable setting,
evocative of Plato’s Phaedrus:
I took him by the hand and led him to a neighbouring grove, where the
plane trees grew thick and plentiful, and the water flowed by cool and
clear, just as it comes from melted snow. I sat him down there on a low
bench, and sat myself next to him. ‘Well, it is time to hear your story,’ I
said. ‘A setting such as this is delightful, and just right for erotic fiction’
(muthōn axios erōtikōn). (1.2.3)
1 1.1.1–1.3.1; there is no first-person form before 1.1.2.

2 Quotations are taken from the translation by T. Whitmarsh 2001b, with minor
adaptations.
494 part nine – chapter thirty-eight
It is clear that what he anticipates from Clitophon’s narration is

pleasure, even though Clitophon has already hinted that his experi-
ences were at least partially negative; his engagement is at the level of
entertainment, not profound human sympathy. In this respect he corre-
sponds closely to the primary narratee of his own narration.
After the opening scene the primary narrator disappears. There is no
suggestion that he has doctored Clitophon’s narrative in any way, and
at no stage during Clitophon’s narrative does he make his presence felt,
either by interrupting the speaker or by inserting an aside to describe,
for example, Clitophon’s body-language or his own reaction to the
narrative.3 The text ends with the conclusion of Clitophon’s narration;
the frame narrative is not resumed. This lack of closure appears to be
deliberate.4 In formal terms Clitophon’s narrative corresponds to the
embedded narrations by secondary internal narrators found in other
Greek novels, and the primary narrator’s frame to the central story into
which they are set. It is this formal arrangement that both allows and
requires Clitophon’s narration to be in a first-person form.5 In Achilles’
novel the embedded internal narrative has expanded to occupy almost
the entire text, but the failure to close the frame leaves the novel in
a pseudo-fragmentary form and allows one to imagine the primary
narrator’s story continuing after his audience with Clitophon.
The greater part of this chapter will be concerned with Clitophon’s
narration, which is imagined as taking place not long after the events,
as Clitophon is still a young man when he meets the primary narrator
(1.2.1). He is an internal narrator, and his narrative is to a large extent
shaped by his role in the story and his character. Although the cir-
cumstances of his narration are very specific, he makes no reference to
them.6 And though a distinction between Clitophon-as-character (the
3 Here one may contrast Clitophon’s own procedures: at 7.11.5 he inserts an edito-
rial parenthesis into a courtroom speech, and at 7.15.1 describes his own reactions at a
crucial juncture in the legal proceedings.
4 On this issue see Nakatani 2003.
5 Leucippe and Clitophon is the only surviving novel written predominantly as an
internal narrative. However, in the Wonders beyond Thule by Antonius Diogenes the
protagonist’s narration of his experiences was also introduced by an authenticating
frame-narrative. There are also some fragmentary fictional narratives with (apparently)
internal primary narrators: the Phoenicica of Lollianus (P. Colon.3328+P. Oxy.1368); and
the so-called Herpyllis Romance (P. Dubl.inv. C3), though we cannot be sure that we are
not dealing with embedded narrative within a more conventional text.
6 At 1.4.3 he compares Leucippe’s beauty to a painting of Selene he had once seen;
a variant reading making this a picture of Europa rather than Selene is sometimes
j. morgan – achilles tatius 495
experiencing-I) and Clitophon-as-narrator (the narrating-I) is implicit

in the whole structure of his narrative, Clitophon himself does not
often draw attention to it. An occasional present tense reminds us
of the act and moment of narration, as when Clitophon exclaims
‘I cannot not express in words how I felt’ (8.14.2), or introduces an
episode by saying ‘I remember Melite making a joke during the fes-
tivities’ (5.14.4), or looks back on his former self from a distance in
time:
We separated, I unwilling and suffering, she—well, I do not know what
her emotions were … I know not of any time before when my heart had
been so joyous: then it was that I first learned that nothing can rival for
pleasure the kiss of desire. (2.8.2–3)7
Similarly he makes one or two passing references to the part of his life
not covered by his narrative:
[The water of the Nile] was sweet to drink, no more cold than was
pleasant (some of the rivers I know in Greece can cause pain; I noted
the contrast with this river). (4.18.4)8
On three occasions Clitophon makes internal cross-references to earlier
phases of his narrative. So at 8.17.2 he summarizes the first part of
an embedded narrative thus: ‘He began by telling everything that I
have already narrated.’ Curiously the other two cross-references are
faulty: at 2.14.2 he introduces Sostratus ‘who was commander of the
war, as I mentioned’, when in fact this is a new piece of information;
and at 7.14.1 he refers to Sostratus’ visit to Tyre ‘as I stated at the
beginning of my story’, though in fact the embassy referred to had
merely been sent on Sostratus’ suggestion. These contradictions may be
due to textual corruption or to authorial incompetence, but may be an
attempt to reproduce the slippages characteristic of an unpremeditated
oral performance.
In the course of his narrative, Clitophon never addresses his spe-
cific narratee, though he makes relatively frequent use of the indefi-
nite second-person singular in descriptive or ekphrastic contexts, as a
taken as a reference to the painting before which he met the primary narrator; however
toiautēn eidon egō pote more naturally suggests an unspecified occasion in the past than the
specific circumstances of a few minutes previously.
7 Note that the first part of this extract is also an explicit example of the limitation
of knowledge of other people’s thoughts discussed below, but unusually (for Achilles)
applied to the present of narration rather than the past of experience.
8 Compare 5.2.2: ‘There was a torchlit procession, that largest I have ever seen.’
rhetorical trope of vivid description, promoting the illusion of presence

or visual perception for his narratee:
As a wave rose high, touching the very clouds, it looked high as a
mountain facing the vessel in the distance; as you watched it approach,
you felt sure the ship would be gulped down. (3.2.7)9
As any narrator might, he also inserts explanatory aside for the benefit
of his narratee: to give a very simple example: ‘Charicles (that was his
boyfriend’s name) came running up in distress’ (1.7.3).10 On a couple
of occasions communication with the narratee is possibly signalled
by exclamations within Clitophon’s narrative: ‘With the gods as my
witnesses, I had no idea what I ate’ (1.5.3).11
A final aspect of Clitophon’s communication with his narratee is the
extreme sententiousness of his narrative.12 Although he is not the only
character in this novel prone to generalizing statements, he is far more
so inclined than any of the external narrators of the other novels. The
large majority of Clitophon’s sententiae concern love, its physiology and
its psychology. The primary narrator/secondary narratee (‘Achilles’)
has already characterized himself as erōtikōs, and these digressive and
often quite lengthy disquisitions on love can be read as playing to his
interests—and through him to those of the primary narratee.
The use of an internal narrator activates a number of protocols. First
there is the question of restriction of knowledge: what the narrator may
be allowed to know and how and when he may plausibly be imagined
to have acquired knowledge of events in which he himself played no
part.13 In the case of Clitophon we find both paralipsis and paralepsis:
sometimes, that is to say, the narrating Clitophon suppresses his own
knowledge of how things turned out and limits himself to what he knew
at the time of the action, but sometimes he appears to know things
for which no plausible channel of information is provided, notably the
thoughts and emotions of characters with whom he had no later con-
9 Note that the narratee is defined as male by the participle blepōn; similar uses
of the second person at 3.7.2, 3.8.4, 4.12.1, 4.19.6 (again with a masculine participle),
5.13.1.
10 Similar naming formulae at 2.17.2, 4.2.1, 6.2.5; explanatory parentheses at 2.16.1,
2.19.1 (a fairly elaborate explanation of the layout of Clitophon’s house), 4.2.1, 7.12.1.
11 Also at 3.17.7.
12 On Achilles’ sententiae and their functions, see Scarcella [1987] 1993, Morales 2000.
13 This issue is discussed at greater length by Hägg 1970: 124–136 and Reardon 1994,
to whose accounts the next few paragraphs are much indebted; also Lowe 2000: 246–
248.
tact or details of scenes and conversations in which neither he nor Leu-

cippe had any part. However, one use to which Achilles puts embedded
narratives is precisely to show how such information came to the sec-
ondary internal narrator. Secondly, the narrator’s intimate involvement
with the story allows (or even, for the sake of plausibility, requires) him
to be partial (in the sense of taking sides): thus Clitophon constantly
makes evaluations of events and persons reflecting his own interests
and attachments. Thirdly, the narrator’s personality may obtrude on
the way in which he narrates his story: if not exactly an unreliable
narrator, Clitophon is one whose judgments and perceptions are not
always coincident with those of the author.14
As an actor in the story, Clitophon lacks the omniscience of the
external narrators of Chariton and Xenophon. This shows itself at a
basic level in his treatment of the divine. Whereas those earlier narra-
tors know exactly what is happening in the superhuman dimensions of
their created world, Clitophon is confined to imprecise deductions:
By chance (some deity had no doubt willed it), there was a festival going
on … I also beheld Zeus Meilichios and the temple of Zeus Ouranios.
I addressed a prayer to the great god, asking that our sufferings should
one day cease … It seemed (eōikei), however, that the god did not nod his
assent to our prayers: Fortune was once again set to put us through our
paces. (5.2.1–3)
The tense locates the uncertainty in the past with the character, rather
than in the present with the narrator, whose surplus of knowledge
is revealed in the prolepsis.15 Similarly Clitophon generally (though
not with complete consistency) is not privy to the inner thoughts and
emotions of other people. This applies particularly to his dealings with
his beloved Leucippe, who reaches us only through the prism of his
perceptions and assumptions. A good example of the effect of this
stance occurs when the lovers experience the generically typical double
dream. Leucippe’s dream can only be known if presented through her
own direct speech, addressed to Clitophon, whereas Clitophon himself,
as narrator, communicates his dream to his narratee without any such
14 By this I mean not the real-life Achilles Tatius, but the controlling intelligence of
the novel, whose careful triangulation and distancing of Clitophon allows the ‘author’s
narratee’ to find a ‘truer’ story than Clitophon is able to narrate. Compare Conte’s
idea of the ‘hidden author’ in Petronius.
15 For similar lack of precision with the divine see 3.5.1 (‘some good spirit’) and
3.23.3 (‘some deity’). The relative paucity of references to divine agency in Clitophon’s
narrative is itself due to his position as internal narrator.
mediation (4.1.2–4). Similarly in the wooing scene of the second book,

Clitophon narrates his own feelings, but only his reading of the visual
indices of Leucippe’s:
I found myself dazed and incapable of speech.
‘Greetings, mistress,’ I said.
She smiled sweetly, and her smile was a coded signal that she had un-
derstood why I had said ‘Greetings, mistress’. (2.6.1–2)
There are a number of other occasions where Clitophon acknowledges
that his version of events is based on surmise or interpretation:
At this point, an emissary arrived from the satrap of Egypt with a letter
for the general. The message, in all probability (hōs eikos), urged him to
war: at any rate, he ordered all his men to arm themselves to fight the
Herdsmen. (4.11.1)16
Particularly in the earlier parts of the narrative, Clitophon generally
tells his story apparently without the benefit of hindsight. This is well
demonstrated by the typical motif of the apparent death of the heroine.
Leucippe has just been kidnapped and is being carried off in a pirate
ship; Clitophon is in pursuit:
When the bandits saw the ship now closing in to attack, they stood the
girl on the deck with her hands tied behind her. One of them shouted
out in a loud voice:
‘Behold your prize!’
Then he cut her head off and shoved the rest of the body into the sea.
When I saw this, I screamed and made to throw myself after her. The
bystanders restrained me, but I begged them to halt the ship and have
someone dive in, in the hope that I might have the girl’s body, even if
only for burial. Two of the sailors hurled themselves overboard, seized
the body and brought it on board. (5.7.4–6)
Clitophon then mourns at length over the headless torso before fading
to a scene six months later. However, Leucippe is not dead; by the time
he is narrating this episode Clitophon knows that she survived, and that
the mistress of one of the pirates had been dressed in her clothes and
beheaded in her stead. The narrator deceptively limits his knowledge
to that of the experiencing-I in order to create effects of suspense and,
when Leucippe reappears, surprise. The event is not disambiguated
16 It is tempting to read the similar qualifications hōs to eikos (6.2.3) and kata to eikos
(6.7.7) as equivalent to a narratorial ‘probably’.

until close to the end of the novel, when Leucippe explains it in the
course of a narrative of her experiences for the benefit of her father.17
In fact she is prompted to include this episode by Clitophon:
Please tell us that fabulous story (ton muthon) about the bandits of Pharos,
and the mystery of the decapitation there, so that your father can hear
it too. That is the only part of the whole plot (tou pantos dramatos) that
remains unheard. (8.15.4)
It seems that Clitophon has heard the story before, but in his narrative
has deliberately suppressed even Leucippe’s first telling of it, for yet
greater effect.
The technique of Clitophon-narrator restricting his knowledge to
that of Clitophon-character is observed rigorously in the first sections of
the novel, and compromised for the first time only at 2.13 ff., where he
uses knowledge gained later to explain the actions of Callisthenes sev-
eral months earlier than the moment being narrated.18 In the complex
intrigue of the second half of the novel, however, Clitophon behaves
more and more like an omniscient narrator, and repeatedly gives ac-
counts of scenes and conversations that he did not witness. Sometimes
he will specify a source of information, as in 4.6–8, where Menelaus
reports to him the substance of his discussions with the general Char-
mides. Sometimes we are left to suppose that Leucippe gave him an
account of her experiences at some time before he communicates them
to his narratee, as with the account of her scenes with Melite at 5.22 ff.
At one point Clitophon specifically presents his account as the prod-
uct of his own deduction: ‘She thought, clearly, that she would not be
believed if she refused: that, I imagine, is why she promised’ (5.22.7).
There are clearer paralepses, however, when secret thoughts and feel-
ings are attributed to characters who cannot be imagined to have told
Clitophon about them, as, for example, in the account of Thersander’s
emotions at 6.17.5–18.2.
Another aspect of the subjectivity of Clitophon’s narrative is that he
does not hesitate to be judgmental: his narrative reflects his sympathies
and personal responses as much as his knowledge. The perspective of
these evaluations can be that of Clitophon as character:
She truly was beautiful: you would have said that her face was daubed
with milk, and that roses grew in her cheeks. Her brilliant eyes scintil-
17 Note that she is deriving ‘great pleasure’ from her own narration (8.15.3).
18 Reardon 1994: 82.
lated with erogenous sparkle, and her hair was thick, long, and golden in
colour. It was not without a certain feeling of pleasure that I beheld the
woman. (5.13.1)
Or it can be a more retrospective one corresponding to that of Cli-

tophon as narrator. Thus before the priest who speaks on behalf of the
protagonists in the courtroom scenes of the final book even begins his
oration, he is introduced as follows:
He was a speaker of no slight ability, and in particular emulated the style
of Aristophanic comedy. He delivered his own speech, opening with an
extremely suave exordium in the style of the comic poets. (8.9.1)
Such judgmental passages seek to compel Clitophon’s narratee to share

his interpretation of the story, and occasionally the narratee is con-
scripted fairly explicitly into the narrator’s attitudes:
Among their slaves was a fellow who was interfering, garrulous, glut-
tonous, and anything else one might want to call him, by the name of
Conops. (2.20.1)19
This last example, however, opens up a crucial issue of reading. The

primary narratee20 can easily see that the bad-mouthing meted out to
Conops is due to the fact that his job is to block the narrator’s access to
the heroine’s bedroom, and that from any other perspective a different
judgment would be drawn. Clitophon is an object of the novel’s irony,
and the novel as a whole invites a different response from the one
Clitophon is represented as trying to elicit from his narratee. The text,
that is to say, invites a critical and distanced reading on the basis of the
narrator’s character.21
A number of points need to be made briefly on this head. First, we
have already noted the sententiousness of Clitophon’s discourse. His
sententiae are merely his, and are a leitmotif of his characterization. Not
only do they mark him and his narratee as erōtikoi. They are also irri-
tating in the delays they impose on the narrative, running on occasion
19 The third-person indefinite inscribes the narratee just as much as a second person.
20 By this I mean the reader of the novel, rather than the person implicitly addressed
by the primary narrator.
21 The two Latin novels by Petronius and Apuleius play more obviously on this
effect: Encolpius and Lucius are both unreliable narrators, for whom corrections must
be made. Note particularly the reading of Petronius developed by Conte 1996 on the
basis that Encolpius is a skholastikos distanced from a ‘hidden author’ critical of all that
the narrator represents. A similar approach to Achilles Tatius needs to be developed;
see, for the first steps, Morgan 1997: 179–186.
to entire pages of digressive and generalizing statements and gener-

ally involving pretentious rhetorical conceits.22 The narrator does not
simply thwart the primary narratee’s desire for narrative pleasure: he
seems to be working to a different, ostensibly more philosophically and
rhetorically elevated agenda. In constructing his secondary narratee,
Clitophon (by the author’s intention) alienates the primary one.
Secondly, at least some of Clitophon’s sententiae are undermined by
his own narrative. For example, Helen Morales has shown how the
sententia on the sexual appetites of barbarians at 5.2.2, applied in the first
instance to Tereus’ rape of Philomela, but hinting at Clitophon’s enemy
Thersander, who is also a Thracian adulterer, turns out in fact to
apply most aptly to Clitophon himself.23 She also draws attention to the
‘tension between the didacticizing form of the sententious declaration
and the hackneyed lessons which it conveys’ and the ‘bathos in the
disjunction between the self-authorizing, self-important pronouncing
of the pseudo-science and the predictable ordinariness’ of the events
narrated.24 At the same time that Clitophon is parading his knowledge
of the generalizations of love, few, if any, of which are grounded in
narrated incident, he signally fails to learn from the lessons that a
careful reader can see the plot is offering him.25
Thirdly, the events of the narrative suggest a very different charac-
terization of Clitophon from that of his self-representation. This is a
man who emerges as weak, cowardly, pompous, self-serving, gullible,
wilfully blind about himself and others. The point is that he is shown as
shaping his autobiographical narrative so as to assimilate it to literary
norms and generic expectations. His narrative dances over the chasm
between the person he is and the person he ought, as the hero of a
novel, to be, the person his narratee wants him to be. This tension is
palpable even before the beginning of Clitophon’s narrative. He tells
the primary narrator:
That is a swarm of stories that you are stirring up … My tale is like a
fictional adventure. (1.2.2)26
22 1.6.2–4 on the effects of night on the pain of love; 2.29.1–5 on conflict of emotions;
3.11.1–2 on tears; 5.13.3–4 on the exclusivity of love; 6.7.1–2 on tears; 6.19.1–7 on desire
and anger. Morales 2000: 67ff. provides a conspectus of critical views.
23 Morales 2000: 79–80.
24 Morales 2000: 83–84.
25 Morgan 1997: 182–186.
26 ta gar ema muthois eoike, picked up in the primary narrator’s choice of a setting
muthōn axios erōtikōn (1.2.3); muthos is used of the plot of the novel at 8.4.2, 8.4.3, 8.5.9,
And soon thereafter he refers to his experiences as a drama27 and reflects

on the effects of erotic stories:
Erotic stories fuel the appetite. Even if you school yourself into self-
control, an example incites you to imitate it. (1.5.5)28
Clitophon is thus set up as a self-conscious narrator, whose story is
mediated through literary artifice. A major task of the primary narratee
is to disentangle ‘fact’ from the spin put on it by the secondary narra-
tor.
Embedded narratives
As we have already seen, Clitophon’s own narration is strictly speak-

ing an embedded or secondary narrative in the open-ended primary
frame story. It has embedded within it further or tertiary narratives.
Not unexpectedly these cover a wider range of forms and functions
than in the simpler novels of Chariton and Xenophon. Some are exter-
nal analepses, generally dealing with mythical or aetiological material.
A few of these cases are digressions in Clitophon’s own voice, and can
be seen as another index of his conceitedly scholastic persona. They are
prompted by objects or events in the plot, but have no organic connec-
tion with it. Thus at 2.2.2–6 he interrupts his narrative of the meal at
which he made his first advance on Leucippe to give a narrative on the
origin of wine, and a few lines later, when he is on the point of being
reluctantly married to his half-sister Calligone, he digresses into an aeti-
ology of Tyrian purple, prompted by the bridal gown; this digression
is said to derive from ‘Tyrian mythology’.29 Two of these narratorial
digressions take the form of narrative ekphraseis of paintings of Perseus
and Andromeda and the rape of Philomela; the thematic and allegor-
ical connection of these stories to the plot that frames them is fairly
clear, but seems to elude Clitophon himself.30 In other cases the narra-
8.15.4, 8.17.1; at 7.4.1 muthos denotes a false story.

27 1.3.3; the same word is similarly used by the narrator at 1.9.1, 6.3.1, 8.5.9; and by
Clitophon as character at 8.15.4; at 7.2.1 drama denotes a fictitious charade.

28 Similar sententiae about the reception of narratives at 5.5.1, 6.2.3, 8.4.2.
29 2.11.4–8; similar aetiologies from the narrator at 7.13.2–3 on the sanctuary of
Artemis in which Leucippe takes refuge; 8.12.1–8 on the myth of Rhodopis, introducing
the rite of trial of ordeal to be undergone by Melite.
30 On thematic relationships between digression and plot, see Bartsch 1989: 40ff.
tive is put in the mouth of a character, including Clitophon-as-character

when (at 5.5.1–9) he offers Leucippe an exegesis of the painting of
Philomela he has already described in narratorial ekphrasis.31
Two of these embedded narratives form a witty and interesting pair
that playfully further the plot, as Conops, a slave detailed to guard
Leucippe, and Clitophon’s slave Satyrus engage in a duel of fables,
playing on Conops’ name (‘Gnat’). Conops’ fable (2.21) concerns an
elephant consoling a lion afraid of cockerels by telling him that a gnat
can kill an elephant by flying into its ear; Satyrus replies with a fable
of a gnat who defeated a lion by buzzing around his head, only to
be caught in the moment of triumph by a spider’s web (2.22). The
alternative views of the strength and weakness of the gnat conceal
threat and counter-threat.
A final external analepsis occurs when, on his first appearance in
the plot, Clitophon’s new friend Menelaus narrates the death of his
beloved boyfriend (2.34). Although the content of this narrative is not
directly connected to the main story, it serves both to characterize
Menelaus (not least in respect of his homosexual preferences) and to
offer thematic parallels both to the main story (love, loss, exile, and
return) and to the experiences of Clitophon’s other homosexual friend,
Clinias, who also lost his beloved in tragic circumstances.
Other embedded narratives bear directly on the plot. They fall
into two main groups: those recapitulating events already narrated
and those introducing new material. Many of the recapitulations are
very short and serve merely to record the transmission of informa-
tion between the characters. Only in the most marginal sense are they
directed at the primary narratee. Thus, when Clitophon first encoun-
ters Menelaus, there is an exchange of stories. Menelaus’, which is
new, is recorded in full, but those of Clinias and Clitophon, which are
already in the primary narratee’s possession, are simply mentioned:
With a groan, Clinias recounted the story of Charicles and the horse. I
then told mine. (2.34.7)32
31 Similarly digressive are 4.4.7–8 (Charmides relates an event he witnessed, as part
of a disquisition on the sweetness of the breath of the elephant) and 8.6.7–11 (the priest’s
aetiology of the syrinx, just before it plays an important part in Leucippe’s trial by
ordeal).
32 Similar brief transfers of information at 3.14.1, 5.8.3, 5.11.4; cf. Puccini-Delbey
2001: 97–98.
The recapitulations are fuller when information is shared selectively.

A particularly interesting case occurs after the court-case towards the
end of the novel, when Clitophon has to clear the air with Leucippe’s
father, Sostratus. First the priest addresses Sostratus, stressing the plea-
sure to be had from the narration of vicissitudes; Sostratus then passes
the baton to Clitophon, again stressing the entertainment-value of a
narrative of painful events:
After all, if events have caused me a certain amount of grief, it is certainly
not your fault but Fortune’s. And, anyway, a narrative of events past
provides more entertainment than grief for one whose sufferings are over.
(8.4.4)
Since the story being elicited is essentially the same as the one being
narrated, the emphasis on narrative pleasure here inscribes the con-
cerns of both secondary and primary narratees. Clitophon’s response
lists the salient episodes of the novel in such a way that the primary
narratee can review the whole plot, but also understand that Cli-
tophon as a narrator is liable to manipulate and embellish his mate-
rial:
I narrated everything: the passage from Tyre, the sea voyage, the ship-
wreck, Egypt, the Herdsmen, the abduction of Leucippe, the fake belly
beside the altar, Menelaus’ artifice, the general’s infatuation and Chaere-
as’ drug, the bandits’ abduction, the wound in my thigh. I showed them
the scar. When I came to the part about Melite, I omitted my perfor-
mance of the act, reshaping the story into one of chaste self-control,
although I told no actual lies: I told of how Melite was smitten with
desire, and how I controlled myself, of all the time she spent beseeching
me, and failing, of all her promises and all her laments. I narrated the
part about the ship, the sea voyage to Ephesus, how we had both passed
the night together, and how—‘with Artemis here as my witness!’—she
had risen as if from another woman’s bed. Only one of my actions in the
course of the drama did I overlook, namely the services I subsequently
rendered Melite. (8.5.1–3)
He then turns to Leucippe’s story, which he claims outdoes his own;

again his partiality and embellishment of the truth are openly acknowl-
edged:
When I reached the part about Sosthenes and Thersander, I elevated
(exēiron) her story even more than I had done mine, in an amorous
attempt to gratify her, given that her father was listening. I told of how
she had endured having every kind of outrage inflicted upon her body
except one, and that it was for the sake of the last-mentioned that she
had tolerated all the others. (8.5.5)
Finally the reaction of the tertiary narratees is recorded: a mixture of

wonder and close emotional involvement, which may be suspected of
closely cuing the response both of Clitophon’s (secondary) narratee and
the primary narratee:
As they listened to this, the priest sat agape with wonder at each of
the events in the story, while Sostratus wept whenever the plot included
Leucippe. (8.5.9)
The complex process is completed by Leucippe’s own narration to her
father (discussed above), where pleasure is again identified as a product
of the narrative.
On other occasions, selective or distorted recapitulations made by
characters in the story are given in direct speech; such, for example
are Melite’s denial of Thersander’s accusations of infidelity, Clitophon’s
speech of self-accusation in the belief that Leucippe is dead, or Clinias’
judicial defence of Clitophon.33
Where an embedded narrative is used to introduce new material, the
most common structure is that retrospective narrative disambiguating
an earlier event is introduced at the point where Clitophon himself was
first made aware of it. In other words, the order of the narrative reflects
the narrator’s processes of cognition, not the story itself (though Leu-
cippe’s narration to her father of her apparent decapitation seems to be
an important exception). We have already discussed this issue in con-
nection with the narrator’s suppression of subsequently acquired knowl-
edge. Examples are Menelaus’ explanation about the apparent death of
Leucippe as a victim of human sacrifice (3.19.1–22.6); Chaereas’ reve-
lation that Leucippe’s madness is the result of an aphrodisiac admin-
istered to her by the now dead Gorgias, sourced by a confession of
Gorgias’ servant (4.15.4–6); Clinias’ account of his survival of the ship-
wreck in which he had been presumed to have died and including the
news that Clitophon’s father has agreed to his betrothal to Leucippe
and is searching for him (5.9.1–10.7); Sostratus telling Clitophon the
ending of the story of Calligone and Callisthenes (8.17.2–18.4), com-
pleting the narration of events given earlier by Clitophon himself (‘he
began by telling everything that I have already narrated’; in fact Cli-
33 6.9.2–7, 7.7.3–6, 7.9.9–13.

tophon’s knowledge of Callisthenes’ actions leading up to the abduction

of Calligone can only derive from Sostratus’ version here).34
Two other embedded narratives merit special comment. The first
occurs at 5.11.4, where Melite is first mentioned, when Satyrus explains
Clitophon’s situation to Clinias and Menelaus. This masks an odd
paralipsis, since Clitophon has just passed over the whole six-month
period following the apparent beheading of Leucippe, in the course of
which he must have encountered Melite, in a single clause. Even more
blatantly than Leucippe’s narrative at the end of the novel, this is a case
where the embedded narrative does not represent the channel through
which Clitophon became aware of the facts, but channels new material
to the primary narratee in a way calculated to maximize effects of
surprise and suspense.
The second is a lying narrative at 7.3, arranged by Thersander to
convince Clitophon that Leucippe is dead. This, of course, has impor-
tant functions in the story, motivating Clitophon’s behaviour in the
courtroom. It is worth mentioning, however, for the extreme complex-
ity of the narrative structure. Thersander has planted someone in the
same prison as Clitophon. The stoolpigeon strikes up a conversation
with a third prisoner, on which Clitophon eavesdrops, as an unintended
narratee; he spins a false story of a chance encounter with a young
man whose extorted confession to the murder of Leucippe he over-
hears. The kernel of the lying narrative is thus doubly embedded in a
sort of Russian-doll effect, and this arrangement is specifically chosen
in order to lend authenticity to a fiction.
In comparison with the pre-sophistic novels of Chariton and Xeno-
phon of Ephesus, Achilles’ is formally more complex and requires more
sophisticated interpretation by its primary narratee. Of course, narra-
tology insists that a separation between narrator and author should
always be made, but Achilles’ is the first novel to make of that sep-
aration an important component of its signification. In his case, the
fact that the bulk of the story is given to a secondary internal narrator
makes the situation reasonably clear. Longus explores a similar effect
with a primary internal narrator.
34 On the effect of dividing the sub-plot in this way, see Morgan 1997: 185–186.
chapter thirty-nine
LONGUS
J. Morgan
Like Leucippe and Clitophon, Longus’ Daphnis and Chloe is introduced by a

brief framing narrative by an anonymous masculine internal primary
narrator; in each case the story that forms the bulk of the novel is
formally a secondary narrative. But whereas Achilles’ primary narra-
tor quickly becomes the silent and invisible external narratee of a sec-
ondary narration by an internal narrator (Clitophon), the first-person
voice of Longus’ prologue continues as that of the external narrator of
the story of Daphnis and Chloe. And whereas Achilles’ primary nar-
rator is scarcely characterized, Longus’ is carefully triangulated in a
way that ironically distances him from the author and invites an un-
straightforward response to his telling of the story:1
On Lesbos, while hunting, in a grove of the Nymphs, I saw the most
beautiful sight I have ever seen, a depiction of an image, a history of love
… I looked and I wondered, and a desire seized me to respond to the
painting in writing. I found someone to interpret the picture, and have
laboured hard to create four books, an offering to Love, the Nymphs and
Pan, a possession to delight all mankind, which will heal the sick and
comfort the distressed, stir the memory of those who have been in love,
and give preparatory instruction to those who have not. For certainly no
one has ever escaped Love, nor ever shall, so long as beauty exists and
eyes can see. For ourselves, may the god grant us to remain chaste in
writing the story of others. (proem 1; 3–4)2
Several significant features of the narrator are established in this open-
ing paragraph. First, the fact that he is in the countryside to hunt aligns
him with a series of wealthy urban characters within the story who
also use the country for pleasure. Like them he has a palpably urban
perspective and aesthetic, which affect the way he tells his story; the
novel’s deconstruction of their ‘soft’ pastoral fantasy extends to its own
narrator and his narratee. Second, while the derivation of the narrative
1 This idea is explored at greater length in Morgan 2003.

