George Boys-Stones, Dimitri El Murr, Christopher Gill - The Platonic Art of Philosophy-Cambridge University Press (2013)

more information - www.cambridge.
org/9781107038981
T h e P l ato n i c A rt o f P h i l o s o p h y
This is a collection of essays written by leading experts in honour of

Christopher Rowe, and inspired by his groundbreaking work in the
exegesis of Plato. The authors represent scholarly traditions which are
very different in their approaches and interests, and rarely brought
into dialogue with each other. This volume, by contrast, aims to
explore synergies between them. Key topics include: the literary
and philosophical unity of Plato’s works; the presence and role of
his contemporaries in his dialogues; the function of myth (especially
the Atlantis myth); Plato’s Socratic heritage, especially as played
out in his discussions of psychology; his views on truth and being.
Prominent among the dialogues discussed are Euthydemus, Phaedo,
Phaedrus, Republic, Theaetetus, Timaeus, Sophist and Laws.
g e o rg e b oy s - s to n e s is Professor of Ancient Philosophy at

Durham University.
d i m i t r i e l m urr is Lecturer in Ancient Philosophy at the
University of Paris 1 Panthéon-Sorbonne and Junior Member of the
Institut Universitaire de France.
c h r i s to p h e r g i l l is Professor of Ancient Thought at the
University of Exeter.
Professor Christopher Rowe, OBE
T h e P l ato n i c A rt
of Philosophy
edi ted by
G e o rg e B oy s - S to n e s
D i m i t r i E l Murr
and
C h r i s to p h e r G i l l
University Printing House, Cambridge CB2 8BS, United Kingdom
Published in the United States of America by Cambridge University Press, New York
Cambridge University Press is part of the University of Cambridge.
It furthers the University’s mission by disseminating knowledge in the pursuit of
education, learning and research at the highest international levels of excellence.
www.cambridge.org
Information on this title: www.cambridge.org/9781107038981
© Cambridge University Press 2013
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2013
Printed in the United Kingdom by CPI Group Ltd, Croydon CR0 4YY
A catalogue record for this publication is available from the British Library
Library of Congress Cataloguing in Publication data
The platonic art of philosophy / edited by George Boys-Stones,
Dimitri El Murr and Christopher Gill.
pages cm
Includes bibliographical references and index.
isbn 978-1-107-03898-1
1. Plato. I. Boys-Stones, G. R., editor of compilation.
b395.p5375 2013
184–dc23 2013008186
isbn 978-1-107-03898-1 Hardback
Cambridge University Press has no responsibility for the persistence or accuracy of
URLs for external or third-party internet websites referred to in this publication,
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.
Contents
List of contributors page vii

Preface ix
List of abbreviations and conventions x
Introduction 1
George Boys-Stones
1 Macrology and digression 10
Monique Dixsaut
2 Two conceptions of the body in Plato’s Phaedrus 27
María Angélica Fierro
3 Socrates in the Phaedo 51
Noburu Notomi
4 Socratic intellectualism in the Republic’s central digression 70
David Sedley
5 Timaeus in the cave 90
Thomas Johansen
6 Reflective commentary (1): ‘Socratic’ psychology in Plato’s
Republic 110
Christopher Gill
7 Reflective commentary (2): appearance, reality and the
desire for the good 122
Dimitri El Murr
8 Waving or drowning? Socrates and the sophists on
self-knowledge in the Euthydemus 130
M. M. McCabe
v
vi Contents
9 Why was the Theaetetus written by Euclides? 150
Michel Narcy
10 The Wooden Horse: the Cyrenaics in the Theaetetus 167
Ugo Zilioli
11 The Wax Tablet, logic and Protagoreanism 186
Terry Penner
12 A form that ‘is’ of what ‘is not’: existential einai in
Plato’s Sophist 221
Denis O’Brien
13 Truth and story in the Timaeus-Critias 249
Sarah Broadie
14 The Atlantis poem in the Timaeus-Critias 269
Mauro Tulli
15 Friendship and justice in the Laws 283
Malcolm Schofield
Bibliography 298
Christopher Rowe publications 1969–2012 312
Index of ancient passages 325
Index of topics 338
Contributors
Ge orge Boy s- Stones is Professor of Ancient Philosophy at Durham

University.
S arah Broad i e is Professor of Moral Philosophy and Wardlaw
Professor at the University of St Andrews.
M oniqu e Dixsaut is Professor Emerita of Ancient Philosophy at the
University of Paris 1 Panthéon-Sorbonne.
D i mit ri El M urr is Lecturer in Ancient Philosophy at the University
of Paris 1 Panthéon-Sorbonne and Junior Member of the Institut
Universitaire de France.
M aría Angé l i ca F i erro is Associate Researcher at CONICET,
Argentina.
C hristo ph e r G i ll is Professor of Ancient Thought at the University
of Exeter.
T h o mas J o h ansen is Reader in Ancient Philosophy at the University
of Oxford and a Fellow of Brasenose College.
M . M . M cCab e is Professor of Ancient Philosophy at King’s College
London.
Mic h e l Narc y is Emeritus Directeur de recherches, Centre National
de la Recherche Scientifique, Centre Jean Pépin UPR 76.
N obu ru Noto mi is Professor of Ancient Philosophy at Keio University,
Tokyo.
D e nis O’ Brie n is Emeritus Directeur de recherches, Centre National
de la Recherche Scientifique, Centre Jean Pépin UPR 76.
vii
viii List of contributors
Te rry Pe nner is Emeritus Professor of Philosophy at the University of
Wisconsin, Madison.
M alcol m Schofi eld is Emeritus Professor of Ancient Philosophy at
the University of Cambridge and a Fellow of St John’s College.
David Se d l ey is Laurence Professor of Ancient Philosophy at the
University of Cambridge and a Fellow of Christ’s College.
M au ro Tu l li is Professor of Classical Philology at the University of
Pisa.
Ug o Z il io l i is Marie Curie Research Fellow at the University of Pisa.
Preface
In 2009 Christopher Rowe retired from his last full-time academic post,
the Chair of Greek in the Department of Classics and Ancient History at
Durham University, which he had held since 1995. In 2010 the Northern
Association for Ancient Philosophy (an association of which Christopher
has been a keen and loyal supporter) dedicated its annual meeting to
his work, in recognition of his groundbreaking contributions to ancient
philosophy, notably in the field of Plato studies. The meeting was held
over two days in April, in University College, Durham (Christopher’s col-
lege), with eight papers given by friends, colleagues and former students.
Between them, they represented both the variety of themes addressed in
Christopher’s work, and his international network of collaborations and
conversation partners. Their work was the basis for this volume, which
was compiled in a similar spirit: with the aim, that is, of engaging with
and building upon themes and approaches to understanding Plato devel-
oped in Christopher’s work, and in doing so to bring into conversation
voices from a wide variety of perspectives and traditions, traditions which
too often seem reluctant to acknowledge how much they have to learn
from one another.
We would like to thank all those who took part in the original confer-
ence, and the institutions that gave it financial support: the Aristotelian
Society, the Classical Association, Durham University, the University of
Paris I Panthéon-Sorbonne and the Society for the Promotion of Hellenic
Studies. We would also like to thank all those who made the timely
appearance of this volume not only possible but a pleasant experience:
all of the contributors, and Michael Sharp and his team at Cambridge
University Press. And, as Christopher shows no signs of retiring from
the conversation himself, we hope that this volume will serve for him
not just as a heartfelt gift, but as a tool and spur for his (and others’)
future work.
ix
Abbreviations and conventions
Abbreviations for ancient authors and works (including Plato) are those
of LSJ or the OLD. All Greek and Latin is translated into English, and
unless otherwise specified, the translation is by the contributor. All sec-
ondary works cited by author and date are included in the Bibliography at
the end of the volume.
The following abbreviations for modern works are used:
DK H. Diels and W. Kranz, Die Fragmente der Vorsokratiker. 6th
edn. 3 vols. Dublin / Zurich, 1954.
LSJ H. G. Liddell, R. Scott, and H. S. Jones, A Greek-English
Lexicon, 9th edn., Oxford, 1940.
ML Meiggs and Lewis 1988.
OED Oxford English Dictionary, 2nd edn. 20 vols. Oxford, 1989.
OLD P. Glare (ed.), Oxford Latin Dictionary, Oxford, 1982.
SSR Giannantoni 1990.
x
Introduction
George Boys-Stones
According to an anecdote preserved by Athenaeus, Socrates once reported