2 Translations are taken from Morgan 2004.
508 part nine – chapter thirty-nine
from a painting provides it with a fictional provenance and fictionally

exonerates the narrator from the charge of having invented the story, at
the same time it exposes his limitations. His description of the painting
as a series of unconnected scenes enacts his initial failure to construe
it as a narrative, until an exegete instructs him. The story, within the
fiction, has an existence independent of this particular telling of it: the
narrator is not its controlling intelligence, but rather a not particularly
good reader (of the painting) or narratee (of the exegete),3 reliant on
a third party’s exposition and driven by a potentially irrational desire,
whose take on the story need be no more authoritative than any other
reader’s, on either the factual or the interpretive level. Third, he spec-
ifies the kind of people he envisages as his narratees: they share his
sentimental preconceptions, and notably his view that love is a disease
in need of a cure, a view that the story itself will comprehensively dis-
prove. Finally, the prayer for self-control is double-edged. It voices the
narrator’s fear of losing artistic distance and ending up with mere por-
nography, and the very fact that the narrative begins immediately after
implies that the narrator believes his prayer has been answered and his
text will resist misreading. However, his prayer has drawn attention to
the possibility of such misreading; it is a challenge that the reception
history of Daphnis and Chloe shows has often been accepted.4
It is obvious that the narrator of this novel is just as much a fictional
creation as its protagonists; he is carefully distinguished and distanced
from the author.5 In the following paragraphs, therefore, it will be nec-
essary to approach him and his narratee(s) at two distinct levels. The
first takes him at his face value and examines how he tells his story. The
second recognizes the narrator as a literary device of the author; if not
exactly unreliable, he is at least a distorting lens, for whom the reader
must make corrections in order to realize the text fully. Correspond-
ingly, we must also distinguish two narratees: the narrator’s narratee, at
whom the fictitious narrator directs his discourse and who is imagined
to respond as the narrator intends; and the author’s narratee or reader,
who reads through the narrator to uncover the author’s communica-
tion. However, although the narrator is the target of the author’s irony,
3 The exegete’s narration is of course only implicit in the text to which it gave rise;
the narrator is not directly represented as the exegete’s narratee.

4 Fascinating material on reception history is presented by Barber 1989.
5 In speaking of the ‘author’ here I mean not the real-life Longus, but the voice
of the novel as a whole, a figure sometimes termed ‘the implied author’ or ‘abstract
author’; see Morgan 2003: 175 n. 9.
j. morgan – longus 509
it is not possible or methodologically desirable systematically to disen-

tangle these two levels. Indeed, the uncertainty as to where authority is
located in this text is both an element of its power and fascination and
wholly characteristic of Longus’ Hellenistic poetics.
The narrator’s act of narration implicitly takes place soon after the
discovery of the painting that it expounds. The impulse to narrate
was immediate, but composition was laborious and lengthy.6 For the
narrator the events of the story already belong in an indeterminately
distant past.
Although the author’s narratee is led to see that the narrator does
not really fully ‘know’ his story, he presents himself as a conventionally
omniscient narrator, with access to the entire action, including the
unspoken thoughts, emotions and dreams of his characters. In fact he
knows more about the characters’ feelings than they do themselves;
a pose of ironic superiority to which we shall return. The narrator
limits this omniscience, however, in two interesting ways. He avoids
claiming knowledge of action on the divine plane, except when the
gods intrude physically or through dreams into the world of mortals
and so become open to human perception. For example, after Chloe
has been abducted by the Methymnaeans, the Nymphs appear in a
dream to Daphnis and tell him that they took pity on Chloe when
she was exposed as a baby, have been taking care of her ever since,
and have now asked Pan to intervene to save her (2.23.2–5). None of
these divine actions has been previously narrated: they are revealed
achronically at the point where the divinities tell the characters what
they have done. Furthermore the narrator’s knowledge of the Nymphs’
state of mind is limited to what a human observer could infer from
their appearance: ‘it was as if they were sorry for Daphnis’ (2.23.1).
Similarly, Pan’s intervention is narrated only as it would have appeared
to those involved; his agency is revealed only when he informs the
Methymnaean commander Bryaxis of it in a dream (2.27.1–3). The
supernatural events of the intervention itself are carefully relegated to
the realm of appearances:
… all the land suddenly seemed ablaze with fire, and there came a noise
of the splashing of oars, as if a great fleet were sailing in to the attack.
Someone gave the call to arms, another shouted for the commander.
6 This is implied by the verb exeponēsamēn (pr. 3), with its resonance of Alexandrian
painstaking craftsmanship. We recall for instance that Catullus’ friend Cinna took nine
years to compose his epyllion Zmyrna (Cat. 92).
One man appeared to have been wounded, one lay on the ground in a
semblance of death. One might have thought one was watching a night-time
battle, but there was no foe there. (2.25.3)
Eros controls events,7 but again the narrator disclaims direct knowl-
edge, retreating into formulations such as ‘as if Love had taken pity on
him, the following happened’ (3.6.5).8 This reticence about the actions
of the gods contrasts with the practice of the embedded mythical nar-
ratives, whose narrators are both narratologically more primitive and
theologically more privileged: when Daphnis tells Chloe the story of
Echo, for example, he is able to tell her not just of Pan’s actions but of
his emotional states too, and even the motives of Earth in hiding the
scattered limbs of Echo (3.23.3–4).
At the opposite end of creation, the narrator distances himself from
the pathetic fallacy, and avoids ascribing human thoughts and emotions
to the animals. For example Daphnis’ goats ‘seemed to be listening’
to the music of his pipes (1.13.4), and after Dorcon’s death his cows’
mournful mooing and aimless running was their way of lamenting their
dead herdsman ‘in the estimation of shepherds and goatherds’ (1.31.4).
In communicating with his narratee, the narrator uses the first per-
son of himself only once, in a context like those just discussed, to qual-
ify the attribution of human emotions to animals which were ‘pining, I
think (oimai), for Daphnis and Chloe’ (1.32.3). The narratee is addressed
directly just once, when the narrator is describing Mytilene, which ‘will
give you the impression of an island rather than a city’ (1.1.2).9 Else-
where, however, the narrator invokes the reactions of a hypothetical
tis to the events he is describing, to similar effect. So, for example,
when Chloe watches Daphnis bathing, the narrator says of his suntan
‘one might have supposed its colour came from the shadow of his hair’
7 At 2.27.2 Pan refers to Chloe as ‘a maiden from whom Love intends to make a
story’, the story being the very one narrated.

8 The one exception to this is at 1.11.1, where the narrator says ‘Love plotted
something serious’. I read this as an intended metaphor on the narrator’s part, which
the author’s narratee can see is the literal truth. On the other hand, when the narrator
says (1.15.4) ‘now Daphnis too had to know the deeds of love’, his omniscience is simply
that of one who knows the whole story, not of one who knows the minds of the gods;
similarly 2.2.6.
9 nomiseis ou polin horan alla nēson. If we accept this reading from the better of the
two primary manuscripts (V), the future tense may be understood as the narrator
anticipating the narratee’s reaction to his vivid description. However, the other MS
(F) has the more expected optative nomisais implying a condition such as ‘if you were to
visit it’.
(1.13.2); an instance of the ‘anonymous focalizer’ device. The indefinite

pronoun is distinct from Chloe, whose reactions are described omni-
sciently in the next sentence, and also, formally, from the narrator, who
is describing events directly. It provides a locus for the narratee within
the frame of the fiction, but one that allows the narrator simultaneously
to impose on him a self-consciously romantic or literary response and
to distance it from himself.10
Another simple form of communication between the narrator and
his narratee is the explanatory aside, which often, as with Chariton,
does not even disrupt the syntax of the sentence in which it is included.
A good example is at 2.21.2: ‘So they sailed away, toiling at the oars
(for there was no wind), but Daphnis …’ A particular use of this device
is in naming characters, the narrator’s prerogative in this text: ‘Greatly
alarmed, Bryaxis (this was the commander’s name) leapt up’ (2.28.1).11
This narrator is not a particularly sententious one, but his senten-
tiae do further establish the basis of shared interests and assumptions
on which the narrator’s narratee is constructed. For instance, on three
occasions a sententia is used as a psychological explanation, embodying a
general truth of human nature, unquestioningly acceptable to both nar-
rator and narratee. After Daphnis has been recognized by his natural
parents
he hugged them to his breast as if he had known them all his life, and
refused to leave their embrace. So quickly does nature win credence.
(4.23.2)
A few chapters later he devotes his pastoral gear to the deities of the
countryside:
But there is more pleasure in what is familiar than in unaccustomed
prosperity, so much so that he wept over each of these objects as he
parted with it. (4.26.3)
The romantic proclivities of the narrator and his narratee are most
clearly shown by the explicative function of the sententia at 3.5.4:
But for love all ways are passable, through fire, water, and the snows of
Scythia. So (oun) he ran all the way to the yard.12
10 The same effect occurs at 1.23.2, 2.25.4, 2.35.2, 4.2.1, 4.4.4, 4.15.4, 4.32.2.
11 At 4.5.2 one MS (V) includes in the parenthesis an etymology of the runner
Eudromus’ name. If this reading is retained, it will be an example of the kind of
characterization of the narrator discussed below.
12 This seems to be a proverb or cliché; a very similar formulation occurs in Lucian’s
Occasionally the narrator comments on the action in a way that again

presupposes some community of thought or experience with his nar-
ratee: ‘Such sights would have turned even old men’s thoughts to sex’
(3.13.3). But these comments are not always easy to interpret. For exam-
ple, after Daphnis is rescued from a wolf-trap, the goat he was chasing
is also hauled out, with both its horns broken. The narrator comments:
‘So did Justice punish him for his defeat of the other goat’ (1.12.5).
The evocation of providential Justice in such an incongruous context
is clearly ironic, but it is impossible to decide whether the irony is that
of the narrator sneering at his rustic subject matter (as he often does;
see below), or of the author mocking a sentimental view of the world
voiced in all seriousness by the narrator.
Often the narrator’s comments make his narratee complicit in a
view of normality that naturalizes the action of the story in spite of
its unfamiliar or fantastic quality:
Astonished, as was only natural, he drew closer and discovered a bonny
baby boy, in swaddling clothes. (1.2.3)13
Or:
The dogs that accompanied her to protect the sheep and goats, with
typical canine keenness to pick up a scent, got wind of Dorcon as he
moved to attack the girl. (1.21.2)
Or:
As old men do when they have had a drop or two, they started telling
each other lots of stories. (2.32.2)14
More ambiguously, the narrator trades on assumptions shared with his
narratee to adopt an overtly evaluative or ironical stance towards his
characters and their actions. His high profile in this regard is an ele-
ment of the author’s characterization of him. The very partiality that
binds the narrator to his narratee often distances him from the author’s
narratee and exposes the limitations of the narrator’s understanding. A
very simple manifestation of this evaluative stance is the specific state-
ment of moral qualities in narratorial character-sketches: the narrator
tells rather than shows what sort of people, for instance, Astylus and
encomium of Demosthenes (14), and the thought is a familiar one from erotic poetry.
This self-conscious use of proverbs etc. is an aspect of the narrator’s persona.
13 Similar phrases denote ‘naturally’ at 2.2.1, 2.10.1, 4.27.1.
14 Similar formulations at 3.21.2 (what sailors always do), 3.27.1 (what lovers with no
money always do).

Gnathon are.15 Likewise he expresses sympathy, approval and so on

in quite straightforward ways, referring, for example, to ‘poor Dorcon’
after his death (1.31.4), and describing Dionysophanes’ formal garden
as ‘absolutely superb’ (4.2.1).16 More significant are comments the nar-
rator passes on rusticity from an urban perspective, or on sexual inno-
cence from a standpoint of knowledge and experience. For example,
the developing beauty of Daphnis and Chloe is described as ‘too fine
for the countryside’ (1.7.1).17 Chloe believes Daphnis’ oaths of devotion
sworn on the animals ‘for she was but a young girl and a shepherdess
and thought the goats and sheep were the special gods of shepherds
and goatherds’ (2.39.2). Daphnis ‘like the rustic and goatherd he was’
has no suspicions of Lycaenion’s intentions (3.18.1). This attitude of
ironic superiority by the narrator towards the characters is an essen-
tial aspect of the dynamic of his narrative. It is, for instance, central to
the account of the protagonists’ parallel inamorations that they do not
know what their emotions are, whereas the narrator and his narratee
instantly recognize the symptoms of love. A single word can encapsu-
late the stance of superiority: the narrator introduces Daphnis’ puzzled
soliloquy on the effects of love with the verb apelērei (1.17.4) with its overt
connotation of foolishness, and at the end, with patronising humour,
calls him ‘Sir Daphnis’ (1.19.1).18 This amused and distanced superior-
ity runs throughout the narrative. When Daphnis and Chloe are first
sent out to tend the flocks, the narrator comments that ‘they took on
this job with great delight, as if it were a major office’ (1.8.3), indicating
that from his sophisticated perspective it is nothing of the sort. When
Lycaenion offers to teach Daphnis how to do what he wants to do to
Chloe, the narrator comments that he responds ‘as if he was about to
15 Respectively ‘a large-hearted young man and not unacquainted with the pain of
love’ (4.17.1) and someone ‘whose accomplishments comprised eating, getting drunk,
and drunken fornication, and who consisted of nothing more than jaws, a stomach,
and the parts below the stomach’ (4.11.2).
16 In each of these cases he is distanced from the author’s narratee: Dorcon is a
complex and ambivalent figure, while the symbolic structure of the novel invites a
comparison of Dionysophanes’ park with Philetas’ cottage-garden, not to its advantage,
when the author’s narratee’s intertextual awareness of Hellenistic poetry comes into
play; see Morgan 2004: 223–225.
17 Closely similar is the description of Lycaenion as ‘by country standards rather
glamorous’ (3.15.1). Such judgments are echoed by the urban characters in the novel,
confirming the author’s alignment of the narrator with them.
18 Ho beltistos Daphnis; compare ‘Sir Cricket’ at 1.26.3.
be taught something important, something truly heaven-sent’ (3.18.2),

implying this time that for men of the world such as he and his narratee
sex holds no mysteries.
The narrator’s frequent tone of sardonic humour also has the effect
of belittling the story. As an example we can cite the episode where
Daphnis tumbles into a pit dug to trap a predatory she-wolf. The nar-
rator milks the slapstick, first by equating Daphnis to the rutting goat
he is chasing and then in a po-faced description of the pair disappear-
ing into a big hole one at a time, capped by Daphnis’ perpendicular
flight on goat-back, like an inept Bellerophon on a caprine Pegasus:
What with the one desperate to get away and the other pursuing furi-
ously, they paid scant attention to the path beneath their feet, but both
fell into a pit, first the goat and then Daphnis. What saved Daphnis’ life
was that he used the goat as a steed on his way to the bottom.
(1.21.1–2)
There seems to be more narrator’s humour at 4.12.3, where Daphnis

knocks over the pederastic parasite Gnathon and leaves him sprawled
on the ground: ‘It was not a boy he needed now to lend a helping hand
but a man.’ The narrator’s weak but innocent aphorism is subverted
by the author’s game of luring his narratee into seeing a smuttier
sense than the narrator’s words will literally bear.19 Often the narrator’s
humour resides in a single pointedly chosen word, as when he refers
to Lycaenion’s sexual ‘tutorial’ (paidagōgia, 3.19.1) of Daphnis, or uses a
technical agricultural term of the beating up dealt to Lampis and his
rustic chums by the effete Gnathon and his gang of city-boys (4.29.2
sunēloēse ‘he completely threshed’).
Before we examine how the author’s narratee might construe these
comments, we must note some other facets of this narrator’s visible
persona. He consistently glosses the action, with phrases like ‘this was
the first time they had ever heard Love’s name’ (2.8.1), displaying an
omniscience extending beyond the substance and chronological limits
of the narrative itself.20 At the very end he includes an external prolep-
sis, which he marks as such with a degree of self-consciousness about
the act of narration: ‘But these names and acts came later’ (4.40.1).
19 This effect depends on sensing the possibility of an erotic double entendre in the
word cheiragōgia. Gnathon had been seeking erotic cheiragōgia from Daphnis but now
literally needs a strong arm to lean on. The weakness of the narrator’s clean joke may
be a deliberate element in the author’s characterization of him.
20 Similar glosses are frequent; as at 1.22.3, 2.9.1 etc.
This self-awareness is at its most apparent on three occasions when he

proudly draws attention to the paradoxical novelty of his material,21 but
also functions in a more complex way, as when he advertises the double
unexpectedness of Daphnis’ escape from piracy and shipwreck (1.31.1).
The complexity here is that for anyone who knows the conventions of
the Greek novel (as the author’s narratee does but the narrator’s narra-
tee does not) these incidents are anything but unexpected: they are the
canonical generic clichés. The narrator’s pride in his plotting is under-
cut by the author’s ironical critique of generic conventions. One striking
way in which the narrator signals his control of the narrative is through
the use of ‘if-not situations’ (→ Homer), especially at transitions into
new episodes:
Perhaps they would have done the real thing, had not trouble of the
following sort overtaken that entire area of the countryside. (2.11.3)22
The narrator signalled his concern for sōphrosunē in the prayer at the
end of the prologue, and although the story of Daphnis and Chloe is the
most overtly sexual of any of the Greek novels, the narrator certainly
keeps within the bounds of propriety. Neither of the novel’s two acts of
sexual intercourse (Lycaenion’s ‘tutorial’ and the protagonists’ wedding
night) is narrated, the narrator drawing a curtain of prim euphemism
at the crucial moment. Other sexually charged incidents, such as the
famous scene where Daphnis retrieves a cicada from Chloe’s bosom
are also narrated with a sort of artificial innocence. At such moments
the story resists the narrator’s telling of it.23
Last but not least, the narrator is prone to displays of pedantic
erudition. This is exhibited in his bookishly humorous use of proverbs:
Dorcon, after coming so close to danger and escaping from the jaws not
of a wolf, as in the proverb, but of a dog, was nursing his wounds.
(1.22.1)24
21 3.30.2 ‘asking them for their son’s hand in marriage, a quite unprecedented thing’;
4.7.5 ‘mourning for flowers, a thing without precedent’; 4.22.3 ‘Daphnis might have
been lost in being found, an event without precedent’; in this last example the unreal
condition, and the elaborate wordplay draw further attention to the narrator.
22 Compare 3.24.3 ‘Chloe might easily have become a woman, had not the thought
of blood scared Daphnis’; 4.22.3 ‘Daphnis might have been lost in being found—an
event without precedent—had not Astylus realised what was happening and called out
again.’
23 Morgan 2003: 185ff.
24 The proverb is found in most Greek paroemiographers, who refer it to those who
get something unexpectedly. The humour lies in using the literal sense of the proverb in
But most striking are passages where the narrator intrudes himself
directly into the text. When pirates abduct Daphnis, he is rescued when
Chloe plays the pipes given her by Dorcon, whose cows, which the
pirates have stolen, respond by jumping overboard and capsizing the
pirates’ ship. This already ludicrous sequence is expanded by a minute
account of the physics of the sinking of the ship and rounded off with a
little paradoxographical excursus:
In fact, a cow swims even better than a human being, and comes second
only to water-fowl and, of course, fish. A cow would never drown while
swimming, were it not for the fact that the ends of its hoofs drop off if
saturated with water. Evidence to this effect is provided by the existence
to this day of a large number of places by the sea named ‘Oxford’.
(1.30.6)
The absurdity of this has dismayed scholars, and even led to a proposal
to delete the whole passage as a copyist’s addition.25 However, there is
no good reason to do so, and from our narratological viewpoint we can
see that the joke is on the narrator himself, whose ridiculous pedantry
distances him from the author and the best reader of the novel. Other
discursive intrusions by the narrator26 lack the obvious irony of this one,
but nevertheless position him as an eager purveyor of erudite detail
from an urban perspective. They too have aroused the unjust suspicions
of textual critics.
It is time to turn to the effects of a narrator being so clearly dis-
tinguished from his author. It would be an overstatement to term him
unreliable, but he should not be taken as possessing ultimate authority:
the deep-level meanings articulated by the symmetry and structure of
the plot and its recurrent symbolism do not coincide with those voiced
by the narrator.27 In a general way, the novel’s evident awareness of its
own artificiality and its deconstructive play with generic conventions
belong to the author and are at the expense of the narrator, who buys
into the stereotypes at face value (as we have seen in the case of his
boast about the unexpectedness of events at 1.31.1).
a context where its accepted meaning is inappropriate. A similar oddity occurs at 2.2.6,
where Daphnis and Chloe ‘like proverbial dogs off the leash, frolicked, piped, sang,
played rough and tumble with the goats and sheep’, though Aristotle (Rhet. 3.1406b28)
applies the proverb to sudden and petty viciousness.
25 Castiglioni 1906; Utinam recte (says Reeve 1982) sed Longum sapit.
26 Such as those on Lesbian wine at 2.1.4 and 4.10.3.
27 Morgan 2004: 10–20.
More specifically, despite the narrator’s parading of the tropes of

omniscience, the author’s text supplies material suggesting that there
are aspects to the story (which, remember, fictionally existed before this
telling of it) that have eluded the narrator. So, for example, the narra-
tor does not make the connection, obvious to many readers, between
the painted image whose discovery he relates in the prologue, and the
images dedicated by Daphnis and Chloe at the end of the story.28 At
one level this failure heightens the reality effect: the narrator’s failure
to see everything implies that there ‘really’ is something there to see.
But equally the connection gives the story a depth and resonance that
bypass its narrator. Similarly, Philetas’ account of his meeting with Eros
and his encomium of the god’s powers are marked both by their place
in a clearly signed structure and by the rhetoric of their delivery as cen-
tral to the text’s meaning, but are treated by the narrator as a decora-
tive interlude.29 Even at the level of detail, one constantly encounters a
significance that is not narrated. For example, Daphnis is rescued from
the wolf-trap after Chloe removes her breast-band and enlists Dor-
con’s help to pull him out. The narrator does not mention the effect
of Chloe’s innocently bared breasts on Dorcon, but when Dorcon re-
enters the plot a few paragraphs later as a suitor sexually obsessed with
Chloe the cause of his infatuation is obvious enough to the author’s nar-
ratee.30 Here the story itself gives a good reader material from which he
can reach a rounder, deeper and more explicit understanding than the
narrator, with his worries about sōphrosunē, allows. Similarly, the narra-
tor appears not to grasp the full significance of the character of Lycae-
nion, whose ‘unwritten story’ the author has carefully planted in the
details of the story.31
The narrator makes Daphnis and Chloe a less profound and more con-
ventional story than it really is. This is partly because, in accordance
with the persona constructed in the prologue, he is made to impose on
the story two opposite but inseparable urban preconceptions. The first
idealizes the countryside, through the sentimental fantasies of simplic-
28 Wouters 1989–1990; Hunter 1983: 42–43; Imbert 1980.

29 This is particularly true if the rhetoric contains a level of specific intertextual
engagement, now lost to us, with the poetry of Philetas of Cos. Such implicit intertex-
tuality would obviously belong to the author rather than the narrator.
30 The author even drops hints behind his narrator’s back: the name Dorcon is
cognate with derkomai: the experienced Dorcon ‘sees’ Chloe more clearly than the
innocent Daphnis can.
31 Dorcon and Lycaenion are discussed at greater length in Morgan 2003: 182–184.
ity and innocence that constitute ‘soft’ pastoral. The story itself resists
this simplification, repeatedly exposing the falsity of the easy antithe-
sis between ‘good’ country and ‘bad’ city.32 Simultaneously, the urban
perspective also entails disdain, manifested as an amused superiority or
even disgusted hostility towards the lack of sophistication in the coun-
tryside and its inhabitants. Within the story the poles of this ambiva-
lence are perfectly figured by the attitudes of the young Methym-
naeans, who begin by taking an idyllic holiday and playing at rustic self-
sufficiency, but end with contemptuous aggression when reality intrudes
too far into their fantasies. One way in which the author communi-
cates with his narratee is precisely by providing such foci within the
fictional frame that mirror and make visible the narrator and his nar-
ratee. This complex literary technique, of course, distances the author’s
narratee from the responses that the narrator seems to invite. It is a
recurrent trope of the novel that the narrator’s irony is turned back on
himself and that in assuming himself more sophisticated than his char-
acters he reveals himself as less profound than his story. To return to
two examples from our earlier discussion: the narrator smiles at Daph-
nis and Chloe’s naivety in supposing the care of the flock to be ‘a major
office’, but the story itself teaches us that shepherding is the analogue of
Love’s providential care for humankind, and so, yes, it is indeed a major
office. And the narrator’s amusement at Daphnis’ enthusiastic response
to Lycaenion’s offer of tuition is undermined by a Platonic allusion
reminding the author’s narratee that Love truly is heaven-sent, and that
the sexual act that Daphnis is about to learn is the outward and visible
sign of the benevolent dispensation driving the whole of creation.
Finally we turn to the embedded tertiary33 narratives in Daphnis and
Chloe. These fall into three distinct groups. First, explanatory analepses
conveying between the characters information already known to the
narratees; second, explanatory analepses introducing new information,
particularly to fill in the section of the story before the narrative begins;
and third, a series of three mythological narratives told by characters.
The first group is technically straightforward, but Longus can use
these simple internal repeating analepses for subtle literary effects. For
example, the primary narrator’s account of Lamon’s discovery of the
infant Daphnis is immediately followed by a reported narrative of
Lamon to his wife Myrtale:
32 This is argued out in Morgan 1994a.

33 The main story itself of course being formally a secondary narrative.
He told her the whole story: how he discovered it lying abandoned, how
he saw it being suckled, how he was ashamed to leave it there to die.
(1.3.2)
This reinforces the characterization of Lamon by singling out the qual-

ities that distinguish him from Dryas: his altruism and the small impor-
tance he attaches to the valuables exposed with the child. Similarly
when Daphnis returns to Chloe’s arms after the wreck of the pirate’s
ship, the important thing is not what Chloe tells Daphnis, but what
she conceals: ‘… the kiss was the only thing she did not mention, out
of modesty’ (1.31.2). After Chloe’s restoration from the Methymnaeans,
the summary reported narratives of Chloe and Daphnis emphasize the
numinous aspects of the story: Chloe the miracles of Pan’s intervention,
Daphnis his vision of the Nymphs (2.30.3–4).
Two of these recapitulatory narratives merit further attention: those
of Lamon and Dryas revealing to Dionysophanes the truth of their
discoveries of the infant Daphnis and Chloe (4.19.3 ff., 4.30.3–4). These
are constructed with exact symmetry and each has a crucial plot-
function, the first leading directly to the recognition of Daphnis, the
second establishing Chloe’s social standing as a suitable wife for him
and paving the way to her own recognition; each is tailored to its
narrator’s character and needs. The slave Lamon gives an adept little
speech, whose narrative is designed to win his master’s good will and
persuade him that he has not overstepped the limits of his discretion.
The free Dryas is less deferential and more proudly rhetorical. The
importance of these speeches is stressed by the astonishment and joy of
their narratees.
These repeating internal analepses are symmetrically answered by
two external analepses spoken by the biological fathers of the protago-
nists, Dionysophanes and Megacles, recounting the circumstances lead-
ing to the exposure of the children (4.24.1–4, 4.35.3–5). As with the nar-
ratives of the two foster-fathers there is contrast within the symmetry:
the motives for exposure differ, and whereas Dionysophanes was led to
regret his exposure of his son by a change of fortune for the worse,
Megacles experienced an immediate change of fortune for the better.
A final internal analepsis that delivers new information to the pri-
mary narratee is Philetas’ account to Daphnis and Chloe of his encoun-
ter with the god Eros in his garden (2.3.2–2.6.2). Although the primary
narrator does not register its thematic importance, the significance of
this episode is stressed for the author’s narratee by its pivotal place in
the novel’s didactic structure. It articulates large truths about Love and
its place in the benevolent governance of the world; although the details
elude us, it seems clear that the poet Philetas of Cos is a major inter-
textual presence. The author makes Philetas and the narrator use the
same form of the same verb (exeponēsamēn, 2.3.3, pr. 3) respectively of
their horticultural work and effort in producing the narrative: the effect
is to make garden and text analogous. So we need not be surprised
to find the character acting as the author’s mouthpiece: his narrative
encapsulates the major authorial themes of the entire text: the mean-
ing of Love, the relationship of art and nature, truth and fiction. This
last surfaces in the responses of Philetas’ narratees who ‘were greatly
delighted, as if the story they were being told was fiction, not fact’
(2.7.1). The allusion to the Platonic distinction between logos and muthos
indicates to the author’s narratee that muthos here means much more
than Daphnis and Chloe intend by it: Philetas’ narrative is a myth in
the Platonic sense, imparting truths truer than mere fact. By thus fore-
grounding the relationship of fiction to transcendent Truth, the author
prepares his narratee for the startling revelation that the whole novel is
a muthos whose author is Eros (2.27.2).
The episode is also remarkable for the interaction between the em-
bedded narrator and his narratees. Philetas has come to teach Daphnis
and Chloe something of great importance. His first attempt to do so
is the narrative of his encounter with Eros, whether we read it as
truth or fiction within the fiction. However, they receive it with simple
unreflective pleasure, albeit in a way which points the author’s narratee
to a deeper understanding of the text’s philosophy. Philetas responds by,
as it were, decoding the allegory, and describing in elevated but explicit
language the nature and power of Eros. However, Daphnis and Chloe
seem no more capable of understanding explicit philosophical discourse
than poetically nuanced allegory: so Philetas resorts to a description
of his own emotions when in love with Amaryllis, now his wife. This
at last enables the lovers to connect their own experiences with Love.
Meanwhile the two failed attempts at communication within the fiction
have allowed the author’s most explicit communication with his own
narratee, without the involvement of the narrator.
Finally, there are three mythological narratives embedded at
roughly the same point in each of the first three books: Daphnis’ narra-
tive to Chloe about an unnamed girl who was transformed into a dove
(1.27.2–4); Lamon’s narrative of the story of Syrinx to a rustic gather-
ing (2.34.1–3); and Daphnis’ narration of the story of Echo to Chloe
(3.23.1–5). These narratives all have an explanatory function as aetiolo-

gies of phenomena in the main story: the dove’s song, the music of the
pipes, the echo; and the second is obviously distractive as well, filling
the time taken for Tityrus to fetch Philetas’ pipes. But it has long been
recognized that their main importance lies in their complex thematic
mirror-functions relative to the main narrative.34
The three myths are linked by common themes. Each is an aetiology
of musical beauty produced by the metamorphosis of a central female
character as the result of an act of aggression by a male. However,
they also form an escalating series: the levels of sexuality and violence
increase, counter-pointing the increasingly sexual relationship of Daph-
nis and Chloe. Metamorphosis is an extreme form of transition, the
loss of one’s previous self, but in these myths it is the doorway to a
kind of immortality. As a series the myths teach that violence and loss
are necessary preparation for the emergence of a new harmony, and in
this respect foreshadow and lend significance to the metamorphosis of
Chloe from virgin to bride and mother: she too must lose her old self,
her childhood innocence, in order to effect the transition from charm-
ing sterility to fully socialized fertility. The myths construct paradig-
matic gender roles, casting man as aggressor and woman as victim for
her own good, and identify these paradigms with the Nymphs and Pan,
who function within the story as the female and male principles of Eros,
overseeing the lovers’ experience and guiding it towards the correct
outcome.35 They are other tellings of the story of Daphnis and Chloe,
as is made clear not just by detailed resemblances between the mythical
characters and the lovers but also by the fact that after Lamon’s narra-
tion Daphnis and Chloe cast themselves as Pan and Syrinx in a danced
version of the story. At the same time, however, the myths also distance
the mutual and consensual love of Daphnis and Chloe from less ideal
forms of erotic action. In this respect they rehearse negative possibilities
that the main story is able to avoid.
It is interesting to note how the myths are adapted to their narrators
and narratees. The lack of overt eroticism in the first myth reflects the
protagonists’ ignorance even of the name of Love; the story of Syrinx
34 There is a large literature on these myths, summarized and discussed by Morgan
1997: 2238–2241; for fuller discussion of their interpretation Morgan 2004: 171–172, 195–
198, 213–216.
35 The characters of the first myth are unnamed mortals, but within the narrative
is a reference to the story of Pan and Pitys, a story of erotic violence analogous to the
other two but appropriately displaced at this point in the protagonists’ development.
comes at the point where they have learned the name of Love but not
the mechanics of sex; the violence and emphasis on virginity in the
story of Echo correspond to a stage in the story where Chloe is in
real danger of losing her virginity, while the high blood quotient of
the story echoes the warning Daphnis has just received from Lycae-
nion about the bleeding that first intercourse will cause Chloe. There
is also evidence of characterization through prose style. Lamon’s nar-
rative is marked by asyndeton, within and between sentences, and by
paratactic brevity, but over his head, as it were, is woven a prose poetry
full of assonance and symmetry, reinforcing the competitively intertex-
tual relationship with the classics of Hellenistic pastoral poetry already
signed in Lamon’s statement that he learned the story from a Sicilian
goatherd (alluding to Theocritus). Daphnis too narrates in an artificially
naïve style.
All three of these narratives are designated explicitly as muthoi, just
as the story of the novel is described as a muthos made out of Chloe
by Love. They are all received with pleasure by their narratees, just as
the novel will bring pleasure to its readers. These things are the mark
of fiction. Nevertheless they serve a vital educative function, counter-
pointing the protagonists’ affective development, just as the prologue
of the novel promises its narratee preparatory education about Love.
The series of aetiological metamorphoses of young heroines into objects
of immortal beauty and pleasure is thus really completed by Chloe’s
transformation into muthos and novel. The series of myths stakes the
author’s claim for his own work, self-referentially mythic, uniquely
combining pleasure and utility.
chapter fourty
HELIODORUS
J. Morgan
Introduction
Chronologically the last of the novels, Heliodorus’ Ethiopian Story is also