a dream he had had in which he saw Plato as a crow, jumping on his head
and cawing. ‘I think, Plato,’ Socrates concluded, ‘that you have heaped a
lot of untruths on my head.’1
The anecdote apparently has anti-Platonic roots,2 but underneath the
polemic there is an important question. Ought we to expect ‘truths’ about
Socrates from Plato? Or rather, perhaps: what sort of truth ought we to
expect? After all, no one seriously believes that Plato’s dialogues are meant
as historical records in the most literal sense. Yet the attempt to disen-
tangle the historical Socrates from them (to see as it were where Plato is
telling the truth and where he is interposing his own philosophical char-
acter) has been a lively, even dominant tradition within recent Platonic
scholarship.3
The honorand of this volume, Christopher Rowe – represented on the
cover of this volume by another ‘CRow(e)’ – has been at the forefront
of attempts in recent decades to rethink our approach to Plato, and to
recognise that his philosophical artistry involves a more nuanced engage-
ment with his teacher, one that cannot be usefully unpacked at any point
in terms of ‘inaccuracies’ in his portrayal of Socrates – as if portraying
Socrates was ever his point. Rather, we are to think of Plato as philoso-
phising in the tradition of Socrates so that, as he puts it in Rowe 2007a,
Plato is in every relevant sense consistently ‘true’ to Socrates through his
work. From this perspective, Rowe rejects the ‘developmentalist’ reading
1
Deipnosophistae 11, 507cd.
2
It seems to be a ‘vicious parody’ (Riginos 1976: 54–5) of another story in which Socrates dreamed
that he saw Plato as a swan (e.g. Apuleius, On Plato 1.1 [182]). Other complaints about Plato’s ‘mis-
representation’ of historical individuals are recorded at DL 3.35 (Socrates again, on hearing the Lysis);
Athenaeus 11, 505d and 506a (Gorgias and Phaedo respectively, on hearing the dialogues named for
them).
3
See Dorion 2011, especially 13–14.
1
2 George Boys-Stones
of Plato which is based on an attempt to trace the path from the ‘real
Socrates’ to a ‘mature Plato’ who is someone else altogether. Instead, Rowe
invites us to engage with the dialogues on their own terms: as the philoso-
phy of Plato the Socratic.
Rowe’s determination to understand the dialogues rather than to
judge them has led him to combine, to an unusual degree, two strands in
modern scholarship on Plato (and ancient philosophy generally) which
normally constitute two quite separate traditions. One is the more philo-
logical, or at least textually based, approach which remains strong espe-
cially in continental Europe. The other is an approach centred on themes,
issues and theories, which is broadly informed by analytic philosophy
and is dominant in English-language scholarship. Over the years, Rowe
has published a remarkable series of commentaries or annotated transla-
tions of Plato’s dialogues, a process which is continuing with translations
for Penguin Classics and other series.4 On the other hand, especially in
Penner and Rowe (2005) and his wide-ranging monograph (Rowe 2007a),
as well as in many papers, he has addressed central, fundamental ques-
tions in scholarly debate on Socratic and Platonic ethics, epistemology,
politics and psychology. In general, his intellectual (and personal) outlook
has been resolutely open and international, pressing lines of inquiry in a
way that goes beyond received academic traditions.
Rowe’s work on Plato has not only combined philological and philo-
sophical strands but has also brought out the integral connection between
these two aspects. His translations of Plato aim at a strongly literal ren-
dering (even at the expense of a smoother English style) with a view to
bringing out clearly the salient philosophical point that is being made
in any given passage. This focus on accuracy has generated some excep-
tional insights. For instance, his re-interpretation of a single sentence in
the Statesman, 300c4–6, formed a key part of a radical re-thinking of the
relationship of this dialogue to Plato’s earlier political thought. He showed
how a new understanding of the syntax of the sentence removes a cru-
cial piece of evidence for the common view that the Statesman marks a
transition away from the knowledge-based political theory of the Republic
towards the constitutionalism of the Laws.5 In his 2007 monograph, sig-
nificantly entitled Plato and the Art of Philosophical Writing, Rowe shows
how searching examination of specific passages and fundamental questions
about the core ideas and overall shape of Platonic theory are intimately
4
A full list of Rowe’s publications to the end of 2012 is given in this volume.
5
See further Rowe 1995: 15–18, 230–1; also discussed in Gill 2002a: 156–8.
Introduction 3
interlinked. For example, close attention to Socrates’ comments on the
discarnate psyche in R. 611c–612b, a passage often seen as an afterthought
to the main argument, becomes an integral part of a re-evaluation of the
significance of the idea of the tripartite psychology for Plato.6
Latterly, Rowe’s research has focused more especially on the Socrates–
Plato relationship, and on the question whether there really is a decisive
break between Socratic and Platonic philosophy. A crucial stimulus
for this move has been his collaboration with Penner and their shared
re-examination, in connection with the Lysis, of Socratic ethics and psych-
ology (Penner and Rowe 2005). But in Rowe 2007a and other publica-
tions in this period, this has been extended to a far-reaching re-assessment
of the question whether Plato ever, in fact, repudiates the key Socratic the-
ses on ethics, psychology and epistemology. This line of inquiry, as noted
above, represents a frontal challenge to the developmental approach to
Plato that is widespread especially in English-language scholarship, and
also to Aristotle’s evidence, which is widely taken as supporting a develop-
mental view. This inquiry is very much a part of Rowe’s ongoing work; but
it has already had an energising effect on Platonic scholarship in opening
up a quite new understanding of the nature and direction of Plato’s phil-
osophy in his middle and later periods.
The contributors to this volume were invited to respond in their own
way to these distinctive features of Rowe’s work, but it is fascinating to see
how similar themes recur and intertwine. Monique Dixsaut sets the tone
for the whole volume in the first chapter, which addresses most explicitly
the way in which a need for a multi-dimensional and nuanced approach
to the reading of Plato’s works emerges from their character as dialogues:
Plato has deliberately chosen not to give the kind of ‘linear’ account of a
philosophical ‘system’ which would invite two-dimensional comparisons
with the thought of others (for example, of Socrates). Dixsaut comments
on quite how deliberate that choice is, focusing on moments where the
dialogues invite reflection on how appropriate (for example in terms of
length) the various contributions are. The answer is that a contribution is
appropriate so long as it does not lose sight of the question. This accounts
for the many new beginnings and ‘digressions’ one finds in Plato as well.
Interpreters of Plato need to remember that ‘unity’ is not the same thing
as ‘continuity’.
A ‘dialogical’ conception of unity like this is one that can be extended
beyond the exegesis of individual dialogues to challenge the ‘linear’
6
Rowe 2007a: 140–1, 165, 170–2, a point taken up in Chapters 4, 6 and 7 below.
assumptions made by developmentalist readings of Plato. Rowe has been
a champion of the idea that many of the differences between different dia-
logues can be explained by the different perspectives they take rather than
by a change in Plato’s underlying beliefs. María Angélica Fierro argues
that this may be the case even for one of the most compelling of the exam-
ples stressed by developmentalists: the supposed shift in Plato’s view of
the body. On a traditional account, Plato inclined to see the body as an
impediment to philosophy in relatively ‘early’ works such as the Phaedo or
Symposium, but accorded it a more clearly constructive, teleological role in
later works such as the Timaeus. Fierro uses the Phaedrus to make a case
that it is Plato’s focus, not his view, that has changed. For in the Phaedrus,
these two extremes clearly co-exist: the dangers of bodily distraction are
given full acknowledgement in the speeches on love, but so is the possibil-
ity that the body can support philosophical activity – and even that the
philosopher may ultimately aspire to the bodily condition of the gods.
The developmentalist conviction that there is a sharp break between
earlier ‘Socratic’ dialogues and those in which Socrates increasingly
becomes merely a mouthpiece for Plato’s mature views is also contested
by Noburu Notomi, who adopts Christopher Rowe’s challenging sug-
gestion that Plato’s works are all fundamentally ‘Socratic’. Notomi takes
the Phaedo as his lens on the question, arguing that everything about this
work – including the choice of narrator, Phaedo, and its setting in Phlius,
the ancestral home of Pythagoras – suggests that its unifying theme is the
nature of philosophy as an activity oriented towards the care of the soul
rather than the body. In this light, we can see that the ‘Platonic’ elements
of the dialogue (for example the introduction of ‘forms’) are not a matter
of Plato striking out in new directions, as Aristotle rather implies; on the
contrary, they represent ‘Plato’s way of developing Socrates’ ethical mes-
sage’ (p. 67).
A very similar conclusion is arrived at in David Sedley’s reading of the
moral psychology of the Republic – one which happens to take an explicit
‘digression’ as its unifying perspective on the work. Sedley observes that
the central portion of the Republic (Books 5–7 in the ten-book division –
which, as Sedley cautions, may not have been the original one) appears
for the most part to discard the tripartite soul of Books 4 and 8–9 and
resume the ‘intellectualist’ view normally associated with Socrates. (Sedley
identifies the Euthydemus and Phaedo quite specifically as intertexts here.)
Where Rowe has argued that this is because Plato never really moved away
from Socrates’ position, Sedley rather suggests that his intention is to allow
us to focus on the virtue of the philosopher, which is defined in terms
Introduction 5
of contemplative activity. Sedley also maintains that Plato himself is evi-
dently keen to resist a crude developmentalist view of his works: he makes
his own case for the continuity of his thought with that of Socrates.
The lessons for intellectual progress in the central image of the digres-
sion in the Republic, the Cave, are the subject of more detailed study by
Thomas Johansen. Johansen starts from an old puzzle about the Cave. It
appears to be about the ethical and political redemption of people ‘like
us’; yet it is clearly meant to pick up the lessons of the Sun and Line
concerning the philosopher’s need to be educated in a scientific under-
standing of the cosmos at large. Johansen argues that the puzzle is to
be answered by finding ways in which the Cave itself shows that ethical
progress requires broader cosmological understanding. To take one exam-
ple which the Timaeus, with its emphasis on our perceptual and intel-
lectual engagement with the heavens, might help us to appreciate: the
prisoners can only improve in understanding about anything at all thanks
to the ‘sight’ (that is, cognition) whose ultimate source is the sun. The sun,
then, already mediates between the Good, of which it is an image, and
the prisoners’ ethical progress. (That there is not more detailed emphasis
on cosmological studies is only due to the fact that the prisoners are, ex
hypothesi, not citizens of Callipolis and thus not already enrolled on the
ideal curriculum.)
In addition to their focus on the structural cohesion of the dia-
logues tackled, the chapters so far have been increasingly interested in
the theme of ‘Socratic intellectualism’. ‘Reflective commentaries’ by
two of the co-editors (so-called because they comment reflectively on
other chapters in the volume and on related claims made by Rowe)
highlight this theme by focusing on what we should mean by ‘Socratic’
(Christopher Gill) and on what we should mean by ‘intellectualism’
(Dimitri El Murr). Gill argues that the focus of Rowe and Sedley on
the continuity in Socratic thought (so described) within Plato’s work
might risk underplaying the possible differences between the histori-
cal Socrates and Plato just as much as the traditional developmental-
ist view exaggerates them. Gill suggests that a sympathetic reading of
Plato might entail a yet more radical rejection of the ‘Socratic’ ques-
tion – abandoning the very terms ‘Socratic’ and ‘Platonic’ and the quest
to identify all or some of the dialogues as one or the other. There is
after all no indication from Plato himself that he wished us to do this.
Instead, we should focus on the philosophical positions advanced: in
the case of the Republic, for example, the moral psychology can be read
as a sophisticated, double-jointed combination of tripartition (requiring
harmonisation under the leadership of reason) and a form of intellectu-
alism that is properly understood as its result.
As to ‘intellectualism’: as Johansen and Sedley make clear, the philoso-
pher’s intellectual journey is very clearly oriented towards (and so, pre-
sumably, to be understood in terms of ) the Good. But what arguments
does Plato offer for its pre-eminence? El Murr, developing a suggestion
of Rowe’s, argues that this follows from the – psychological – fact that the
good, as a matter of fact, is something whose mere appearance is never
considered desirable. (Contrast justice, or beauty, for example.) It is what
Plato argues to be the universal recognition that the good exists independ-
ently of opinion that makes it so important for human thought: the con-
cept that can make everything else intelligible, as the sun makes things in
the sensible world visible.
From moral psychology we move to dialectic – beginning once again
with a question about structural unity in a dialogue: for M. M. McCabe
argues that the apparent simplicity of the Euthydemus relative to other
dialogues only makes it harder to see its real unifying purpose. This pur-
pose is found by her in a ‘deep’ epistemological discussion which emerges
from Socrates’ encounter with the sophists, and picks up and develops his
attempt early in the dialogue to reflect on wisdom – rather than taking
(and leaving) it as the springboard for the display of sophistic fireworks.
In particular, McCabe identifies the sophists’ systematic and recurrent
refusal to allow Socrates to qualify his statements as a challenge to our
understanding of how one might reasonably qualify a principle such as
the principle of non-contradiction. It is Socrates’ failure to adduce reasons
which would provide this qualification that ultimately limits the scope of
his inquiry into knowledge in this dialogue and puts him trouble with the
sophists – real trouble, from which it will take the work of further dia-
logues to rescue him.
The prohibition on ‘qualification’ which is a feature of the brothers’
practice of eristic in the Euthydemus is used by Michel Narcy to explain
certain features of Socrates’ engagement with Protagoras in the Theaetetus.
Socrates imagines himself challenged by Protagoras to treat him fairly; but
because he is no longer there to defend himself, this can only mean deal-
ing with his words without allowing further qualification. Socrates thus
proves himself a master of eristic in responding to Protagoras – but pre-
cisely because it is the fairest way of responding to him. (The contrast
between eristic and dialectic is maintained, however, by the fact that this
eristical display is subservient to the dialectical encounter with Theaetetus.)
Narcy suggests that the focus on Socrates’ use of eristic, understood as an
Introduction 7
inflection of the ‘elenchus’, is what explains the choice of Euclides as the
narrator – indeed, as the ‘author’ – of the Theaetetus: eristic characterised
the Megarian school, of which he was the founder.
If the Socrates of the Theaetetus adopts a literalist approach to his
engagement with Protagoras, Ugo Zilioli argues that he is more devious
in his discussion of the ‘subtler thinkers’ supposedly allied to his way of
thinking. Through an analysis of the fundamental features of their episte-
mology, Zilioli concludes that they coincide with the Cyrenaics, who by
Plato’s day had developed a strong version of the identification of knowl-
edge with perception. Plato’s sensitivity to the historical setting of the
Theaetetus does not allow him to have Socrates engage directly with the
school of Aristippus; but the ‘Megarian’ frame (along with the traces of
Megarian thought that colour the narrative) serves duty here too, inviting
us from the beginning to re-read the earlier dialogue between Socrates and
Protagoras as one that plays out in an engagement between Plato and his
Socratic contemporaries in general.
Terry Penner continues the exploration of Plato’s engagement with
theories of the incorrigibility of perception represented by Protagoras, at
least as represented in the Theaetetus. Penner argues that such incorrigible
perceptions have something in common with the ‘propositions’ of mod-
ern semantics: they give content, that is, to our expressions or beliefs by
denying that these expressions or beliefs refer to real things in the world.
This works well, Penner suggests, for the operation of a ‘neutral logic’,
that is, an analysis of meaning that is indifferent to whether a sentence
is, as a matter of fact, true or false. But Plato might have thought that
philosophy ought, rather, to be concerned about things in the world –
the things that speakers refer to, rather than the (hypothetical?) entities to
which sentences refer. In this light, Plato’s analysis of false belief might be
more radical than we generally assume. Suppose that we believe one thing
(say, Theodorus) is another (say, Theaetetus), and apparently find our-
selves committed to the paradoxical claim that Theodorus is Theaetetus.
Plato might really have thought that the problem in this case is not to do
with ‘intentional contexts’ or the mismatch of memory and perception
(the ‘wax tablet’ analogy, compared by Penner with Frege’s move). The
problem is that we both know and do not know the same object – with
implications, of course, for Plato’s understanding of ‘knowledge’.
Penner’s suggestion that Plato remained more focused on real-world
entities than modern semantic theory, and appropriately so, has some
resonance in Denis O’Brien’s argument with what he sees as the preju-
dicial insistence of modern logic that being is never a predicate – and of
commentators, that Plato cannot have thought that it was. At the heart of
O’Brien’s argument is an analysis of the answer to Parmenides that Plato
develops in the Sophist. Parmenides had said that it was not possible to
talk or think of ‘what is not in any way at all’. Surprisingly, perhaps, Plato
agrees – but he also adds crucial specification. Parmenides was right if we
understand ‘what is not in any way at all’ to mean ‘what is contrary to
being’. We can, however, talk of non-being in the sense of what is other
than being. And not only can we talk about it: absolutely everything there
is – except being itself – participates in the ‘form of non-being’. This anal-
ysis requires, and so supports, the idea that being can be predicated: to
say, for example, that ‘non-being is’ (that is, that it participates in being,
one of the ‘great kinds’) is not to say that it ‘is something’.
Reflections on politics and history characterise the remaining papers –
tied as ever, however, to questions of form and unity. We begin with the
Timaeus, which contains one of the most striking and puzzling disconti-
nuities of any of the dialogues. For before we get to Timaeus’ cosmology,
which is the meat of the dialogue, there is a long trailer for the account
of Atlantis and its encounter with ancient Athens that will be the subject
of the Critias. Furthermore, it is puzzling that the characters of the dia-
logue accept the veracity of the Atlantis story every bit as much as they
do that of Timaeus’ cosmology. Sarah Broadie argues that all of this is
meant to focus our attention on the question of why it matters whether
something is true. Socrates and the others can accept Critias’ story as true
precisely because, as uncritical historical description, its truth does not
matter very much to the philosophical among them at all. Conversely,
Critias’ deep concern with the truth of the Atlantis tale, which is allied
to his failure to acknowledge the significance of how Socrates arrived at
the ideal political constitution in the ‘yesterday’ of the dialogue, shows his
own very unphilosophical outlook. This prepares us to see what matters
in the case of the cosmos: although its structure is a matter of evident his-
torical record (unlike the constitution of the Republic), what will matter
in Timaeus’ account are the reasons for its being this way (which Critias’
account of Atlantis pointedly ignores).
Broadie emphasises that the Atlantis narrative, for what it is worth, is
accepted as true by its audience. Mauro Tulli explores further the posi-
tive implications of this observation in his suggestion that Plato has more
interest in, and respect for, history than Aristotle, for example. In insisting
that Solon began to write the story down as a poem, and that the ‘writ-
ings’ exist for Critias to consult, Plato indicates that the historicity of the
story does matter. One reason for this may be that it has an inspirational
Introduction 9
function: if the ideal city existed once, it could exist again. (Note the cru-
cial role of Solon the lawgiver in the transmission of this history.) But at
the same time, Plato has a lesson to teach here about what the best kind
of literature is like. The lesson is indicated by the fact that he is following
in the footsteps of Solon in writing about Atlantis – in the case of the
Critias, starting to do so but not finishing, just as Solon had done. And
Solon’s work (he says) would have been greater than that of Homer and
Hesiod. So this, Plato suggests, is literature suitable for the ideal city. It
is so, presumably, because it is grounded in appropriate education and
understanding – something signalled by its convergence with the political
lessons of the Republic.
Plato’s interest in history is further explored by Malcolm Schofield,
who shows how he puts a revisionist account of two historical states –
Cyrus’ Persia and Athens at the time of Salamis – to the service of the
political theory of the Laws. The key aim of the legislator for Plato is
friendship: this was (he suggests) deliberately produced under Cyrus as a
matter of monarchical policy, and emerged in Athens through collective
fear of invasion. The ‘second-best’ state of the Laws steers a middle path
between these two historical precedents, invoking wisdom and freedom as
ideals supportive of friendship rather than appealing to a benign dictator
on the one hand, or the action of chance on the other.
Schofield chooses ‘friendship’ as a theme appropriate to the occasion.
All of us involved in this volume would wish to concur that Christopher’s
generous friendship is not the least of his moral virtues – and a fitting
end-point to this tribute to his intellectual inspiration.
ch apter one
Macrology and digression

Monique Dixsaut
Introduction
The title of this volume being The Platonic Art of Philosophy, it would
hardly seem appropriate to use a purely stylistic approach. Yet my aim here
is to take what are apparently formal features of speech, namely macrol-
ogy and digression, and to see whether the words gain in some dialogues
a philosophical meaning, and if an analysis of their usage might not even
lead to a better understanding of how to philosophise according to Plato.
In the Platonic dialogues, the opposition between macrology and
brachylogy (speaking at length and speaking briefly) is part of a strategy
to turn these rhetorical categories against rhetoric, a strategy whose evo-
lution can be traced from the Protagoras to the Statesman. But once the
context ceases to be agonistic, and when it is no longer a matter of criti-
cising one type of speech but of worrying about the length of one’s own
speech, whether past or future, these two words lose their technical char-
acter. The fear of an over-long speech is the most frequent cause of the
author’s incursions into his own texts, where his concerns are transposed
into reflections by the protagonist on his own speech. They interrupt the
flow of the discourse in order to inquire whether at some point the logos
has ceased to move in a straight line and therefore ceased answering the
question being discussed, or whether it is opportune to explore a problem
and thus risk being carried too far away. When the very act of putting this
kind of question finds its legitimacy challenged, this requires a distinction
between two kinds of measure and finally the need to go beyond both
kinds. For if a logos is judged too long as regards due measure, is it the
undue length alone that makes it a digression, or is it the fact of deviating
from the subject that produces a sense of undue length?
It is difficult to know what is, and what is not, digressive in Plato,
because we are dealing with dialogues, that is, a kind of logos whose con-
tinuity is, by definition, broken up by the alternation of questions and
10
Macrology and digression 11
answers; but also (unless this amounts to the same thing) because, even
if this alternation is judged to be a literary device, proceeding ‘by ques-
tioning’ (δι’ ἐρωτήσεων)1 implies that one accepts in principle that the
sequence may be disturbed and made to wander. Arguments are linked as
in a chain, but questions pop up and lead to each other. Rather hesitantly,
Socrates states in the Phaedrus that ‘philosopher or something like it’ is
the name most suitable for the man who participates in such discourse
(Phdr. 278d4–6). His reticence may be due to the seemingly contradic-
tory demands made by the very name, ‘philosopher’. To be worthy of his
name, the philosopher must add to his dialectical power to make (δρᾶν)
correct divisions and collections (Sph. 253d1–e9) the courage to speak
without knowing the truth and the courage to explore. ‘To present one’s
arguments at a time when one is in doubt and searching – which is just
what I am doing – is a thing both frightening and risky’ (ἀπιστοῦντα δὲ
καὶ ζητοῦντα ἅμα τοὺς λόγους ποιεῖσθαι, ὃ δὴ ἐγὼ δρῶ, φοβερόν τε καὶ
σφαλερόν, R. 450e1–451a1). The freedom Plato gives himself not to argue
in chains, ‘the freedom to take one logos after the other’ (καὶ μεταλήψει
τῶν λόγων καταχρώμεθα, Tht. 173b6–7), may be the indication of an
essential kinship between the philosophical and the digressive.
Macrology and brachylogy

The last division in the Sophist distinguishes two species of the ‘doxo-
mimetic art’ (the art of imitating not the thing itself but the opinion
one forms of it). One is naive and the other ironical, that is, aware of
not knowing what it pretends to know and to imitate. In the final div-
ision of the ironical species, the man who practises his art ‘publicly, in
long speeches to a large assembly’ (δημοσίᾳ τε καὶ μακροῖς λόγοις πρὸς
πλήθη) is opposed to ‘the man who uses short arguments in private (τὸν
δὲ ἰδίᾳ τε καὶ βραχέσι λόγοις) and forces others to contradict themselves
in conversation’ (Sph. 268b1–5). As for the type of imitator on the left
side of the division, the popular orator, he is called ‘more macrological’:
μακρολογώτερος, 268b7). He speaks at greater length than the other, the
sophist, but too long as well, we may be sure – a meaning suggested by
the tragic sound of the word.2 Macrology is brought back in as a means
of defining the sophist a contrario in a context that is entirely under the
1
NB ἐρώτησις refers to the act of questioning, not to the interrogative utterance.
2
Its opposite, brachulogoteros, is put in the mouth of Gorgias: Socrates will have to acknowledge
that he has never heard a more ‘brachylogical’ speaker (Grg. 449c7). Each of these comparatives is a
hapax legomenon in Plato. For this sense of tragikos, see Meno 76e3.
12 Monique Dixsaut
supervision of the person who knows how to divide according to classes:
the dialectician. Apart from the Sophist, the coupling of macrology and
brachylogy can be found only when Socrates is facing an interlocutor who
boasts that he is equally capable of both: Hippias, Gorgias or Protagoras.3
Whoever claims that he has contraries in his power, both of them or either
at will (ἤτοι ἀμφότερα ἢ τὸ ἕτερον, Hp. Mi. 376a3–4), implies that, in his
mind, these contraries are equivalent and that mastering them is a purely
technical matter.4 This holds true a fortiori of such rhetorical activities as
macrology or brachylogy, and this enables Socrates to bring it about that
his interlocutor is hoist with his own petard. When he blames Protagoras
or Polus for their ‘macrology’, he is not blaming them for their long
speeches but for making a technical mistake and getting the two kinds
mixed up: speaking in private conversation as if addressing a crowd, or
launching into an encomium instead of asking questions and answering
them.5
And yet, if one considers which answers trigger Socrates’ attacks on the
length of his interlocutors’ speeches and his own threats to quit the con-
versation, it is their brevity one finds striking. Why is Socrates reacting
in such a way at Protagoras 334c, whereas after a long speech taking up
eight full Stephanus pages (320c2–328d2), he is depicted as wishing that
Protagoras would continue and as setting great store by what he has just
heard?6 Is Plato being inconsistent? Or is it not that, in his myth as well as
his educational programme, Protagoras is answering the question whether
virtue can be taught (whatever one makes of that answer), whereas when
questioned on the useful and the good, he merely accumulates examples
proving their relativity (334a1–c6). True enough, such a theme offers infin-
ite scope for variation, and there is no reason why Protagoras should stop.
Yet he does stop, and fairly soon, so that it appears that the macrology
Socrates blames him for is not a matter of quantity. Too long a speech is
3
Hp. Mi. 363cd, Grg. 449bc, Prt. 329b.
4
And it raises no moral problems such as those discussed in the Hippias Minor: knowing how to do
good does not prevent one from doing bad, and knowing how to speak what is true does not pre-
vent one from speaking what is false: it is even a prerequisite for being able to do so.
5
‘For I had thought that getting together to enter into dialogue and holding forth to the people (τὸ
δημηγορεῖν) were two different things’ (Prt. 336b3); cf. Grg. 448d7–10: ‘It is obvious to me that
Polus, considering the language he just used, has practised more what is called rhetoric than dia-
lectic’ (my translations). All translations of the Phaedrus will be from Hackforth (1952); those of the
Republic from Bloom (1991); those from the Theaetetus by M. J. Levett as revised in Burnyeat (1990);
those of the Statesman from Rowe (1995); and those of the Timaeus from Cornford (1937). They may
have been (more or less) slightly modified. Translations from other dialogues are mine.
6
ἔτι πρὸς αὐτὸν ἔβλεπον ὡς ἐροῦντά τι, ἐπιθυμῶν ἀκούειν (Prt. 328d5–6); πολλοῦ γὰρ ποιοῦμαι
ἀκηκοέναι ἃ ἀκήκοα Πρωταγόρου (328d9–e1).
not to him a speech lasting too long, but a speech that forgets, and causes
one to forget, the question that it is supposed to answer. Thus, Protagoras’
style of macrology is not digressive; rather, it is made to seem an attempt
at digression. Too long a speech does not speak of something else; it nei-
ther raises nor treats one or more questions different from the one that
was asked at the outset. It talks about nothing and, so to speak, it talks
more and more about nothing: its inner vacuity proliferates unceasingly.
But if the speaker finally ‘persuades a large crowd’ in this way,7 it is not
too long, it has a length suitable for the effect it is aiming at; its only meas-
ure is its efficacy. The distinction in the Statesman (283e–284b) between
two kinds of measure stems directly from the thoughts of the practition-
ers and theoreticians of rhetoric and from the joke made by Prodicus: a
speech should be neither long nor short, but according to due measure.8
The reference to a due measure renders absurd the application of predi-
cates such as ‘long’ or ‘short’, but Prodicus’ assertion remains a joke since
the whole question is to determine what this due measure consists in.9 As
long as no one has found the norm by which to determine this measure,
this amounts to substituting a subjective relativity for a quantitative one:
‘what is required according to you or to me?’ Protagoras asks Socrates (Prt.
334e2–3). The norm will be spelled out in the digression of the Theaetetus
(172d8–9): ‘speaking at length or concisely does not matter at all, provided
only that they hit on that which is’ (καὶ διὰ μακρῶν ἢ βραχέων μέλει
οὐδὲν λέγειν, ἂν μόνον τύχωσι τοῦ ὄντος). When macrology is used
as the technical name for the public speaker’s discourse, neither measure
comes into play, according to Socrates. Macrology errs neither through
excess nor through deficiency nor because it is not fitting or inopportune,
but only through inanity. ‘If we spoke of the Sybil and a number of other
soothsayers … we would be going on at length (μηκύνοιμεν) about some-
thing obvious to every one’ (Phdr. 244b3–5). Going on at length, being
long-winded, means speaking for the sake of speaking.
7
Plato quoting Gorgias: to ‘persuade a large crowd’ is the aim of a long speech, the aim of a brief
speech is to make the interlocutor contradict himself (Helen 13).
8
One day when he was following a debate on conciseness and extension in speeches, Prodicus began
to laugh: ‘Only I,’ he said, ‘have discovered which speeches are demanded by art: none that are
long or short, but those which follow a just measure’ (δεῖν δὲ οὔτε μακρῶν οὔτε βραχέων ἀλλὰ
μετρίων) (Phdr. 267b3–5). See also Gorgias’ comment: ‘There are some answers, Socrates, which
require speaking at length’ (Grg. 449b9–10), or Protagoras’ question: ‘Am I to answer more briefly
than is required?’ (Prt. 334d6–7).
9
The Athenian gives another version of the same joke: it would be foolish to judge a speech by the
number of letters in it, ‘for it is the excellent things, I think, and not the shortest, nor those of great
length which deserve to be praised’ (τὰ γὰρ οἶμαι βέλτιστα, ἀλλ’ οὐ τὰ βραχύτατα οὐδὲ τὰ μήκη
τιμητέον, Lg. 4, 722a8–9).
14 Monique Dixsaut
Such a logos is not a detour or a deviation; it just never ends, which
points not only to an inability to conclude but also to an inability to begin.
One knows that in Plato a logos only begins once a question has been
asked – provided it is put in a certain way. Questions such as ‘Which is
the better man, Achilles or Ulysses?’10 or ‘Is a king happy with all his gold?’
are pretexts for macrology; only those dealing with an essence (ousia) call
for reflection. When facing the great orators of his age, Socrates keeps ask-
ing them if they would be good enough to be brief, but he is not intent on
showing that there is no one like him in brachylogy. He feigns a demand
for laconic brevity, but this is an attempt to enforce a different mode of
discussion. Macrology is monological by nature, and brachylogy is dia-
logical only in appearance. They both differ from dialectic not in scale
but in modality: when the questions are rhetorical or sophistical, they are
questions that do not arise.
Macrology and digression