structurally the most complex and the most polyphonic in its repertoire
of narrators. Unlike Chariton, who names himself as narrator in his
opening paragraph, Heliodorus remains anonymous until the conclu-
sion of his narrative, where he presents himself not as narrator but as
‘composer’:1
So concludes the composition (syntagma) of the Ethiopian story about Thea-
genes and Chariclea, composed (synetaxen) by a Phoenician from Emesa,
one of the race of the Sun, son of Theodosius, Heliodorus. (10.41.4)2
The difference is striking. Heliodorus positions himself as author out-
side the narrative structure of the novel. His usage distinguishes him
from the primary narrator, who is an element of the text that he has
‘composed’.
The primary narrative begins in the middle of the story. The pri-
mary narrator is anonymous and external. The date and setting of
the act of narration remain unspecified, and the primary narratee is
directly addressed only once in the second person (5.13.4). Although
the primary narrator must know the whole of the story, he often con-
ceals his knowledge, compelling the primary narratee to assemble it
from a series of embedded narratives which characters address to one
another. The sections of the story prior to the beginning of the nar-
rative are communicated to the primary narratee through two prin-
1 The verb suntassō and its cognates are a standard term for literary composition,
generally of an historical or factual kind, the central notion being that of compilation,
not invention.
2 Translations are from Morgan 1989a, occasionally adapted.
524 part nine – chapter fourty
cipal secondary narratives. In both cases the secondary narrators are

internal, and their narratees are, strictly speaking, external; they do not
feature as characters within the secondary narratives, but they are inti-
mately affected by the events being narrated, sometimes in ways that
they themselves do not at first realize.
The first secondary narrator is the Athenian Cnemon, who tells
the protagonists of the novel, Theagenes and Chariclea, about his
stepmother’s infatuation for him and her conspiracy with her slave,
Thisbe. His narrative is divided between two occasions: the first session
(1.9.1–18.1) is solicited by his narratees as consolation for their own
sufferings, and takes place during the night in a robber settlement in
the Nile Delta, where they are all captive; it is received with tears,
ostensibly of sympathy but in fact of self-pity (1.18.1). It includes a
tertiary narrative by Cnemon’s friend Charias, bringing him news of
events in Athens following his exile, culminating in the stepmother’s
death (1.14.3–17.6). Charias is an external narrator, but his narrative
is sourced in the narrative once told him by the internal narrator
Thisbe, from which he quotes verbatim conversations between her and
her mistress. The second section of Cnemon’s narrative (2.8.3–9.5)
is delivered over Thisbe’s body, and explains what happened to her
after her mistress’s death and how she came to Egypt. In this section,
Cnemon is an external but interested secondary narrator, paraphrasing
a tertiary narrative by an external narrator, his friend Anticles, bringing
further news from Athens. We can see in this sequence both how
information is conveyed to the primary narratee through embedded
narratives plausibly but variously motivated within the primary frame,
and how these embedded narratives authenticate the story from within,
forging chains of information extending back to first-hand witnesses.
The second secondary narrator bulks even larger in the economy
of the novel as a whole. He is the enigmatic Egyptian priest Calasiris,
mentor and guide of the protagonists. Again the narration is divided
between two separate occasions. The external narratee of the first and
longer session (2.24.5–5.1.3) is Cnemon. It is occasioned by a chance
encounter between Cnemon and Calasiris and the discovery that they
share the acquaintance of the protagonists. Calasiris’ narrative informs
Cnemon about them, and occupies a whole night in the house of the
merchant Nausicles, where Calasiris is lodging. As internal narrator,
Calasiris recounts his discovery of the heroine’s identity at Delphi. This
secondary narrative again contains embedded narratives. Chariclea’s
foster-father Charicles contributes as internal narrator a tertiary narra-
j. morgan – heliodorus 525
tive, with Calasiris as external narratee, of how he was given the infant
girl by an Ethiopian ambassador in Egypt (2.29.2–33.4). This narra-
tive is set immediately after Calasiris has given a public display of his
Egyptian wisdom; its ‘argument’ function is to persuade him to break
down Chariclea’s resistance to marriage. Within Charicles’ narrative is
embedded a fourth-level narrative, with Charicles as external narratee,
by the Ethiopian ambassador as internal narrator, covering his rearing
of the infant Chariclea after her exposure by her parents (2.31.1–5), and
designed to persuade Charicles to take on the care of the child. At this
point the primary narratee is reading a narrative within a narrative
within a narrative within a narrative. Later in the secondary narrative
is another important tertiary narrative (4.8.1–8), a message in Ethiopian
script left beside the exposed child,3 passed by the ambassador to Char-
icles and unread until Calasiris gains access to it, ostensibly to help
him fulfil his promise to Charicles, but in reality because he is seeking
to confirm his suspicions about Chariclea’s true identity. The internal
narrator of this message is Chariclea’s mother, Persinna, the queen of
Ethiopia, and its internal narratee is the exposed child herself, intended
to read the message at some unspecified future time; in fact Chari-
clea only becomes an actual narratee when Calasiris reads the message
to her slightly later.4 Within a page or two, four levels of narratee—
Calasiris, Chariclea, Cnemon, and the primary narratee—share the
revelation that the heroine is an Ethiopian princess. The message has
a vital function within the secondary narrative, motivating Chariclea to
head for the land of her birth; it also has an effect within the primary
narrative, as Cnemon’s knowledge that his companions are bound for
Ethiopia contributes to his decision to return to Athens.
Calasiris’ second session is set at a party in Nausicles’ house after he
is reunited with Chariclea, and is elicited from him as entertainment.
He tells his whole story for Nausicles’ benefit, but the primary narrator
occludes the part already familiar to the primary narratee. Still as
internal narrator, Calasiris relates events subsequent to the protagonists’
departure from Delphi culminating with the scene with which the
primary narrative began (5.17.1–33.3). In this section there is only one
3 The text, of course, presents it in Greek, as it does all non-Greek speech. At this
juncture Calasiris says nothing about translating it, but later (4.11.4) he tells Cnemon
that he translated it for the benefit of Chariclea.
4 4.11.4; by this point, of course, Calasiris already knows the text of the message,
and so it is not repeated.

embedded tertiary narrative (5.20.2–9), when the fisherman Tyrrhenus

warns Calasiris of a pirate plot by means of an internal narrative of his
conversation with the pirate captain Trachinus.
From this point in the novel to the conclusion, the dominant voice
is that of the primary narrator. A few small-scale embedded narratives
pass information around the dramatis personae, but the second half of the
novel is essentially a straightforward communication between primary
narrator and primary narratee. The Ethiopian story is a long as well as a
complex text, and it will not be possible to present a detailed analysis
of all its narrators and narratives. What follows is perforce selective and
suggestive. I begin with the primary narrator and his narratee, then
move to the secondary narrators and their narratees, and finally offer a
brief overall interpretation.
Heliodorus’ primary narrator is an elusive figure. For much of the time

he shows rather than tells: one point of the rigmarole of multiple nar-
rators is precisely to avoid the imposition of an obtrusive primary nar-
rator who mediates all relevant information to the primary narratee.
Once the decision was taken, for whatever reason, to present the story
in a non-linear form, essential elements of the fabula had to be commu-
nicated somehow. Rather than having the primary narrator repeatedly
pause the story to fill in the gaps at appropriate junctures, Heliodorus’
strategy allows information to be ‘overheard’ by the primary narra-
tee as it passes between characters within the frame of the narrative,
by means of plausibly motivated narratives, so preserving the dramatic
illusion.5 Other aspects of the primary narrator cohere with this strat-
egy.
Most importantly, although he is, like the narrators of the other nov-
els, omniscient,6 he most frequently conceals his knowledge, describ-
ing events as they would have appeared to someone present at them.
This protocol is established in the famous opening tableau: the narrator
describes the scene of destruction on the beach through some bandits’
5 The first, and still the fullest, analysis of the novel’s ‘dramatic’ presentation is Hefti
1950.
6 As is shown by the analepses discussed below, and a smaller number of explicit
prolepses, as at 7.12.1–2; 8.17.5.

perceptions of it; he offers no explanations but reports their hypothe-

ses.7 Here the primary narrator abandons his omniscience by present-
ing events through the eyes of secondary focalizers within the text.
More often, however, he achieves much the same effect by restricting
his own focalization and becoming, as it were, merely an eye-witness
of the events he narrates. In the very first sentence he introduces these
observers not as ‘bandits’ but as ‘men in bandit gear’. The sense is
given that he offers not omniscience but easy inference from visible
data. This inferential stance is so prevalent as to defy complete listing:
She blushed, apparently at having allowed sleep to get the better of her.
(5.34.2)
Most could be identified as Persians by their apparel and equipment.
(6.12.2)
The majority of the men of Bessa … seemed unhappy with the proposal.
(7.5.2)
At other times the narrator’s self-imposed restriction to the perceptible

leads him to withhold from his narratee information necessary for
a full understanding of the action. This information emerges later,
through the action and the words of the characters. The procedure
is antithetical to that of the narrators of Chariton and Xenophon
of Ephesus, and analogous to that of an internal narrator such as
Achilles’ Clitophon, who limits his knowledge at any given point to
that he had as a character at the time of the action. By these means
Heliodorus’ external narrator is able to achieve some of the special
effects of internal narration, including surprise and suspense. Three
examples will give the idea of the narrative mode of most of the novel:
(1) Thyamis hides his beloved Chariclea in a cave when his strong-
hold comes under attack. Despairing of survival he returns and
kills a Greek-speaking woman in the darkness. When Theagenes
and Cnemon go to recover Chariclea they stumble on the body.
A voice is heard through the darkness. When Cnemon fetches a
light and looks at the body, he finds that it is his Athenian nemesis,
Thisbe, and as they go further into the cave, they find Chariclea,
who had been calling to them. At every turn in this sequence the
narrator suppresses his knowledge in order to make his narratee
share the experiences of the characters (1.29–2.6).
7 For more detailed discussion see Bühler 1976; Morgan 1991: 86–90.
(2) While in Egypt, Chariclea is sentenced to be burned alive, but

she turns out to be fireproof, and stands unharmed in the middle
of the flames, ‘like a bride in a chamber of flame’. The narrator
withholds the information that one of the recognition tokens that
she thought she was taking with her to her death is a jewel with
special properties, and continues to do so until the protagonists
realize the truth.8
(3) In the battle at Syene, some contingents of the Ethiopian army
act in a way which is, at first sight, mad. Faced with a charge by
the Persian armoured cavalry, they rush towards the advancing
foe. Their strategy only becomes clear when it takes visible effect:
they throw themselves beneath the Persian horses and stab them
upwards in the belly, the one vulnerable spot in their armour
plating (9.17.2–18.2).
It is coherent with this strategy of not imparting full information to his
narratee that some of the primary narrator’s most explicit interventions
are to place limitations on his own knowledge:
They were given the warmest of welcomes by their host’s daughter, a
young lady of marriageable age, and by all the serving-women of the
house, for such, I imagine (oimai) had been their master’s instructions.
(2.22.1)9
Such expressions of uncertainty are a mannerism of this narrator, and
are difficult to parallel in the other novels. Sometimes they take the
simple form of the example just given, but sometimes the narrator’s
incomplete knowledge is enacted through processes of speculation akin
to those of an historian, resulting in multiple explanations:
About midnight a section of the dyke where the previous evening the
Ethiopians had begun to dig an outlet ruptured without warning: it may
be that the earth in that section had been piled up loosely and not
properly tamped down, so that the base gave way as the water soaked
into it; or those excavating the tunnel may have created an empty space
into which the base of the dyke could collapse; or possibly the workmen
had left the place where they had started their digging somewhat lower
than the rest of the dyke, so that as the water level rose during the
night, causing a fresh influx, the water was able to find a way through
the place where the earth had been shovelled away, and, once that
8 8.9.9–11.9. On this example and the strategy in general see Morgan 1994b.
9 Other examples: with oimai (‘I think’) at 1.8.1; 6.5.1; 8.8.2; 10.6.5; with tacha or isōs
(‘perhaps’) at 2.20.2; 9.11.6, 19.5; with hōs eoike (‘so it seems’) or similar at 6.14.6; 7.5.2;
8.10.1; 10.6.5, 28.3.
happened, the channel grew deeper without anyone being aware of the
fact; alternatively one might ascribe the event to divine intervention.
(9.8.2)10
In a work of fiction such uncertainty is itself fictitious. Heliodorus’
narrator is constructed to have—or to feign to have—a relationship
with the material of the story quite different from that of a creating or
controlling author.11 The limitations placed on him make an important
contribution to the effect of the novel as a whole.
However, although the narrator’s subtraction of himself and his om-
niscience from the narrative is an important innovation, it is a stance
that can be abandoned when convenient. For example, between the two
sections of Calasiris’ narrative a complicated intrigue works itself out:
Cnemon is deluded into believing that a woman captured by Nausicles
is Thisbe, whom he knows to be dead, when in fact it is Chariclea.
The situation is set up in a typical scene of direct speech in which the
primary narrator’s only contribution is a recurrent ‘he said’. However,
at the climax the narrator intervenes heavily to explain:
In fact, however, a supernatural power, whose habit it is in general to
make mock of all human life and use it as its plaything, was evidently
playing with Cnemon and refusing to allow him great happiness without
also making him feel some pain. In a short while Cnemon was going to
experience joy, which heaven therefore was now combining with sorrow,
perhaps simply giving another demonstration of its habitual malice—
though possibly human nature cannot admit of pure joy without a taint
of sadness. Thus it was that that day Cnemon was turning away in
fear from that which he wanted above all else, and that that which was
sweetest to him of all things caused him such terror. For the woman he
had heard lamenting was not Thisbe, but Chariclea! (5.4.1–2)
There follows an extended and omniscient analepsis filling in events
between Cnemon’s separation from the protagonists and Chariclea’s
appearance in Nausicles’ house (5.4.3–9.2). This is completely at vari-
ance with the narrator’s normal protocols: the imposition of limits to
his knowledge is replaced by an omniscience extending to the divine
plane of action, and allowing a precise and explicit prolepsis. The inter-
vention of the primary narrator could easily have been avoided: Chari-
clea could have narrated her experiences to Cnemon the next morning,
10 Other examples of multiple explanations at 1.31.4; 2.13.2; 7.6.4; 8.9.2; 10.28.4,

38.3.
11 The forms and functions of the narrator’s uncertainties are discussed in detail in
Morgan 1982.
so that the primary narratee would, as usual, acquire information by

eavesdropping on a secondary narrative; Nausicles is at hand to fill in
any missing details. We can only speculate as to why this course was not
followed: the crucial point is the flexibility of Heliodorus’ primary nar-
rator as a communicative tool. One special effect in the present case is
that he can describe an intimate scene in a way unthinkable from Char-
iclea’s own lips, both explicit and humorously ironic in its displaced use
of sexual vocabulary:
They clasped one another in a prolonged embrace so tight that they
seemed to be of one flesh. But the love they consummated was sinless
and undefiled; their union was one of moist, warm tears; their only
intercourse was one of chaste lips. For if ever Chariclea found Theagenes
becoming too ardent in the arousal of his manhood, a reminder of his
oath was enough to restrain him. (5.4.5)
We move now to some other aspects of the primary narator. Firstly,
he is very sententious; as in the other novels, normative generalizations
delineate areas of community between narrator and narratee. In this
case some of these sententiae focus on a polis-based Hellenocentricity,
emphasizing the otherness of barbarians or outlaws:
Once embarked on a course of action, the heart of a savage brooks
no turning back. And when a barbarian loses all hope of his own
preservation, he will usually kill everything he loves before he dies, either
in the deluded belief that he will be reunited with it beyond the grave or
else to save it from the shameless clutches of his enemies. (1.30.6)
So much more precious, evidently, do brigands consider money than life
itself: friendship and kinship are defined solely in terms of financial gain.
This was certainly the case here. (1.32.4)12
Such community of cultural perspective is also inscribed in the openly
evaluative descriptions of certain characters marked as ‘other’:
He [Thermouthis] had the hot blood of all brigands and the quick tem-
per of all savages, which, aggravated by his frustrated passion, impelled
him to close with the supposed culprits there and then. (2.12.5)
Arsace was a tall, handsome woman, highly intelligent and arrogant
and proud by reason of her noble birth, naturally enough for one who
had been born the sister of the Great King. But the life she led was
disreputable: in particular she was a slave to perverted and dissipated
pleasure. (7.2.1)
12 Other cultural sententiae belonging to the primary narrator at 1.4.3; 5.7.3; 7.26.10;
8.9.4.
This Hellenocentricity is all the more striking because in order to

make such comments the narrator must abandon his usual stance of
writing with the restricted knowledge of an eye-witness and take his
narratee omnisciently into the heart of the other.13 Particularly in the
episodes centred on the carnally sensual Persian court in Egypt, he
opposes the barbarian to the central romantic values. These values and
the psychology of love form the basis of another series of normative
sententiae:
So it is that genuine affection and wholehearted love disregard all exter-
nal pains and pleasures and compel the mind to concentrate thought and
vision on one object: the beloved. (1.2.9)14
Another aspect of the narrator’s Hellenic identity is supplied by a series
of similes and metaphors from the classical Greek theatre.15 These are
the narrator’s metaliterary-referential commentary on the nature of
his plotting and also serve to construct his narratee as a spectator as
opposed to merely a listener, but the area from which they are drawn
polemically locates both him and his narratee in the Greek literary
tradition. Here is a spectacular example:
At that very moment either some divine power or some fortune that
arbitrates over human destiny made the drama take a new and tragic
twist, almost as if bringing a second drama on stage to compete with the
one already in progress: by a miracle of stagecraft it brought Calasiris on
to the scene on that very day, at that very hour … and before the people
of Memphis could recover from their amazement … the drama was
interrupted by another unexpected entry: Chariclea! … all the people
thronging the section of the wall where Arsacae, already swelling with
jealousy at the spectacle of Chariclea, had her throne were enraptured
13 It is also striking, of course, in that the author, identifies himself as non-Greek,
and the novel as a whole seems to propose a renegotiation of Hellenic identity, even
re-centring the world on Ethiopia—which seems to serve as a cipher for the other sun-
kingdom of Syrian Emesa, Heliodorus’ home city.
14 Other erotic and psychological sententiae at 1.4.3; 2.6.4, 15.2; 5.7.3; 6.7.8, 12.2; 7.7.5;
8.6.1, 7.6; 10.16.1.

15 These were first studied by Walden 1894; for more recent discussion see Marino
1990 and Montes Cala 1992. Apart from their metaliterary nature, they establish a
set of intertextual parameters within which the text can signify; they form the basis
of a fine study by Paulsen (1992). These theatrical metaphors are not confined to
the primary narrator; they sit naturally enough in the mouths of Greek characters
(Charicles at 2.29.4; Chariclea at 2.8.3; 5.6.3–4; 6.8.5; Cnemon 1.8.7; 2.11.2, 23.5, 24.4;
3.1.1; Nausicles 5.12.2; Theagenes 2.7.3; 5.6.4); but rather oddly with Egyptians and
Ethiopians (Calasiris 4.5.3; Hydaspes 10.12.2; Sisimithres 10.39.2).
by this miracle of theatrical art … all were agreed that the high point of
the drama was its romantic side. (7.6.4–5; 7.7.4; 7.7.7; 7.8.2)16
Another prominent aspect of the primary narrator is his ability to make

factual digressions from the narrative, on matters such as the geography
of Meroe, Persian armour-plated cavalry, or Egyptian religion.17 These
encyclopedic displays characterize the primary narrator as a repository
of knowledge and the primary narratee as someone interested in such
things. No other primary narrator of a novel is so excursive.18 It is in
these digressions that communication between primary narrator and
primary narratee is most explicit. For example, the only second-person
verb addressed to the primary narratee occurs in the digression on
Ethiopian amethysts (5.13.4); and the account of the theology behind
the festival of the Nile flood concludes with a piety including both
narrator and narratee in a first-person plural pronoun:
Well, may the gods pardon us for saying this much. The greatest myster-
ies may not be spoken of: let us respect their sanctity as we continue our
story of the events around Syene. (9.10.1)19
Finally, we must add a few words about the primary narrator’s prose
style, whose exuberant floridity both characterizes him as narrator
and directs his narratee’s responses. Frequent wordplays inscribe the
narrator’s reactions to his material. To take just a few examples: after
describing how bandits living on rafts protect their babies by tethering
them by the ankle, the narrator ends with a bon mot for his narratee to
savour:
A strange way to keep children in hand, to tie them by the feet!
(1.5.4)
The wordplay highlights the paradox and the irony, but also the textual
pleasure of both narrator and narratee, drawing attention precisely
16 Other notable examples from the primary narrator include 5.11.2 (a ‘theatrical
recognition-scene’); 8.17.5 (a prolepsis couched in abstruse theatrical terms); 10.16.3

(‘Destiny’s stage-management of human life’), 38.3 (divine staging of events).
17 These examples are at, respectively, 10.5; 9.15; 9.9. These take us outside the
narrative into reality, and should be distinguished from extended description of objects
within the narrative.
18 Achilles Tatius contains more and longer digressions, but, of course, they belong
to the secondary internal narrator and are part of his characterization.

19 Note that this digression rests on authority outside the novel (Egyptian ‘natural
philosophers and theologians’) and includes one of the narrator’s conscious omissions
(more logical as this is not invented material).
to the narrator and his role in mediating events to the narratee.20 A

similar effect is achieved by a conceit as when Theagenes serves wine
to Arsace:
As she quaffed the cup, she kept her eyes fixed steadily on Theagenes;
and she drank deeper of the cup of love than of the wine. (7.27.3)
The narrator does not simply narrate: he draws attention to the act of
narration, inviting a response not just to the story but to his telling of it.
A different effect is achieved in the first paragraph of the whole
novel where the description of the scene on the beach is disrupted
by a grammatical anacolouthon, as if the shock of seeing the bodies
disturbs the narrator’s syntax. But as the device is the narrator’s, it is
designed to convey the shock of the thieves through whose eyes the
scene is described. Even at a point where cognitively he has subtracted
himself, the narrator remains as self-advertising manipulative presence.
Secondary narrators
We turn now to the major secondary narrators, beginning with Cala-

siris. The narrative structure of the Ethiopian story recalls that of the
Odyssey; within this structure Odysseus’ role in filling in the earlier
parts of the story by means of a retrospective narrative by an internal
narrator is taken over by Calasiris. Before even beginning his narrative,
Calasiris alludes to the opening words of Odysseus’ narrative (2.21.5).
He is thus doubly cast as an Odyssean figure before his story begins.
His role as an actor involves trickery and duplicity, albeit in the service
of the higher good; his act of narration is similarly less—or more—
than straightforward. This is made explicit right at the beginning of his
narrative, when Calasiris appears to be telling a quite different story
from the account of the protagonists that Cnemon wants. Cnemon
interrupts before he has finished a paragraph, and associates Calasiris
with another Odyssean trickster:
You very nearly succeeded in bringing me straight to the ending of the
story with your talk, before I realized what you were up to, wheeling on
this subplot which, so the saying goes, has nothing to do with Dionysus.
So take your narrative back to what you promised. So far I have found
20 Similar wordplay at, for example, 1.1.6; 5.4.5; 7.12.1; 8.1.5, 9.8, 14.2; 9.11.5, 19.3,
22.2.
you just like Proteus of Pharus, not that you take on false and shifting
forms as he did, but you are forever trying to lead me in the wrong
direction. (2.24.4)
And when Calasiris makes a new start, he apologizes for his apparent
deviousness:
First I shall tell you briefly about myself. This is not, as you think, a
sophist’s trick to avoid telling the story, but the logical way to present my
narrative and an indispensable preliminary. (2.24.5)
At the heart of his narrative lies an apparent contradiction that calls
his reliability as narrator into serious question.21 At the beginning of
his narrative, Calasiris says he will pass over his travels after depar-
ture from Egypt, since they are irrelevant to Cnemon’s enquiry about
Theagenes and Chariclea. He picks up his story with his arrival at Del-
phi, which he chose as an appropriate destination for a priest. There
his help is enlisted by Charicles, the priest of Apollo, to break down
his foster-daughter’s resistance to marriage; Charicles tells him that she
was entrusted to him by an Ethiopian who had raised her after she
was exposed as an infant. Having committed himself to assist, Calasiris
receives an oracle prophesying a journey to the ‘Black Land of the
Sun’, and, after observing the protagonists fall in love, experiences a
vision of Apollo and Artemis telling him to take the lovers to Egypt.
Perplexed as to what the gods want and how to carry out their instruc-
tions, he decides he must see the message from Chariclea’s mother.
But, having read the message and told Chariclea of its contents, he also
tells her that he visited Ethiopia before coming to Delphi, was com-
missioned by Persinna to find her daughter, and came to Delphi in the
knowledge that she was there.
At the very least this exposes Calasiris to the charge of withholding
important information from his narratee, much as the primary narra-
tor does. More importantly, unless a way can be found to reconcile the
two accounts, either the contradiction derives from authorial incompe-
tence, or the primary narratee is intended to understand one of them
as deliberately deceptive. If Calasiris knew all along that Persinna’s
daughter was at Delphi, his aporia about Chariclea’s identity and fail-
ure to understand some pretty transparent divine messages is difficult
21 This has drawn the attention of many scholars. The first full analysis was that
of Hefti 1950; more recently Winkler 1982 has been hugely influential in arguing that
the contradiction can be resolved; see also Futre Pinheiro 1991, Fuchs 1993: 174–188;
Baumbach 1997 argues that Calasiris is lying when he says he went to Ethiopia.
to accept: is the account of his thought-processes he gives Cnemon

therefore less than truthful? Or is the postponed revelation that he
went to Ethiopia a falsehood invented to secure Chariclea’s cooper-
ation in the fulfilment of the divine plan? None of these solutions is
problem-free, but a narratological approach perhaps favours the last.
The novel would be unreadable if the primary narratee were not able
to accept the facts of Calasiris’ narrative to Cnemon, however manipu-
lative its presentation. However, the story of the visit to Ethiopia occurs
in a tertiary narrative by Calasiris embedded within his secondary nar-
rative; the narrative structure requires the primary narratee to accept
that Calasiris tells Cnemon the truth about what he said to Chariclea,
but not necessarily that what he told her was true. In formal terms
his tertiary narrative is equivalent to other deceptions he employs to
achieve his god-sanctioned goals, such as the charlatanry about the Evil
Eye, the blatantly wrong dream-interpretation offered Charicles, or the
various instances of hocus-pocus he employs to represent himself as a
practitioner of Egyptian wisdom.22 Like Odysseus, with whom he is so
programmatically identified, Calasiris is deceitful as an actor but not as
a narrator.23
The other aspect of Calasiris’ secondary narrative which requires
comment is the close interplay between narrator and narratee.24 The
narrative is elicited by Cnemon’s curiosity, even before he realizes that
Calasiris has any connection with his absent friends. Once he knows
of that connection, Cnemon demands the narrative as payment for
news about Calasiris’ ‘children’, but he is motivated as much by the
anticipation of pleasure as by concern for his friends.25 In the course of
his narrative, Calasiris repeatedly addresses Cnemon by name, and the
primary narrator repeatedly records the secondary narratee’s responses
and reactions. The curiosity that led Cnemon to request the story in
the first place drives him on through the narrative: he is impatient to
know the outcome and frustrated by delays:
22 Respectively 3.7.2–8.2; 4.15.1–3; 3.17.1–3; 4.5.3; 4.7.12–13.

23 This reading is confirmed by the fact when the action reaches Meroe, Persinna
shows no sign of expecting Chariclea’s return and does nothing to exculpate Calasiris
from Charicles’ accusation that he is a false prophet who collaborated in a criminal act.
24 This is analysed in detail by Morgan 1991: 95–100; in addition to Winkler, there
are important discussions in Hardie 1998 and Hunter 1998a.

25 At 2.23.5 he associates his desire for stories with the god Dionysus, who ‘has taken
up residence in him’ (in other words he has a drink or two), and envisages the narration
as a dramatic event.
‘It is not surprising,’ interrupted Cnemon, ‘that those who were there
watching should have felt the suspense; even now my heart goes out
to Theagenes. So, I beg you, make haste and tell me whether he was
proclaimed the victor.’ (4.3.4)26
He derives such pleasure from the narrative that he becomes immune
to weariness,27 and he only agrees to the adjournment of the narrative
when external interruptions cannot be ignored (5.1.4). His pleasure is
inextricably linked with the romantic nature of the story:
I cannot agree with Homer, father, when he says that there is satiety of
all things, including love. In my estimation, one can never have a surfeit
of love, whether one is engaged in its pleasures or listening to tales of
it. And if the story being told is the love of Theagenes and Chariclea,
who could be so insensitive, so steely-hearted, that he would not be
spellbound by the tale, even if it lasted a whole year? (4.4.3)
But while he is sometimes impatient, he also sometimes retards the
narrative to elicit a larger dose of enargeia. He wants the illusion of being
present, and compels Calasiris to paint a word-picture of the Delphic
procession and even give a verbatim rendition of the processional hymn:
You have not yet described them so that I can see them for myself. Your
story has me in its power, body and soul, and I cannot wait to have the
pageant pass before my very eyes. Yet you hurry past without a second
thought. I feel like the proverbial guest who has turned up too late for the
feast! You have rung the curtain up and brought it down again all in one
phrase … you are trying to cheat me of the best part of the story by not
giving me all the details of the hymn. It is as if you had only provided me
with a view of the procession, without my being able to hear anything.
(3.1.1; 3.2.3)
Some of his interventions voice his delight in vivid and accurate de-
scription, as with Calasiris’ description of Delphi (2.26.1–2) or of the
protagonists:
Your description portrayed them so vividly (enargōs), so exactly as I know
them from my own experience, that they seemed to be before my eyes.
(3.4.7)
26 At 2.32.3 he barges through two narrative levels to interrupt Charicles’ tertiary
narration, at the point where the Ethiopian ambassador departs before telling Char-
icles the truth about Chariclea’s origins. He expresses his disappointment and seeks
Calasiris’ reassurance that the truth will be told in due course.
27 At 3.4.4 dialogue between Calasiris and Cnemon marks the passing of time; at
4.4.2 Calasiris remarks that he is ‘proof against sleep’.