On the other hand, when the protagonist of a dialogue (Socrates, Timaeus,
the Eleatic or the Athenian Stranger) says that he fears he may have been
going on too long, or may be on the point of doing so, the topos of the
overlong speech loses the strategic function it had in an agonistic con-
text – but perhaps not all strategic value. The fact that we are dealing with
dialogues makes matters more complicated insofar as the pairing of pro-
tagonist and interlocutor is superimposed on the pairing of author and
reader, without entirely coinciding with it. The task of voicing the author’s
doubts is entrusted to the protagonist with whom one may hesitate to
identify the author completely; and the interlocutor cannot be identi-
fied with the reader, though at certain times he does do what the author
expects from his reader. When the progress of the discussion is stopped by
a worry, whether prospective or retrospective, about the excessive length
of a logos, the interlocutor often displays surprising benevolence and even
intelligence. He not only urges the other to proceed, but moralises and
reminds him that the time required for thought has nothing in common
with the time measured by the clepsydra. This is summed up in Glaucon’s
phrase: ‘For intelligent men … the proper measure of listening to such
arguments is a whole life’ (μέτρον δέ γε … τοιούτων λόγων ἀκούειν ὅλος
ὁ βίος νοῦν ἔχουσιν, R. 450b6–7).
10
The Hippias Minor opens with this question, which is soon discarded in favour of the search for a
definition of dunamis.
This too is why Clinias is so obviously irritated by having to reply so
often to the Athenian: ‘But, Stranger, we have often, in so short a while,
declared that in the present situation nothing compels us to hold brachyl-
ogy in higher regard than length’ (ὡς οὐδὲν ἐν τῷ παρόντι δεῖ προτιμᾶν
βραχυλογίαν μᾶλλον ἢ μῆκος, Lg. 10, 887b2–3). Whatever his previous
misunderstandings or reluctance to agree, in a Platonic dialogue no inter-
locutor will be happy to be taxed with impatience or indolence. By mak-
ing a character judge that compliance with rhetorical rules is inadequate,
Plato shows the reader how to read – like an intelligent interlocutor, as
if playing the part written for him. These metatextual incursions tend to
prevent the written text from closing up on itself. They offer it to be read
as if it were in the very process of searching for itself, inventing itself.
Writing becomes contemporaneous with reading and forces its presence
onto the reader. But when it is thinking that hesitates and wonders about
which road has been travelled or is to be travelled, it is not only the move-
ment of writing which is interrupted. It is the continuity of the reasoning
and the argument.
If thinking amounted to no more than demonstrating, the question
of ‘too long’ or ‘too short’ would not arise: there would only be neces-
sity, as in mathematics. The only obstacles to overcome would be logical
ones, and it is well known that logical thought asks only the questions it
can answer. But anyone who reads the dialogues will learn that the logos
seems to cut the path for itself only by overcoming at almost every step
those unpredictable and heterogeneous obstacles: deafness, objections,
humours, passions or the complexity – sometimes the indeterminateness
(to apeiron) – of the object of the inquiry. The very pursuit of the logos
seems to have something heroic about it,11 and in its course the logos accu-
mulates a history, and like any history it is punctuated by events. Being
led away from its path by all these contrary forces is what would threaten
the logos with the danger of being unnecessarily long.
Yet one cannot believe such danger is real in the case of a written work.
Verba volant, scripta manent. There is no going back on speech, not the
slightest fragment of it can be withdrawn. From the moment it is uttered,
it flies away, and resists correction. More precisely, correcting it means
adding to it, and if the problem is that it is too long, adding length to
length. On the other hand, one may well delete or erase what has been
written and thus make undetectable what was found erroneous or faulty.
11
Hence the ‘new equipment’ (ἄλλης μηχανῆς), and the ‘other weapons’ (βέλη … ἕτερα), needed in
Phlb. 23b7–8, or the ‘as wrestlers would do’ (καθάπερ ἀθληταί, 41b8).
16 Monique Dixsaut
If Socrates’ first speech in the Phaedrus had really been found faulty by
Plato – Socrates twice calls it a fault (ἁμάρτημα, 242c6, d2) – it would
not coexist with the second. The palinode exposes it as sacrilegious but
when the outlook changes from ‘poetical’ to ‘technical’ the definition it
includes is not said to be wrong, but incomplete (265a). Likewise, in the
Statesman, the Stranger finds that his myth has corrected the two mistakes
(275b1–6, 277b1–3) made by the previous divisions. Yet it is not the faulty
speech, but its correction, that is judged to be too long (μακροτέραν τὴν
ἀπόδειξιν πεποιήκαμεν, 277b6).
The fear of having gone on too long, when expressed in writing, is a
literary device meant to make one believe that writing is as irreversible
as speaking. In a Platonic dialogue such a device takes on an extra mean-
ing. It is meant to show that thinking does not consist in following a
path already mapped out, but in working one’s way along one path, then
another, then yet another. When these paths end up in the same place, the
difference between them can only be a difference of length, not of value. If
one way was good – good meaning right – and the other bad – bad mean-
ing wrong – clearly they would not end up in the same place.
Allowing the interlocutor to choose between two roads,12 and even
taking him along both, as in the Statesman, sounds like the macro- and
brachy-logical facility the sophist boasts of. The difference is that these
are not two types of speech but two lines of thought, two routes, two
‘methods’. Sometimes the longer version exists only in a passing allusion,
by preterition’13 or potentially, in what purports to be a resumé. The pro-
logue of the Timaeus sets out a summary, not of the Republic as we know
it, but of the Republic that needed to be written as an introduction to a
cosmology and a physics.14 At other times, on the contrary, the shorter
version is brought up only for the sake of being dismissed by the longer
one: ‘so why ever, then, didn’t we immediately (εὐθὺς) reply that weaving
was the intertwining of woof and warp, and instead went round in a circle
(περιήλθομεν ἐν κύκλῳ) defining a whole collection of things with no pur-
pose?’ (Plt. 283b1–3). Or again the two paths taken follow each other, as in
the Statesman again, in order to please Young Socrates, in consequence of
12
As Megillus is offered (Lg. 721d7–722a2).
13
That is, the rhetorical device ‘of making summary mention of something by professing to omit it’
(OED., s.v.). E.g. Phaedrus expected from Socrates a speech balancing his preceding one, i.e. an
enumeration of the benefits a boy gains by yelding to a man who does not love him, and Socrates
answers ‘in one word’ (ἑνὶ λόγῳ): it is enough to say ‘that to each evil for which I have abused the
one party there is a corresponding benefit belonging to the other – who needs a long speech?’ (καὶ
τί δεῖ μακροῦ λόγου; Phdr. 241e5–8).
14
On this prologue, see Broadie in this volume.
which the account of the shorter road makes the journey longer, ‘adding
the digression as a kind of interest’ (προσθεὶς τὴν ἐκτροπὴν οἷον τόκον,
267a2).15 The longer road includes more stages, each of which reveals a
further aspect of the Form to be defined, by making clear a new relation-
ship between that Form and another Form. The two definitions of man
reached by the two ways – a hornless and non-interbreeding two-footed
animal, or a featherless, two-footed one (265a–266e) – are not exactly cap-
able of convincing us of the superiority of one way or the other. Dividing
tame walking animals may not be the best way to be followed in answer-
ing the question ‘What is man?’, though man may be met along that way,
for undeniably he is some kind of two-footed animal. But it is possible
to follow another way, that of the myth that is ‘so very long’, which will
reveal a truth of another sort about man, a truth as contradictory as is the
human animal in the state of nature, since this animal will only become
truly human by dint of an appropriate culture, a paideia.
It is generally safer to bet on the longer road, either because by so doing
one has nothing to lose but false opinions, and therefore everything to
gain,16 or because ‘nothing imperfect can be the measure of anything’
(ἀτελὲς γὰρ οὐδὲν οὐδενὸς μέτρον, R. 504c2–3). But the fear of an over-
long speech and the choice of the shorter road may be justified, espe-
cially in those dialogues in which an account is being given according to
a plan and proceeds without interruption. In his fear of adding ‘to the
length of the argument another secondary length on top of it’ (οὔτ’ ἐπὶ
λόγου μήκει πάρεργον ἄλλο μῆκος ἐπεμβλητέον, Tim. 51c7–d1), Timaeus
succeeds in finding a remarkably swift way to reach a decision in favour
of the existence of intelligible Forms (51c–52a). The logos that follows the
plan determined by Timaeus is entitled to be as long as it must be, but
a longer speech on the difference between knowledge and true opinion,
between an unchanging and a changing reality, and between philosophers
and non-philosophers, would a priori have been too long because second-
ary with regard to the question raised. In other words, a speech is too long
if it digresses.
15
Interest, tokos, also means ‘offspring’. ‘Aristotle explains the double sense of tokos as follows: “Usury
is hated with good reason because of its being acquisition from currency itself and not from which
currency is provided. For currency came into being for the sake of exchange. The tokos makes it
more. And it is from that that it got its name. For as the offspring are like the parents, so tokos
breeds currency from currency”’ (Bloom 1991: 463–4 n. 31). To say that an ektropē (digression) is
a sort of tokos means that it is a logos breeding logos from logos – and so adding logos to logos. This
roundabout way provides some extra logos, not more knowledge.
16
‘Either we’ll find what we are after, or we’ll be less inclined to think we know what we don’t in fact
know at all; and such a reward wouldn’t be anything to complain about’ (Tht. 187c1–3).
18 Monique Dixsaut
Digressive is progressive
But what is meant by calling a speech a digression? That it is straying
from the subject. Whose decision is it? That of the man who has some
knowledge of the subject, and therefore of what is relevant or not. But in
Plato it is precisely that knowledge that is under inquiry, so that ruling on
what is digressive or not would at the same time be a decision about the
nature of what is debated. In the dialogues, digressing cannot mean speak-
ing about something else, because a logos must speak of something else in
order to define what it is about, and even to speak at all. So a digression
is rather a logos answering a question unconnected with the previous one,
and the compound words used to name such a logos insist either on its
superfluous (periergos)17 or its secondary, incidental (parergos)18 character;
ektropē, detour, is like periergos a hapax;19 periodos (probably like periergos)
has a circular connotation: what ‘revolves around’ instead of moving in a
straight line (Plt. 286e5).
Whether it moves from the hypothesis to the consequences or works
back from the effects to their causes, the straight line admits of no discon-
tinuity, and in that sense it is the very image of method. There is some-
thing normative about the rectilinear, and when Sterne in Tristram Shandy
(VI, 40) announces his ‘chapter upon straight lines’, he states its univer-
sally acknowledged superiority – a superiority which becomes slightly sus-
pect with the list of its advocates: Cicero, cabbage-planters, Archimedes
… Obviously, to these we need to add Descartes and the third part of
his Discourse on the Method. Following the order of reasons is a safeguard
against rambling and wandering; it enables one to overcome difficulties and
reach the solution as easily and quickly as possible. But according to the
Stranger, the question as to ‘how we must find it most easily and quickly’
(ὡς ἂν ῥᾷστα καὶ τάχιστα εὕροιμεν) is on the contrary ‘a secondary one’
(δεύτερον ἀλλ’ οὐ πρῶτον, Plt. 286d7–8). Taking as the premises of an
argument the conclusion of the preceding argument is almost unknown
in Plato: either the connection is a narrative one, as in the Phaedo, or
the passage from one logos to another is, as Socrates says to Theaetetus,
17
LSJ s.v. II, 2. Plt. 286c1: περίεργα (hapax).
18
πάρεργος is most frequently used: see e.g. Tim. 51c7, Plt. 302b8. It is the only term to be used by
Aristotle, Rhet. 3, 1404a5–7: ‘for in all fairness, the only weapons with which one must fight are
facts (τοῖς πράγμασιν), so that whatever is not used to prove them is digression (πάρεργα)’, cf.
Metaph. 1089a11. It is parekbasis, used by Isocrates and chosen by Quintilian, which will become
the technical term.
19
ἐκτροπή, ‘diversion of water from its channel, turning off or aside’ is the only word translated by
‘digression’ in LSJ with a reference to Plt. 267a2.
‘starting all over again from the beginning, what precedes being deleted’
(ἐξ ἀρχῆς, πάντα τὰ πρόσθεν ἐξαλείψας, Tht. 187a9–b1). The only choice
of a deductive, because economical, procedure is presented by Socrates as
a concession he should never have made to his interlocutors (R. 504b1–5):
We were, I believe, saying that in order to get the finest possible look at
these things [sc. the virtues] there was another and longer road around
(ἄλλη μακροτέρα εἴη περίοδος), and that to the men following it all
the way they would become evident, but that proofs in accordance with
what had been said up to then could be tacked on (τῶν μέντοι ἔμπροσθεν
προειρημένων ἑπομένας ἀποδείξεις οἷόν τ’ εἴη προσάψαι). And you all
said that that would suffice. And so, you see, the statements made at that
time were deficient in precision.
Clinias is for his part delighted to find the Athenian determined to pro-
ceed. He urges him to tack on the beginning of what remains to be said to
the end of what has been said, and, if he agrees, to take it as a starting point
(καὶ ἀπ’ ἐκείνων ἀρχώμεθα, εἴ σοι φίλον, ὧν οὐχ ὡς προοιμιαζόμενος
εἶπες τότε); but becomes aware that this was just a preamble, which
amounts to giving to what remains to be said on the same subjects ‘the
status of a preamble instead of talking haphazardly, as was the case until
then’ (ὡς προοίμιον ἀλλ’ οὐ τὸν τυχόντα λόγον περαίνοντες, καθάπερ
ἄρτι, Lg. 4, 723d5–e3). Being thus encouraged, the Athenian immediately
adds an extra preamble (735a–738e), and yet another one, which will mis-
takenly be found surprising, for a classification of constitutions will make
it necessary to justify the rank granted to that whose model is now being
sketched – so that one cannot determine clearly what is still a preamble
and what is no longer one. Even the Laws is no exception: the different
logoi that make up the dialogue do not rest upon each other, but come
one after another, one ‘waking up’ the other,20 or ‘springing from it’, and
sometimes break up in succession like waves.
Thinking is not geometrical and linear, but astronomical and circular.
It is well known that ‘god invented vision in order that we might observe
the circuits of intelligence in the heaven (τὰς ἐν οὐρανῷ τοῦ νοῦ …
περιόδους) and profit by them for the revolutions of our own thought,
which are akin to them’ (τὰς περιφορὰς τὰς τῆς παρ’ ἡμῖν διανοήσεως,
συγγενεῖς ἐκείναις οὔσας, Tim. 47b7–c1). If a circular line is as continuous
as a straight one, and more akin to the truth, every logos should go round.
20
‘The true opinions existing in the young slave become knowledge when awakened by questioning’
(ἐνέσονται αὐτῷ ἀληθεῖς δόξαι, αἳ ἐρωτήσει ἐπεγερθεῖσαι ἐπιστῆμαι γίγνονται, Men. 86a7);
‘by inviting me to this, you don’t know what great swarm of arguments you are awakening’ (ὅσον
ἑσμὸν λόγων ἐπεγείρετε, R. 450b1).
20 Monique Dixsaut
The man who ‘goes round in circles’ is only treating different subjects in
order to progress, and the dialogue is the total of those various circles. But
it also happens that the same question, forcing its way through erroneous
or partial answers, compels one to revert to it. In those cases (for instance
in the Theaetetus),21 the only motion is that of ‘let us start afresh’ (πάλιν
ἐξ ἀρχῆς). Therefore a logos is not digressive because it is circular; circular-
ity and wandering are not signs by which a digression may be recognised,
since they can also characterise the movements of a logos that proceeds
in series of questions. A logos is a digression when the protagonist says it
would be one and will not accept it, or when that logos is to be skipped so
that the continuity of the inquiry may be restored. So Plato presents ‘as a
digression’ what had been introduced in Books 6, 7 and 8 of the Republic.22
But early in Book 6 Socrates had said he had completed a long difficult
logos about those who are philosophers and those who are not, and he was
sorry he had not stressed its difficulty even more. He would have done so
‘if he had had nothing else to speak about, and if many other matters had
not needed setting out in order to show how the just life differs from the
unjust life’ (484a6–b1). The end of Book 5 is like a smaller circle inside the
large one, and it is certainly not the only circle.
Early in Book 8 of the Republic, Socrates reviews the situation and
finds that, since the description of the beautiful city has been carried
through, he must, with Glaucon, ‘recall where we took the detour that
brought us here, so that we can go back to the same way’ (ἀναμνησθῶμεν
πόθεν δεῦρο ἐξετραπόμεθα, ἵνα πάλιν τὴν αὐτὴν ἴωμεν, 543c5–6). In
the Laws the Athenian finds it necessary to keep back what he has left
aside until he has completed his account (ταύταις μὲν οὖν εἰρήσθω πρὸς
τῷ τέλει περιμένειν ἡμᾶς, Lg. 6, 768c8–d1), for only then will the point
pending become clear.23 For, as he knows very well, what he is reverting
21
The whole dialogue is structured by these new beginnings: πάλιν δὴ οὖν ἐξ ἀρχῆς (151d3); πάλιν
ἐπανασκεψόμεθα (154e8); πάλιν ἐξ ἀρχῆς (164c1); πάλιν ἀντιλαβώμεθα (169d3); πάλιν ἐπὶ τὸν
λόγον τρεπώμεθα (173b5); πάλιν ἐπανίωμεν (177c5); πάλιν ἐξ ἀρχῆς (187a9–b1); ἴσως γὰρ οὐκ
ἀπὸ καιροῦ πάλιν ὥσπερ ἴχνος μετελθεῖν (187e1–2); Οὐκοῦν μακρὰν περιελθόντες πάλιν ἐπὶ
τὴν πρώτην πάρεσμεν ἀπορίαν (200a11–12). At this point Socrates becomes concerned, and asks
(200c3–4): ‘are you going to let yourselves be forced, in that way, to keep perpetually coming round
to the same point, without making any progress?’ (καὶ οὕτω δὴ ἀναγκασθήσεσθε εἰς ταὐτὸν
περιτρέχειν μυριάκις οὐδὲν πλέον ποιοῦντες;) – hence his final invitation: ‘tomorrow morning,
Theodorus, let us meet here again’ (ἕωθεν δέ, ὦ Θεόδωρε, δεῦρο πάλιν ἀπαντῶμεν, 210d4).
22
Cf. Leroux 2002: 691 n. 2.
23
Cf. Tht. 173b8–c4: ‘for you were quite right: we who belong to this philosophic chorus [sc. of free
men] are not the servants of the arguments, on the contrary, the arguments are, so to speak, slaves
to us, and each of them has to stand and wait to get finished off when we choose’ (πάνυ γὰρ εὖ
τοῦτο εἴρηκας, ὅτι οὐχ ἡμεῖς οἱ ἐν τῷ τοιῷδε χορεύοντες τῶν λόγων ὑπηρέται, ἀλλ’ οἱ λόγοι
ἡμέτεροι ὥσπερ οἰκέται, καὶ ἕκαστος αὐτῶν περιμένει ἀποτελεσθῆναι ὅταν ἡμῖν δοκῇ).
to might not be waiting quietly – in fact one never returns to the same
point. Indeed, it becomes clear in Book 8 of the Republic that the best city
(kallipolis) is no longer the just political regime to which inferior ones are
opposed: the aristoi (‘the best’) corresponding to an aristocratic constitu-
tion have been identified, and they are the philosophers. The philosopher
has become the true political actor; his knowledge of the Good guarantees
the value of the city he institutes and governs; his absence would bring
about gradual degeneracy; and the perversion of philosophy is one of the
main causes of the passage to tyranny. Similarly, after the ‘digression’ on
non-being, the sophist is no longer a man capable of imitating any technē;
he is an impersonator of opinions who creates semblances, that is, false-
hood. As for the herd governed by the good statesman, it is no longer
a gathering of hornless or featherless creatures but a community of ani-
mals humanised and made partly virtuous by education. But from that
moment onwards, the circles grow into spirals, for every digression is a
progression (Tristram Shandy again).24 Circular routes bring about change
in the point they started from and return to, so that they may fit into the
new field and tone resulting from the digression.
The problem, then, is not a matter of continuity but of unity: a dia-
logue is not one great circle; it is a series of logoi, each being a smaller
circle. One might be tempted to think that Socrates’ inviting the Stranger,
early in the Sophist (217c3–6), to choose the manner of speaking he usually
prefers is a renewal of the old macrology / brachylogy alternative, since
he asks him to choose ‘to discourse at length by yourself on any mat-
ter you wish to make clear’ (μακρῷ λόγῳ διεξιέναι λέγων τοῦτο ὃ ἂν
ἐνδείξασθαί τῳ βουληθῇς), or alternatively to use the method of ‘pro-
ceeding by questions’ (δι’ ἐρωτήσεων) – as Parmenides himself did on
one occasion, ‘developing some magnificent arguments’ (διεξιόντι λόγους
παγκάλους). This time, ‘long’ is not the opposite of ‘short’. In the alter-
native put forward by Socrates, the same verb describes the progression
(διεξιέναι, διεξιόντι); neither is the difference between ‘by yourself or with
someone else’ (the Stranger finds it insignificant in 217e1–3). The opposi-
tion lies between uttering one logos, in the singular, or several logoi, in the
plural. One logos when developed at full length and in detail is necessarily
a long one (makros), but several logoi answering several different questions
will not be short, they will be ‘magnificent’.
24
‘By this contrivance the machinery of my work is of a species by itself; two contrary motions are
introduced into it, and reconciled, which were thought to be at variance with each other. In a
word, my work is digressive and it is progressive too, – and at the same time’ (L. Sterne, Tristram
Shandy, Book I, chapter 22).
22 Monique Dixsaut
Magnificent as they may be, their plurality has ever since antiquity been
held responsible for the disconnected character of the dialogues. There is
hardly a single Platonic work about which the question of unity has not
been raised, and from the nineteenth century onwards a lack of unity has
become an argument for inauthenticity. Almost every single dialogue has
caused attempts ranging from a partial emendation to the restoration of
a passage, or even to the idea of ‘primitive’, ‘early’ versions of a dialogue
littered with clumsy interpolations (the romance of the two Republics
drags on as late as Ryle).25 The Platonic text is found chaotic, inconsistent,
unconnected, ill put together, full of ‘gaps, hasty links and glaring con-
tradictions’,26 lacking a centre, branching off in all directions. In a word:
digressive.
No doubt, dialectical writing, if it keeps flaunting the rules of rhetorical
art, risks becoming a rhetoric of anti-rhetoric – using the words ‘which
occur to you at the moment’ (Ap. 17c2), going where the wind of the logos
takes you27 – and it guarantees neither the truth nor even the sincerity
of a speech. Besides, this erratic spontaneity seems to be at variance with
Plato’s own conception of a logos: a living creature endowed with a head,
chest and feet, or with a beginning, a middle and an end (Phdr. 264c2–5).
A living unity, however, has little in common with a logical and static
one. A logos having a unity is not a logos sticking to a plan with conscien-
tious obstinacy but a logos which, while proliferating and going round in
circles, feeds on its wanderings. The staging of these returns, reversals and
diversions reveals what any method by definition seeks to obfuscate (if
by method is meant a set of rules aiming at the direction of the mind):28
that thinking is an adventure in constant danger of meeting with failure,
inconsistency and nonsense.
When the logos relating these adventures is worried about its own
consistency, it does not worry about its logical connections but about
its length. The question of macrology is for this logos an opportunity for
self-reflexion. The quantitative criterion is challenged and replaced by that
of the fitting (πρὸς τὸ πρέπον, Plt. 286d1–2), but one must add at once
that this is no general rule, for the new word is relative as well:
25
Ryle 1966.
26
Krohn 1876: 27, cited by Diès: 1932, cxxiii. Diès’ treatment of this problem in his Introduction to
the Republic (1932: cxxii–cxxxviii) stands as a model of precision and moderation.
27
‘But wherever the argument, like a wind, tends, thither must we go’ (ἀλλ’ ὅπῃ ἂν ὁ λόγος ὥσπερ
πνεῦμα φέρῃ, ταύτῃ ἰτέον, R. 394d8–9).
28
To paraphrase Descartes’ Regulae ad directionem ingenii.
There will of course in principle be different criteria of what is ‘fitting’,
dependent on what one’s goal is; the Eleatic Stranger now reiterates what
the appropriate goal is in this case – not pleasure, nor ease and speed of
progress, but an increase in one’s dialectical capacities.29
‘In this case’ probably refers to the problem of the suitable length of a
speech. A dialectical inquiry seeks to display in a logos the things that are;
it attempts to achieve the explanation (dēlōsis) of one of the finest and
greatest things, ‘and everything that is now being said is for the sake of
these things’ (τούτων δὲ ἕνεκα πάντ’ ἐστὶ τὰ νῦν λεγόμενα, Plt. 286a7).
The Form examined by a logos regulates it by orienting it, without, how-
ever, depriving it of its irregular speed, its breaks, its reversals; but it guar-
antees a kind of consistency, since what it is proper to speak of is what is
required by the object of the discussion, or rather by what the dialectician
thinks proper, for the time being. And yet the Stranger had just asked
this question in connection with his inquiry about the statesman (Plt.
285d5–6): ‘Has it been set before us for the sake of that very thing, or
for the sake of our becoming more able dialecticians in relation to all the
subjects?’ (ἕνεκα αὐτοῦ τούτου προβέβληται μᾶλλον ἢ τοῦ περὶ πάντα
διαλεκτικωτέροις γίγνεσθαι;).
The goal of an argument is different from the end pursued by the dia-
lectician: he expects his logos to have an effect on himself and on any lis-
tener. The same distinction is made in the Theaetetus: contrary to his habit,
Socrates refuses to deal with the theory that the universe stands still. But
if he thinks that ‘the question deserves much more than to be treated as a
side-issue’ (εἴτε τις ἐν παρέργῳ σκέψεται, ἀνάξι’ ἂν πάθοι), and above
all if he fears it causes the question ‘what is science?’ to be forgotten, his
refusal is not justified by the need to answer the question: what he must
do is to make use of his midwife’s art to deliver Theaetetus of the thoughts
that he has conceived about the nature of knowledge (Θεαίτητον ὧν κυεῖ
περὶ ἐπιστήμης πειρᾶσθαι ἡμᾶς τῇ μαιευτικῇ τέχνῃ ἀπολῦσαι, 184a3–
b2). Succeeding at least in saying what it is not will deliver Theaetetus
and render him gentler and less tiresome. It is its manner of progressing
(his methodos) which gives this speech its power to ‘render the reader bet-
ter at finding things’ and ‘a better dialectician’.30 About every subject it
should then be said that it is not for its sake that the dialectician goes on
inquiring.
29
Rowe 1995: 213, note on 286d4.
30
Methodos ‘means the way one takes, a pursuit, a search, and the manner of that search. “Method”
has taken too rigid a sense in modern science to be used as an accurate translation’ (Bloom 1991:
457 n. 22).
24 Monique Dixsaut
What Plato is saying here is fairly outrageous, to such an extent that
it is usually found preferable to give a kindly pedagogical rendering of it,
such as: what matters is the heart you put into it, don’t let it get you down
if you don’t succeed, you’ll improve … eventually. A similar interpretation
can be found in commentaries on the Parmenides: it is just a kind of gym-
nastics, an exercise. But this amounts to toning down what Zeno asserts:
‘For want of this rambling in every direction, this wandering, it is impos-
sible to meet with truth and gain intelligence’ (ἀγνοοῦσιν γὰρ οἱ πολλοὶ
ὅτι ἄνευ ταύτης τῆς διὰ πάντων διεξόδου τε καὶ πλάνης ἀδύνατον
ἐντυχόντα τῷ ἀληθεῖ νοῦν σχεῖν, Parm. 136e1–3).
It is for the man saying that ‘he will not put up with this going round
in circles’ (τὰς ἐν κύκλῳ περιόδους, Plt. 286e5) to demonstrate that the
same effect would have been reached by a shorter speech, that is, a speech
moving in a straighter line. The criterion of effect makes it possible to get
rid of quantitative measure applied to the logos, but also calls into question
the legitimacy of applying a normative measure to a dialectical speech.
This kind of logos should not be considered as an object (as the product of
a technē) to which a measure can and must be applied. It is the expression
of ἡ τοῦ διαλέγεσθαι δύναμις, the capacity to discern resemblances and
differences between beings, to divide according to classes, and so to find in
each case the best way to do so. The man with this power, the dialectician,
has no other end but to think better, and become more and more able to
do so. But if this is the case, a digression is not the result of an opposition
between what is required or not, between the necessary and the superflu-
ous, and this, I think, is precisely what the three examples provided by the
Stranger aim to show. The reason for all such reflections on measure, he
says, is (Plt. 286b7–11):
the disagreeableness we felt there was in our macrology about weaving, and
about the reversal of the universe, and about the sophist – the being of
non-being (περὶ τὴν μακρολογίαν τὴν περὶ τὴν ὑφαντικὴν ἀπεδεξάμεθα
δυσχερῶς, καὶ τὴν περὶ τὴν τοῦ παντὸς ἀνείλιξιν καὶ τὴν τοῦ σοφιστοῦ
πέρι τῆς τοῦ μὴ ὄντος οὐσίας).
The examples provided by the Stranger – two of them, at least – are
those that seem least open to criticism of this kind. Besides, in his digres-
sion on digressions, the Stranger is not denying that the myth of the two
states of nature, or that the ontological part of the Sophist, are digres-
sive; he fears or pretends to fear they are too long, in order to justify the
criterion he chose. He is not defending them because of their decisive
part in the discovery of the definition, for this would be using the wrong
criterion, namely ‘what is fitting to the subject’. He is trying to make it
clear that the power of a logos lies elsewhere. It does not lie in the fact of
reaching a goal; it is measured by its power to increase dialectical power.
Linking the other two examples with the truly painful division of weaving
means that the criterion stated must be taken seriously.
Platonic digressions are not parentheses: they open and close, but by
doing so they open up the logos to a new dimension. What is pointed out
as digressive often brings about a change to a different level as if, by ‘evolv-
ing around’, thought had gained height. The previous question, being
now viewed from higher up, becomes wider and deeper. The ‘digression’
comprising the middle books of the Republic clearly is such an ‘elevation’,31
just like the praise of divine madness in the Phaedrus, the ontological part
of the Sophist and the myth of the Statesman. And this also holds good of
the division of weaving since its boring and excessive meticulousness has
brought about the issue of the true goal of a dialectical discussion and has
made it clear that, for a dialectician, digression is progression. Not because
this would amount to some extra and nevertheless rewarding gymnastics,
but because these shifting perspectives, these unexpected bright intervals,
are like tributes to memory ‘which, by inspiring us with regret for the past,
now causes us to speak too long’ (δι’ ἣν πόθῳ τῶν τότε νῦν μακρότερα
εἴρηται, Phdr. 250c7–8). A dialectical logos is an exercise in reminiscence,
and each question becomes an opportunity for the philosopher to learn
again what he had forgotten he knew. Sometimes one word is enough.
At one point in the Theaetetus Socrates is ‘seized’ by a new logos:
‘But it is one logos after another, Theodorus, a greater emerging from a
lesser, which seizes us’ (λόγος δὲ ἡμᾶς, ὦ Θεόδωρε, ἐκ λόγου μείζων
ἐξ ἐλάττονος καταλαμβάνει, 172b8–c1) and at the end he describes what
he has just said as digressive (ἐπειδὴ καὶ πάρεργα τυγχάνει λεγόμενα),
accompanied by the ritual ‘let’s go back to what we were saying before’
(ἐπὶ δὲ τὰ ἔμπροσθεν ἴωμεν, 177c1–2). A digression, therefore, but the
digression of the Theaetetus has something particular about it: its pretext
is not a question, and it raises no question. Socrates is speaking of ‘what
he had thought many times and in many other circumstances’ (πολλάκις
… ἄλλοτε κατενόησα, 172c3–4), and what he has been examining at
the end was the nature of the likely (εἰκότως). It is one word spoken by
31
‘The “longer way” is the way of Books VI and VII, and the difference between the two ways lies in
this. In Book IV the method employed in studying the virtues was a psychological method … In
Books VI and VII, however, he has moved from the plane of moral psychology to philosophy’ (Cross
and Woozley 1964: 200–1).
26 Monique Dixsaut
Theodorus – leisure (σχολή) – which provides the opportunity to set up
two models: the free man, the philosopher, as opposed to a servant, a
slave – the rhetor. We are thus asked to choose between two kinds of life.
Now this choice, together with the consequences that follow from it, are
usually in Plato the theme of an eschatological myth: the digression in the
Theaetetus is, in that sense, a makros logos, and it works as an eikōs muthos.
But this long speech does not aim to persuade; its purpose is to show what
kind of life is chosen by a man addressing only his fellow slaves, and what
kind of life a man lives who is talking to gods (Phdr. 273e).
One can talk to the gods when one has been brought up ‘in leisure
and freedom’ (Tht. 175e1). Leisure is one kind of freedom, and it gives
rise to another freedom, probably the most fundamental one, the freedom
consisting in freeing oneself from the immediacy of our fluctuating per-
ceptions and from the opinions deriving from them. The digression does
not raise the question of knowledge, but it states in which kind of time
our thinking must move in order to know. So insofar as it spells out the
conditions for dialectical thinking it is also a piece of metaphilosophy.
Taking all the time needed, freeing oneself from the constraints created
by a time physiologically punctuated and socially articulated, yielding
neither to the clepsydra nor to the heat of noon, such is the condition
of a science of which the Stranger says in the Sophist that it is the ‘free
men’s science’ (τὴν τῶν ἐλευθέρων… ἐπιστήμην, Sph. 253c7–8). Time is
thought’s greater obstacle and greater ally. In order to free oneself from
the burden of opinions and of their values, one must be at leisure to think
and to come to realise that thinking does not consist in having the greatest
possible number of true opinions accompanied with their logos. It means
having trod the circular arguments of the Theaetetus. Judging that the
digression in the Theaetetus is irrelevant, something like a long footnote,32
amounts to what is said by a character in Cervantes: ‘You, be silent – and
go on with your story!’
In other words: what a pity Plato has not silenced his dialogues and
expounded his system – or written his unwritten doctrines.33
32
‘A present-day footnote or appendix’ (McDowell 1973: 174); see the discussion in Burnyeat 1990:
34–9.
33
My heartfelt thanks to Denis O’Brien, Kenneth Quandt and Jean Dixsaut for their invaluable help
with the English version of this paper. All remaining mistakes and imperfections are mine.
ch apter t wo
Two conceptions of the body in Plato’s Phaedrus