On other occasions, Cnemon demands encyclopedic digressions,

pressing Calasiris on the Homeric method of recognizing gods (3.12.1)
or the geography of the Gulf of Corinth (5.17.2). But while eager for
detail, he has a clear sense of priorities, suppressing questions and
keeping Calasiris to the matter in hand, even when the latter mentions
names—Thisbe, Thyamis—which are of great interest to his narratee.
This sense of what is important and what is secondary also ostensibly
characterizes Calasiris as a narrator:
The last thing I want to do, Cnemon, is to bore you with such irrele-
vancies (tois exōthen), so I was confining myself to the central theme (ta
kairiōtera) of my tale and the answers to your original questions.
(3.1.2)28
However, the irrelevancies that Calasiris here proposes to omit include
the Delphic rite at which Theagenes and Chariclea fell in love at first
sight. This is hardly irrelevant to any conception of the central theme,
and it could not have been Calasiris’ intention to omit it. Unless the
author is not fully in control of his text, we must assume that Calasiris
is provoking Cnemon to ask for a more expansive treatment, as he
certainly does elsewhere, dropping tantalizing references to ‘Egyptian
Homer’ into his narrative, a bait that Cnemon resists a few times before
he eventually swallows it.29
Calasiris the narrator of course knows more than Calasiris the char-
acter knew, and generally he observes the protocol of not exploiting
knowledge gained subsequently. Attention is even drawn to the conven-
tion, when Calasiris is at a loss to understand the instructions given him
by Apollo and Artemis; Cnemon interrupts: ‘I am sure you learned the
answer afterwards, father, and will tell me in due course’ (3.12.1). He
also observes the documentary convention to the extent of recording an
action in its proper chronological place but noting that he only came
to know of it later (4.15.4). More frequently he narrates in an inferential
way, avoiding omniscience about, for example, other people’s thoughts
and emotions:
28 Cf. 3.10.3 ‘there is no point in boring you with all the details …’, a bait to which
Cnemon does not rise.

29 Although Calasiris is in a hurry to get to the end of the story so that he can hear
Cnemon’s news of Theagenes and Chariclea, he also has a psychological compulsion to

tell his tale to someone, that existed even before he knew that Cnemon was acquainted
with his ‘children’ (2.21.6).
My words were followed by a short silence, but the look in her eyes was
enough to tell me that her mind was a turmoil of shifting emotions.
(4.6.1)30
Nevertheless, he sometimes conceals from his narratee thoughts and
knowledge that he had at the time of the action, and which are thus
perfectly narratable. At 3.18.3 he says that he dissembled to Charicles,
acknowledging that he is now concealing his true thoughts also from his
narratee. A little later (4.5.1) he tells Cnemon that he had already begun
to form suspicions about Chariclea’s parentage, but does not reveal
what they were or how they formed him. This reticence heightens the
dramatic effect of the revelation about Chariclea’s birth when Calasiris
reads the message from her mother.
Calasiris shares some features of the primary narrator. He too uses
normative sententiae both to explain the action and create an affec-
tive community with his narratee. Because he is Egyptian, but mostly
because of the subject matter of his narrative, his generalizations tend
to concern erotic psychology rather than Hellenic identity:
The mind of a person in love is rather like that of a drunkard: volatile
and completely unstable, since in both cases the soul is riding on a tide of
emotional fluidity, which is why lovers are prone to heavy drinking and
drunkards to falling in love. (3.10.5)31
Like the primary narrator he is aphoristic and stylistically self-con-
scious:
The following day was the last of the Pythian tournament, but for the
young couple another tournament was still at its height, one presided
over and refereed, it seems to me, by Love, who was determined to use
these two contestants, in the only match he had arranged, to prove that
his particular tournament is the greatest of all. (4.1.1)
As a very Hellenized Egyptian, he can quote Homer directly (3.4.1),
compare Theagenes explicitly to the Homeric Achilles (4.3.1) and al-
lude to other Greek literary texts.32
However, Calasiris narrates as Egyptian wise man. This persona is
established right at the start of his narrative:
30 Thus the central scene of inamoration is restricted to Calasiris’ observations and
inferences (3.5.4–7).
31 Cf. 3.10; 4.1.2, 3.2, 4.4.; non-erotic sententiae at 2.24.7; 3.4.8; 5.25.2, 3.
32 Calasiris’ narrative includes consciously intended allusions to both the Iliad and
the Odyssey, to Moschus and Euripides.

After [my wife’s] release from this life, I lived untroubled for some time,
complacently proud of the two sons she had borne me; but before many
years had passed the pre-ordained celestial cycle of the stars turned the
wheel of our fortunes; the eye of Cronus lit upon my house and brought
a change for the worse. My science had given me warning of this, but not
the ability to escape it, for while it is possible to foresee the immutable
dispensations of fate, it is not permitted to evade them. (2.24.6)
Encyclopedic digressions on matters such as the Nile floods (2.28),
Homer’s Egyptian background, and the nature of the gods’ appearance
(3.13) confirm and renew it. It is his wisdom that enables him to
form suspicions about Chariclea’s true identity before, as it were, the
correct time to narrate them to someone less wise than he. However,
his holiness produces a potential mismatch with his narratee. Briefly
put, the story Cnemon wishes to hear is not necessarily the one that
Calasiris wishes to tell. Cnemon’s ‘interest in incidental spectacle’ leads
Calasiris to call him a ‘true Athenian’ (3.1.2) and in the exchange
about divine epiphanies in Homer,33 Cnemon cheerfully classes himself
among ‘the ignorant majority’ who fail to read beyond the surface of
the text (3.12.3). The climax of Calasiris’ narrative is the moment when
he reads Persinna’s message to her daughter; for him it is a moment
of religious epiphany, and there is latent in the whole of Calasiris’
narrative the sense that for him this is more of a religious story than
a romantic one, both in the way he understands and in the way he tells
it:
On reading this, Cnemon, I perceived the hand of the gods and mar-
velled at the subtlety of their governance. I was filled with a mixture of
pleasure and sadness, and had the peculiar experience of being moved
simultaneously to joy and tears. My heart was thankful that the mystery
had been explained, that the riddle of the oracle had been solved, but it
was sorely troubled about the course the future might take and filled with
pity for the life of man, whose instability and insecurity, whose constant
changes of direction were made all too manifest in the story of Chari-
clea. (4.9.1)34
33 The secondary narratee takes this stuff seriously, but the primary narratee surely
does not, the exegesis of Iliad 13.71–72 at 3.12–13 being comically forced. The question
of whether Calasiris himself takes it seriously is left open, but the primary narratee is, I
think, intended to see that Calasiris is having fun at Cnemon’s expense, just as, within
his narrative, he acts and speaks duplicitously to a number of other characters. But as
this is a digression from his narrative, it does not invalidate the principle that the facts
of his narration are to be taken as true.
34 This is argued in more detail in Winkler 1982; one does not need to accept the
whole of Winkler’s thesis to agree with this point.

Cnemon’s responses and interventions, on the other hand, concern

spectacle and emotion. What he says of his knowledge of Homer can
well describe his understanding of Calasiris’ deeper meanings:
I know that I was taught the superficial purport of the lines when the
text was first expounded to me, but I am totally unaware of the religious
teaching embedded in them. (3.12.3)
If Calasiris’ secondary narrative formally serves an explanatory func-
tion, Cnemon’s is distractive, in two senses. For the secondary narra-
tees, Theagenes and Chariclea, ‘a story of woes like their own would
be a great consolation’ (1.9.1.), and divert them from a night of tears; in
fact at the end of Cnemon’s narrative they still weep for their own dis-
tress. For the primary narratee Cnemon’s narrative is a postponement;
at this stage he is being tantalized with hints and scraps of information
about the protagonists; it is clear that the primary narrator is not going
to explain who they are, and also, from two failed attempts to com-
municate with Egyptians, that the presence of a Greek-speaking audi-
ence is needed before the protagonists can make any revelations about
themselves. When the Egyptian robber-captain assigns Theagenes and
Chariclea to the care of a young Greek captive, ‘so that they might have
someone to talk to’ (1.7.3), the expectation is that this will allow them
to make their revelations. However, they end up as embedded narratees
rather than embedded narrators, and the primary narratee must wait
for Calasiris’ narrative.35
Cnemon introduces his narrative in tragic terms:
Why do you batter and prise open these doors, to borrow a phrase from
the tragedians? This is no time to introduce a new theme into your own
tragedy in the form of my misfortunes. (1.8.7)36
A careful reading of his story, however, suggests that his portrayal of
himself in tragic terms is at variance with the comic intrigue that forms
his narrative.37 Like Clitophon in Achilles Tatius, he is an internal
narrator, whom the primary narratee must learn to interpret. It is not
just that he is understandably partial, leaving no doubt as to where
his perceptions of good and evil lie; his whole take on his story is
35 This and other approaches to Cnemon’s narrative are discussed in Morgan 1989b.
36 Cf. 2.11.1, where he fears that Thisbe has come to make him ‘the victim of another
Attic tragedy, but in an Egyptian setting’.
37 Paulsen 1992: 85–102 is excellent on this point. Even Cnemon’s name marks him
out as a figure from the comic stage.

different from the one which the primary narratee is invited to take by
the novel’s system of dramatic imagery and allusion.
Although they interrupt him less than he will interrupt Calasiris,
Cnemon engages constantly with his narratees, addressing them in the
second person plural, omitting boring details (1.9.4), worrying about
their need for sleep (1.14.2). Exclamations display his emotional invest-
ment in re-enacting past experience:
With these words I stepped forward to dispatch the pair of them. But—o
gods!—it was my father who slid from the bed and fell at my feet.
(1.12.3)
An interestingly piquant narrative situation arises much later in the

novel (6.2–3), when Cnemon tells the same story to Nausicles to provide
a diversion during a journey.38 The primary narrator presents a sum-
mary in indirect speech, recapitulating the story with a minimal surplus
of new information. The spice here is that the narratee of this occluded
secondary narrative is an internal one, but the equally internal narrator
does not realize it; it is only later that he learns that Thisbe’s anony-
mous lover was Nausicles. The complex reactions of this unrecognised
internal narratee are written out in full in the text:
Cnemon’s story left Nausicles in a state of utter indecision, by turns
disposed to admit the truth about himself and Thisbe and then inclined
to put off doing so to another time. In the end he kept his own counsel,
though only with difficulty, partly by his own decision, partly because
something else occurred to prevent his telling all. (6.3.1)
The main importance of Cnemon’s story, however, is as analogue and

antithesis to the primary narrative.39 Thematically, Cnemon’s story—
one of illicit and degrading sexual attractions, egocentric intrigues,
meaningless casual erotic and merely physical extra-marital liaisons,
relationships based on inequalities of power and status—counterpoints,
detail by detail, the ideal love story of the protagonists, whose love is
mutual, exclusive, chaste, selfless, spiritual, permanent, and profoundly
meaningful. The figure of Thisbe in particular is constructed as Char-
iclea’s antitype, and the plot repeatedly contrives situations where the
one is somehow exchanged for the other. This moral antithesis between
the base love of Cnemon’s Athens and the ideal love of Theagenes
38 Note that Cnemon can tell the same story for radically different purposes; in
neither is he successful.
39 The theme of this paragraph is treated at greater length by Morgan 1989b.
and Chariclea, between the world of the secondary narrative and the
direction of the primary narrative, represents the novel’s most profound
statement of values.
Among other embedded narratives, we must note Chariclea’s lying
account of her origins in response to Thyamis’ proposal of marriage
(1.21–22). This has an ‘argument’ function, to buy time by getting him
to agree to a postponement, and to safeguard the protagonists’ posi-
tion by convincing him that she and Theagenes are brother and sister.
Chariclea is (supposedly) an internal narrator, and Thyamis is (truly) an
internal narratee, since his discovery of the couple on the beach forms
the last detail of the narrative. It is this appeal to the narratee’s own
knowledge that lends the whole narrative its credibility for him. Chari-
clea’s performance of the narrative is scripted with unusual precision:
For a while she stood with her eyes fixed on the ground, repeatedly
shaking her head, apparently gathering her thoughts to say something.
Eventually she looked Thyamis full in the face. Her beauty dazzled him
even more now, for her reflections had brought a special blush to her
cheeks and there was fire in her eyes. With Cnemon interpreting she said
… (1.21)
Her body language, including the tears with which she concludes, is
integral to her persuasion. The effect of this narrative on the primary
narratee is interesting. At this early stage, he probably realizes that
Theagenes and Chariclea are the protagonists, and suspects that, by the
rules of the genre, they are lovers; but as yet no firm information has
been forthcoming. The primary narratee thus cannot judge the truth
or otherwise of Chariclea’s narrative; its apparently solid detail must
prompt at least a reconsideration as to whether the generic assumptions
really will apply in this novel. It is only a few chapters later that the
mendacity of the narrative is definitively established, as so often without
the intervention of the primary narrator.
Conclusion
The Ethiopian story is a novel in which much narrating is done. This is

clearly the product of its plot-structure, but, equally clearly, Heliodorus
is interested in the narrative act for itself. The secondary narrative of
Calasiris in particular explores the interaction of narrator and narra-
tee, and it is difficult not to conclude that a paradigm of some sort
is being offered to the primary narratee for his relation to the novel,
though whether he sees Cnemon as a qualified portrait of an ideal

narratee or as an ironically distanced deterrent example is a ques-
tion still open to argument. The point perhaps is that each view is
both correct and wrong; the mise en scène is contrived to make the pri-
mary narratee aware that there is an issue concerning the correct way
to respond to narrative, to distinguish between central and incidental,
between superficial and profound understandings. Cnemon’s sympathy
and engagement are hardly things that the primary narratee will want
to abjure, but they may not be the complete or exclusive qualifications
for a good reader of a romantic novel. Similarly, the plot’s emphasis on
cognition and interpretation is enacted at all its narrative levels. Both
Cnemon’s and Calasiris’ narratives are introduced in such a way as to
highlight the hermeneutic problems they pose. Calasiris’ account of his
own re-assembly of Chariclea’s true story from multiple fragments mir-
rors the primary narratee’s activity in synthesizing the whole plot of the
Ethiopian Story out of its multiple, sometimes partial and limited, some-
times deliberately elusive component narratives. Above all, Heliodorus
presents a veritable menagerie of narrating specimens, and compels his
reader to consider and compare their forms, functions and effects; the
primary narrator and narratee are not exempted from this scrutiny: as
the author distinguishes himself from the primary narrator, the reader
must think what it is to be a primary narratee. If we are looking for a
narratological study from the ancient world, we are more likely to find
it in Heliodorus’ novel than anywhere else.
epilogue

I.J.F. de Jong and R. Nünlist
Narratology and genre
One important conclusion that emerges from the preceding overview

of narrators in ancient Greek literature is that there is no direct corre-
lation between genre and type of narrator. For example, epic narrators,
who by virtue of their subject matter are external, may be either covert
(Homer, Homeric hymns) or overt (Theogony, Apollonius of Rhodes).
Narrators of historiographical and biographical texts can be either
external (the majority) or—partially—internal (Thucydides, Polybius,
Cassius Dio). Similarly, in the novels we find both external (Chariton)
and internal (Achilles Tatius) narrators, as well as overt (Chariton) and
covert (Xenophon of Ephesus) ones.
Nor does single authorship necessarily mean a preference for one
type of narrator. The Hesiodic narrator is external in the Theogony
(excluding the introductory ‘Hymn to the Muses’) and internal in Works
& Days. The Platonic dialogues feature both primary and secondary
narrators (depending on whether the dialogue has a narrative frame,
or is purely dialogic), and both overt and covert narrators (depending
on whether it is Socrates or one of his disciples who is reporting). The-
ocritus and Callimachus produced poems with a narrative frame and
purely ‘mimetic’ ones. And a disparate corpus such as that of Lucian
contains virtually every conceivable type of narrator. Within a text, the
narrator’s status may be constant, but one may also see a shift from
external to internal (Thucydides, Polybius, Cassius Dio),1 or an alterna-
1 Admittedly, in the case of historians this is not really a matter of choice. As soon as
they embark on contemporary history, they immediately become potentially internal.

546 part nine – epilogue
tion between overt and covert (e.g. Appian and Herodian). Similarly, in
the course of his narrative the secondary narrator Clitophon (Achilles
Tatius) displays a transition from restricted knowledge to omniscience.
That the choice of narrator is relatively independent of the genre
can also be deduced from the fact that several a priori non-narrative
genres such as drama, oratory, and choral lyric nevertheless tend to
incorporate narrative, sometimes extensively.
Consequently, one of the more striking insights gained by the exer-
cise undertaken in this volume is that, from a narratological point of
view, lines can be drawn between various authors and texts which dif-
fer considerably from those normally drawn on the basis of a generic
approach. To give only a few examples: the narrator and narrative style
in Herodotus are not only strongly indebted to Homer, but also dis-
play features of the contemporaneous scientific, epideictic genre. The
narrator in Apollonius of Rhodes, who has thus far been brought into
connection with Homer and drama, reveals upon closer narratological
examination a surprisingly Herodotean outlook. More in general, the
influence exercised by the Herodotean narrative style is much greater
than is generally acknowledged: not only Xenophon (in his historio-
graphical and biographical works) and Plutarch are indebted to him,
but also Aristides, Dio Chrysostom, and Lucian.
This said, it is nevertheless possible to point to a few tendencies. His-
toriography, biography, and choral lyric tend to have external narrators
(they recount stories from the past) who are overt (they want to teach
a lesson); forensic oratory and drama have narrators who are internal
(they are to a greater or lesser extent—and in the case of oratory even
critically—involved in the events) and overt (they want, indeed need, to
persuade their addressees).
Narrators
Leaving aside generic considerations and looking at the types of nar-

rators as they are found in the texts, the following observations can be
made. Among the primary narrators, by far the most frequent type is
the external and overt narrator. The overtness usually takes the form of
narratorial interventions, not dramatization; no primary external nar-
rator ever develops a full-fledged personality.
Next in frequency among the primary narrators is the internal nar-
rator, whose prominence can vary considerably: he may be a protago-
de jong & nünlist – narrators, narratees, and narratives 547
nist (e.g. the Socratic narrator or the narrator in oratory),2 a more or

less active participant in, and witness to, contemporary history (overt:
Polybius, Cassius Dio, Herodian, covert: Thucydides), or a silent inter-
locutor and witness to a philosophical dialogue (the disciple narrator in
Plato and Xenophon). Special cases are the internal narrator of Achilles
Tatius, who disappears from the stage after a couple of paragraphs
never to return again, his place being taken by another—secondary—
narrator, and Longus, where again a frame narrative quickly yields to a
secondary narrative, one in which the primary narrator remains centre
stage, only switching from being internal to external.
The least frequent among the primary narrators is external and
covert (Homer, the Homeric Hymns, Moschus, Xenophon of Ephesus).
This observation leads to the somewhat paradoxical conclusion that
at this point the most influential storyteller of Greco-Roman antiquity,
Homer, has not set the tone.
Interestingly enough, the signs by which all these types of narrat-
ors—explicitly or implicitly—reveal their presence as narrating agents
are universal and not restricted to a single genre, though they may be
exploited for different effects in the various genres: first-person refer-
ences, references to the narrator’s own time, apostrophes, gnomic utter-
ances, comments (both metanarrative or evaluative), ‘if not’ situations,
interactional particles, cross-references, prolepsis and analepsis, descrip-
tions, similes, suggestive juxtaposition, motifs, proofs, visualization, and
appeals to what is natural.
Finally, it is worth remembering (→ Introduction) that there is one
group of texts which are best narratologically analysed if one assumes
a suppressed primary narrator: the mimetic dialogues of Plato and
Lucian, the mimetic hymns of Callimachus, and the mimetic idylls
of Theocritus. Their closest parallels in this respect are such texts as
Achilles Tatius and Longus, where the embedded narrative predomi-
nates, and the primary narrative provides only an introductory frame,
which may or may not be resumed at the end. Conversely, a primary
narrator may display a general reluctance to make use of secondary
narrative (Hesiod, Homeric Hymns, Apollonius of Rhodes, Arrian,
2 From a modern perspective it is perhaps surprising that the corpus under con-
sideration does not contain a single example of a text with primary internal narration
from beginning to end, in the style of Moby Dick or A la recherche du temps perdu. The
closest ancient parallel seems to be Augustus’ Res Gestae. This type of narrative can be
found in pre-Greek literature (e.g. Hittite).
Appian, the orators). An intermediate position is occupied by texts such

as Homer’s, where the primary narrator predominates, but which nev-
ertheless incorporates numerous extended secondary narratives.
When we turn to secondary narrators, it is more difficult to draw
clear quantitative conclusions. Here we should keep in mind that every
narrative text has one primary narrator but often several secondary
narrators. We find external and internal narrators in about equal num-
bers, whereby Homer and Herodotus display a predominance of exter-
nal narrators, who recall events from the mythic past by means of paral-
lels to the present, while drama features mainly internal narrators. But
there are no strict rules, since we find the occasional internal narrator
in Homer (Achilles, Odysseus) and Herodotus (Cambyses recounting
why he—mistakenly, as he now realizes—killed his brother Smerdis), or
an external narrator in drama (the chorus). All of these secondary nar-
rators, whether external or internal, are by definition dramatized: they
have a personality, in that each is a character in the primary narrative
or the dramatic universe. However, being an internal narrator does not
automatically mean that one is an overt narrator. This may be the case
(Clitophon in Achilles Tatius, the oratorial narrators), but the messen-
gers of the Attic stage are highly invisible as characters in their own
stories (something which allows them to see things hidden to others),
but highly overt as narrators, commenting on what they are reporting
and expressing their emotions.
In addition to secondary narrators, certain texts favour reported nar-
rators: collective or anonymous spokesmen whose stories are quoted
by the narrator in indirect speech (‘the Spartans say’, ‘they say’). Not
surprisingly, we encounter these reported narrators with particular fre-
quency in historiographical, biographical, and rhetorical narratives,
where the primary narrator feels the need to authenticate his own story,
or disqualify other people’s stories, by introducing sources, authorities,
alternative versions, witnesses and the like.
A final device which enables authors of narrative texts to introduce
some sort of voice is the narratorial alter ego: characters, who, because
of their profession, function, opinions, name, etc. invite a comparison
with the primary narrator. Such alter egos are found in Homer (the
singers Phemius and Demodocus), Apollonius of Rhodes (the singer
Orpheus and the seer Phineus), Herodotus (the wise advisers Solon,
Artabanus, Demaratus, the incredulous enquirer Polycrates, and the
historian Hecataeus), Polybius (the character ‘Polybius’), and Lucian
(Lycinus, Momus, Parrhesiades, etc.).
All types of narrators, whether primary or secondary, internal or

external, may at times turn to second-person narration. As noted in
the Introduction (→), this is in general a rare phenomenon, but it
does occur with some frequency in ancient Greek narrative, mainly in
the honorific context of the apostrophe of heroes in Homer and—less
often—Apollonius of Rhodes, and of gods in the Homeric and Calli-
machean hymns, and in the accusatory context of narrators reminding
their internal narratees of what they have done (notably Menelaus and
Helen in Odyssey 4, ‘Hesiod’ and Perses in the Works and Days, and Hyl-
lus and Deanira in Sophocles Trachiniae). We find a special and rather
eerie use of second-person narration in Cassandra’s extended apostro-
phes of Troy and of certain of its inhabitants (Paris, Hector), doomed to
be destroyed or die, in Lycophron.
Considering the power of the spoken word in antiquity, it is hardly
surprising that almost all primary narrators present themselves as
speakers rather than writers. In the cases where the texts are likely
to have been intended for a reading audience (Apollonius of Rhodes,
Callimachus), we are perhaps dealing with the device of feigned oral-
ity.3 Embedded narratives also tend to be oral, but documents such as
written sources or letters do occasionally occur (Thucydides, Polybius,
Philostratus, novels).
Narratees
The primary narratees of ancient Greek narrative are generally covert

and considerably less visible than the narrators. They are often anony-
mous, unspecified as to gender, age, or even home city, and only rarely
explicitly addressed. Hardly ever do narratees have an actual name
(Perses in Hesiod’s Works and Days, the victors of the epinician odes, the
dedicatees of Theocritus or Philostratus), or a more or less specific iden-
tity (the Athenian jurors). This unemphatic and largely unspecified sta-
tus of the narratee, which may be seen as the counterpart of the avoid-
ance to dramatize the external narrator, may perhaps be due to the
largely public nature of ancient Greek literature: the audience was—at
any rate in pre-hellenistic times—physically present, which invited an
3 As a minor caveat one should perhaps take into account that even modern
authors, who clearly write for readers, frequently present their narrators as telling a
story.
immediate identification of audience and narratee. At the same time,

too specific a characterization of the narratee might have led to an
undesirable discrepancy with the actual audience. On a slightly differ-
ent level, the unspecified status of the narratee may also have to do
with the universalistic claim of the narrators (first witnessed in Homer):
they often envisage the whole world as their intended or at least poten-
tial recipient. This claim is particularly palpable in the Greek authors
writing under the Roman empire, who hope to serve both Greek and
Roman audiences. In any case, the texts examined in this volume seem
to indicate that ancient Greek authors were reluctant to make extensive
use of the possibilities inherent in a more specific narratee (contrast, for
example, Tristram Shandy or Tom Jones).
Though largely invisible, these narratees are nevertheless implicitly
present as the beneficiaries of the narrator’s story, indeed of the many
devices which he employs to tell his story in the precise manner and
with the precise effect he wants. In addition to the signs of the nar-
rator already discussed, which in their effect are of course all aimed
at the narratees, there are numerous less prominent indications which
testify to the communication between narrator and narratee: rhetori-
cal questions, the ‘presentation through negation’ device, explanations
(often in the form of gar-clauses), the ‘indefinite second person’ device
(‘there you/someone could have seen’), the ‘anonymous witness’ device
(‘this is what each citizen said as he saw the Argonauts rushing for-
ward with their weapons’) and the ‘anonymous interlocutor’ device (‘if
someone would desire more exact measurements …’). Certain narra-
tors even engage their narratees in a more active participation in the
(re)construction of the story, which thereby becomes part of their ‘edu-
cation’.
As has been argued in the introduction to this volume, narratees,
both primary and secondary, are important instruments by which a
narrator steers the intellectual, moral, and emotional reception of his
story. Sometimes this works in a straightforward manner, the narratees
being obviously set up as models for the readers (e.g. the secondary nar-
ratees of Homer, and the narratees of Chariton, who obviously share
the beliefs and knowledge of the narrator). But sometimes the process is
more complicated, in that the text constructs a narratee who is incredu-
lous, misinformed, philosophically uneducated, or even downright crit-
ical or dissonant, and hence still must be persuaded or educated by the
narrator. Examples include: Perses in Works and Days, the narratees in
Plutarch and in Xenophon’s works, both historical and philosophical.
Needless to say, the construction of such dissonant narratees forms part

of the text’s rhetoric, in that they provide a negative foil for an ideal pri-
mary narratee. As such they steer the reception in an equally efficient
way.
Another complication—and source of narrative exploitation—is the
fact that primary and secondary narratees need not agree in their reac-
tion to, and evaluation of, the story: the primary narratees may know
or understand more about what is happening than the characters, and
this discrepant awareness often results in pathos or dramatic irony. A
related phenomenon, found mainly in drama, is the case of multiple
narratees, when the same story has a different effect on different recipi-
ents. Here again, the primary narratees are led to compare the various
reactions and in the end distil their own.
The functions of storytelling
From the intended recipients of narrative it is a small step to the

question of why stories are told in the first place. Here it is easiest to
start with embedded narratives, because the reason for telling them is
often explicitly indicated.
Only rarely are embedded stories narrated solely with a view to
entertainment or in order to ‘kill time’ (Theocritus, some of the Home-
ric instances, although these always have a key function for the pri-
mary narratees). Much more often they serve to inform characters
(Xenophon Anabasis and Hellenica, Chariton, and most dramatically
Cassandra in Lycophron) or primary narratees (Homeric Hymns, Apol-
lonius of Rhodes, Callimachus, where the embedded narratives convey
aetiological, mythological, or descriptive material; Homer, Pindar, Bac-
chylides, and Euripidean prologues and epilogues, sketching the events
which preceded or followed the main story); in other words, here the
embedded narratives function as analepses or prolepses which supple-
ment the main story. A special category is formed by the false stories of
Homer, Sophocles, and Aristophanes, which serve to mislead charac-
ters (while the primary narratees are expected to note the discrepancy).
A very common function is that of the paradigm or instruction: events
from the past are held up like a mirror, in order to persuade or dis-
suade. Though the interaction here primarily concerns the characters
within the story, it is clear that the primary narratees are also meant
to benefit from this form of instruction. As primary narratees, they
are privileged in that they have a multiple perspective: they both hear
the paradigmatic story and witness the reaction of the characters, who
may or may not get the message. By and large, the points made about
embedded narratives also apply to narrative texts in general. The genre
which comes closest to being written primarily with a view to entertain-
ment is perhaps the novel; the other narrative genres and texts show a
stronger tendency to inform or even instruct the audience in one way
or another.
A history of ancient Greek narrative
Taking a final, bird’s eye view of the history of narrators, narratees, and
narratives in ancient Greek literature, it is clear that we are not dealing
with a development in the sense of a steadily increasing refinement and
expansion of what was initially a modest set of simple instruments. The
history of Greek literature begins with a ‘big bang’, in that the first texts
we have, the Homeric epics, display much of the narratorial repertoire
and handle it in a virtuoso manner, while at its end the relatively simple
Chariton and the extremely sophisticated Heliodorus exist side by side.
Nevertheless—and though being aware of the teleological fallacy—
it is possible to see certain trends in the course of the centuries. On
the formal level, there seems to be a tendency towards experimen-
tation with the number of narrative levels (e.g. Plato’s Symposium and
Lycophron’s Cassandra with four narrative levels each). A similar devel-
opment may be seen when the distinction between the different narra-
tive voices is deliberately blurred, for example, when the primary nar-
rator usurps a character’s story (Apollonius of Rhodes and Jason) or
interferes in the stories of secondary or reported narrators (Herodotus,
Thucydides). This whole question must also be seen in connection
with the boundaries of narrative. Clear demarcation (e.g. Homer, tragic
messengers, and the orators) contrasts with blurred demarcation (e.g.
Pindar and Aristophanes), while Cassandra again forms an experimen-
tal extreme. Aristophanes’ case also shows the development of heavily
punctuated narrative taking the place of uninterrupted narrative.
On the conceptual level, there seems to be an increased problema-
tization of the narrator’s omniscience and thus his narrative author-
ity. While almost taken for granted by early Greek poets as part of
the Muses’ inspiration, omniscience becomes qualified and modified by
historiography and other types of scientific writing, in which narrators
authenticate their accounts and attest to the limitations of their knowl-

edge. One result is the mildly paradoxical situation of a poet, Apollo-
nius of Rhodes, who invokes the Muses, but nevertheless qualifies his
omniscience. Another is the downright dismissal of autopsy and other
forms of authentication (Lucian). A third is the ironic undermining of
an omniscient narrator (Longus).
All in all, it is best not to try to ‘emplot’ the story of ancient Greek
narrative in terms of progress or decline, but rather, borrowing Quintil-
ian’s metaphor of Homer, to see it as an ocean, a large reservoir from
which streams draw their water or to which they return it again. It
thereby increases or decreases but flows neither forward nor backward.
It is to be expected that the later volumes of the series will further sup-
plement and refine this slowly evolving picture of the ‘ocean’ of ancient
Greek narrative techniques.
BIBLIOGRAPHY
Adams, C.D., The Speeches of Aeschines (Cambridge Mass./London 1919).