María Angélica Fierro
For Eros is frankly rooted in what man shares with the animals, the
physiological impulse of sex … yet Eros also supplies the dynamic
impulse which drives the soul forwards in its quest of a satisfaction
transcending earthly experience.
(Dodds 1951: 218)
Introduction
Prima facie there are two main, antithetical conceptions of the body (sōma)
that can be traced to different stages in the development of Plato’s philoso-
phy. On the one hand, the body is understood as an ‘obstacle’ (empodion)
to rationality in (what we might call) ‘middle-dogmatic’ dialogues such as
the Phaedo. On the other hand, according to ‘late-critical’ dialogues such as
the Timaeus, the sōma has a positive role in the exercise of our rationality,
as some recent studies have shown.1 Plato’s change of mind in this respect
could be linked to the more general assumption that in his last period he
overcame his rejection of the sensible world to which the body belongs.
Against this interpretative background, I will argue that the Phaedrus
presents itself as a puzzle in that here there are two notions of sōma that
roughly overlap with the two mentioned above, namely: (a) the mortal
body which is a ‘tomb’ (sēma), and which serves as an obstacle to gaining
full knowledge of the eidetic realm;2 and (b) the body which functions
as a ‘vehicle’ (ochēma) or ‘carriage’ (harma) and is instrumental for ideal
love.3 I will also claim that, according to the Phaedrus, this optimal kind
of body is only possessed by the gods in a permanent and paradigmatic
1
For this view see Ostenfeld 1982: 132–3; 1987: 26–7; Joubaud 1991: 19–21; Johansen 2000: 85; Carone
2005a: 50.
2
For this view of the body in the Phaedrus, see Rowe 1988 on Phdr. 250c4–6; Brisson 1991: 11–12.
3
For the idea that the body serves a neutral or even a positive role in the Phaedrus, see Griswold
1986: 136; Eggers Lan 1995; Nussbaum 1986: 220; Brisson 1991: 15; Rowe 1988 on Phdr. 250c4–6;
Nightingale 2004: 154–68.
27
28 María Angélica Fierro
way. However, philosophers are, in principle, able to transform to a cer-
tain degree their mortal bodies into an ochēma in their present, incarnated
existence and might even acquire a god-like, perfect body when their souls
leave the cycle of reincarnations.
I intend to show that the coexistence of these two (apparently contra-
dictory) views of the body in the Phaedrus should be explained not just by
seeing it as a transitional dialogue, in which Plato is fluctuating between
two different approaches to the body,4 but also by exploring the role of the
body in the dialectic of the relevant dialogues.
First, I will consider Plato’s conception of the body in middle dialogues
such as the Phaedo and Symposium. I will argue that, although in all these
works negative comments are made about the body, we also find indica-
tions that the body can play a positive role in cognitive and ethical devel-
opment. I will also suggest that the negative comments about the body
are informed by the dramatic framework, structure and subject of each
dialogue (the imminence of death in the Phaedo, the presentation of erōs
in the drinking party in the Symposium). Secondly, I will argue that in
the Timaeus the cosmological focus of the dialogue explains the positive,
teleological approach to the human body, which is viewed as part of the
body of the universe. I will emphasise the idea that the ‘divine body’ of the
universe functions as a perfect vehicle which human beings might ideally
acquire.
Subsequently, I go back to the Phaedrus in order to gain a better under-
standing of Plato’s puzzling conception of the body in this dialogue by ref-
erence to its overall framework and structure. I will show how the literary
artifice of the myth in Socrates’ palinode allows Plato to counteract the
positive, though uncritical, attitude towards physical love and the body in
the first two speeches by bringing together different elements. The myth
of the Phaedrus, by combining different views on the body, allows Plato
to suggest that the sōma is something that can either hinder philosophical
development or be transformed into an instrument for it.5
This discussion, as well as treating a specific topic, is also intended as
a contribution towards scholarly debate on the way that Plato presents
4
For this view, see Ostenfeld 1982: 234.
5
My approach assumes that in Plato’s philosophy there is a soul–body dualism, but not of a Cartesian
type, especially as the soul is always conceived by Plato as animating a body, whether mortal or
immortal. On this point see Ostenfeld 1987: 28–30; Broadie 2001; Carone 2005a: 5. Carone 2005b
even defends a kind of reductionist monism, arguing that in Plato the soul is corporeal; but against
this see Fronterotta 2007. The reference at Phd. 114c to those who live without any body at all could
be taken as meaning living without a ‘mortal’ body, rather than without any body.
Two conceptions of the body in Plato’s Phaedrus 29
philosophical ideas in his dialogues, in line with the broader aims of this
volume. As noted earlier, Plato’s attitude towards the body has some-
times been seen as forming part of the chronological development of his
thought which scholars attempt to trace within the whole sequence of
dialogues.6 My discussion, however, points in a different direction. I sug-
gest that the seemingly divergent treatment of this topic in different dia-
logues represents, rather, the overall dialectical project or objective of each
dialogue, taken on its own. If the dialogues are interpreted in this way, it
is not obvious that we can actually find the sharp differences of doctrine
that have been explained by positing development in Plato’s thought.7 To
this extent, my approach is also in line with the more unified (or ‘unitar-
ian’) way of reading the dialogues that Christopher Rowe has explored in
his most recent work and it chimes in with other examples of this line of
interpretation in this volume.8
The negative attitude to the mortal body in