Albis, R.V., Poet and Audience in the Argonautica of Apollonius (Lanham 1996).
Alexiou, E., ‘Die Funktion der Homerzitate in Plutarchs Biographien’, in
A. Haltenhoff, F.-H. Mutschler (eds.), Hortus litterarum antiquarum: Festschrift
für Hans Armin Gärtner zum 70. Geburtstag (Heidelberg 2000) 51–65.
Alföldy, G., Die Krise des römischen Reiches (Stuttgart 1989).
Andersen, O., ‘Myth, Paradigm, and “Spatial Form” in the Iliad’, in J.M. Bre-
mer, I.J.F. de Jong, J. Kalff (eds.), Homer, Beyond Oral Poetry. Recent Directions in
Homeric Interpretation (Amsterdam 1987a) 1–13.
———, ‘The Making of the Past in the Iliad’, HSPh 93 (1987b) 25–45.
Anderson, G., Philostratus: Biography and Belles-Lettres in the Third Century A.D.
(London 1986).
———, ‘Philostratus on Apollonius of Tyana: the Unpredictable on the Unfath-
omable’, in G. Schmeling (ed.), The Novel in the Ancient World (Leiden 1996)
613–618.
———, ‘Some Uses of Storytelling in Dio’, in S. Swain (ed.), Dio Chrysostom:
Politics, Letters, and Philosophy (Oxford 2000) 143–160.
Anderson, J.K., Xenophon (London 1974).
Andrews, N.E., ‘Narrative and Allusion in Theocritus, Idyll 2’, in M.A. Har-
der, R.F. Regtuit, G.C. Wakker (eds.), Theocritus (Groningen 1996) 21–53.
Asper, M., Onomata allotria: zur Genese, Struktur und Funktion poetologischer Metaphern
bei Kallimachos (Stuttgart 1997).
Badian, E., From Plataea to Potidaea: Studies in the History and Historiography of the
Pentecontaetia (Baltimore 1993).
Bain, D., Actors and Audience. A Study of Asides and Related Conventions in Greek Drama
(Oxford 1977).
Bakker, E.J., Poetry in Speech. Orality and Homeric Discourse (Ithaca N.Y. 1997).
Bal, M., Narratology. Introduction to the Theory of Narrative (Toronto–London [1985]
1997).
Ballabriga, A., ‘Le deutéro-Hésiode et la consécration de l’hésiodisme’, in Le
métier du mythe. Lectures d’Hésiode (Lille 1996) 71–82.
Barber, G.L., The Historian Ephorus (Cambridge 1935).
Barber, G., Daphnis and Chloe. The Markets and Metamorphoses of an Unknown
Bestseller (London 1989).
Barrett, J., Staged Narrative: Poetics and the Messenger in Greek Tragedy (Berkeley
2002).
Barth, H., ‘Zur Bewertung und Auswahl des Stoffes durch Herodot (Die
Begriffe thôma, thômazô, thômasios, thômastos)’, Clio 50 (1968) 93–110.
Barthes, R., ‘An Introduction to the Structural Analysis of Narrative’, in
S. Heath (ed.), Image—Music—Text ([1966] London 1977) 79–124.
556 bibliography
———, ‘Historical Discourse’, in M. Lane (ed.), Introduction to Structuralism

([1967] New York 1970) 145–155.
Bartsch, S., Decoding the Ancient Novel. The Reader and the Role of Description in
Heliodorus and Achilles Tatius (Princeton 1989).
Baumbach, M. ‘Die Meroe-Episode in Heliodors Aithiopika’, Rheinisches Museum
140 (1997) 333–341.
Baur, D., ‘Chor und Theater: Zur Rolle des Chores in der griechischen Tra-
gödie unter besonderer Berücksichtigung von Euripides’ Elektra’, Poetica 29
(1997) 26–47.
Behr, C.A., Aelius Aristides and the Sacred Tales (Amsterdam 1968).
———, P. Aelius Aristides, The Complete Works, Vol. 2: Orations xvii–liii (Leiden 1981).
Belardinelli, A.M., Menandro. Sicioni (Bari 1994).
Belloni, L., Eschilo: I Persiani, 2nd ed. (Milano 1994).
Berge, H.M. ten, Antiphon’s Zesde Rede (Nijmegen 1948).
Billault, A., ‘Les formes romanesques de l’héroïsation dans la Vie d’Apollonios de
Tyane de Philostrate’, BAGB (1991) 267–274.
———, ‘Le personnage de Philostrate dans la Vie d’Apollonius de Tyane: auto-
portrait de l’auteur en biographe’, in M.-F. Baslez, P. Hoffmann, L. Pernot
(eds.), L’invention de l’autobiographie d’Hésiode à Saint Augustin (Paris 1993) 271–
278.
Bing, P., The Well-Read Muse (Göttingen 1988).
Blank, D., ‘The Arousal of Emotion in Plato’s Dialogues’, CQ 43 (1993) 428–
439.
Blondell, R. The Play of Character in Plato’s Dialogues (Cambridge 2002).
Blundell, J., Menander and the Monologue (Göttingen 1980).
Boedeker, D., ‘Herodotus’ Genre(s)’, in M. Depew, D. Obbink (eds.), Matrices of
Genre (Cambridge Mass. 2000) 97–114.
Bompaire, J., ‘Le sacré dans les discours d’Aelius Aristide (XLVII–LII)’, REG
102 (1989) 28–39.
———, ‘Quatre styles d’autobiographie au IIe siècle apès J.-C.: Aelius Aristide,
Lucien, Marc-Aurèle, Galien’, in M.-F. Baslez, P. Hoffmann, L. Pernot
(eds.), L’invention de l’autobiographie d’Hésiode à Saint Augustin (Paris 1993) 199–
209.
Booth, W.C., The Rhetoric of Fiction (Chicago [1961] 1983).
Bordes, J., Politeia dans la pensée grecque jusqu’à Aristote (Paris 1982).
Bosworth, A.B., A Historical Commentary on Arrian’s History of Alexander, 2 vols.
(Oxford 1980–1995).
———, From Arrian to Alexander: Studies in Historical Interpretation (Oxford 1988).
Bowersock, G.W., ‘Philostratus and the Second Sophistic’, in Easterling–Knox
1985, 655–658.
Bowie, A.M., Aristophanes: Myth, Ritual, and Comedy (Cambridge 1993).
Bowie, E.L., ‘Apollonius of Tyana: Tradition and Reality’, ANRW 2.16.2 (1978)
1652–1699.
———, ‘Lies, Fiction and Slander in Early Greek Poetry’, in C. Gill, T.P. Wise-
man (eds.), Lies and Fiction in the Ancient World (Exeter 1993) 1–37.
———, ‘Philostratus: Writer of Fiction’, in J. Morgan, R. Stoneman (eds.), Greek
Fiction: the Greek Novel in Context (London/New York 1994) 181–199.
bibliography 557
———, ‘Ancestors of Historiography in Early Greek Elegiac and Iambic Poet-

ry?’, in N. Luraghi (ed.), The Historian’s Craft in the Age of Herodotus (Oxford
2001) 45–66.
Branham, R.B., ‘Introducing the Sophist: Lucian’s Prologues’, TAPhA 115
(1985) 237–243.
———, Unruly Eloquence: Lucian and the Comedy of Traditions (Cambridge Mass.
1989).
Braswell, B.K., A Commentary on Pindar Nemean One (Fribourg 1992).
Breitenbach, H.R., Historiographische Anschauungsformen Xenophons (Freiburg i.Ü.
1950).
Broadhead, H.D., Aeschylus. Persae. Edited with Introduction, Critical Notes and Com-
mentary (Cambridge 1960).
Brodersen, K., Appian und sein Werk, ANRW 2.34.1 (1993) 339–363.
Bruns, I., Das literarische Porträt der Griechen im fünften und vierten Jahrhundert vor
Christi Geburt (Berlin 1896).
Brunt, P.A., Arrian. With an English Translation, 2 vols. (London/Cambridge
Mass. 1976–1983).
Bühler, W., ‘Das Element des Visuellen in der Eingangsszene von Heliodors
Aithiopika’, WS 10 (1976) 177–185.
Bulloch, A.W., Callimachus. The Fifth Hymn (Cambridge 1985).
Bürger, K., ‘Zu Xenophon von Ephesus’, Hermes 27 (1892) 36–67.
Burr, V., Nostrum mare: Ursprung und Geschichte der Namen des Mittelmeeres und seiner
Teilmeere im Altertum (Stuttgart 1932).
Burton, R.W.B., The Chorus in Sophocles’ Tragedies (Oxford 1980).
Busch, S., ‘Orpheus bei Apollonios Rhodios’, Hermes 121 (1993) 301–324.
Bushnell, R.W., Prophesying Tragedy: Sign and Voice in Sophocles’ Theban Plays (Itha-
ca N.Y./London 1988).
Byre, C.S., ‘The Narrator’s Addresses to the Narratee in Apollonius Rhodius’
Argonautica’, TAPhA 121 (1991) 215–227.
Cairns, F., ‘Theocritus, Idyll 26’, PCPhS 38 (1992) 1–38.
Calame, C., ‘From Choral Poetry to Tragic Stasimon: The Enactment of
Women’s Song’, Arion 3 (1994/95) 136–154.
Cameron, A., Callimachus and his Critics (Princeton 1995).
Canfora, L., Storia della letteratura greca (Roma 1986).
Carey, C., A Commentary on Five Odes of Pindar: Pythian 2, Pythian 9, Nemean 1,
Nemean 7, Isthmian 8 (New York 1981).
———, Lysias. Selected Speeches (Cambridge 1989).
———, ‘Pindar and the Victory Ode’, in L. Ayres (ed.), The Passionate Intellect.
Essays on the Transformation of Classical Traditions Presented to Professor I.G. Kidd
(New Brunswick/London 1995) 85–103.
———, ‘Ethos and Pathos in Bacchylides’, in I.L. Pfeijffer, S.R. Slings (eds.), One
Hundred Years of Bacchylides. Proceedings of a Colloquium Held at the Vrije Universiteit
Amsterdam (Amsterdam 1999) 17–29.
———, Aeschines (Austin 2000).
Carey, C., Reid, R.A., Demosthenes. Selected Private Speeches (Cambridge 1985).
Carlier, P., La royauté en Grèce avant Alexandre (Strasbourg 1984).
Cary, E., Dio’s Roman History, 9 vols. (London/Cambridge. Mass. 1914–1927).
558 bibliography
Casertano, G. (ed.), La struttura del dialogo platonico (Napoli 2000).

Castiglioni, L., ‘Osservazioni critiche a Longo sofista, Senofonte Efesio e
Caritone I: per un’ edizione di Longo’, RF 34 (1906) 293–320.
Ceserani, R., Raccontare la letteratura (Torino 1990).
Chantraine, P., ‘Les verbes grecs significant “lire” (anagignōskō, epilegomai, en-
tunkhanō, analegomai)’, in Pagkarpeia, Mélanges Henri Grégoire ii, AIPhO 10 (1950)
115–126.
Chaplin, J.D., Livy’s Exemplary History (Oxford 2000).
Chatman, S., Story and Discourse: Narrative Structure in Fiction and Film (Ithaca N.Y.
1978).
———, Coming to Terms: The Rhetoric of Narrative in Fiction and Film (Ithaca N.Y.
1990).
Christ, M.R., ‘Herodotean Kings and Historical Inquiry’, ClAnt 13 (1994) 167–
202.
Clare, R.J., The Path of the Argo: Language, Imagery and Narrative in the Argonautica of
Apollonius Rhodius (Cambridge 2002).
Clarke, K.J., ‘In Search of the Author of Strabo’s Geography’, JRS 87 (1997)
92–110.
———, Between Geography and History: Hellenistic Constructions of the Roman World
(Oxford 1999).
Clauss, J.J., The Best of the Argonauts. The Redefinition of the Epic Hero in Book One of
Apollonius’ Argonautica (Berkeley 1993).
Clay, D., ‘Plato’s First Words’, YCS 29 (1992) 113–129.
———, ‘The Origins of the Socratic Dialogue’, in P. Vander Waerdt (ed.), The
Socratic Movement (Ithaca N.Y. 1994) 23–47.
Clay, J. Strauss, The Politics of Olympus. Form and Meaning in the Major Homeric
Hymns (Princeton 1989).
———, ‘The Education of Perses: From “Mega Nepios” to “Dion Genos” and
Back’, MD 31 (1993) 23–33.
———, ‘The Homeric Hymns’, in I. Morris, B. Powell (eds.), A New Companion
to Homer (Leiden 1997) 489–507.
Cobet, J., Herodots Exkurse und die Frage der Einheit seines Werkes (Wiesbaden 1971).
Cohn, D., Transparent Minds: Narrative Modes for Presenting Consciousness in Fiction
(Princeton 1978).
———, The Distinction of Fiction (Baltimore–London 1999).
Connor, W.R., ‘A Post Modernist Thucydides?’, CJ 72 (1977) 289–298.
———, Thucydides (Princeton 1984).
———, ‘Narrative Discourse in Thucydides’, in M.H. Jameson (ed.), The Greek
Historians: Literature and History. Papers Presented to A.E. Raubitschek (Stanford
1985) 1–17.
Conte, G.B., The Hidden Author. An Interpretation of Petronius’ Satyricon, trans.
E. Fantham (Berkeley 1996).
Cook, J.M., The Persian Empire (London 1983).
Crane, G., The Blinded Eye: Thucydides and the New Written Word (Lanham 1996).
———, Thucydides and the Ancient Simplicity (Berkeley 1998).
D’Alessio, G.B., Callimaco. Introduzione, traduzione e note (Milano 1996).
Darbo-Peschanski, C., ‘L’historien grec ou le passé jugé’, in N. Loraux, C. Mi-
bibliography 559
ralles (eds.), Figures de l’intellectuel en Grèce ancienne (Paris 1998) 143–189.

Davidson, J., ‘The Gaze in Polybius’ Histories’, JRS 81 (1991) 10–24.
DeForest, M.M., Apollonius’ Argonautica: A Callimachean Epic (Leiden 1994).
Denniston, J.D., The Greek Particles (Oxford [1934] 1954).
Depew, M., ‘Mimesis and Aetiology in Callimachus’ Hymns’, in M.A. Harder,
R.F. Regtuit, G.C. Wakker (eds.), Callimachus (Groningen 1993) 57–77.
Derow, P.S., ‘Polybius, Rome, and the East’, JRS 69 (1979) 1–15.
———, ‘Historical Explanation: Polybius and his Predecessors’, in S. Horn-
blower (ed.), Greek Historiography (Oxford 1994) 73–90.
Detienne, M., The Masters of Truth in Archaic Greece, trans. J. Lloyd (New York
1996).
Dewald, C., ‘Practical Knowledge and the Historian’s Role in Herodotus and
Thucydides’, in The Greek Historians: Literature and History (Saratoga 1985) 47–
63.
———, ‘Narrative Surface and Authorial Voice in Herodotus’ Histories’, Arethusa
20 (1987) 147–170.
———, ‘Wanton Kings, Pickled Heroes, and Gnomic Founding Fathers: Strate-
gies of Meaning at the End of Herodotus’ Histories’, in D. Roberts,
F. Dunn, D. Fowler (eds.), Classical Closure. Reading the End in Greek and Latin
Literature (Princeton 1997) 62–82.
———, ‘“I Didn’t Give My Own Genealogy”: Herodotus and the Authorial
Persona’, in E.J. Bakker, I.J.F. de Jong, H. van Wees (eds.), A Companion to
Herodotus (Leiden 2002) 267–289.
———, ‘Form and Content: The Question of Tyranny in Herodotus’, in K. Ha-
gan (ed.), Popular Tyranny (Austin, forthcoming).
DeWitt, N.W., DeWitt, N.J., Demosthenes, vol. 7 (Cambridge Mass./London
1949).
Dihle, A., Griechische Literaturgeschichte (Stuttgart 1967).
———, Die griechische und lateinische Literatur der Kaiserzeit (München 1989).
Dillery, J., Xenophon and the History of his Times (London 1995).
Doody, M.A., The True Story of the Novel (London 1996).
Dover, K.J., ‘Anthemocritus and the Megarians’, AJPh 87 (1966) 203–209, repr.
in The Greeks and their Legacy: Collected Papers II (Oxford 1988) 181–186.
———, Lysias and the Corpus Lysiacum (Berkeley 1968).
———, Aristophanes: Frogs (Oxford 1993).
Dubel, S., ‘Dialogue et autoportrait: les masques de Lucien’, in A. Billault
(ed.), Lucien de Samosate (Lyon 1994) 19–26.
Dubischar, M., Die Agonszenen bei Euripides (Stuttgart/Weimar 2001).
Dubuisson, M., Le latin de Polybe (Paris 1985).
Duckworth, G.E., Foreshadowing and Suspense in the Epics of Homer, Apollonius, and
Vergil (diss. Princeton 1933).
Due, B., The Cyropaedia. Xenophon’s Aims and Methods (Aarhus 1989).
Duff, T., Plutarch’s Lives. Exploring Virtue and Vice (Oxford 1999).
Duhoux, I., ‘Grec écrit et grec parlé. Une étude contrastive des particules
aux Ve–IVe siècles’, in A. Rijksbaron (ed.), New Approaches to Greek Particles
(Amsterdam 1997) 15–48.
Dunbar, N., Aristophanes: Birds (Oxford 1995).
560 bibliography
Dunn, F.M., Tragedy’s End: Closure and Innovation in Euripidean Drama (New York/
Oxford 1997).
Dzielska, M., Apollonius of Tyana in Legend and History, trans. P. Pienkowski
(Roma 1986).
Easterling, P.E., ‘Philoctetes and Modern Criticism’, ICS 3 (1978) 27–39.
———, Knox, B.M.W. (eds.), The Cambridge History of Classical Literature. Vol. 1:
Greek Literature (Cambridge 1985).
Ebert, T., ‘Una nuova interpretazione del Fedone platonico’, in G. Casertano
(ed.), La struttura del dialogo platonico (Napoli 2000) 51–73.
Eckstein, A.M., Moral Vision in the Histories of Polybius (Berkeley 1995).
Edwards, M.J., ‘Damis the epicurean’, CQ 41 (1991) 563–566.
Edwards, M., The Attic Orators (London 1994).
———, Andocides (Warminster 1995).
———, Lysias. Five Speeches (London 1999).
Edwards, M., Usher, S., Antiphon and Lysias (Warminster 1985).
Elsner, J., ‘From the Pyramids to Pausanias and Piglet: Monuments, Travel
and Writing’, in S. Goldhill, R. Osborne (eds.), Art and Text in Ancient Greek
Culture (Cambridge 1994) 224–254.
———, ‘Hagiographic Geography: Travel and Allegory in the Life of Apollonius
of Tyana’, JHS 117 (1997) 22–37.
———,‘Describing Self in theLanguage of Other:Pseudo (?) Lucian at theTem-
ple of Hierapolis’, in S. Goldhill (ed.), Being Greek under Rome: Cultural Identity,
the Second Sophistic and the Development of Empire (Cambridge 2001) 123–153.
Erbse, H., Studien zum Prolog der euripideischen Tragödie (Berlin/New York 1984).
———, Studien zum Verständnis Herodots (Berlin/New York 1992).
Erskine, A., Troy Between Greece and Rome: Local Tradition and Imperial Power
(Oxford 2001).
Fantham, E., Ovid. Fasti Book IV (Cambridge 1998).
Fantuzzi, M., ‘Theocritus and the “Demythologizing” of poetry’, in M. De-
pew, D. Obbink (eds.), Matrices of Genre. Authors, Canons, and Society (Cam-
bridge Mass./London 2000) 135–151.
———, Hunter, R., Muse e modelli. La poesia ellenistica da Alessandro Magno ad
Augusto (Roma/Bari 2002).
Feeney, D.C., The Gods in Epic: Poets and Critics of the Classical Tradition (Oxford
1991).
Finkelberg, M., ‘A Creative Oral Poet and the Muse’, AJPh 111 (1990) 293–303.
Fisher, N.R.E., Aeschines. Against Timarchos (Oxford 2001).
Flinterman, J.-J., Power, Paideia & Pythagoreanism: Greek Identity, Conceptions of the
Relationship between Philosophers and Monarchs and Political Ideas in Philostratus’ Life
of Apollonius (Amsterdam 1995).
Flory, S., ‘Who read Herodotus’ Histories?’, AJPh 101 (1980) 12–28.
———, ‘Thucydides’ Hypotheses about the Peloponnesian War’, TAPhA 118
(1988) 43–56.
Flower, M.A., Theopompus of Chios: History and Rhetoric in the Fourth Century BC
(Oxford 1994).
Fludernik, M., ‘Introduction: Second-Person Narrative and Related Issues’,
Style 28 (1994a) 281–328.
bibliography 561
———, ‘Second-Person Narrative as a Test-Case for Narratology: the Limits of

Realism’, Style 28 (1994b) 445–479.
———, Towards a ‘Natural’ Narratology (London 1996).
Foley, H., ‘Tragedy and Politics in Aristophanes’ Acharnians’, JHS 108 (1988)
33–47 (reprinted in Segal 1996: 117–142).
Ford, A., Homer. The Poetry of the Past (Ithaca N.Y./London 1992).
Forster, E.M., Aspects of the Novel ([1927] Harmondsworth 1979).
Foucault, J.-A. de, Recherches sur la langue et le style de Polybe (Paris 1972).
Foucault, M., The History of Sexuality, vol 3: The Care of the Self, trans. R. Hurley
([1984] London 1990).
Fountoulakis, A., ‘On the Literary Genre of Lycophron’s Alexandra’, AAntHung
38 (1998) 291–295.
Fowler, R.L., ‘Herodotos and his Contemporaries’, JHS 116 (1996) 62–87.
Fraenkel, E., De media et nova comoedia quaestiones selectae (diss. Göttingen 1912).
———, Aeschylus: Agamemnon (Oxford 1950).
Frazier, F., ‘Remarques à propos de l’usage des citations en matière de chrono-
logie dans les Vies’, ICS 13.2 (1988) 297–309.
———, ‘Contribution à l’étude de la composition des “Vies” de Plutarque:
l’élaboration des grandes scènes’, ANRW 2.33.6 (1992) 4487–535.
———, Histoire et morale dans les Vies parallèles de Plutarque (Paris 1996).
Friedemann, K., Die Rolle des Erzählers in der Epik (Leipzig 1910).
Friedländer, P., ‘Zum Proömium der Theogonie’, Hermes 49 (1914) 1–16.
Fuchs, E. Pseudologia. Formen und Funktionen fiktionaler Trugrede in der grieschischen
Literatur der Antike (Heidelberg 1993).
Fuhrer, Th., Die Auseinandersetzung mit den Chorlyrikern in den Epinikien des Kallima-
chos (Basel 1992).
Fusillo, M., ‘L’Alessandra di Licofrone: racconto epico e discorso “dram-
matico”’, ASNP 14 (1984) 495–525.
———, Il tempo delle Argonautiche. Un’analisi del racconto in Apollonio Rodio (Roma
1985).
———, ‘The Mirror of the Moon: Lucian’s A True Story—from Satire to Utopi’,
in S. Swain (ed.), Oxford Readings in the Greek Novel (Oxford 1999) 351–381.
———, ‘Apollonius Rhodius as “Inventor” of the Interior Monologue’, in T.D.
Papanghelis, A. Rengakos (eds.), A Companion to Apollonius Rhodius (Leiden
2001) 127–146.
———, Hurst, A., Paduano, G. (eds.), Licofrone: Alessandra (Milano 1991).
Futre Pinheiro, M., ‘Calasiris’ Story and its Narrative Significance in Helio-
dorus’ Aethiopica’, Groningen Colloquia on the Novel 4 (1991) 69–83.
Gabba, E., Appiano e la storia delle guerre civili (Firenze 1956).
Gagarin, M., Antiphon. The Speeches (Cambridge 1997).
Gagarin, M., MacDowell, D.M., Antiphon and Andocides (Austin 1998).
Gantz, T., Early Greek Myth: A Guide to Literary and Artistic Sources (Baltimore/
London 1993).
Gardiner, C.P., The Sophoclean Chorus. A Study of Character and Function (Iowa City
1987).
Garvie, A.F., Aeschylus: Choephori (Oxford 1986).
———, Sophocles Ajax (Warminster 1998).
562 bibliography
Gellie, G.H., Sophocles. A Reading (Melbourne 1972).

Gelzer, T., ‘Some Aspects of Aristophanes’ Dramatic Art in the Birds’, BICS 23
(1976) 1–14 (reprinted in Segal 1996: 194–215).
Genette, G., Introduction à l’architexte (Paris 1979).
———, Narrative Discourse ([Paris 1972] Ithaca N.Y./London 1980).
———, Narrative Discourse Revisited ([Paris 1983] Ithaca N.Y./London 1988).
———, Fiction et Diction (Paris 1991)
Georgiadou, A., Larmour, D.H.J., Lucian’s Science Fiction Novel ‘True Histories’.
Interpretation and Commentary (Leiden 1998).
Gera, D., Xenophon’s Cyropaedia: Style, Genre, and Literary Technique (Oxford 1993).
Gill, C., ‘Dialectic and the Dialogue Form’, in J. Annas, C. Rowe (eds.), New
Perspectives on Plato, Modern and Ancient (Cambridge Mass. 2002) 145–171.
Goldhill, S., ‘Reading Differences: the Odyssey and Juxtaposition’, Ramus 17
(1988a) 1–31.
———, ‘Battle Narrative and Politics in Aeschylus’ Persae’, JHS 108 (1988b) 189–
193.
———, ‘The Paradigms of Epic: Apollonius Rhodius and the Example of the
Past’, in The Poet’s Voice: Essays on Poetics and Greek Literature (Cambridge 1991)
284–333.
———, ‘Collectivity and Otherness—The Authority of the Tragic Chorus:
Response to Gould’, in M.S. Silk (ed.), Tragedy and the Tragic: Greek Theatre
and Beyond (Oxford 1996) 244–256.
———, Who Needs Greek? Contests in the History of Hellenism (Cambridge 2002).
Goldmann, B., Einheitlichkeit und Eigenständigkeit der Historia Romana des Appian
(Hildesheim 1988).
Gomme, A.W., Andrewes, A.A., Dover, K.J., A Historical Commentary on Thucy-
dides, vol. V (Oxford 1981).
González, J.M., ‘Mousai hypophetores: Apollonius of Rhodes on Inspiration and
Interpretation’, HSPh 100 (2000) 269–292.
Gould, J., Herodotus (London 1989).
———, ‘Tragedy and Collective Experience’, in M.S. Silk (ed.), Tragedy and the
Tragic: Greek Theatre and Beyond (Oxford 1996) 217–243.
———, ‘… And Tell Sad Stories of the Death of Kings: Greek Tragic Drama
as Narrative’, in Myth, Ritual, Memory, and Exchange (Oxford 2001) 319–
334.
Gow, A.S.F., Theocritus (Cambridge [1950] 1952).
Goward, B., Telling Tragedy. Narrative Technique in Aeschylus, Sophocles and Euripides
(London 1999).
Gowing, A.M., The Triumviral Narratives of Appian and Cassius Dio (Ann Arbor
1992).
Graf, F., ‘Maximos von Aigai: ein Beitrag zur Überlieferung über Apollonios
von Tyana’, JbAC (1984–1985) 27–28, 65–73.
Gray, V.J., The Character of Xenophon’s Hellenica (London 1989a).
———, ‘Xenophon’s Defence of Socrates: the Rhetorical Background to the So-
cratic Problem’, CQ 39 (1989b) 136–140.
———, ‘The Moral Interpretation of the “Second Preface” to Arrian’s Anaba-
sis’, JHS 110 (1990) 180–186.
bibliography 563
———, ‘Continuous History and Xenophon Hellenica 1–2.3.10’, AJPh 112 (1991)
201–228.
———, The Framing of Socrates (Stuttgart 1998).
———, ‘Short Stories in Herodotus’ Histories’, in E.J. Bakker, I.J.F. de Jong, H.
van Wees (eds.), A Companion to Herodotus (Leiden 2002) 291–317.
———, ‘Interventions and Citations in Xenophon, Hellenica and Anabasis’, CQ
53.1 (2003) 111–123.
———, Review of Christopher Nadon, Xenophon’s Prince, AncPhil (forthcoming).
Green, P., Alexander to Actium (Berkeley/London 1990).
Gribble, D., ‘Narrator Interventions in Thucydides’, JHS 118 (1998) 41–67.
———, Alcibiades and Athens. A Study in Literary Presentation (Oxford 1999).
Griffin, J., Homer on Life and Death (Oxford 1980).
———, ‘Homeric Words and Speakers’, JHS 106 (1986) 36–57.
Griffith, M., Aeschylus: Prometheus Bound (Cambridge 1983a).
———, ‘Personality in Hesiod’, CA 2 (1983b) 37–65.
van Groningen, B.A., La composition littéraire archaïque grecque: procédés et réalisations
(Amsterdam 1958).
Grossardt, P., Die Trugreden in der Odyssee und ihre Rezeption in der antiken Literatur
(Bern 1998).
Groten, F.J., ‘Herodotus’ Use of Variant Versions’, Phoenix 17 (1963) 79–87.
Gutzwiller, K., ‘The Evidence for Theocritean Poetry Books’, in M.A. Har-
der, R.F. Regtuit, G.C. Wakker (eds.), Theocritus (Groningen 1996) 119–
148.
Hägg, T, ‘Die Ephesiaka des Xenophon Ephesios—Original oder Epitome?’,
C&M 27 (1966) 118–161.
———, Narrative Technique in Ancient Greek Romances (Stockholm 1971).
———, The Novel in Antiquity (Oxford 1983)
Hahn, I., Németh, G., ‘Appian und Rom’, ANRW 2.34.1 (1993) 364–402.
Halfmann, H., Die Senatoren aus dem östlichen Teil des Imperium Romanum bis zum
Ende des 2. Jh. n. Chr. (Göttingen 1979).
Hall, E., Aeschylus: Persians. Edited with an Introduction, Translation, and Commentary
(Warminster 1996).
Halperin, D.M., ‘Plato and the Erotics of Narrativity’, in J.C. Klagge, N.D.
Smith (eds.), Methods of Interpreting Plato and his Dialogues (Oxford 1992) 93–
129.
Hamilton, R., ‘Bacchae 47–52: Dionysus’ Plan’, TAPhA 104 (1974) 139–149.
———, ‘Prologue, Prophecy, and Plot in Four Plays of Euripides’, AJPh 99
(1978) 277–302.
Hammond, N.G.L., ‘The Speeches in Arrian’s Indica and Anabasis’, CQ 49
(1999) 238–253.
Hamon, P., Du descriptif (Paris 1993).
Hanson, V.D., review of J.P. Hallett, T. Van Nortwick (eds.), Compromising
Traditions: The Personal Voice in Classical Scholarship (London 1997), Arion 6
(1998/99) 131–168.
Harder, M.A., ‘Callimachus and the Muses’, Prometheus 14 (1988) 1–14.
———, ‘Untrodden Paths: Where Do They Lead?’ HSPh 93 (1990) 287–309.
———, ‘Insubstantial Voices’, CQ 42 (1992) 384–394.
564 bibliography
———, ‘Generic Games’ in Callimachus’ Aetia’, in M.A. Harder, R.F. Regtuit,