some middle dialogues
The sōma as an ‘obstacle’ (empodion) in the Phaedo

The Phaedo has always been considered as the clearest expression of Plato’s
negative view on the sōma in that he presents the mortal body mainly as
an obstacle (empodion) which diverts the soul from its primary concern:
to know things by themselves and to attain the truth. For instance, sense
perceptions, and especially sight, are described as not being ‘accurate’ or
‘clear’ (Phd. 65b) because, if the soul uses them for inquiry, ‘it is dragged
by the body (sōma) to things which never remain the same, and it wanders
about and is confused and dizzy like a drunken man’ (Phd. 79c).9 Thus,
the soul needs to consider things ‘itself by itself ’ and to remove itself ‘so
far as possible, from eyes and ears, and, in a word, from the whole body’
(Phd. 66a) in order to get things right and attempt to attain wisdom.10
Furthermore, it is also stated in the Phaedo that, because of the body’s
6
A problem for this approach, not always fully acknowledged, is the speculative nature of any
chronological arrangement of the dialogues. See Kahn 2002.
7
For recent volumes relevant to this kind of approach, see Gill and McCabe 1996, and Annas and
Rowe 2002.
8
See esp. Rowe 2007a; also, in this volume, the reading of the psychology of the Republic by Sedley,
stressing continuity with the Socratic approach found in earlier dialogues.
9
Translations from Phaedrus are taken from Rowe 1988, sometimes modified slightly; text (with
references) from Burnet 1900–1907.
10
See further Nightingale 2004: 154.
connection with the sensible world, different kinds of appetites and emo-
tions arise within us, which distract us from the pursuit of the truth (Phd.
66b5–d3). This negative view of the body is underlined by the claim that
it is only in death that the philosopher, who has purified himself from
bodily things as far as possible and is now released from the body’s chains,
might attain and possess complete knowledge.11
However, even in this dialogue, a more careful analysis reveals that the
body can have a positive role in triggering eidetic knowledge if it is used
appropriately. Sensible stimuli constitute a necessary, though not sufficient,
condition for starting the process of gaining knowledge of the forms, which
is described as a process of ‘recollection’ (anamnēsis). In fact, this attempt to
gain knowledge of the forms in our incarnate existence by correct use of bod-
ily sense perceptions is an essential part of the process of purification which
the philosopher must go through in order to gain full knowledge after death.
The mortal body is thus a hindrance for those whose motivation is focused
on it, but for those who transform their attitude to the body in the process of
gaining knowledge, it acquires a beneficial function. It is our own desires that
make our sōma a prison or a medium for releasing our soul: it is up to us to
use the body to attain wisdom rather than to attach ourselves to bodily pleas-
ures (Phd. 82e–83c). In other words, the body acquires its real significance
when it is transformed into a vehicle for the expression of the soul.12
Thus, the overall framework of the Phaedo explains the emphasis on the
negative effects of the mortal body. The dramatic context, above all, the
impending death of Socrates, requires the main focus to be eschatologi-
cal. Since the stress falls on the body’s absolute separation from the soul
in death, and, along with this, the imminent corruption and annihilation
of the mortal body (Phd. 88b), the possibility of a constructive use of the
body in our present existence is not underlined. The emphasis falls rather
on the limitations that the body imposes on the soul’s rational activity and
the advantages for a purified soul in leaving behind such a problematic
dwelling.13 The final myth takes this perspective further and also provides
11
See Phd. 66e2–5, 68b4. The distinction in the Phaedo between the quasi-virtues, which concern the
body and belong to the body-lover, and authentic virtue (wisdom), which concerns the soul and
belongs to the wisdom-lover, is based on this negative view of the body and positive view of the
soul. See Sedley below, pp. 82–4.
12
Thus in the godlike condition in which the philosopher acquires wisdom in this life ‘and tran-
scends the body and its concerns’ (see Sedley below, p. 84) the body would not be eliminated but
transformed, as far as possible, into a vehicle for dialectical understanding and the performance of
practical actions according to this knowledge.
13
By contrast, non-purified souls, since they are attached to the bodily, remain wandering about the
sensible realm (Phd. 81c–82a).
a kind of cosmological framework for conjectures on our post mortem
existence without the mortal body.14
The negative attitude to physical love in the Symposium

In Diotima’s speech the account of the ‘higher mysteries’ of the erotic
ascent (Smp. 210a–212a) implies in certain respects a negative attitude to
physical love as regards the goal of coming to understand Beauty itself.
The love of beautiful bodies is placed at the very bottom of the scala amoris
and is presented as a stage that one needs to go beyond (Smp. 210ab, 211c).
Also, in the succeeding stages increasing understanding of the unity and
nature of beauty implies an increasing rejection of love for individuals
who are physically beautiful. Beautiful souls must be preferred to beauti-
ful bodies (Smp. 210bc); subsequently, beautiful adolescents who drive us
mad are considered as nothing in comparison with beautiful sciences and
Beauty itself (Smp. 210d, 211d).
However, there are also signs of a positive attitude to the bodily. First,
love for beautiful bodies is a prerequisite in order to trigger erōs and to
motivate the erotic ascent. Further, the ‘lesser mysteries’ of Diotima’s
speech offer a positive view of physical love in that erōs, as desire for the
good and for immortality, is standardly expressed in the form of biological
procreation. The sexual union of man and woman is presented as some-
thing ‘divine’ (Smp. 206c), although it is an inferior manifestation of erōs
in comparison to cultural procreation (Smp. 206a–209e).
Here too, a better understanding of Plato’s (seemingly ambivalent) atti-
tude towards the body in this dialogue is gained by putting it into context.
The dramatic context of the Symposium, the drinking party at Agathon’s,
has as its starting point the normal conception of erōs of the time, that is,
the sexual drive for a particular individual with whom one is in love. In
Greek culture erōs, which denotes the god as well as the emotion, refers to
the sexual desire for a particular individual and, specifically, to the experi-
ence of falling in love.15 The earlier speeches on erōs mainly represent a
rhetorical competition between different acknowledged experts of this
period;16 but they also anticipate uncritically some of the ideas of Diotima’s
speech. This is also the case as far as the body is concerned. Thus, in that
Pausanias’ speech (Smp. 180c–185c) places a low value on standard, het-
erosexual love and praises ‘heavenly’, homoerotic love which is directed
14
See further Sedley 1989. 15 Dover 1980: 1; Hunter 2004: 15–20.
16
See Hunter 2004: 29–37.
at the moral improvement of the beloved and not just at physical inter-
course, it anticipates to some degree the subordination of physical love to
higher goals in Diotima’s speech. Similarly, in Eryximachus’ speech (Smp.
185e–188e), where erōs concerns essentially the body and operates at all lev-
els of the cosmos, two different kinds of erōs are distinguished: an exces-
sive, harmful erōs, which brings illnesses and all sorts of disturbance, and a
temperate, healthy erōs, which medicine helps to re-establish by bringing
harmony in cases where this is disturbed by excessive erōs. Again, although
Aristophanes’ discourse (Smp. 189c–193e) expresses a positive evaluation of
sexuality overall, ordinary physical attraction towards someone else seems
to be inferior to the mutual erōs felt by the halves that fall in love and were
originally one, insofar as in the latter case what the soul of each half wants
from the other is something beyond sexual intercourse, although they
cannot explain it (Smp. 192de). The discourse of Socrates and Diotima
(Smp. 201d–212a) offers a kind of dialectical synthesis-cum-replacement
of the previous speeches; physical love is not just rejected but transcended
by higher expressions of erōs. The spectrum of types of erōs is expanded,
and erōs is presented as the force that motivates each of us towards the
good and towards procreation in beauty (Smp. 204d–206b). Physical erōs
is important as a necessary stage to start the process and so complete rejec-
tion of it is not appropriate; rather, it should be incorporated in a tran-
scendental perspective. Alcibiades’ speech (Smp. 215a–222b) shows, in a
specific case, both how a philosopher, Socrates, loves a specific beautiful
individual while transcending physical attraction, and how someone like
Alcibiades is not able to transcend the physical understanding of love.17
The positive view of the body in the Timaeus: the body

as a ‘vehicle’(ochēma)
In the Timaeus, although some comments are made about the negative
influence of the mortal body and the parts of the soul connected to it, the
cosmological focus of this dialogue brings with it a positive, teleological,
approach both to the body of the universe and to the human body. The
gods, such as the heavenly bodies and the universe itself, have an optimal
kind of body, which works as a vehicle or ochēma for the performance
of intelligent activity. The mortal body of human beings has also been
17
For this reading of the dialogue, see Fierro 2006, which, like that of Nussbaum 1986: 229–68,
allows the possibility of Socrates falling in love with a particular individual while still serving as an
exemplar for the description of Eros as lover of wisdom in Diotima’s speech (cf. Bury: 1932; Rowe
1998: 1–2).
designed for the best purpose and in this sense it is also an appropriate
vessel for the soul. However, it is up to us to use it rightly and to organise
the parts of the soul harmoniously.
The imperishable body of the gods as a vehicle for rational activity

In spite of his criticisms in the Republic of many anthropomorphic fea-
tures of the Homeric gods (R. 2, 377b–378e), Plato retains in subsequent
dialogues the traditional view that the gods have a body as well as that
their physical condition constitutes a projection of the human aspiration
to rise above our bodily restrictions, especially its inescapable deteriora-
tion and annihilation in time. However, he reconceives these ideas and
reformulates them in his own philosophical terms. Thus, the gods’ bod-
ies, as in the Homeric world, are perfect, immortal, beautiful and free
from the multiple demands of mortal bodies but, at the same time, in line
with the Republic’s critique of traditional religion, their anthropomorphic
characteristics are removed, and the gods are identified, ultimately, with
the heavenly bodies and the physical universe itself. In this respect, the
physical state that is characteristic of planets and stars and which enables
their orderly movements represents a paradigm of the ideal kind of body,
which human beings, if they emulate the gods, should wish to acquire
after death.18
In the Timaeus, all bodies are analysed ultimately in terms of the four
basic elements (fire, air, water and earth), which in turn are reducible to
two basic kinds of triangle (isosceles and scalene) which provide mathem-
atical organisation for the ‘receptacle’ of the world, the chōra. By deriv-
ation, any existing body consists of any or all these elements, which ensure
that the bodily is visible and tangible (Ti. 31b, 28b).
The gods are presented as living creatures who possess a body animated
by a soul, and three types of divine body can be distinguished: (a) the
body of the universe, composed of the four elements as these constitute
the totality of things in the world; (b) the body of the stars and planets;19
(c) the body of the traditional gods, whatever they represent.20
18
This interpretation implies, then, an extension of the ideal of godlikeness in Plato’s philosophy (see
Sedley 1999; Annas 1999: 52–71; Fierro 2001) to the body.
19
See Ti. 33b–34a; 38c–39e. While our world is mostly made of earth, the heavenly bodies of the stars
are made of fire (Brisson 2003: 17) or mostly so (Cornford 1937: 118).
20
Cf. Ti. 40d–41a. Plato’s inclusion of some traditional, anthropomorphic gods may be an indication
that he thinks that human rationality is limited in its capacity to understand the true nature of the
gods (Ti. 28c, 29cd, 30d; Phdr. 246c). See Cornford 1937 on Ti. 40d–41a. See also n. 58 below.
In principle the bodies of the gods are perishable, in that they belong
to the realm of what is generated and comes to be. The reason why their
bodies are everlasting and not subject to dissolution, as are those of mortal
living creatures such as human beings, inheres, according to the myth, in
the ‘will’ of the demiurge (Ti. 32c, 41a, 34ab). However, the text suggests
that this will is based on the fact that the demiurge (who can be taken as a
symbol of ‘craftsmanly’ reason),21 has established the quantities of the four
primary elements for the world-body as well as for the other gods, in per-
fect harmony and proportion, and thus it would be ‘unreasonable’ for him
to dissolve them (Ti. 31c, 41b). Also, we can infer that there is a relation-
ship between the immortality of the gods’ bodies and their self-sufficiency.
In the case of the world-body, since it embraces all that exists and nothing
is left out, nothing can affect it from outside and produce, in consequence,
decay or illness. Moreover, the world-body comprises all the shapes there
are inside its spherical boundaries. Also, according to the myth, it is free
from organs designed for the satisfaction of typical corporeal demands,
such as the mouth, designed to take in food for nourishment or air for
breathing, or eyes (to see) or ears (to hear). It does not have hands to catch
things, or feet to walk either, since the universe does not need to per-
form rectilinear movements (Ti. 33bc). Thus, the body of the universe is
such that it is not subject to any kind of physical instability which might
threaten its integrity; and similar characteristics can plausibly be attrib-
uted to the body of the rest of the gods, namely the heavenly bodies.22
The Timaeus develops in various ways the idea that the indestructibil-
ity, stability and self-sufficiency of the gods’ bodies make them an ideal
medium of transportation for the soul in its performance of different
rational movements, which can be summarised as follows.23 Concerning
the world-soul, there are two main movements. On the one hand, there is
21
From antiquity onwards there have been literal and non-literal interpretations of the eikōs muthos
of the creation of the cosmos in the Timaeus, and especially the figure of the demiurge. For a sum-
mary of the issue, see Carone 2005a: 31–5. Some authors (e.g. Moreau 1939: 35–6, 43–5) think that
the demiurge is a symbol of the forms; others (e.g. Brisson 1974: 81–4) take him to be an intellect,
distinct from the forms and the universe; others still (e.g. Ostenfeld 1982: 245–7; Carone 2005a:
24–52) argue that he can be identified with the world-soul – more specifically, the reason or nous
of the world-soul – which governs the universe in accordance with eidetic order. I adopt here this
last line of interpretation. Nevertheless, all interpretations share the fundamental truth that in the
Timaeus the cosmos is presented as an organised whole with a teleological order, although this
truth can only be really achieved through philosophical understanding. (Cf. Broadie below in this
volume, on truth in the Timaeus and Critias.)
22
See Brisson 2003: 17–19.
23
Here I follow mainly the table of the celestial motions given by Cornford 1937: 136–7. He also
explains how the earth is able to rotate on its axis (Ti. 40bc) and at the same time remain still (cf.
Cornford 1937: 120–34).
the axial rotation of the whole universe according to the circle of the same
(Ti. 37c), which causes the diurnal revolution from east to west of all heav-
enly bodies, namely the stars and planets (Ti. 40b, 36c, 39a). On the other
hand, there is the movement of the seven ‘errant’ bodies according to ‘the
circle of the other’ from west to east (Ti. 36c, 37b, 38c), which is opposite
to the movement of ‘the circle of the same’, and is subdivided into six sec-
tions comprised of seven orbits (Ti. 36d, 37c).24 Regarding the soul of each
star and planet, the movements take the form of axial rotation (40a) and,
as regards the planets, revolution at different speeds (Ti. 36d, 38d, 40c) as
well as reverse movement (Ti. 38d), except in the case of the sun and the
moon. The constant performance of these movements with unfailing cor-
rectness is linked to the fact that the world-soul (Ti. 37b) and, derivatively,
the soul of the heavenly bodies (Ti. 38e) always succeed in knowing the
truth and so in making true judgements, whether about intelligible or per-
ceptible things. For this reason the world-soul never falls short in know-
ing, so to speak, ‘the right thing to do at the right moment’.25 Moreover, its
direct access to the intelligible, without the need for the effort to recollect
from perceptible instances as a starting point, eliminates any gap between
theoretical understanding and the practical implementation of this know-
ledge. It is thanks to these orderly movements of their souls that the body
of the world and the heavenly bodies are said to trace ‘a moving image of
eternity’ (Ti. 37d) and be considered as ‘visible gods’ (Ti. 40d).26 Also, in
their case, although they do have affections that come from their bodies
and respond to the physical-mechanical causation of ‘necessity’(anankē),27
they are easily subordinated by noetic activity to a teleological purpose.28
Thus the soul only needs to employ a minimum amount of its activity in
keeping these bodies in order and moving them skilfully. In this way, in
the performance of these rational, astronomical movements, the body sim-
ply plays the role of being a physical support for the soul, so that nothing
interrupts, interferes with or delays the intelligent activity of the gods.
The gods, then, represent what it would mean to have a perfect somatic
condition. This condition must be immortal insofar as it is indissoluble
24
That is, the moon, the sun, Venus, Mercury, Mars, Jupiter and Saturn.
25
As Brisson 1974: 340 states: ‘La fonction cognitive de l’âme du monde découle directement de
sa fonction motrice’ (‘the cognitive function of the world-soul results straightforwardly from its
kinetic function’). See all of Brisson 1974: 340–52; also Johansen 2000: 89–90.
26
Cf. Nightingale 2004: 172.
27
Also called the ‘errant cause’ (Ti. 48a) or ‘necessary cause’ (Ti. 68e–69a).
28
On this point see Ostenfeld: 1987: 33–7 and Carone 2005a: 35–42. On the way the body becomes a
‘vehicle’ and instrument of the intelligent action of the soul, esp. the world-soul, in the Timaeus see
Carone 2005a: 50.
and maintains a perfect, indestructible proportion between its compo-
nents. Moreover, thanks to these characteristics, the body is able to work
solely as a support or ‘vehicle’ for the understanding and performance of
intelligent action.29 A body with these features is one that, as will be shown,
only the philosopher (among human beings) could hope to attain – par-
tially in his incarnate existence, and perhaps wholly after death.
The teleological conception of the mortal body of human beings

This idea that our mortal body can play a positive role in the attainment
of truth is also developed in the Timaeus, as others have shown.30 On the
one hand, as in other dialogues, the human soul’s implantation in a mortal
body is presented as bringing irrational disturbances such as perceptions
that come from violent impressions, desires mixed with pain and pleasure,
fears and anger (Ti. 42ab, 49c). These motions of the mortal body inter-
fere with rational activity insofar as the circular motions of the same and
the different, which are always coordinated in non-mortally incarnated
human souls, such as the world-soul and the souls of the heavenly bod-
ies, are disrupted in different ways.31 However, according to the Timaeus’
general project of providing a teleological explanation for natural events,
this necessary factor (namely, the irrational affections required by the
human soul’s attachment to a mortal body) is made to co-operate with
reason.32 Similarly, the human soul is deliberately distributed in different
parts of the body so that any necessary influence which comes from the
body helps, or at least does not hinder, the work of intellect.33 In fact there
are suggestions that even the appetitive part of the soul can participate,
though distantly, in some kind of rational activities.34 In other words, the
‘errant cause’ (Ti. 48a), which operates mechanically and unintelligently,
29
On the way that intelligent action involves contemplation and action, see Nightingale 2004: 127–
31; Armstrong 2004.
30
On what follows see Johansen 2000.
31
The circle of the same stops altogether and does not coordinate the circle of the different any
longer. The circle of the different deflects into rectilinear movements (Ti. 43a).
32
See Carone 2005a: 61.
33
The immortal part, the intellect, is put in the head to maintain the circular movements (Ti. 44de);
the spirited part is between the head and the midriff to be able to follow reason’s advice; the appeti-
tive is in the lower abdomen to keep its influence far from reason’s abode (Ti. 69c–73a).
34
See Lorenz 2006: ch. 7. Some interpreters also attribute some kind of cognitive capacities to the
appetitive part in the Republic, especially on the basis of R. 10, 602c–608b (see Penner 1971: 100–
8; Moline 1978: esp. 10–15; Annas 1981: 338–44; Janaway 1995: 143–57; Price 1995: 48–72; Cooper
1984; Lorenz 2006: ch. 5); others such as Murphy (1951: 239–43) think this cognitive capacity is an
inferior version of the rational part (on this point see also Sedley 2004: 113). On either view, the
is also in this case directed by the mind as far as possible.35 In this way, our
mortal body, which transports the soul in our mortal existence, and espe-
cially our head where intellect dwells, is conceived as a ‘vehicle’ (ochēma,
Ti. 44e, 69c) designed for the best purpose. However, the Timaeus also
stresses that part of the precondition of a just life is keeping the body
in the right proportion in relation to the soul, as well as calibrating its
impact on the different aspects of the soul, especially the irrational parts.36
This should lead to the practice of philosophy or, at least, to an amateur
contemplation of the order of the universe in order to achieve full har-
mony in soul and body.37
The possible acquisition of a god-like body by philosophers

In addition to this positive, teleological approach to the body, there are
suggestions at Ti. 41d–42d regarding the possible acquisition by the phil-
osopher of a god-like type of body. First of all, at the beginning of the
world the demiurge provides as a corporeal support for each human soul a
different star;38 and ‘mounting them [i.e. the human souls] as in a vehicle
(ochēma)’, he showed them ‘the nature of the whole’ (Ti. 41de). Thus, before
mortal embodiment, the soul is imagined as possessing a heavenly body as
a medium of transportation, and being able to travel with the gods across
the universe. The soul is also seen as capable of achieving an unmediated
understanding of ‘the whole’ (τοῦ παντός), which probably includes the
physical as well as the metaphysical realm.39 However, human-souls-to-be
are unable to stay in such condition and must be reincarnated in mortal
bodies. Even so, according to Ti. 42b any human soul can hope to return
at some time to its native star by acquiring again an ochēma that allows the
performance of uninterrupted noetic activity – provided that the person
has lived his mortal embodied life well. This includes the stipulation that
appetitive part (like the spirited part) is amenable to being shaped by reason, either through acquir-
ing correct beliefs or through habituation by reason.
35
See Ostenfeld 1982: 132–41 and 1987: 33–7; also Carone 2005a: 37–42.
36
See Brisson 1991: 14.
37
See Carone 2005a: 12, 70–8 on these two different ways of gaining access to virtue; also Johansen
below in this volume on how, in the Republic, intellectual engagement with the heavens improves
understanding. On the interpretative issues raised by Plato’s presentation of embodied psychology
in the Timaeus, see also Gill 2000b.
38
ἄστρα here refers to the stars, not the planets where, subsequently, the demiurge will sow the
human souls before they enter into mortal bodies (cf. Taylor 1928 on Ti. 42a). Del Forno 2007:
287–8 summarises the different interpretations of this problematic passage.
39
τὸ πᾶν might refer just to the whole physical universe but, as Nightingale 2004: 178 suggests, could
also mean the totality of what exists.
during the span of his present existence he has been able to achieve a good
management of his body so that, as far as possible, it does not obstruct
but rather helps him in the acquisition of knowledge and in intelligent
behaviour.40 However, it is not at all clear what it would mean to have a
star as a body or to be the soul of such a body. Each star already has its
own soul; and, besides, it is difficult to imagine what kind of personal sur-
vival would be constituted by inhabiting a star.41 The whole idea sounds,
then, more like the expression of a human aspiration to overcome in some
way the vulnerability and limitations of our mortal body by gaining some
kind of perfect physical support than the expression of confidence or cer-
tainty about this possibility.
Understanding the Phaedrus’ puzzling attitude to the body