G.C. Wakker (eds.), Genre in Hellenistic Poetry (Groningen 1998) 95–113.
———, ‘Intertextuality in Callimachus’ Aetia’, Entretiens sur l’antiquité classique 48
(Genève 2002) 189–233.
Hardie, P., ‘A Reading of Heliodorus, Aithiopika 3.4.1–5.2’, in R. Hunter (ed.),
Studies in Heliodorus (Cambridge 1998) 19–39.
Harmon, A., Lucian, vol. 5 (Cambridge Mass./London 1936).
Harris, E.M., ‘The Authenticity of Andokides’ De Pace’, in P. Flensted-Jensen,
T.H. Nielsen, L. Rubinstein (eds.), Polis and Politics. Studies in Ancient Greek
History (Copenhagen 2000) 479–505.
Harrison, T., Divinity and History. The Religion of Herodotus (Oxford 2000).
Harvey, D., ‘The Sykophant and Sykophancy: Vexatious Redefinition?’, in
P. Cartledge, P. Millett, S. Todd (eds.), Nomos. Essays in Athenian Law, Politics
and Society (Cambridge 1990) 103–121.
Haslam, M.W., ‘Plato, Sophron, and the Dramatic Dialogue’, BICS 19 (1972)
17–38.
Hefti, V., Zur Erzählungstechnik in Heliodors Aethiopica (Wien 1950).
Heiden, B., Tragic Rhetoric. An Interpretation of Sophocles’ Trachiniae (New York
1989).
Hellwig, B., Raum und Zeit im homerischen Epos (Hildesheim 1964).
Helmbold, W.C., O’Neil, E.N., Plutarch’s Quotations (Baltimore 1959).
Hemmerdinger, B., ‘La division en livres de l’oeuvre de Thucydide’, REG 61
(1948) 104–117.
Henrichs, A., ‘“Why Should I Dance?”: Choral Self-Referentiality in Greek
Tragedy’, Arion 3 (1994/95) 56–111.
Henry, W.P., Greek Historical Writing (Chicago 1966).
Hidber, Th., ‘Zeit und Erzählperspektive in Herodians Geschichtswerk’, in
M. Zimmermann (ed.), Geschichtsschreibung und politischer Wandel im 3. Jh. n.
Chr. (Stuttgart 1999) 145–167.
Higgins, W.E., Xenophon the Athenian (Albany N.Y. 1977).
Highet, G., ‘The Huntsman and the Castaway’, GRBS 14 (1973) 35–40.
Hinds, S., Allusion and Intertext: Dynamics of Appropriation in Roman Poetry (Cam-
bridge 1998).
Hirsch, S., The Friendship of the Barbarians (Hanover N.H. 1985).
Hock, R.F., O’Neil, E.N., The Chreia in Ancient Rhetoric, vol.1: Progymnasmata
(Atlanta 1986).
Hollis, A.S., ‘Teuthis and Callimachus, Aetia Book 1’, CQ 32 (1982) 117–120.
———, Callimachus. Hecale (Oxford 1990).
Holzberg, N., Menander. Untersuchungen zur dramatischen Technik (Nürnberg 1974).
Hopkinson, N., Callimachus. Hymn to Demeter (Cambridge 1984).
Hornblower, J., Hieronymus of Cardia (Oxford 1981).
Hornblower, S., A Commentary on Thucydides, 2 vols. (Oxford 1991–1996).
———, ‘Narratology and Narrative Techniques in Thucydides’, in S. Horn-
blower (ed.), Greek Historiography (Oxford 1994) 131–166.
Hose, M., Studien zum Chor bei Euripides (Stuttgart 1990–1991).
———, Erneuerung der Vergangenheit. Die Historiker im Imperium Romanum von Florus
bis Cassius Dio (Stuttgart/Leipzig 1994).
bibliography 565
Hubbard, T.K., The Mask of Comedy: Aristophanes and the Intertextual Parabasis
(Ithaca N.Y./London 1991).
Hulton, A.O., ‘The Prologues of Sophocles’, G&R 16 (1969) 49–59.
Hunter, R.L., A study of Daphnis and Chloe (Cambridge 1983).
———, The Argonautica of Apollonius: Literary Studies (Cambridge 1993a).
——— (tr.), Apollonius of Rhodes, Jason and the Golden Fleece (The Argonautica) (Oxford
1993b).
———, Theocritus and the Archaeology of Greek Poetry (Cambridge 1996).
———, ‘The Aithiopika of Heliodorus: Beyond Interpretation?’, in R. Hunter
(ed.), Studies in Heliodorus (Cambridge 1998a) 40–59.
———, ‘Before and after Epic: Theocritus (?), Idyll 25’, in M.A. Harder, R.F.
Regtuit, G.C. Wakker (eds.), Genre in Hellenistic Poetry (Groningen 1998b) 115–
132.
———, Theocritus: A Selection (Cambridge 1999).
———, ‘The Poetics of Narrative in the Argonautica’, in T.D. Papanghelis,
A. Rengakos (eds.), A Companion to Apollonius Rhodius (Leiden 2001) 93–
125.
———, ‘The Sense of an Author: Theocritus and [Theocritus]’, in R. Gibson,
C. Kraus (eds.), The Classical Commentary (Leiden 2002) 89–108.
———, ‘Reflecting on Writing and Culture: Theocritus and the Style of Cul-
tural Change’, in H. Yunis (ed.), Written Texts and the Rise of Literate Culture in
Ancient Greece (Cambridge 2003) 213–234.
Hutchinson, G.O., Aeschylus: Septem Contra Thebas (Oxford 1985).
———, Hellenistic Poetry (Oxford 1988).
Ibendorff, E., Untersuchungen zur darstellerischen Persönlichkeit des Polybios (Rostock
1930).
Imbert, C., ‘Stoic Logic and Alexandrian Poetics’, in M. Schofield et al. (eds.),
Doubts and Dogmatism: Studies in Hellenistic Epistemology (Oxford 1980) 182–216.
Imhof, M., Bemerkungen zu den Prologen der sophokleischen und euripideischen Tragödien
(Winterthur 1937).
Immerwahr, H.R., ‘Thucydides’, in Easterling–Knox 1985, 441–456.
Jackson, S., ‘Apollonius and the Emperors’, Hermathena 137 (1984) 25–32.
Jaeger, W., Paideia, vol. 1 (Berlin [1933] 1954).
Janko, R., ‘The Structure of the Homeric Hymns: A Study in Genre’, Hermes
109 (1981) 9–24.
———, Homer, Hesiod and the Hymns. Diachronic Development in Epic Diction (Cam-
bridge 1982).
Jens, W. (ed.), Die Bauformen der griechischen Tragödie (München 1971).
Johnson, W.A., ‘Oral Performance and the Composition of Herodotus’ Histo-
ries’, GRBS 35 (1994) 229–254.
Jones, C.P., ‘Towards a Chronology of Plutarch’s Works’, JRS 56 (1966) 61–74
(reprinted in Scardigli 1995: 95–123).
———, ‘Sura and Senecio’, JRS 60 (1970) 98–104.
———, Plutarch and Rome (Oxford 1971).
———, Culture and Society in Lucian (Cambridge Mass. 1986).
de Jong, I.J.F., Narrators and Focalizers. The Presentation of the Story in the Iliad
(Amsterdam 1987).
566 bibliography
———, ‘Homeric Words and Speakers: an Addendum’, JHS 108 (1988) 188–
189.
———, Narrative in Drama: the Art of the Euripidean Messenger-Speech (Leiden 1991a).
———, ‘Tijdsaspecten in Pindarus’ Pythische vier’, Lampas 24 (1991b) 199–210.
———, ‘The Subjective Style in Odysseus’ Wanderings’, CQ 42 (1992) 1–11.
———, ‘gar Introducing Embedded Narratives’, in A. Rijksbaron (ed.), New
Approaches to Greek Particles (Amsterdam 1997) 175–185.
———, ‘Aspects narratologiques des “Histoires” d’Hérodote’, Lalies 19 (1999)
217–275.
———, A Narratological Commentary on the Odyssey (Cambridge 2001).
———, ‘Narrative Unity and Units’, in E.J. Bakker, I.J.F. de Jong, H. van Wees
(eds.), A Companion to Herodotus (Leiden 2002) 245–266.
Jörgensen, O., ‘Das Auftreten der Götter in den Büchern i–m der Odyssee’,
Hermes 39 (1904) 357–382.
Kacandes, I., ‘Narrative Apostrophe: Reading, Rhetoric, Resistance in Michel
Butor’s La Modification and Julio Cortazar’s Graffiti’, Style 28 (1994) 329–349.
Kahn, C., ‘Aeschines on Socratic Eros’, in P. Vander Waerdt (ed.), The Socratic
Movement (Ithaca N.Y. 1994) 87–106.
Kahn, C.H., Plato and the Socratic Dialogue (Cambridge 1996).
Kaimio, M., The Chorus of Greek Drama within the Light of the Person and Number
Used (Helsinki 1970).
Käppel, L., ‘Die Rolle des Chores in der Orestie des Aischylos: Vom epischen
Erzähler über das lyrische Ich zur dramatis persona’, in P. Riemer, B. Zim-
mermann (eds.), Der Chor im antiken und modernen Drama (Stuttgart 1998) 61–
88.
Katsouris, A.G., Prologos: dramatikê tekhnê [in Greek] (Ioannina 1997).
Katz, J.T., Volk, K., ‘“Mere Bellies”?: A New Look at Theogony 26–28’, JHS
120 (2000) 122–131.
Kayser, W., ‘Wer erzählt den Roman?’, in Die Vortragsreise. Studien zur Literatur
(Bern 1958) 90–101.
Keller, J., Struktur und dramatische Funktion des Botenberichtes bei Aischylos und Sopho-
kles (diss. Tübingen 1959).
Kennedy, G.A. (ed.), The Cambridge History of Literary Criticism, vol. 1: Classical
Criticism (Cambridge 1989).
———, A New History of Classical Rhetoric (Princeton 1994).
Kerkhecker, A., Callimachus’ Book of Iambi (Oxford 1999).
Kindstrand, J.F., Homer in der zweiten Sophistik (Uppsala 1973).
Köhnken, A., ‘Hellenistic Chronology: Theocritus, Callimachus, and Apollo-
nius Rhodius’, in T. Papanghelis, A. Rengakos (eds.), A Companion to Apollo-
nius Rhodius (Leiden 2001) 73–92.
Konstan, D., Sexual Symmetry. Love in the Ancient Novel and Related Genres (Princeton
1994).
Korenjak, M., Publikum und Redner: ihre Interaktion in der sophistischen Rhetorik der
Kaiserzeit (München 2000).
Koskenniemi, E., Der philostrateische Apollonios (Helsinki 1991).
Kosmetatou, E., ‘Lycophron’s “Alexandra” Reconsidered: The Attalid Con-
nection’, Hermes 128 (2000) 32–53.
bibliography 567
Kraus, C.S., ‘Logos men est’ archaios: Stories and Story-Telling in Sophocles’
Trachiniae’, TAPhA 121 (1991) 75–98.
Krentz, P., Xenophon Hellenika I–II.3.10 (Warminster 1989).
———, Xenophon Hellenika II.3.11-IV.2.8 (Warminster 1995).
Kurke, L., ‘Charting the Poles of History: Herodotos and Thoukydides’, in
O. Taplin (ed.), Literature in the Greek and Roman Worlds: A New Perspective
(Oxford 2000) 133–155.
Kurz, A., Le corpus Theocriteum et Homère. Un problème d’authenticité (Idylle 25)
(Bern/Frankfurt 1982).
Laird, A., Powers of Expression, Expressions of Power: Speech Presentation in Latin
Literature (Oxford 1999).
Lamb, W.R.M., Lysias (Cambridge Mass./London 1930).
Lane, M., ‘Socrates and Plato’, in C. Rowe, M. Schofield (eds.), The Cambridge
History of Greek and Roman Political Thought (Cambridge 2000) 155–163.
Lannoy, L. de, ‘Le problème des Philostrate (état de la question)’, ANRW
2.34.3 (1997) 2362–449.
Lardinois, A., ‘Modern Paroemiology and the Use of Gnomai in Homer’s
Iliad’, CPh 92 (1997) 213–234.
Lateiner, D., ‘Pathos in Thucydides’, Antichthon 11 (1977) 42–51.
———, The Historical Method of Herodotus (Toronto 1989).
Leeman, A.D., Orationis ratio. The Stylistic Theories and Practice of the Roman Orators,
Historians and Philosophers (Amsterdam 1963).
Van der Leest, J., ‘Appian’s References to his Own Time’, AHB 3 (1989) 131–
133.
Lefkowitz, M.R., First-Person Fictions: Pindar’s Poetic ‘I’ (Oxford 1991).
Lehmann, G.A., ‘The “Ancient” Greek History in Polybios’ Historiae: Tenden-
cies and Political Objectives’, SCI 10 (1989–1990) 66–77.
Leidl, Chr., Appians Darstellung des 2. Punischen Krieges in Spanien (Iberike c.1–38
§§ 1–158). Text und Kommentar (München 1996).
Lesky, A., Geschichte der griechischen Literatur (Bern [1957] 1971).
Lewis, J., Speaking of Gods: Representing Oracles in the Roman Greek World, PhD thesis
(Exeter forthcoming).
Lightfoot, J.L., Lucian: On the Syrian Goddess (Oxford 2003).
Lintott, A.W., ‘Cassius Dio and the History of the Late Roman Republic’,
ANRW 2.34.3 (1997) 2497–523.
Lloyd, M., The Agon in Euripides (Oxford 1992).
Lloyd-Jones, H., Parsons, P.J., Supplementum Hellenisticum (Berlin/New York
1983).
Long, T., Repetition and Variation in the Short Stories of Herodotus (Frankfurt a. Main
1987).
Loraux, N., ‘Thucydide a écrit la guerre du Péloponnèse’, Metis 1 (1986) 139–
161.
Louden, B., ‘Pivotal Contrafactuals in Homeric Epic’, ClAnt 12 (1993) 181–198.
Lowe, N.J., ‘Tragic and Homeric Ironies: Response to Rosenmeyer’, in M.S.
Silk (ed.), Tragedy and the Tragic: Greek Theatre and Beyond (Oxford 1996) 520–
533.
———, The Classical Plot and the Invention of Western Narrative (Cambridge 2000).
568 bibliography
———, Introduction to G.W. Mooney (ed.), The Alexandra of Lycophron (Bristol

2004).
Lubbock, P., The Craft of Fiction (London [1921] 1926).
Lynn, J.K., Narrators and Narration in Callimachus (diss. New York 1995).
MacDowell, D.M., Andocides. On the Mysteries (Oxford 1962).
———, Demosthenes. On the False Embassy (Oration 19) (Oxford 2000).
MacLaren, M., ‘Xenophon and Themistogenes’, TAPhA 65 (1934) 240–247.
———, ‘A Supposed Lacuna at the Beginning of Xenophon’s Hellenica’, AJPh
100 (1979) 228–238.
Macleod, C.W., Homer Iliad XXIV (Cambridge 1982).
MacLeod, L., Dolos and Dike in Sophokles’ Elektra (Leiden 2001).
Maehler, H., Die Lieder des Bakchylides, II. Die Dithyramben und Fragmente (Leiden
1997).
Maidment, K.J., Minor Attic Orators, vol. 1 (Cambridge Mass./London 1941).
Makkink, A.D.J., Andokides’ Eerste Rede (Amsterdam 1932).
Malitz, J., ‘Das Interesse an der Geschichte. Die griechischen Historiker und
ihr Publikum’, in H. Verdin, G. Schepens, E. de Keyser (eds.), Purposes of
History. Studies in Greek Historiography from the 4th to the 2nd Centuries BC (Leuven
1990) 323–349.
Manakidou, F., ‘Die Seher in den Argonautika des Apollonios Rhodios’, SIFC 13
(1995) 192–208.
Manuwald, B., Cassius Dio und Augustus (Wiesbaden 1979).
Marasco, G., ‘Erodiano e la crisi dell’impero’, ANRW 2.34.4 (1998) 2837–927.
Marincola, J., ‘Herodotean Narrative and the Narrator’s Presence’, Arethusa 20
(1987) 121–137.
———, ‘Thucydides 1. 22. 2’, CPh 84 (1989a) 216–223.
———, ‘Some Suggestions on the Proem and “Second Preface” of Arrian’s
Anabasis’, JHS 109 (1989b) 186–189.
———, ‘Odysseus and the Historians’, Histos 1 (October 1996; www.dur.ac.uk/
Classics/histos/1997.marincola.html).
———, Authority and Tradition in Ancient Historiography (Cambridge 1997).
Marino, E., ‘Il teatro nel romanzo: Eliodoro e il codice spettacolore’, Materiali
e Discussioni per l’Analisi dei Testi Classici 25 (1990) 203–218.
Markantonatos, A., Tragic Narrative: A Narratological Study of Sophocles’ Oedipus at
Colonus (Berlin/New York 2002).
Martin, W., Recent Theories of Narrative (Ithaca N.Y./London 1987).
McGing, B.C., ‘Appian’s “Mithridateios”’, ANRW 2.34.1 (1993) 496–522.
Merkle, S., ‘Telling the True Story of the Trojan War’, in J. Tatum (ed.), The
Search for the Ancient Novel (Baltimore 1994) 183–196.
Michelini, A.N., Tradition and Dramatic Form in the Persians of Aeschylus (Leiden
1982).
Millar, F., A Study of Cassius Dio (Oxford 1964).
———, ‘Polybius between Greece and Rome’ in J.T.A. Koumoulides (ed.), Greek
Connections: Essays on Culture and Diplomacy (Notre Dame Ind. 1987) 1–18.
Miller, A.M., From Delos to Delphi. A Literary Study of the Homeric Hymn to Apollo
(Leiden 1986).
———, ‘Pindaric Mimesis: The Associative Mode’, CJ 89 (1993) 21–53.
bibliography 569
Miller, N., Menander. Plays and Fragments (London 1987).

Mineur, W.H., Callimachus. Hymn to Delos (Leiden 1984).
Mirhardy, D., Too, Y.L., Isocrates I (Austin 2000).
Misch, G., A History of Autobiography in Antiquity, 2 vols., trans. E.W. Dickes
(London 1950).
Moles, J.L., ‘The Interpretation of the “Second Preface” in Arrian’s Anabasis’,
JHS 105 (1985) 162–168.
———, ‘The Kingship Orations of Dio Chrysostom’, PLLS 6 (1990) 297–375.
———, ‘Dio Chrysostom, Greece, and Rome’, in D. Innes et al. (eds.), Ethics and
Rhetoric: Classical Essays for Donald Russell on his Seventy-Fifth Birthday (Oxford
1995) 177–192.
———, ‘The Dionian Charidemus’, in S. Swain (ed.), Dio Chrysostom: Politics,
Letters, and Philosophy (Oxford 2000) 187–210.
———, ‘Herodotus and Athens’, in E.J. Bakker, I.J.F. de Jong, H. van Wees
(eds.), A Companion to Herodotus (Leiden 2002) 33–52.
Möllendorff, P. von, Auf der Suche nach der verlogenen Wahrheit: Lukians Wahre
Geschichten (Tübingen 2000).
Momigliano, A., ‘Per l’unità logica della Lakedaimoniôn Politeia di Senofonte’,
RFIC 14 (1936) 170–173.
———, The Development of Greek Biography: Four Lectures (Cambridge Mass. 1971).
Montes Cala, J.G., ‘En torno a la “Imposta dramática” en la novella griega:
comentario a una écfrasis de espectáculo en Heliodoro’, Habis 23 (1992)
217–235.
Morales, H., ‘Sense and Sententiousness in the Greek Novels’, in A. Sharrock,
H. Morales (eds.), Intratextuality. Greek and Roman Textual Relations (Oxford
2000) 67–88.
Morgan, J.R., ‘History, Romance and Realism in the Aithiopika of Helio-
dorus’, CA 1 (1982) 221–265.
———, ‘Heliodorus: Ethiopian Story’, in B.P. Reardon (ed.), Collected Ancient Greek
Novels (Berkeley 1989a) 349–588.
———, ‘The Story of Knemon in Heliodoros’ Aithiopika’, JHS 109 (1989b) 99–
113 (reprinted in S. Swain [ed.], Oxford Readings in the Greek Novel [Oxford
1999] 259–285).
———, ‘Reader and Audiences in the Aithiopika of Heliodoros’, Groningen Collo-
quia on the Novel 4 (1991) 85–103.
———, ‘Daphnis and Chloe: Love’s own Sweet Story’, in J.R. Morgan, R. Stone-
man (eds.), Greek Fiction. The Greek Novel in Context (London/New York 1994a)
64–79.
———, ‘The Aithiopika of Heliodoros: Narrative Riddle’, in J.R. Morgan and
R. Stoneman (eds.), Greek Fiction: The Greek Novel in Context (London/New
York 1994b) 97–113.
———, ‘Erotika mathemata: Greek Romance as Sentimental Education’, in
A.H. Sommerstein, C. Atherton (eds.), Education in Greek Fiction (Bari 1996)
163–189.
———, ‘Longus’ Daphnis and Chloe: a Bibliographical Survey, 1950–1995’, ANRW
2.34.3 (1997) 2208–2276.
———, ‘Nymphs, Neighbours and Narrators: a Narratological Approach to
570 bibliography
Longus’, in S. Panayotakis, M. Zimmerman, W. Keulen (eds.), The Ancient

Novel and Beyond (Leiden 2003) 171–189.
———, Longus. Daphnis and Chloe (Warminster 2004).
Morgan, K., Myth and Philosophy from the Presocratics to Plato (Cambridge 2000).
———, ‘Plato’s Dream: Philosophy and Fiction in the Theaetetus’, in S. Panay-
otakis, M. Zimmerman, W. Keulen (eds.), The Ancient Novel and Beyond (Lei-
den 2003) 101–113.
Morrison, D.R., Bibliography of Editions, Translations, and Commentary on Xenophon’s
Socratic Writings 1600–present (Pittsburgh 1988).
Mossman, J.M., ‘Is the Pen Mightier than the Sword? The Failure of Rhetoric
in Plutarch’s Demosthenes’, Histos 3: http://www.dur.ac.uk/Classics/Histos/
1999/ Mossman.html.
Most, G., ‘Hesiod and the Textualization of Personal Temporality’, in G. Arri-
ghetti, F. Montanari (eds.), La componente autobiografica nella poesia Greca e Latina
fra realtà e artificio letterario (Pisa 1993) 73–92.
Moulton, C., Similes in the Homeric Poems (Göttingen 1977).
———, ‘Comic Myth-making and Aristophanes’ Originality’, from ‘Aristoph-
anes’, in T.J. Luce (ed.), Ancient Writers: Greece and Rome (New York 1982)
291–312 (reprinted in Segal 1996: 216–228).
Müller, C.W., ‘Chariton von Aphrodisias und die Theorie des Romans in der
Antike’, A&A 22 (1976) 115–136.
Müller, F.L., Herodian. Geschichte des Kaisertums nach Marc Aurel (Stuttgart 1996).
Murari Pires, F., ‘The Rhetoric of Method (Thucydides, 1. 22 and 2. 35)’, AHB
12 (1998) 106–112.
Murison, C.L., Rebellion and Reconstruction: Galba to Domitian. An Historical Com-
mentary on Cassius Dio’s Roman History Books 64–67 (A.D. 68–96) (Atlanta 1999).
Murnaghan, S., Disguise and Recognition in the Odyssey (Princeton 1987).
Murray, A.T., Demosthenes, vols. 4–6 (Cambridge Mass./London 1936–1939).
Murray, P., ‘Poetic Inspiration in Early Greece’, JHS 101 (1981) 87–100.
Nadon, C., Xenophon’s Prince (Berkeley 2001).
Nagy, G., ‘Transformations of Choral Lyric Traditions in the Context of
Athenian State Theater’, Arion 3 (1994/95) 41–55.
Naiden, F.S., ‘The Prospective Imperfect in Herodotus’, HSPh 99 (1999) 135–
149.
Nakatani, S., ‘A Re-Examination of some Structural Problems in Achilles
Tatius’ Leucippe and Clitophon’, Ancient Narrative 3 (2003).
Nelis, D.P., ‘Demodocus and the Song of Orpheus’, MH 49 (1992) 153–170.
Nesselrath, H.G., ‘Lucian’s introductions’, in D.A. Russell (ed.), Antonine Litera-
ture (Oxford 1990) 111–140.
———, Ungeschehenes Geschehen. ‘Beinahe-Episoden’ im griechischen und römischen Epos
(Stuttgart 1992).
———, ‘Herodot und die Enden der Erde’, MH 52 (1995) 20–44.
Nestle, W., Die Struktur des Eingangs in der attischen Tragödie (Stuttgart 1930).
Norlin, G., Isocrates, vols. 1 and 2 (Cambridge Mass./London 1928–1929).
Nouhaud, M., L’utilisation de l’histoire par les orateurs attiques (Paris 1982).
Nünlist, R., Poetologische Bildersprache in der frühgriechischen Dichtung (Stuttgart/
Leipzig 1998).
bibliography 571
———, ‘Speech within Speech in Menander’, in A. Willi (ed.), The Language of

Comedy (Oxford 2002) 219–259.
O’Sullivan, J.N., Xenophon of Ephesus. His Compositional Technique and the Birth of
the Novel (Berlin/New York 1995).
Packman, Z.M., ‘The Incredible and the Incredulous: the Vocabulary of Dis-
belief in Herodotus, Thucydides, and Xenophon’, Hermes 119 (1991) 399–
414.
Padilla, C., Los milagros de la Vida de Apolonio di Tiana: morfología del relato de milagro
y géneros afines (Cordoba n.d.).
Palm, J., Über Sprache und Stil des Diodoros von Sizilien: Ein Beitrag zur Beleuchtung der
hellenistischen Prosa (Lund 1955).
Panagl, O., Die ‘dithyrambischen Stasima’ des Euripides: Untersuchungen zur Komposition
und Erzähltechnik (Wien 1971).
Parlavantza-Friedrich, U., Täuschungsszenen in den Tragödien des Sophokles (Berlin
1969).
Paulsen, T., Inszenierung des Schicksals. Tragödie und Konödie im Roman des Heliodor
(Trier 1992).
Pearcy, L.T., ‘Theme, Dream, and Narrative: Reading the Sacred Tales of Aelius
Aristides’, TAPhA 118 (1988) 377–391.
Pearson, L., The Lost Histories of Alexander the Great (New York 1960).
———, The Art of Demosthenes (Meisenheim 1976).
———, The Greek Historians of the West: Timaeus and his Predecessors (Atlanta 1987).
Pédech, P., La méthode historique de Polybe (Paris 1964).
Pelling, C.B.R., ‘Plutarch’s Method of Work in the Roman Lives’, JHS 99
(1979) 74–96 (repr. with a postscript in Scardigli 1995: 265–318 and in Pelling
2002).
———, ‘Aspects of Plutarch’s Characterisation’, ICS 13.2 (1988) 257–274 (repr. in
Pelling 2002).
———, ‘The Moralism of Plutarch’s Lives’, in D.C. Innes, H.M. Hine, C.B.R.
Pelling (eds.), Ethics and Rhetoric. Classical Essays for Donald Russell on his Seventy-
Fifth Birthday (Oxford 1995) 205–220 (repr. in Pelling 2002).
———, ‘Is Death the End? Closure in Plutarch’s Lives’, in D.H. Roberts, F.M.
Dunn, D.P. Fowler (eds.), Classical Closure. Reading the End in Greek and Latin
Literature (Princeton 1997a) 228–250 (repr. in Pelling 2002).
———, Plutarco: Filopemene e Tito Flaminino, with an Italian translation by E. Me-
landri (Firenze 1997b).
———, ‘Making Myth Look like History: Plato in Plutarch’s Theseus-Romulus’,
in Pérez Jiménez-García López-Aguilar 1999: 431–443 (expanded version in
Pelling 2002).
———, ‘Fun with Fragments: Athenaeus and the Historians’, in D. Braund,
J. Wilkins (eds.), Athenaeus and his World: Reading Greek Culture in the Roman
Empire (Exeter 2000a) 171–190.
———, Literary Texts and the Greek Historian (London 2000b).
———, Plutarch and History (London 2002).
Pérez Jiménez, A., García López, J., Aguilar, R.M. (eds.), Plutarco, Platón, y
Aristoteles. Actas del V Congreso Internacional de la I.P.S. Madrid-Cuenca, 4–7 de
Mayo de 1999 (Madrid 1999).
572 bibliography
Perkins, D., Is Literary History Possible? (Baltimore/London 1992).

Perkins, J., The Suffering Self: Pain and Narrative Representation in the Early Christian
Era (London 1995).
Pfeiffer, R., Callimachus (Oxford 1949–1953).
Pfeijffer, I.L., Three Aeginetan Odes Of Pindar. A Commentary on Nemean V, Nemean
III, & Pythian VIII (Leiden 1999a).
———, ‘Bacchylides’ Homer, His Tragedy & His Pindar: Intertextuality in Bac-
chylides’, in I.L. Pfeijffer, S.R. Slings (eds.), One Hundred Years of Bacchylides.
Proceedings of a Colloquium Held at the Vrije Universiteit Amsterdam (Amsterdam
1999b) 43–60.
Pfister, M., The Theory and Analysis of Drama, trans. J. Halliday ([München 1977]
Cambridge 1988).
Pietsch, C., ‘Weltdeutung im Orpheusgesang: Zur Bedeutung von Apollonios
Rhodios, Arg. 1.496–511’, Gymnasium 106 (1999) 521–539.
Pomeroy, S., Xenophon Oeconomicus. A Social and Historical Commentary (Oxford
1994).
Pratt, L.H., Lying and Poetry from Homer to Pindar. Falsehood and Deception in Archaic
Greek Poetics (Ann Arbor 1993).
Prince, G., ‘Introduction to the Study of the Narratee’, in J.P. Tompkins (ed.),
Reader-Response Criticism: From Formalism to Post-Structuralism ([1973] London
1980) 7–25.
———, Narratology. The Form and Functioning of Narrative (Berlin 1982).
———, Dictionary of Narratology (Lincoln/London 1987).
Puccini-Delbey, G., ‘Figures du narrateur et du narrataire dans les oeuvres
romanesques de Chariton d’Aphrodisias, Achille Tatius et Apulée’,
in B. Pouderon (ed.), Les personnages du roman grec (Lyon 2001) 87–100.
Puskás, I., ‘Myth or reality? Apollonius of Tyana in India’, ACD 27 (1991) 115–
123.
Raynor, D.H., ‘Moeragenes and Philostratus: Two Views of Apollonius of
Tyana’, CQ 34 (1984) 222–226.
Reardon, B.P., Courants littéraires grecs des IIe et IIIe siècles après J.-C. (Paris 1971).
——— (ed.), Collected Ancient Greek Novels (Berkeley 1989).
———, The Form of Greek Romance (Princeton 1991).
———, ‘Achilles Tatius and Ego-Narrative’, in J.R. Morgan, R. Stoneman
(eds.), Greek Fiction. The Greek Novel in Context (London/New York 1994) 80–
96.
Reeve, M.D., Longus. Daphnis et Chloe (Leipzig 1982).
Rehm, R., ‘Performing the Chorus: Choral Action, Interaction, and Absence
in Euripide’, Arion 4 (1996/97) 45–60.
Reinhold, M., From Republic to Principate. An Historical Commentary on Cassius Dio’s
Roman History Books 49–52 (36–29 B.C.) (Atlanta 1988).
Reitzenstein, R., Hellenistische Wundererzählungen (Leipzig 1906).
Richardson, J.S., Appian. Wars of the Romans in Iberia (Iberike) (Warminster 2000).
Richardson, N.J., The Homeric Hymn to Demeter (Oxford 1974).
Richardson, S., The Homeric Narrator (Nashville 1990).
Ricoeur, P., Time and Narrative. Volume I ([Paris 1983] Chicago–London 1984).
Rijksbaron, A., ‘How Does a Messenger Begin his Speech? Some Observa-
bibliography 573
tions on the Opening Lines of Euripidean Messenger-Speeches’, in J.M.

Bremer, S. Radt, C.J. Ruijgh (eds.), Miscellanea tragica in honorem J.C. Kamer-
beek (Amsterdam 1976) 293–308.
———, ‘Chariton 8, 1, 4 und Aristot. Poet. 1449b28’, Philologus 128 (1984) 306–
307.
Roberts, D.H., ‘Different Stories: Sophoclean Narrative(s) in the Philoctetes’,
TAPhA 119 (1989) 161–176.
Romberg, B., Studies in the Narrative Technique of the First-Person Novel (Stockholm
1962).
Romm, J., The Edges of the Earth in Ancient Thought: Geography, Exploration, and
Fiction (Princeton 1992).
Rommel, H., Die naturwissenschaftlich-paradoxographischen Exkurse bei Philostratos,
Heliodoros und Achilleus Tatios (Stuttgart 1923).
Rood, T.C.B., Thucydides: Narrative and Explanation (Oxford 1998).
———, ‘Thucydides’ Persian Wars’, in C.S. Kraus (ed.), The Limits of His-
toriography: Genre and Narrative in Ancient Historical Texts (Leiden 1999) 141–
168.
Roques, D., Hérodien. Histoire des empereurs romains. De Marc Aurèle à Gordien III
(Paris 1990).
Rosenmeyer, T.G., The Art of Aeschylus (Berkeley 1982).
———, ‘Beginnings in Plutarch’s Lives’, YCS 29 (1992) 205–230.
Rösler, W., ‘Die “Selbsthistorisierung” des Autors. Zur Stellung zwischen
Mündlichkeit und Schriftlichkeit’, Philologus 135 (1991) 215–220.
Rowe, C., ‘Platonic Irony’, Nova Tellus 5 (1987) 83–101.
Russell, D.A., Dio Chrysostom, Orations 7, 12, 36 (Cambridge 1992).
———, ‘Self-disclosure in Plutarch and in Horace’, in G.W. Most, H. Peters-
mann, A.M. Ritter (eds.), Philanthropia kai Eusebeia. Festschrift für Albrecht Dihle
zum 70. Geburtstag (Göttingen 1993) 426–437.
Rutherford, R.B., ‘Learning from History: Categories and Case-Histories’, in
R. Osborne, S. Hornblower (eds.), Ritual, Finance, Politics: Athenian Democratic
Accounts Presented to David Lewis (Oxford 1994) 54–68.
———, The Art of Plato (London 1995).
Rütten, U., Phantasie und Lachkultur: Lukians ‘Wahre Geschichten’ (Tübingen 1997).
Sacks, K., Polybius on the Writing of History (Berkeley 1981).
———, Diodorus Siculus and the First Century (Princeton 1990).
Saïd, S., ‘Dio’s Use of Mythology’, in S. Swain (ed.), Dio Chrysostom: Politics,
Letters, and Philosophy (Oxford 2000) 161–186.
———, Trédé, M., le Boulluec, A., Histoire de la littérature grecque (Paris 1997).
Scarcella, A.M., ‘La struttura del romanzo di Senofonte Efesio’, in La struttura
della fabulazione antica (Genoa 1979) 89–113 (repr. in Romanzo e romanzieri. Note
di narratologia greca [Napoli 1993] 165–184).
———, ‘Caratteri e funzione delle gnomai in Achille Tazio’, Euphrosyne 15 (1987)
269–280 (repr. in Romanzo e romanzieri. Note di narratologia greca [Napoli 1993]
153–164).
Scardigli, B. (ed.), Essays on Plutarch’s Lives (Oxford 1995).
Schade, G.A., Lykophrons ‘Odyssee’: Alexandra 648–819 (Berlin/New York 1999).
Schadewaldt, W., Iliasstudien ([1938] Darmstadt 1966).
574 bibliography
Schäublin, C., ‘Konversionen in antiken Dialogen?’, in C. Schäublin (ed.),

Catalepton: Festschrift für Bernhard Wyss zum 80. Geburtstag (Basel 1985) 117–
131.
Schepens, G., L’‘autopsie’ dans la méthode des historiens grecs du Ve siècle avant J.-C.
(Bruxelles 1980).
———, ‘Polemic and Methodology in Polybius’ Book XII’, in H. Verdin,
G. Schepens, E. de Keyser (eds.), Purposes of History: Studies in Greek Histori-
ography from the Fourth to the Second Centuries BC (Leuven 1990) 39–61.
Schmid, W., Der Atticismus in seinen Hauptvertretern (Stuttgart 1887–1897).
Schmidt, H.W., ‘Die Struktur des Eingangs’, in W. Jens (ed.), Die Bauformen der
griechischen Tragödie (München 1971) 1–46.
Schmidt, J.-U., Adressat und Paraineseform. Zur Intention von Hesiods ‘Werken und
Tagen’ (Göttingen 1986).
Schmidt, M.G., ‘Die “zeitgeschichtlichen” Bücher im Werke des Cassius Dio—
von Commodus zu Severus Alexander’, ANRW 2.34.3 (1997) 2591–2649.
———, ‘Politische und persönliche Motivation in Dios Zeitgeschichte’, in:
M. Zimmermann (ed.), Geschichtsschreibung und politischer Wandel im 3. Jh. n.
Chr. (Stuttgart 1999) 93–117.
Schmidt, W., Der Deus ex machina bei Euripides (diss. Tübingen 1963).
Schmitz, T.A., ‘“I Hate All Common Things”: The Reader’s Role in Calli-
machus’ Aetia Prologue’, HSPh 99 (1999) 152–178.
———, ‘Plausibility in the Greek Orators’, AJPh 121 (2000) 47–77.
———, ‘Narrator and Audience in Philostratus’ Lives of the Sophists’, in E. Bowie,
J. Elsner (eds.), Philostratus (Cambridge forthc.).
Scholes, R., Kellogg, R., The Nature of Narrative (Oxford 1966).
Scodel, R., ‘Self-Correction, Spontaneity, and Orality in Archaic Poetry’, in
I. Worthington (ed.), Voice into Text. Orality and Literacy in Ancient Greece (Leiden
1996) 59–79.
Seeck, G.A., ‘Dion Chrysostomos als Homerkritiker (Or. 11)’, RhM 133 (1990)
97–107.
Segal, C.P., ‘Tragic Beginnings: Narration, Voice, and Authority in the Pro-
logues of Greek Drama’, YCS 29 (1992) 85–112.
Segal, E. (ed.), Oxford Readings in Aristophanes (Oxford 1996).
Sens, A., Theocritus: Dioscuri (Idyll 22) (Göttingen 1997).
Shapiro, S.O., ‘Herodotus and Solon’, ClAnt 15 (1996) 348–364.
Shrimpton, G.S., ‘Accuracy in Thucydides’, AHB 12 (1998) 71–82.
Sidebottom, H., ‘The Date of the Composition of Herodian’s History’, AC 66
(1997) 271–276.
———, ‘Herodian’s Historical Methods and Understanding of History’, ANRW
2.34.4 (1998) 2775–2836.
Silk, M.S. (ed.), Tragedy and the Tragic: Greek Theatre and Beyond (Oxford 1996).
———, Aristophanes and the Definition of Comedy (Oxford 2000).
Sommerstein, A.H., Aeschylus: Eumenides (Cambridge 1989).
Speyer, W., Die literarische Fälschung im heidnischen und christlichen Altertum: ein Versuch
ihrer Deutung (München 1971)
Stadter, P.A., Arrian of Nicomedia (Chapel Hill 1980).
———, ‘The Proems of Plutarch’s Lives’, ICS 13.2 (1988) 275–295.
bibliography 575
———, ‘Paradoxical paradigms: Lysander and Sulla’, in P.A. Stadter (ed.),

Plutarch and the Historical Tradition (London/New York 1992) 41–55.
———, ‘The Rhetoric of Virtue in Plutarch’s Lives’, in L. van der Stockt (ed.),
Rhetorical Theory and Praxis in Plutarch (Louvain/Namur 2000) 493–510.
Stanzel, F.K., Theorie des Erzählens (Göttingen [1979] 1982).
Stehle, E., Performance and Gender in Ancient Greece: Nondramatic Poetry in its Setting
(Princeton 1997).
Stoessl, F., ‘Prologos II’, RE 23.2 (1959) 2312–2345.
Strasburger, H., Homer und die Geschichtsschreibung (Heidelberg 1972).
Strohm, H., ‘Zur Gestaltung Euripideischer Prologreden’, GB 6 (1977) 113–132.
Sturrock, J., The Language of Autobiography: Studies in the First Person Singular (Cam-
bridge 1993).
Stylianou, P.J., A Historical Commentary on Diodorus Siculus, Book 15 (Oxford 1998).
Svenbro, J., La parole et le marbre. Aux origines de la poétique grecque (Lund 1976).
Swain, S., ‘The Reliability of Philostratus’ Lives of the Sophists’, ClAnt 10 (1991)
148–163.
———, ‘Dio and Lucian’, in J. Morgan, R. Stoneman (eds.), Greek Fiction: the
Greek Novel in Context (London/New York 1994) 166–180.
———, Hellenism and Empire: Language, Classicism, and Power in the Greek World, A.D.
50–250 (Oxford 1996).
———, ‘Defending Hellenism: Philostratus, In honour of Apollonius’, in M. Ed-
wards et al. (eds.), Apologetics in the Roman Empire: Pagans, Jews, and Christians
(Oxford 1999) 157–196.
Swan, P.M., ‘How Cassius Dio Composed his Augustan Books: Four Studies’,
ANRW 2.34.3 (1997) 2524–2557.
Taplin, O., The Stagecraft of Aeschylus (Oxford 1977).
———, Homeric Soundings. The Shaping of the Iliad (Oxford 1992).
——— (ed.), Literature in the Greek and Roman Worlds (Oxford 2000).
Tarrant, H., ‘Orality and Plato’s Narrative Dialogues’, in I. Worthington
(ed.), Voice into Text: Orality and Literacy in Ancient Greece (Leiden 1996) 129–
147.
Tatum, J., Xenophon’s Imperial Fiction (Princeton 1989).
Thalmann, W.G., Conventions of Form and Thought in Early Greek Epic Poetry
(Baltimore/London 1984).
Theodorakopoulos, E.M., ‘Epic Closure and its Discontents in Apollonius’
Argonautica’, in M.A. Harder, R.F. Regtuit, G.C. Wakker (eds.), Genre in
Hellenistic Poetry (Groningen 1998) 187–204.
Thesleff, H., Studies in the Styles of Plato (Helsinki 1967).
———, Studies in Platonic Chronology (Helsinki 1982).
Thomas, R., Herodotus in Context. Ethnography, Science, and the Art of Persuasion
(Cambridge 2000).
Tigerstedt, E.N., The Legend of Sparta in Classical Antiquity: I (Stockholm 1965).
Todd, S.C., The Shape of Athenian Law (Oxford 1993).
———, Lysias (Austin 2000).
Tonnet, H., Recherches sur Arrien, sa personnalité et ses écrits atticistes, 2 vols. (Amster-
dam 1988).
Trapp, M.B., ‘Sense of Place in the Orations of Dio Chrysostom’, in D. Innes
576 bibliography
et al. (eds.), Ethics and Rhetoric: Classical Essays for Donald Russell on his Seventy-
Fifth Birthday (Oxford 1995) 163–175.
———, ‘Plato in Dio’, in S. Swain (ed.), Dio Chrysostom: Politics, Letters, and
Philosophy (Oxford 2000) 213–239.
Tuplin, C., The Failings of Empire (Stuttgart 1993).
Usher, S., Isocrates: Panegyricus, To Nicocles (Warminster 1990).
———, Demosthenes On the Crown (Warminster 1993).
———, Greek Oratory. Tradition and Originality (Oxford 1999).
Van Hook, L., Isocrates, vol. 3 (Cambridge Mass./London 1945).
Varcoe, G., The Intrusive Narrator. Fielding, Thackeray and the English Novel (diss.
Uppsala 1972).
van der Veen, J.E., The Significant and the Insignificant. Five Studies in Herodotus’
View of History (Amsterdam 1996).
Velardi, R., ‘Scrittura e tradizione dei dialoghi in Platone’, in G. Casertano
(ed.), La struttura del dialogo platonico (Napoli 2000) 108–139.
Vercruysse, M., ‘A la recherche du mensonge et de la vérité: La fonction des
passages méthodologiques chez Polybe’, in H. Verdin, G. Schepens, E. de
Keyser (eds.), Purposes of History: Studies in Greek Historiography from the Fourth to
the Second Centuries BC (Leuven 1990) 17–38.
Verdenius, W.J., Commentaries on Pindar, Volume I: O. 3, 7, 12, 14 (Leiden 1987).
Verity, A., Hunter, R., Theocritus, Idylls (Oxford 2002).
Vince, C.A., Vince, J.H., Demosthenes, vol. 2 (Cambridge Mass./London 1926).
Vince, J.H., Demosthenes, vols. 1 and 3 (Cambridge Mass./London 1930–1935).
Wakker, G.C., ‘Emphasis and Affirmation. Some Aspects of mēn in Tragedy’, in
A. Rijksbaron (ed.), New Approaches to Greek Particles (Amsterdam 1997) 209–
231.
Walbank, F.W., ‘The Geography of Polybius’, C&M 9 (1948) 155–182.
———, A Historical Commentary on Polybius, 3 vols. (Oxford 1957–1979).
———, ‘Polemic in Polybius’, JRS 52 (1962) 1–12.
———, Polybius (Berkeley 1972).
———, ‘Athenaeus and Polybius’, in D. Braund, J. Wilkins (eds.), Athenaeus and
his World: Reading Greek Culture in the Roman Empire (Exeter 2000) 161–170.
Walker, A.D., ‘Enargeia and the Spectator in Greek Historiography’, TAPhA 123
(1993) 353–377.
Walsh, G.B., The Varieties of Enchantment: Early Greek Views of the Nature and
Function of Poetry (Chapel Hill 1984).
Wardman, A.E., Plutarch’s Lives (London 1974).
Wellek, R., ‘The Fall of Literary History’, in R. Kosselleck, W.-D. Stempel
(eds.), Geschichte—Ereignis—Erzählung (München 1973) 427–440.
West, M.L., Hesiod. Theogony (Oxford 1966).
———, ‘Cynaethus’ Hymn to Apollo’, CQ 25 (1975) 161–170.
———, Hesiod. Works and Days (Oxford 1978).
———, ‘An Unrecognized Injunctive Usage in Greek’, Glotta 67 (1989) 135–138.
———, Aeschyli Tragoediae (Stuttgart 1990).
———, The East Face of Helicon. West Asiatic Elements in Greek Poetry and Myth
(Oxford 1997).
West, S.R., ‘Notes on the Text of Lycophron’, CQ 33 (1983) 114–135.
bibliography 577
———, ‘Lycophron Italicised’, JHS 104 (1984) 127–151.

———, ‘Christa Wolf ’s Kassandra: A Classical Perspective’, Oxford German Studies
20/21 (1991/2) 164–185.
———, ‘Lycophron’s Alexandra: “Hindsight as Foresight Makes No Sense?”’, in
M. Depew, D. Obbink (eds.), Matrices of Genre: Authors, Canons, and Society
(Cambridge Mass./London 2000) 153–166.
Westlake, H.D., ‘legetai in Thucydides’, Mnemosyne 30 (1977) 345–362.
Wheeler, G., ‘“Sing, Muse …”: the Introit from Homer to Apollonius’, CQ 52
(2002) 33–49.
White, H., Appian’s Roman History. With an English Translation, 4 vols. (Lon-
don/Cambridge Mass. 1912–1913).
White, H., ‘Notes on Theocritus’, Myrtia 14 (1999) 39–56.
Whitman, C.H., Homer and the Heroic Tradition (Cambridge Mass. 1958).
Whitmarsh, T., Greek Literature and the Roman Empire: the Politics of Imitation (Ox-
ford 2001a).
———, Achilles Tatius. Leucippe and Clitophon (Oxford 2001b).
———, ‘Performing Heroics: Language, Landscape and Identity in Philostratus’
Heroicus’, in E. Bowie, J. Elsner (eds.), Philostratus (Cambridge forthc.).
Whittaker, C.R., Herodian, 2 vols. (London/Cambridge 1969–1970).
Williams, B., Shame and Necessity (Berkeley 1993).
Wilson, N.G., ‘Indications of Speaker in Greek Dialogue Texts’, CQ 20 (1970)
305.
Winkler, J.J., ‘The Mendacity of Kalasiris and the Narrative Strategy of Helio-
doros’ Aithiopika’, YCS 27 (1982) 93–158 (repr. in S. Swain [ed.], Oxford
Readings in the Greek Novel [Oxford] 286–350).
Winnington-Ingram, R.P., Sophocles. An Interpretation (Cambridge 1980).
Wouters, A., ‘The eikones in Longus’ Daphnis and Chloe IV 39,2: ‘Beglaubi-
gungsapparat’?’, SEJG 31 (1989–1990) 465–479.
Wray, D., ‘Apollonius’ Masterplot: Narrative Strategy in Argonautica 1’, in
M.A. Harder, R.F. Regtuit, G.C. Wakker (eds.), Apollonius Rhodius (Leuven
2000) 239–265.
Yamagata, N., ‘The Apostrophe in Homer as Part of the Oral Technique’,
BICS 36 (1989) 91–103.
Young, D.C., Three Odes of Pindar. A Literary Study of Pythian 11, Pythian 3, and
Olympian 7 (Leiden 1968).
Yunis, H., Demosthenes, On the Crown (Cambridge 2001).
Zanker, G., ‘Enargeia in the Ancient Criticism of Poetry’, RhM 124 (1981) 297–
311.
———, Realism in Alexandrian Poetry (London 1987)
———, ‘Pictorial Description as a Supplement for Narrative: the Labour of
Augeas’ Stables in Heracles Leontophonos’, AJPh 117 (1996) 411–423.
Zanker, P., The Mask of Socrates: the Image of the Intellectual in Antiquity, trans.
A. Shapiro (Berkeley 1995).
Zetzel, J.E.G., ‘On the opening of Callimachus, Aetia II’, ZPE 42 (1981) 31–33.
Ziegler, K., rev. H. Gärtner, Plutarchi Vitae Parallelae iv: Indices (Leipzig 1980).
Zimmermann, M., ‘Der Verlust und die Neuerfindung von Wirklichkeiten:
Zur Geschichtsdeutung und Darstellung bei Herodian’, in H. Brandt (ed.),
578 bibliography
Gedeutete Realität. Krisen Wirklichkeiten, Interpretationen (3.–6. Jh. n. Chr.) (Stutt-

gart 1999a) 13–46.
———, ‘Herodians Konstruktion der Geschichte und sein Blick auf das stadt-
römische Volk’, in M. Zimmermann (ed.), Geschichtsschreibung und politischer
Wandel im 3. Jh. n. Chr. (Stuttgart 1999b) 119–143.
———, Kaiser und Ereignis. Studien zum Geschichtswerk Herodians (München 1999c).
THEMATIC INDEX
The index comprises the principal narratological questions discussed in the

present volume. Its main purpose is to enable readers to follow these questions
through the various chapters. Occasionally, the index also refers to discussions
of the topic which do not expressly use the actual catchword. References are
to pages, footnotes are not specifically singled out.
Analepsis, XV–XVII, 7–10, 28, 33, Ellipsis, XVIII, 310

37–39, 112, 133, 134, 140, 160, Evaluation (implicit), 3, 17, 134, 137,
226, 269, 275, 283, 297–300, 302, 141, 205, 273
304, 305, 321, 322, 329, 360, 362, Evaluative Comments (also: Judg-
486, 491, 492, 503, 518, 519, 526, mental Language), 16, 19, 20, 29,
529, 533, 547, 551 34, 37, 51, 52, 65, 66, 68, 84, 104,
Argument Function, XV–XVII, 10, 105, 118, 125, 130, 132, 138–140,
21, 39, 46, 263, 265, 459, 492, 147, 150, 172, 177, 180, 181, 191,
525, 542 192, 197, 205, 237, 262, 303, 321,
Authentication of Knowledge, 14, 330, 381, 397, 414, 484, 497, 499,
15, 46, 62, 106, 129, 266, 379, 395, 500, 512, 513, 530, 547
548, 553 Explanation, 28, 30, 46, 47, 65, 76,
Autopsy, 50, 57, 106, 114, 117, 118, 85, 94, 117, 150, 151, 154, 184, 194,
143, 144, 153, 167, 170, 183, 189, 208, 398, 405, 408, 496, 511, 529,
198, 201, 202, 206, 238, 239, 245, 538, 550; (see also ‘Narratorial
247, 250, 258, 259, 261, 262, 273, devices, gar-clause’)
274, 377, 382, 386, 424, 426, 434,
460, 463, 466, 467, 475, 476, 553; Fabula, XVI–XVIII, 9, 220, 222,
(see also ‘Omniscience’) 223, 227, 239, 309, 428, 429, 436,
444, 445, 526
Blending of Narratorial Voices Fabula-Time, XVIII, 226
(also: Merging), 21, 22, 57–59, Focalization, XII, 53, 60, 269, 273,
66, 67, 290, 386, 441, 552; (see 276, 277, 424, 428, 430, 463, 527
also ‘Narrative, Coordination of Secondary (also: Embedded),
Primary and Secondary’) XVI, 9, 14, 53, 57, 102, 196,
206
Characterization, XII, XV, 7, 17, 47, Focalizer, XVI, 7, 511, 527
310, 338, 352, 362, 363, 383, 406, Frame, Narrative, 2, 32, 33, 85,
522 128, 357, 358, 360, 361, 364, 365,
368, 378, 380, 387, 493, 526, 545,
Description, XII, XV, 8, 17, 28, 46, 547
76, 78, 87, 88, 96, 113, 118, 142, No Return to, 84, 90, 366, 383,
148, 204, 208, 213, 236, 261, 285, 452, 453, 456, 494, 547
434, 502, 532, 547 Frequency, XVI, XVIII
580 thematic index
Implied Author, 3, 4, 508 334, 336, 338, 341, 361, 367,

Interlace Technique, XVI, 482, 490 368, 380, 384, 386, 388, 399,
404, 407, 408, 411, 417, 423,
Key Function, XV, XVI, 10, 21, 265, 424, 430–433, 435, 436, 441,
459, 551 442, 451–464, 466, 467, 469–
474, 479, 486, 487, 490–494,
Metanarrative Comments, 16, 29, 496, 500–508, 510, 518, 519,
59, 65, 68, 117, 240, 247, 248, 523–526, 530–532, 534, 535,
251, 253, 254, 257, 262, 285, 299, 539–543, 549, 550, 551
301, 303, 319, 321, 326, 327, 329, Secondary, XVI, 4–6, 8, 10, 22–
330, 333, 334, 336, 337, 340, 343, 24, 60, 68, 70, 72–75, 124, 126,
344, 346, 349, 352, 461, 482, 515, 173, 209, 226, 227, 248, 249,
547; (see also ‘Narrator, Self- 263, 265, 269, 272, 277, 279,
conscious’) 297, 299, 308, 312, 326, 380,
Cross-references, 49, 103, 117, 383, 384, 399, 453–455, 457–
152, 154, 171, 179, 192, 203, 459, 462, 463, 466, 467, 469,
298, 299, 495, 547 471–473, 487, 492, 493, 495,
Structurizing Comments (also: 496, 500, 501, 504, 505, 524,
Narrator as Organizer), 29, 48, 534, 535, 539–541, 550, 551
49, 103, 107, 112, 117, 131, 134, Story Adapted to, 23, 252, 522
151, 171, 177, 178, 180, 185, 191, Suppressed Primary, 8
203, 204, 262, 284, 286, 289, Tertiary, XVI, 24, 308, 383, 384,
319, 341, 481, 490 434, 453, 454, 462, 463, 471,
Motivation, 505, 520
Actorial, XV, 486 Narration,
Narratorial, XV, XVII Iterative, 26, 27, 40, 41, 77, 126
Omnitemporal, 26, 27
Narratee, passim Prior, XVI, 3, 71, 77, 307, 455
Covert, XVI, 5, 22, 30, 442, 453, Second-person, 3, 5, 16, 20, 36,
479, 493, 507, 549 40, 65, 74, 263, 264, 272, 290,
External, 5, 6, 68, 91, 219, 318, 311, 328, 330, 487, 549
319, 322, 353, 470, 486, 493, Simultaneous, XVI, 3, 26, 27, 32,
507, 524, 525 34, 40, 41, 77, 237, 285, 361
Internal, 5, 6, 22, 32, 34, 36, 161, Subsequent, XVI, 3, 13, 26, 27,
227, 318, 327, 328, 330, 347, 32, 36, 40, 41, 64, 76, 237, 361
361, 525, 541, 542, 549 Narrative (also: Main Story), passim
Overt, XVI, 5, 31, 34, 74, 110, Boundaries of (also: Demarca-
173, 219, 319, 321, 347, 371 tion, Beginning/End), 18, 21,
Primary, XV, XVI, 4, 5, 7, 8, 22, 30, 35, 43–45, 75, 77, 78, 84–
23, 36, 40, 60, 67, 69, 72–75, 86, 89, 91, 93, 95, 111, 131, 133,
77, 83, 90, 93, 109, 110, 112, 156, 190, 214, 216, 235, 237, 238,
115, 121–123, 125, 126, 128, 149, 244, 251, 254, 257, 262–264,
156, 157, 159, 173, 183, 196, 273, 276, 285, 286, 299, 302,
198, 207, 209, 261, 218, 219, 303, 312, 318, 319, 321, 326–
224, 228, 237, 241, 249, 250, 328, 333, 334, 337, 340, 343,
265, 279, 292, 299, 308, 312, 344, 346, 347, 349, 352, 361,
314, 318, 319, 326–328, 330, 369, 371, 373, 374, 378, 552
thematic index 581
Coordination of Primary and 28, 36, 39, 41, 72, 95, 96, 102,
Secondary Narratives, 23, 24, 120, 205, 366, 439, 458, 489,
60, 74, 124, 126, 228, 230–232, 490, 494, 545, 546, 547
521, 525; (see also ‘Blending of Dramatized, 2, 132, 546, 548, 549
Narrative Voices’) External, XVII, 1–3, 5, 6, 9, 14,
Dialogical, 235, 236, 243, 249, 19, 20, 22, 27, 35, 43, 63, 65,
250, 255, 256, 264, 265, 267, 68, 72, 84, 85, 88, 89, 93, 95,
276, 286, 287, 297, 298, 301– 101, 107, 116, 149, 170, 173, 176,
304, 305 189, 218, 219, 235, 236, 239,
Functions of Secondary Narra- 244–246, 249, 258–260, 288,
tive, 10 (general) 298, 319, 329, 339, 340, 342,
explanatory, 10, 58, 143, 235, 360, 385, 392, 424, 458, 465,
243, 304, 305, 521, 540, 551 466, 468, 470, 479, 489, 496,
predictive, 10 497, 506, 523, 524, 527, 545–
thematic, 10, 491, 502, 541 548, 549
persuasive, 10, 21, 39, 108, 126, Internal, XVII, 1, 3, 5, 6, 9, 19,
243, 304, 506, 525, 551 25, 31, 32, 34, 35, 67, 68, 85,
distractive, 10, 70, 248, 540, 88, 90–93, 101, 108, 116, 148,
551 149, 173, 189, 191, 201, 226–
Secondary (also: Embedded), 228, 236, 238, 239, 244, 245,
XI, XV, XVI, XVII, XVIII, 258, 261, 265, 288, 289, 298,
7, 8, 10, 18–21, 23, 29, 32, 33, 310, 318, 319, 320, 321, 322,
57, 67, 71, 74, 83, 93, 108, 119, 326, 330, 333, 334, 338, 339,
123, 124, 126, 143, 149, 160– 344, 345, 347, 352, 360, 361,
162, 173, 226, 227, 243, 254, 379, 380, 382, 386, 387, 446,
269, 280, 281, 297, 314, 357, 451, 454, 459, 460, 462, 463,
361, 365, 367–371, 374–376, 465–468, 470, 493, 494, 496,
382, 384, 385, 400, 403, 404, 497, 506, 507, 524–527, 532,
431, 432, 441, 454, 455, 459, 533, 540–542, 545–548, 549
461, 462, 466, 467, 469–472, Overt, XVII, 2, 13, 25, 29, 31,
486, 487, 490, 491, 494, 497, 34, 41, 43, 61, 63–68, 71, 80,
502, 503, 507, 510, 518–520, 84, 85, 102, 110, 120, 132, 148–
523–526, 530, 535, 540–542, 151, 156, 163–165, 173–175, 178,
546–549, 551, 552; (see also 183, 185, 187, 190, 198, 203,
‘Narrator, Secondary’ and 206, 210, 216, 217, 219, 250,
‘Tertiary’) 262, 308, 310, 319–322, 330,
Monological, 235, 236, 255, 267, 334, 338–340, 344, 345, 347,
282, 287, 297–300, 302, 305 350, 351, 353, 361, 364, 381,
Punctuated, 93, 235, 237, 243, 387, 396, 423, 424, 436, 439,
250, 267, 277, 281, 282, 291, 442, 479, 481, 533, 545–547,
293, 297, 300–303, 305, 318, 548
330, 339, 344, 347, 369, 552 Own Time of, References to, 13,
Narrator, passim 30, 33, 36, 56, 65, 68, 101, 102,
Alter ego of, 22, 58–60, 113, 114, 134, 157, 184, 197, 382, 391,
164, 198, 451, 463, 468–472, 392, 394, 398, 405, 406, 547
548 Primary, XV, XVII, 2, 4, 5, 7, 8,
Covert, XVII, 2, 15, 17, 24, 26, 10, 13, 19, 21, 23, 24, 36, 60,
582 thematic index
63, 65, 67–69, 71, 72, 74, 75, 427, 431, 432, 454, 459, 461,
83, 85, 88–92, 95, 109, 110, 112, 462, 520, 524
115, 123, 125, 129, 132, 143, 144, Narratorial Devices,
149, 161, 163, 165, 173, 183, Anonymous Interlocutor, 5, 53,
198, 203, 205, 210, 216, 218, 110, 120, 137, 139, 167, 174, 184,
219, 223–228, 230, 232, 249, 196, 208, 375, 376, 381, 386,
276, 299, 308, 314, 317, 318, 394, 399, 415, 446, 550
327–330, 336, 359–361, 377, Anonymous Spokesmen, 14,
380, 382, 383, 385, 392, 394, 28, 50, 56, 65, 67, 109, 114,
400, 401, 403, 424, 426–430, 141, 144–146, 168, 193, 206,
433–435, 441, 451–456, 459, 221, 249, 379, 391, 395, 426,
462, 463, 465, 467–469, 472, 439, 453; (see also ‘Narrators,
473, 479, 486, 487, 489–496, Reported’)
501, 506–508, 510, 511, 514, Anonymous Witness, 5, 22, 53,
516, 518, 519, 523, 525–532, 73, 74, 110, 137, 144, 173, 184,
534, 538, 540–543, 545–549, 387, 445, 550
552 ‘aporia’ motif, 16, 21, 65, 67, 87,
Reported, 69, 107–109, 123, 143, 92, 301, 348, 442–445, 447, 482
327, 328, 330, 335, 346, 383, Apostrophe, 3, 16, 27, 31, 33,
385, 427, 428, 431, 434, 439, 35, 45, 48, 64–68, 72, 74,
487, 548, 552; (see also ‘Nar- 85, 88, 216, 263, 264, 272,
ratorial Devices, Anonymous 285, 311, 318, 322, 353, 547,
Spokesmen’) 549
Secondary, XVII, 2, 4, 6, 8, 10, gar-clause, 22, 110, 173, 183, 196,
18–21, 23, 24, 39, 57, 58, 60, 208, 217, 238, 242, 248, 253,
63, 65, 66, 68, 69, 71, 73, 257, 302, 319, 321, 326, 327,
77, 83, 85, 88, 90, 107, 108, 329, 330, 333, 337, 340, 343,
123, 128, 132, 143, 160, 173, 349, 437, 483, 550
183, 198, 209, 218, 223, 226– Generalization, 118, 134, 135, 139,
230, 232, 238, 244, 255, 267, 174, 192, 213, 217, 258, 350,
269, 288, 297, 300, 301, 308, 398, 483, 484, 490, 496, 500,
326, 329, 331, 335, 351, 360, 501, 511, 512, 530, 531, 538,
361, 366, 368, 379, 382–384, 547
386, 388, 399, 400, 403, 426, Gnome, 14, 16, 29, 53, 68, 90,
441, 455, 458, 459, 461, 463, 105, 136, 196, 208, 214, 217,
467–473, 494, 497, 502, 506, 547
507, 524, 532, 533, 545–549, ‘If not’-situation, 28, 37, 38, 41,
552 54, 118, 138, 173, 184, 196, 208,
Self-conscious, XVII, 2, 25, 35, 209, 515, 547
43, 61, 63, 80, 87, 92, 165, 240, Indefinite Second Person, 22, 54,
251, 254, 299, 300, 409, 425, 74, 96, 110, 158, 399, 493, 495,
442, 444, 445, 502; (see also 496, 550
‘Metanarrative Comments’) Muse invocation, 13–15, 18, 27,
Suppressed Primary, 8, 63, 83, 28, 31, 43, 46–48, 50, 56, 58,
254, 358, 467–469, 547 64, 68, 95, 101, 366, 443, 553
Tertiary, XVII, 2, 66, 69, 71, 89, Negation, Presentation Through,
300, 308, 365, 366, 383, 403, 5, 22, 28, 37, 41, 54, 111, 118,
thematic index 583
137–139, 141, 184, 203, 242, Paralepsis, XVII, 496, 499

243, 248, 253, 320, 336, 386, Paralipsis, XVII, 496, 506
550 Pause, XVIII, 8
Questions (incl. Rhetorical), 16, Periphrastic Denomination, XVII
22, 41, 49, 56, 65, 68, 70, 75, Prolepsis, XVI, XVII, 7–9, 14, 28,
77, 78, 80, 93, 107, 110, 121, 33, 37, 38, 46, 60, 95, 102, 106,
132, 136, 137, 141, 148, 150, 151, 112, 131, 133–135, 143, 204, 226,
157, 158, 196, 208, 241–243, 227, 269–271, 275, 278–280, 287,
248, 251, 253, 263, 272, 301, 288, 293, 297, 298, 304, 305, 307,
321, 340, 347, 348, 351, 369, 308, 313, 334, 497, 514, 526, 529,
381, 415, 420, 473, 483, 485, 532, 547, 551
550
Narrator-Text, XVI, XVII, 28, 29, Retardation, XVIII
32, 36, 46, 51, 111, 139 Rhythm, XVIII, 7
Ring-composition, 18, 21, 112, 244
Omission, Explicit, 37, 49, 104, 178,
194, 248, 256, 319, 340, 351, 428, Scene, XVIII, 44, 112
429, 532, 534 Seed, XVIII, 320
Abbruchsformel, 21, 27, 49, 68, 214, Sources, 47, 50, 56, 61, 62, 67–70,
215, 217, 220, 256 78, 80, 106, 109, 117, 120, 123,
‘recusatio’ motif, 16, 21, 165 143, 144, 153, 157, 162, 167, 168,
Omnipresence, 14, 19, 20, 43, 95, 170, 174, 182, 189, 193, 194, 205,
101, 102, 106, 196, 206, 252 221, 308, 309, 371, 386, 394, 396,
Omniscience, 14, 15, 19, 20, 28, 33, 406, 424, 425, 427, 430, 435, 439,
43, 51, 60, 64, 68, 80, 85, 86, 95, 499, 524, 548, 549
101, 102, 106, 196, 206, 239, 246, Speech, XII, XVI, XVII, 2, 7–9, 15,
260, 278, 298, 307, 363, 364, 374, 20, 29, 31–33, 38, 39, 57, 72, 87,
427, 479, 481, 489, 491, 497, 499, 88, 91, 111, 112, 117, 118, 123, 127,
509–511, 514, 517, 523, 526, 529, 139, 144, 151, 158, 160–162, 173,
531, 546, 552, 553 183, 198, 203, 204, 209, 226–228,
Qualified Omniscience (also: 232, 245, 251, 253, 261, 282, 283,
Restricted Knowledge), 19, 286, 300, 308, 312, 326–328, 330,
50, 51, 88, 106, 118, 119, 135, 331, 350, 366, 370, 372, 373, 387,
145, 156, 169, 182, 193, 238, 466, 467, 469, 486, 490, 497, 505,
245, 246, 252, 276, 298, 344, 519, 524, 529
351, 374, 396, 397, 424, 427, Indirect, XVI, 21, 38, 57, 59, 69,
491, 492, 495, 496, 498, 499, 108, 109, 161, 173, 245, 328,
505, 509, 526–529, 537, 546, 329, 366, 373, 387, 404, 439,
552, 553 466, 486, 541, 548
Ex eventu knowledge, 3, 19, 239,
246, 247, 252, 347, 463, 537 Story-Time, XVIII
Omnitemporality, 14, 85 Summary, XVIII, 44, 112
SUPPLEMENTS TO MNEMOSYNE
EDITED BY H. PINKSTER, H.S. VERSNEL,
I.J.F. DE JONG and P. H. SCHRIJVERS
Recent volumes in the series
100. OPHUIJSEN, J.M. VAN. Hephaistion on Metre. Translation and Commentary. 1987. ISBN
90 04 08452 5
101. VERDENIUS, W.J. Commentaries on Pindar. Vol. II. Olympian Odes 1, 10, 11, Nemean 11,
Isthmian 2. 1988. ISBN 90 04 08535 1
102. LUSCHNIG, C.A.E. Time holds the Mirror. A Study of Knowledge in Euripides’ ‘Hippolytus’. 1988.
ISBN 90 04 08601 3
103. MARCOVICH, M. Alcestis Barcinonensis. Text and Commentary. 1988.
ISBN 90 04 08600 5
104. HOLT, F.L. Alexander the Great and Bactria. The Formation of a Greek Frontier in Cen-tral
Asia. Repr. 1993. ISBN 90 04 08612 9
105. BILLERBECK, M. Seneca’s Tragödien; sprachliche und stilistische Untersuchungen. Mit Anhängen
zur Sprache des Hercules Oetaeus und der Octavia. 1988. ISBN 90 04 08631 5
106. ARENDS, J.F.M.Die Einheit der Polis. Eine Studie über Platons Staat.1988.ISBN 90 04 08785 0
107. BOTER, G.J. The Textual Tradition of Plato’s Republic. 1988. ISBN 90 04 08787 7
108. WHEELER, E.L. Stratagem and the Vocabulary of Military Trickery.1988.ISBN 90 04 08831 8
109. BUCKLER, J. Philip II and the Sacred War. 1989. ISBN 90 04 09095 9
110. FULLERTON, M.D. The Archaistic Style in Roman Statuary. 1990.ISBN 90 04 09146 7
111. ROTHWELL, K.S. Politics and Persuasion in Aristophanes’ ‘Ecclesiazusae’. 1990.
ISBN 90 04 09185 8
112. CALDER, W.M. & A. DEMANDT. Eduard Meyer. Leben und Leistung eines
Universalhistorikers. 1990. ISBN 90 04 09131 9
113. CHAMBERS, M.H. Georg Busolt. His Career in His Letters. 1990. ISBN 90 04 09225 0
114. CASWELL, C.P. A Study of ‘Thumos’ in Early Greek Epic. 1990. ISBN 90 04 09260 9
115. EINGARTNER, J. Isis und ihre Dienerinnen in der Kunst der Römischen Kaiserzeit. 1991.
ISBN 90 04 09312 5
116. JONG, I. DE. Narrative in Drama. The Art of the Euripidean Messenger-Speech. 1991.
ISBN 90 04 09406 7
117. BOYCE, B.T. The Language of the Freedmen in Petronius’ Cena Trimalchionis. 1991.
ISBN 90 04 09431 8
118. RÜTTEN, Th. Demokrit — lachender Philosoph und sanguinischer Melancholiker. 1992.
ISBN 90 04 09523 3
119. KARAVITES, P. (with the collaboration of Th. Wren). Promise-Giving and Treaty-Making.
Homer and the Near East. 1992. ISBN 90 04 09567 5
120. SANTORO L’HOIR, F. The Rhetoric of Gender Terms. ‘Man’, ‘Woman’ and the portrayal of
character in Latin prose. 1992. ISBN 90 04 09512 8
121. WALLINGA, H.T. Ships and Sea-Power before the Great Persian War. The Ancestry of the
Ancient Trireme. 1993. ISBN 90 04 09650 7
122. FARRON, S. Vergil’s Æneid: A Poem of Grief and Love. 1993. ISBN 90 04 09661 2
123. LÉTOUBLON, F. Les lieux communs du roman. Stéréotypes grecs d’aventure et d’amour.
1993. ISBN 90 04 09724 4
124. KUNTZ, M. Narrative Setting and Dramatic Poetry. 1993. ISBN 90 04 09784 8
125. THEOPHRASTUS. Metaphysics. With an Introduction, Translation and Commentary by
Marlein van Raalte. 1993. ISBN 90 04 09786 4
126. THIERMANN, P. Die Orationes Homeri des Leonardo Bruni Aretino. Kritische Edition der
lateinischen und kastilianischen Übersetzung mit Prolegomena und Kommentar. 1993.
ISBN 90 04 09719 8
127. LEVENE, D.S. Religion in Livy. 1993. ISBN 90 04 09617 5
128. PORTER, J.R. Studies in Euripides’ Orestes. 1993. ISBN 90 04 09662 0
129. SICKING, C.M.J. & J.M. VAN OPHUIJSEN. Two Studies in Attic Particle Usage. Lysias and
Plato. 1993. ISBN 90 04 09867 4
130. JONG, I.J.F. DE, & J.P. SULLIVAN (eds.). Modern Critical Theory and Classical Literature.
1994. ISBN 90 04 09571 3
131. YAMAGATA, N. Homeric Morality. 1994. ISBN 90 04 09872 0
132. KOVACS, D. Euripidea. 1994. ISBN 90 04 09926 3
133. SUSSMAN, L.A. The Declamations of Calpurnius Flaccus. Text, Translation, and Com-
mentary. 1994. ISBN 90 04 09983 2
134. SMOLENAARS, J.J.L. Statius : Thebaid VII. A Commentary.1994.ISBN 90 04 10029 6
135. SMALL, D.B. (ed.). Methods in the Mediterranean. Historical and Archaeological Views on
Texts and Archaeology. 1995. ISBN 90 04 09581 0
136. DOMINIK, W.J. The Mythic Voice of Statius. Power and Politics in the Thebaid. 1994.
ISBN 90 04 09972 7
137. SLINGS, S.R. Plato’s Apology of Socrates. A Literary and Philosophical Study with a Running
Commentary. Edited and Completed from the Papers of the Late E. De Strycker, s.j. 1994.
ISBN 90 04 10103 9
138. FRANK, M. Seneca’s Phoenissae. Introduction and Commentary. 1995.
ISBN 90 04 09776 7
139. MALKIN, I. & Z.W. RUBINSOHN (eds.). Leaders and Masses in the Roman World. Studies in
Honor of Zvi Yavetz. 1995. ISBN 90 04 09917 4
140. SEGAL, A. Theatres in Roman Palestine and Provincia Arabia. 1995. ISBN 90 04 10145 4
141. CAMPBELL, M. A Commentary on Apollonius Rhodius Argonautica III 1-471. 1994.
ISBN 90 04 10158 6
142. DeFOREST, M.M. Apollonius’ Argonautica: A Callimachean Epic. 1994.
ISBN 90 04 10017 2
143. WATSON, P.A. Ancient Stepmothers. Myth, Misogyny and Reality. 1995.
ISBN 90 04 10176 4
144. SULLIVAN, S.D. Psychological and Ethical Ideas. What Early Greeks Say. 1995.
ISBN 90 04 10185 3
145. CARGILL, J. Athenian Settlements of the Fourth Century B.C. 1995. ISBN 90 04 09991 3
146. PANAYOTAKIS, C. Theatrum Arbitri. Theatrical Elements in the Satyrica of Petronius.
1995. ISBN 90 04 10229 9
147. GARRISON, E.P. Groaning Tears. Ethical and Dramatic Aspects of Suicide in Greek
Tragedy. 1995. 90 04 10241 8
148. OLSON, S.D. Blood and Iron. Stories and Storytelling in Homer’s Odyssey. 1995.
ISBN 90 04 10251 5
149. VINOGRADOV, J.G.& S.D. KRYZICKIJ (eds.). Olbia. Eine altgriechische Stadt im
Nordwestlichen Schwarzmeerraum. 1995. ISBN 90 04 09677 9
150. MAURER, K. Interpolation in Thucydides. 1995. ISBN 90 04 10300 7
151. HORSFALL, N. (ed.) A Companion to the Study of Virgil. 1995 ISBN 90 04 09559 4
152. KNIGHT, V.H. The Renewal of Epic. Responses to Homer in the Argonautica of Apollo-nius.
1995. ISBN 90 04 10386 4
153. LUSCHNIG, C.A.E. The Gorgon’s Severed Head. Studies of Alcestis, Electra, and Phoenissae.
1995. ISBN 90 04 10382 1
154. NAVARRO ANTOLÍN, F. (ed.). Lygdamus. Corpus Tibullianum III. 1-6: Lygdami
elegiarum liber. Translated by J.J. Zoltowski. 1996. ISBN 90 04 10210 8
155. MATTHEWS, V. J. Antimachus of Colophon. Text and Commentary. 1996.
ISBN 90 04 10468 2
156. TREISTER, M.Y. The Role of Metals in Ancient Greek History. 1996. ISBN 90 04 10473 9
157. WORTHINGTON, I. (ed.). Voice into Text. Orality and Literacy in Ancient Greece. 1996.
ISBN 90 04 10431 3
158. WIJSMAN, H. J.W. Valerius Flaccus, Argonautica, Book V. A Commentary. 1996.
ISBN 90 04 10506 9
159. SCHMELING, G. (ed.). The Novel in the Ancient World. 1996. ISBN 90 04 09630 2
160. SICKING, C.M. J. & P. STORK. Two Studies in the Semantics of the Verb in Classical Greek.
1996. ISBN 90 04 10460 7
161. KOVACS, D. Euripidea Altera. 1996. ISBN 90 04 10624 3
162. GERA, D. Warrior Women. The Anonymous Tractatus De Mulieribus. 1997.
ISBN 90 04 10665 0
163. MORRIS, I. & B. POWELL (eds.). A New Companion to Homer. 1997. ISBN 90 04 09989 1
164. ORLIN, E.M. Temples, Religion and Politics in the Roman Republic. 1997.ISBN 90 04 10708 8
165. ALBRECHT, M. VON. A History of Roman Literature. From Livius Andronicus to Boethius
with Special Regard to Its Influence on World Literature. 2 Vols.Revised by G.Schmeling
and by the Author. Vol. 1: Translated with the Assistance of F. and K. Newman, Vol. 2:
Translated with the Assitance of R.R. Caston and F.R. Schwartz. 1997.
ISBN 90 04 10709 6 (Vol. 1), ISBN 90 04 10711 8 (Vol. 2), ISBN 90 04 10712 6 (Set)
166. DIJK, J.G.M. VAN. A‰noi, LÒgoi, MÇuyoi. Fables in Archaic, Classical, and Hellenistic
Greek Literature. With a Study of the Theory and Terminology of the Genre. 1997. ISBN
90 04 10747 9
167. MAEHLER, H. (Hrsg.). Die Lieder des Bakchylides. Zweiter Teil: Die Dithyramben und
Fragmente. Text, Übersetzung und Kommentar. 1997. ISBN 90 04 10671 5
168. DILTS, M. & G.A. KENNEDY (eds.). Two Greek Rhetorical Treatises from the Roman Empire.
Introduction, Text, and Translation of the Arts of Rhetoric Attributed to Anonymous
Seguerianus and to Apsines of Gadara. 1997. ISBN 90 04 10728 2
169. GÜNTHER, H.-C. Quaestiones Propertianae. 1997. ISBN 90 04 10793 2
170. HEINZE, T. (Hrsg.). P. Ovidius Naso. Der XII. Heroidenbrief: Medea an Jason. Einleitung, Text
und Kommentar. Mit einer Beilage: Die Fragmente der Tragödie Medea. 1997. ISBN 90
04 10800 9
171. BAKKER, E. J. (ed.). Grammar as Interpretation. Greek Literature in its Linguistic Contexts.
1997. ISBN 90 04 10730 4
172. GRAINGER, J.D. A Seleukid Prosopography and Gazetteer. 1997. ISBN 90 04 10799 1
173. GERBER, D.E. (ed.). A Companion to the Greek Lyric Poets. 1997. ISBN 90 04 09944 1
174. SANDY, G. The Greek World of Apuleius. Apuleius and the Second Sophistic. 1997.
ISBN 90 04 10821 1
175. ROSSUM-STEENBEEK, M. VAN. Greek Readers’ Digests? Studies on a Selection of
Subliterary Papyri. 1998. ISBN 90 04 10953 6
176. McMAHON, J.M. Paralysin Cave. Impotence, Perception, and Text in the Satyrica of
Petronius. 1998. ISBN 90 04 10825 4
177. ISAAC, B. The Near East under Roman Rule. Selected Papers. 1998.
ISBN 90 04 10736 3
178. KEEN, A.G. Dynastic Lycia. A Political History of the Lycians and Their Relations with
Foreign Powers, c. 545-362 B.C. 1998. ISBN 90 04 10956 0
179. GEORGIADOU, A. & D.H.J. LARMOUR. Lucian’s Science Fiction Novel True Histories.
Interpretation and Commentary. 1998. ISBN 90 04 10667 7
180. GÜNTHER, H.-C. Ein neuer metrischer Traktat und das Studium der pindarischen Metrik in der
Philologie der Paläologenzeit. 1998. ISBN 90 04 11008 9
181. HUNT, T.J. A Textual History of Cicero’s Academici Libri. 1998. ISBN 90 04 10970 6
182. HAMEL, D. Athenian Generals. Military Authority in the Classical Period. 1998.
ISBN 90 04 10900 5
183. WHITBY, M. (ed.).The Propaganda of Power.The Role of Panegyric in Late Antiquity. 1998.
ISBN 90 04 10571 9
184. SCHRIER, O.J. The Poetics of Aristotle and the Tractatus Coislinianus. A Bibliography from
about 900 till 1996. 1998. ISBN 90 04 11132 8
185. SICKING, C.M.J. Distant Companions. Selected Papers. 1998. ISBN 90 04 11054 2
186. SCHRIJVERS, P.H. Lucrèce et les Sciences de la Vie. 1999. ISBN 90 04 10230 2
187. BILLERBECK M. (Hrsg.). Seneca. Hercules Furens. Einleitung, Text, Übersetzung
und Kommentar. 1999. ISBN 90 04 11245 6
188. MACKAY, E.A. (ed.). Signs of Orality. The Oral Tradition and Its Influence in the
Greek and Roman World. 1999. ISBN 90 04 11273 1
189. ALBRECHT, M. VON. Roman Epic. An Interpretative Introduction. 1999.
ISBN 90 04 11292 8
190. HOUT, M.P.J. VAN DEN. A Commentary on the Letters of M. Cornelius Fronto. 1999.
ISBN 90 04 10957 9
191. KRAUS, C. SHUTTLEWORTH. (ed.). The Limits of Historiography. Genre and
Narrative in Ancient Historical Texts. 1999. ISBN 90 04 10670 7
192. LOMAS, K. & T. CORNELL. Cities and Urbanisation in Ancient Italy.
ISBN 90 04 10808 4 In preparation
193. TSETSKHLADZE, G.R. (ed.). History of Greek Colonization and Settlement Overseas.
2 vols. ISBN 90 04 09843 7 In preparation
194. WOOD, S.E. Imperial Women. A Study in Public Images, 40 B.C. - A.D. 68. 1999.
ISBN 90 04 11281 2
195. OPHUIJSEN, J.M. VAN & P. STORK. Linguistics into Interpretation. Speeches of War
in Herodotus VII 5 & 8-18. 1999. ISBN 90 04 11455 6
196. TSETSKHLADZE, G.R. (ed.). Ancient Greeks West and East. 1999.
ISBN 90 04 11190 5
197. PFEIJFFER, I.L. Three Aeginetan Odes of Pindar. A Commentary on Nemean V, Nemean
III, & Pythian VIII. 1999. ISBN 90 04 11381 9
198. HORSFALL, N. Virgil, Aeneid 7. A Commentary. 2000. ISBN 90 04 10842 4
199. IRBY-MASSIE, G.L. Military Religion in Roman Britain. 1999.
ISBN 90 04 10848 3
200. GRAINGER, J.D. The League of the Aitolians. 1999. ISBN 90 04 10911 0
201. ADRADOS, F.R. History of the Graeco-Roman Fable. I: Introduction and from the
Origins to the Hellenistic Age. Translated by L.A. Ray. Revised and Updated by the
Author and Gert-Jan van Dijk. 1999. ISBN 90 04 11454 8
202. GRAINGER, J.D. Aitolian Prosopographical Studies. 2000. ISBN 90 04 11350 9
203. SOLOMON, J. Ptolemy Harmonics. Translation and Commentary. 2000.
ISBN 90 04 115919
204. WIJSMAN, H.J.W. Valerius Flaccus, Argonautica, Book VI. A Commentary. 2000.
ISBN 90 04 11718 0
205. MADER, G. Josephus and the Politics of Historiography. Apologetic and Impression
Management in the Bellum Judaicum. 2000. ISBN 90 04 11446 7
206. NAUTA, R.R. Poetry for Patrons. Literary Communication in the Age of Domitian.
2000. ISBN 90 04 10885 8
207. ADRADOS, F.R. History of the Graeco-Roman Fable. II: The Fable during the Roman
Empire and in the Middle Ages. Translated by L.A. Ray. Revised and Updated by
the Author and Gert-Jan van Dijk. 2000. ISBN 90 04 11583 8
208. JAMES, A. & K. LEE. A Commentary on Quintus of Smyrna, Posthomerica V. 2000.
ISBN 90 04 11594 3
209. DERDERIAN, K. Leaving Words to Remember. Greek Mourning and the Advent of
Literacy. 2001. ISBN 90 04 11750 4
210. SHORROCK, R. The Challenge of Epic. Allusive Engagement in the Dionysiaca of
Nonnus. 2001. ISBN 90 04 11795 4
211. SCHEIDEL, W. (ed.). Debating Roman Demography. 2001. ISBN 90 04 11525 0
212. KEULEN, A. J. L. Annaeus Seneca Troades. Introduction, Text and Commentary.
2001. ISBN 90 04 12004 1
213. MORTON, J. The Role of the Physical Environment in Ancient Greek Seafaring. 2001.
ISBN 90 04 11717 2
214. GRAHAM, A. J. Collected Papers on Greek Colonization. 2001. ISBN 90 04 11634 6
215. GROSSARDT, P. Die Erzählung von Meleagros. Zur literarischen Entwicklung der
kalydonischen Kultlegende. 2001. ISBN 90 04 11952 3
216. ZAFIROPOULOS, C.A. Ethics in Aesop’s Fables: The Augustana Collection. 2001.
ISBN 90 04 11867 5
217. RENGAKOS, A. & T.D. PAPANGHELIS (eds.). A Companion to Apollonius Rhodius.
2001. ISBN 90 04 11752 0
218. WATSON, J. Speaking Volumes. Orality and Literacy in the Greek and Roman World.
2001. ISBN 90 04 12049 1
219. MACLEOD, L. Dolos and Dike in Sophokles’ Elektra. 2001. ISBN 90 04 11898 5
220. MCKINLEY, K.L. Reading the Ovidian Heroine. “Metamorphoses” Commentaries
1100-1618. 2001. ISBN 90 04 11796 2
221. REESON, J. Ovid Heroides 11, 13 and 14. A Commentary. 2001.
ISBN 90 04 12140 4
222. FRIED, M.N. & S. UNGURU. Apollonius of Perga’s Conica: Text, Context, Subtext.
2001. ISBN 90 04 11977 9
223. LIVINGSTONE, N. A Commentary on Isocrates’ Busiris. 2001. ISBN 90 04 12143 9
224. LEVENE, D.S. & D.P. NELIS (eds.). Clio and the Poets. Augustan Poetry and the
Traditions of Ancient Historiography. 2002. ISBN 90 04 11782 2
225. WOOTEN, C.W. The Orator in Action and Theory in Greece and Rome. 2001.
ISBN 90 04 12213 3
226. GALÁN VIOQUE, G. Martial, Book VII. A Commentary. 2001. ISBN 90 04 12338 5
227. LEFÈVRE, E. Die Unfähigkeit, sich zu erkennen: Sophokles’ Tragödien. 2001.
ISBN 90 04 12322 9
228. SCHEIDEL, W. Death on the Nile. Disease and the Demography of Roman Egypt.
2001. ISBN 90 04 12323 7
229. SPANOUDAKIS, K. Philitas of Cos. 2002. ISBN 90 04 12428 4
230. WORTHINGTON, I. & J.M. FOLEY (eds.). Epea and Grammata. Oral and written
Communication in Ancient Greece. 2002. ISBN 90 04 12455 1
231. McKECHNIE, P. (ed.). Thinking Like a Lawyer. Essays on Legal History and General
History for John Crook on his Eightieth Birthday. 2002. ISBN 90 04 12474 8
232. GIBSON, R.K. & C. SHUTTLEWORTH KRAUS (eds.). The Classical Commentary.
Histories, Practices, Theory. 2002. ISBN 90 04 12153 6
233. JONGMAN, W. & M. KLEIJWEGT (eds.). After the Past. Essays in Ancient History in
Honour of H.W. Pleket. 2002. ISBN 90 04 12816 6
234. GORMAN, V.B. & E.W. ROBINSON (eds.). Oikistes. Studies in Constitutions,
Colonies, and Military Power in the Ancient World. Offered in Honor of A.J. Graham.
2002. ISBN 90 04 12579 5
235. HARDER, A., R. REGTUIT, P. STORK & G. WAKKER (eds.). Noch einmal zu....
Kleine Schriften von Stefan Radt zu seinem 75. Geburtstag. 2002. ISBN 90 04 12794 1
236. ADRADOS, F.R. History of the Graeco-Latin Fable. Volume Three: Inventory and
Documentation of the Graeco-Latin Fable. 2002. ISBN 90 04 11891 8
237. SCHADE, G. Stesichoros. Papyrus Oxyrhynchus 2359, 3876, 2619, 2803. 2003.
ISBN 90 04 12832 8
238. ROSEN, R.M. & I. SLUITER (eds.) Andreia. Studies in Manliness and Courage in
Classical Antiquity. 2003. ISBN 90 04 11995 7
239. GRAINGER, J.D. The Roman War of Antiochos the Great. 2002. ISBN 90 04 12840 9
240. KOVACS, D. Euripidea Tertia. 2003. ISBN 90 04 12977 4
241. PANAYOTAKIS, S., M. ZIMMERMAN & W. KEULEN (eds.). The Ancient Novel and
Beyond. 2003. ISBN 90 04 12999 5
242. ZACHARIA, K. Converging Truths. Euripides’ Ion and the Athenian Quest for
Self-Definition. 2003. ISBN 90 0413000 4
243. ALMEIDA, J.A. Justice as an Aspect of the Polis Idea in Solon’s Political Poems. 2003.
ISBN 90 04 13002 0
244. HORSFALL, N. Virgil, Aeneid 11. A Commentary. 2003. ISBN 90 04 12934 0
245. VON ALBRECHT, M. Cicero’s Style. A Synopsis. Followed by Selected Analytic
Studies. 2003. ISBN 90 04 12961 8
246. LOMAS, K. Greek Identity in the Western Mediterranean. Papers in Honour of Brian
Shefton. 2004. ISBN 90 04 13300 3
247. SCHENKEVELD, D.M. A Rhetorical Grammar. C. Iullus Romanus, Introduction
to the Liber de Adverbio. 2004. ISBN 90 04 133662 2
248. MACKIE, C.J. Oral Performance and its Context. 2004. ISBN 90 04 13680 0
249. RADICKE, J. Lucans Poetische Technik. 2004. ISBN 90 04 13745 9
250. DE BLOIS, L., J. BONS, T. KESSELS & D.M. SCHENKEVELD (eds.). The
Statesman in Plutarch’s Works. Volume I: Plutarch’s Statesman and his Aftermath:
Political, Philosophical, and Literary Aspects. ISBN 90 04 13795 5. Volume II: The
Statesman in Plutarch’s Greek and Roman Lives. 2004. ISBN 90 04 13808 0
251. GREEN, S.J. Ovid, Fasti 1. A Commentary. 2004. ISBN 90 04 13985 0
252. VON ALBRECHT, M. Wort und Wandlung. 2004. ISBN 90 04 13988 5
253. KORTEKAAS, G.A.A. The Story of Apollonius, King of Tyre. A Study of Its Greek Origin
and an Edition of the Two Oldest Latin Recensions. 2004. ISBN 90 04 13923 0
254. SLUITER, I. & R.M. ROSEN (eds.). Free Speech in Classical Antiquity. 2004.
ISBN 90 04 13925 7
255. STODDARD, K. The Narrative Voice in the Theogony of Hesiod. 2004.
ISBN 90 04 14002 6
256. FITCH, J.G. Annaeana Tragica. Notes on the Text of Seneca’s Tragedies. 2004.
ISBN 90 04 14003 4
257. DE JONG, I.J.F., R. NÜNLIST & A. BOWIE (eds.). Narrators, Narratees, and Narratives
in Ancient Greek Literature. Studies in Ancient Greek Narrative, Volume One. 2004.
ISBN 90 04 13927 3
258. VAN TRESS, H. Poetic Memory. Allusion in the Poetry of Callimachus and the
Metamorphoses of Ovid. 2004. ISBN 90 04 14157 X

De Jong (2004) PDF

Uploaded by

Copyright:

Available Formats

De Jong (2004) PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

De Jong (2004) PDF

Uploaded by

Copyright:

Available Formats

NARRATORS, NARRATEES, AND NARRATIVES

IN ANCIENT GREEK LITERATURE

I.J.F. DE JONG • P. H. SCHRIJVERS

BIBLIOTHECAE FASCICULOS EDENDOS CURAVIT

H. PINKSTER, KLASSIEK SEMINARIUM, SPUISTRAAT 134, AMSTERDAM

SUPPLEMENTUM DUCENTESIMUM QUINQUAGESIMUM SEPTIMUM

IRENE DE JONG, RENÉ NÜNLIST, and ANGUS BOWIE

NARRATORS, NARRATEES, AND NARRATIVES

IRENE DE JONG, RENÉ NÜNLIST and ANGUS BOWIE

Library of Congress Cataloging-in-Publication Data

Introduction. Narratological theory on narrators, narratees, and

part one. epic and elegiac poetry

Chapter One. Homer (I.J.F. de Jong) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

part two. historiography

Chapter Seven. Herodotus (I.J.F. de Jong) . . . . . . . . . . . . . . . . . . . . . . . . . . 101

part three. choral lyric

Chapter Fifteen. Pindar and Bacchylides (I.L. Pfeijﬀer) . . . . . . . . . . . 213

part four. drama

Chapter Sixteen. Aeschylus (J. Barrett) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235

part five. oratory

Chapter Twenty-Two. Antiphon (M. Edwards) . . . . . . . . . . . . . . . . . . . . 317

part six. philosophy

Chapter Twenty-Eight. Plato (K.A. Morgan) . . . . . . . . . . . . . . . . . . . . . . . 357

part seven. biography

Chapter Thirty. Xenophon (V. Gray) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391

part eight. between philosophy and rhetoric

Chapter Thirty-Four. Dio Chrysostom (T.J.G. Whitmarsh) . . . . . . . . 451

part nine. the novel

Chapter Thirty-Six. Chariton (J. Morgan) . . . . . . . . . . . . . . . . . . . . . . . . . . 479

Epilogue. Narrators, narratees, and narratives in ancient Greek

The germ of this book lies in a research project that I developed in

Someone interested in the history of ancient Greek literature has at

Saïd, Trédé and Boulluec 1997.

of literary history’ was proclaimed,4 is making a come-back.5 One

history of ancient Greek narrative is deﬁned and the foundation laid

7 Martin 1987: 28–29.

actorial motivation: the ‘why’ of the development of the story anal-

analepsis (ﬂashback, ‘Rückwendung’): the narration of an event which

argument function: the function or signiﬁcance which an embedded

characterization: characterization may be explicit or implicit (when

description: descriptions can be static (the story comes to a halt and

embedded narrative: a narrative which is embedded in the main

acting as secondary narrator. It usually takes the form of an analepsis

embedded or secondary focalization: when the narrator repre-

focalizer: the person (the narrator or a character) through whose ‘eyes’

frequency: events may be told singulatively (telling once what happened

interlace technique: the technique of interweaving diﬀerent story-

key function: the signiﬁcance which an embedded narrative has for

narratees: the addressees of the narrator. We may distinguish between

narration: we may distinguish between subsequent narration (following

turns them into a text. We may distinguish between external narrators

narrator-text: those parts of the text which are presented by the

narratorial motivation: the ‘why’ of the development of the story

order: the chronological order of the fabula may be changed in the

paralepsis: a speaker provides more information than, strictly speak-

paralipsis: a speaker provides less information than he could; details

periphrastic denomination: a reference to a character not by prop-

prolepsis (foreshadowing, ‘Vorauswendung’): the narration of an

rhythm: the relation between story-time and fabula-time. An event