We can now return to the Phaedrus and form a better understanding of
its apparently inconsistent attitude to the body by placing this feature in
the context of the dialogue as a whole. The three speeches about erōs in
the first part of the dialogue are intended to function both as forms of
teaching about love adapted to Phaedrus’ mentality, and as examples of
rhetoric, the topic of discussion in the second part of the dialogue (Phdr.
257b–279b). Within this framework, we can find a positive, though uncrit-
ical, attitude towards physical love and the body in Socrates’ first speech
(Phdr. 237b–241d), which challenges Lysias’ speech, a speech that had fas-
cinated Phaedrus (Phdr. 230e–234c). A more critical approach to physical
love and the body is expressed in Socrates’ palinode (Phdr. 243e–257b)
where the body is viewed positively but only if used for cognitive and eth-
ical development. The complexity of the myth of the winged chariot in the
palinode allows Plato to bring several different elements together. These
are an account of the soul and an account of erōs which describe simul-
taneously our present existence, as in the Symposium, our after-death exist-
ence, as in the Phaedo, and human life in a cosmological context, like the
Timaeus. These different approaches to our psychosomatic structure, which
are encapsulated in the literary artifice of the myth, are needed as back-
ground for the second part of the dialogue, where Plato develops the idea
that speech should be adapted to the kind of soul to which it is addressed
(Phdr. 270b). The combination of these different views results in the appar-
ently puzzling view of the body offered in the myth. Erōs is aroused in us
40
On this psychosomatic therapy see Ti. 87c–90d, and further Joubaud 1991: 194–261.
41
See Taylor 1928 on Ti. 41d8–e3.
through the appearance of a beautiful body; if we remain at this level, there
is no opportunity for us to attain wisdom; but if we use this as a stepping-
stone towards recollecting Beauty in itself, the body becomes more like a
vehicle for the soul, enabling us to emulate the gods. This also allows us to
contribute to the development of winged erōs at a cosmic level in our pre-
sent and after-death existence. This approach in the myth of the Phaedrus,
which combines different perspectives, allows us to think that, although
for Plato the mortal body is in principle something negative, insofar as
it hinders the soul’s full attainment of knowledge (see Phdr. 250c), it can
nevertheless actually help philosophical development, if the soul uses it
appropriately.42
The negative dimension: the mortal body as a ‘tomb’ (sēma) for the soul
According to Phdr. 245e, bodies can be divided into two main kinds
according to the way they ‘are moved’ (κινεῖσθαι): ‘For all body which
has its source of motion outside itself is soulless, whereas that which has
it within itself and from itself is ensouled, this being the nature of the
soul.’ The difference between one group and the other lies in whether they
are moved ‘from outside’ (ἔξωθεν) or ‘from inside’ (ἔνδοθεν αὐτῷ) and
‘from itself ’ (ἐξ αὑτοῦ). In the case of the second class of bodies this inner
movement is attributed to the fact that they have a soul, whose nature has
been previously defined as ‘self-moving’ (Phdr. 245c–246a), and for this
reason they are animated by it (ἔμψυχον).
Animated bodies which are moved from inside by the soul are also sub-
divided in the following way at Phdr. 246cd:
(a) the body that provides a location for a ‘mortal living being’ (ζῷον
θνητόν), or the compound of soul and body insofar as this is
‘terrestrial’(literally ‘made of earth’) and thus perishable (σῶμα
γήϊνον);
(b) the body that gives place to an ‘immortal living’ (ἀθάνατόν τι ζῷον)
insofar as in this case soul and body are grown together for the whole
of eternity.43 This kind of body is possessed in a permanent way only
by the gods.
42
For the idea that the soul, although it is normally trapped inside the mortal body, can use the body
for the attainment of its goal of eidetic knowledge, cf. Brisson 1991: 15 (‘the body is neutral; only
the soul can be good or bad’ [my translation]).
43
I return to this point below, p. 42.
By contrast with the gods, the body that all human beings have in our
current existence belongs to the kind that, being animated (Phdr. 245e), is
subject to death (Phdr. 246c). In this respect, we are in the same situation
as any other mortal living creature.44 In principle, the mortal body is seen
as a hindrance to the attainment of absolute knowledge in our present life,
as Phdr. 250c suggests:
[In earlier times] we celebrated [the most blessed mysteries], whole in our-
selves, and untouched by the evils (ἀπαθεῖς κακῶν) which awaited us in a
later time, with our gaze turned in our final initiation towards whole, sim-
ple, unchanging and blissful revelations, in a pure light, pure ourselves and
not entombed in this thing which we now carry round with us and call body
(σῶμα), imprisoned like oysters. (Trans. Rowe 1988, slightly modified; my
italics)
According to this passage, the main disadvantage of our mortal body is
that it is like a tomb in which we are trapped because it blocks direct access
to the eidetic realm and brings us all kinds of evils, which did not exist in
the pure, primordial existence without a body posited by the myth.45 This
passage evokes Plato’s characterisation of the body as an ‘obstacle’ (empo-
dion) to noetic activity (Phd. 65a, 66b–d), noted earlier.46 It also brings
to mind the passage at Republic 10 (611b–d), where our body is described
as an ‘oyster’ (ostreon) whose thousands of evils conceal the soul’s rational
nature. The fact that reincarnation in a mortal body is presented at Phdr.
248c–e, as in other dialogues, as a punishment for those who, even after
death, have not been able to see the forms properly, reinforces the idea that
our present corporeal state impedes full development of rationality. How
the mortal body constitutes an obstacle to noetic activity is presented in
the Phaedrus in two ways.
First, in the case of human beings, there is an obscure aspect of our
soul which is closely connected with our mortal body and which hinders
reason’s rule over the soul. To make sense of this idea we need to refer to
an image of the soul offered at Phdr. 246ab. According to this image, the
44
Phdr. 249b seems to refer to the existence of animal souls which are not those of human beings.
45
McGibbon 1968 maintains, against Bluck 1958, that at Phdr. 248a–249b there is a reference to the
‘first fall’ of the soul. On this view, the passage refers to the pure existence of all souls before the
fall.
46
At Phd. 82e, the soul is also described as bound fast into the body, like a ‘prison’ (eirgmos). At
Cra. 400c, three possible etymologies of sōma are presented. Some people associate sōma with sēma
because they think that the body is a ‘tomb’ for the soul (cf. also Grg. 493a); others make the
same association, but take sēma to mean a ‘sign’ through which the soul expresses itself; followers
of Orpheus think that it is an ‘enclosure’ (peribolos) or ‘prison’ (desmōtērion) in which the soul is
securely kept (sōzetai). In this passage Socrates presents this third etymology as the most likely.
soul, which has been shown to be essentially self-moving and the cause
of the inner movements of any living creature, must be compared to the
dunamis (‘power’) that results from the combination of a yoke of winged
horses and a charioteer. While the charioteer represents reason, specifically
the element in the soul which yearns for truth, and drives the carriage
according to what he knows to be best, the horses symbolise irrational
‘pulls’. These are forces which derive from the soul but are incapable of
determining by themselves the most appropriate direction and itinerary.
More precisely, at Phdr. 246b, the irrational drives of our souls are repre-
sented, unlike those of the gods, by two different kinds of horse: a white
one which is willing to obey reason’s directions, like the divine horses,
and a black one which derives from contrary elements, specifically bad
and ugly ones, and is opposed to reason (Phdr. 246b). It is clear that the
‘black horse’ aspect of the soul, which humans have and gods do not, has
a natural tendency to be attached to a terrestrial, perishable body.47 The
black horse, which participates in badness, hinders the leadership of rea-
son (the charioteer) because of its attachment to earth.48 It is because of
its intrinsic ‘weightiness’, combined with reason’s enfeeblement, since the
wings cannot be nourished from the ‘plain’ of truth, that the soul, filled
with ‘forgetfulness and evil / incompetence’,49 becomes heavy (βαρυνθῇ)
and without feathers, and so is reincarnated in a mortal body (cf. Phdr.
47
The black horse can, then, be described as ‘intrinsically evil’ (see McGibbon 1968: 60) because
its tendencies are always opposite to reason’s yearning for the truth. Right training can reduce its
influence on the dynamics of the soul so that reason’s desire for wisdom (the wings) can strengthen
and grow. However, this element will always drag the soul in the opposite direction to the real
good. I consider that the depiction of the human soul as a charioteer driving a chariot pulled by
two horses with contrasting natures (the evil, black one and the good, white one) matches the tri-
partite soul in the Republic: reason and the two irrational parts (the appetitive and the spirited). As
I have argued elsewhere (Fierro 2003 and 2008), in the Republic the development of erōs in the right
or wrong direction depends on the configuration of the three primary sources of motivation. Also
the non-rational parts in the Republic are linked to the body.
48
At Phdr. 247b, βρίθει may refer to ‘a natural tendency to move downwards’ (thus, a possible transla-
tion is ‘heavy’), while ῥέπων denotes the ‘actual motion downwards’; see McGibbon 1968: 58 n. 2.
49
λήθης τε καὶ κακίας πλησθεῖσα (248c7). The translation of kakia in this passage is controversial:
‘incapacity’ (LSJ s.v.); ‘evil’ (Fowler 1925); ‘wrongdoing’ (Hackforth 1952); ‘weakness’ (De Vries
1969); ‘incompetence’ (Rowe 1988). The real issue is how kakia is connected with lēthē, which refers
here to the fact that soul forgets its divine origin. According to the principle of Socratic intellec-
tualism, if someone knows what is good – in this context, fully ‘remembers’ the forms – he should
act rightly and produce good actions; similarly, if someone does not know what is good – i.e. ‘for-
gets’ the forms – he should act badly and produce evil actions. (I follow e.g. Rowe 2007a: 37–9, 171
in assuming that Socratic intellectualism is not rejected but reformulated with the tripartite theory
of the soul in ‘later’ dialogues such as the Republic and the Phaedrus; cf. variously in this volume the
discussions of Sedley, Gill and El Murr.) If τε καὶ is epexegetical, then ‘kakia would be almost iden-
tical with lēthē’ (De Vries 1969 ad loc.), so that we would have here the Socratic-Platonic equation
of forgetfulness, ignorance and evil.
248c–e). The description of the sexual drive at Phdr. 250e, which is pre-
sented as the typical appetite of the black horse and which clearly stems
from the mortal body, illustrates graphically how and to what extent the
black horse’s bestial demands can persistently drag the soul against rea-
son’s advice. Similarly, in a hypothetical afterlife without the mortal body,
its residual effects persist in the black horse of the soul, which continues
drawing the soul heavily towards earth and interfering with the chariot-
eer’s guidance.50
Secondly, in the Phaedrus, except in the case of those who have been
initiated into philosophy, the view of the physical beauty of the beloved
which comes through the body (διὰ τοῦ σώματος), and especially through
the sight (ὄψις), is described as the fundamental cause of sexual desire
which provokes commotion and disturbance in the lover’s soul (Phdr.
250d) and incites him to assault the beloved wildly in order to obtain
physical pleasure and gratification (Phdr. 251a).51 As in the Phaedo, the
vision of corporeal beauty in the Phaedrus is presented as treacherously
carrying away the soul from reason’s command to restrain from merely
sexual gratification.
The positive dimension: the mortal body as a medium

for cognitive and ethical development
However, according to the Phaedrus, if both sexual appetite and the per-
ception of beauty which awakes bodily desire are handled appropriately,
they do not impede rational activity but become the starting point for
stimulating in the soul erōs for the truth. It is noteworthy that at Phdr.
246de the ‘wings of desire’ which lift the soul upwards towards the truth
are described as related to the body:52
The natural property of a wing is to carry what is heavy upwards, lifting it
aloft to the region where the race of the gods resides, and in a way, of all the
things belonging to sphere of the body, it has the greatest share in the divine,
the divine being beautiful, wise, good and everything which is of that kind.
(Trans. Rowe 1988, slightly modified; my italics)
50
See Phdr. 248a: ‘as the horses force them’ (that is, the non-divine souls). This does not contradict
the point that the soul loses its wings because of ‘the evil / incompetence of their charioteers’ (Phdr.
248b, with n. 49 above), as the drivers are ultimately responsible for not giving appropriate training
to the black horse as well as for failing to practise philosophy.
51
See Griswold 1986: 124–5.
52
See De Vries 1969 ad loc., and his quotation of Cushman 1958: 208 n. 62: ‘The phrase περὶ τὸ σῶμα
strongly suggests that, however transcendent in its reference and impulse, the higher erōs is, never-
theless, grounded in and is perhaps continuous to physical desire.’
The connection of the wings with the body seems to be this. Although
long ago all souls were winged (Phdr. 251c), only the gods’ souls are always
perfectly winged (Phdr. 246c), while those of human beings, which are
incarnated in a mortal body, face instead the challenge of ‘becoming
winged’ (249a). In this process, the reception of the ‘flow of beauty’ on
contemplating the physical beauty of the beloved is what actually revives
the natural power of the wings as it makes the ‘feathers’ (that is, the rise
of erotic desire) grow. This produces physical reactions such as sweating,
irritation and tickling, while the absence of the beloved, together with
the beauty that he embodies, dries and blocks the holes through which
the wings (that is, desire) finds its way out (Phdr. 251bc, 255c). Thus, the
experience of falling in love with a particular individual, who is a concrete
corporeal instantiation of Beauty, constitutes the first ‘kick’ towards arous-
ing erotic desire.53
In this respect, as in the Symposium, the first step in turning on the
engine of desire is to fall in love with one beautiful body (Smp. 210d),
which in the Phaedrus even becomes a ‘divine statue’ to adore.54 However,
in order to become really winged and be able to fly towards the truth, it is
necessary to move beyond physical beauty and use it as a trigger to recol-
lect the Beautiful itself (Phdr. 251a). As in the Phaedo, the sensible stimuli
are a necessary, but not a sufficient condition for starting the process of
gaining knowledge of the forms, which is also described here as a process
of ‘recollection’.55 As in the Symposium, one must not remain attached to
one beautiful individual. The erotic drive prompted by a vision of corpo-
real beauty should ideally end in contemplation of the form of Beauty –
although this does not necessarily preclude personal love, but rather puts
it into the right perspective.56 Erotic matters need to be conducted in the
right way and the lover needs to be initiated in the mysteries of love so
that, in that he is able to recollect the eidetic realm, the wings of desire can
receive, as far as possible, the appropriate nourishment which makes them
more robust, namely access to the ‘meadow’ of truth (Phdr. 247e, 248c).
Besides, in order to help the current of desire flow in the right direction
and in order to make it stronger, appropriate training of the black horse is
necessary. We need to train the appetites that derive from our mortal body
so as to diminish the strength of their pull downwards, in the opposite
direction to the meadow of true reality where the wings of desire long
53
See further Nussbaum 1986: ch. 7 (also Griswold 1996: ch. 3). But I disagree with her view that this
marks a radical change from Diotima’s conception of love in the Symposium; see Fierro 2006.
54
Nightingale 2004: 164–8. 55 Phdr. 249c. See Scott 1987: 355.
56
See Fierro 2006.
to carry the soul. As is clearly illustrated at Phdr. 254a–e, the lascivious
craving for merely sexual pleasure represented by the black horse can be
constrained and made to accept reason’s guidance, but only after repeated
experiences of pain and punishment (Phdr. 247b, 254e). Thus, if the bodily
appetites are confined within appropriate limits, they still, in spite of their
tendency to disturb, allow the soul to ascend towards the eidetic realm.
The philosopher’s acquisition of a god-like kind of body in the Phaedrus

The ideal body of the gods in the Phaedrus as the optimal vehicle for intelligent
activity In this section, I shall show that in the Phaedrus the gods who
lead the caravan of the souls towards the eidetic realm possess imperish-
able bodies which function as an appropriate medium for the noetic soul.
In this respect, apart from their resemblance to the Olympian, anthropo-
morphic gods, they are quite similar to the gods of the Timaeus (that is,
the heavenly bodies), and some features of the myth of the Phaedrus even
evoke their astronomical movements. There are also suggestions that phi-
losophers might acquire in time an ideal body of this kind, though it is
not quite clear what this would consist in.
The imperishable nature of the body of the gods As stated at Phdr. 246c,
and by contrast with mortal creatures who possess a perishable, dispos-
able body, in an immortal being, soul and body are fused together for
ever. We are told that only conjecture about the gods, and not fully rea-
soned argument, is available to us. However, it can be inferred that, as in
both mortal and immortal beings the soul is immortal (245c–246a),57 the
immortal character of the second compound must imply the possession of
an imperishable kind of body.58
57
However, only the souls of the gods are properly called ‘immortal’ (Phdr. 247c) because they have
never been contaminated by what is earthly and mortal.
58
I take it that in this passage the gods are said to be an immortal compound of soul and body in
a way similar to Timaeus 41ab, where the stars and planets are conceived of as (a) alive, (b) gods
and (c) immortal – although not in the fullest sense, as they are able to die. The assertion at 246c
(‘immortal it is not, on the basis of any argument which has been reasoned through, but because
we have not seen or adequately conceived of a god’) can be interpreted as a reference to the limi-
tations of human discourse to give a complete and coherent account about this point (see n. 20
above). According to Rowe 1988 (ad loc.), Plato here evokes the popular, anthropomorphic concep-
tion that the gods are, like us, ‘living creatures’ (and this would be, then, the conception that is not
based on a sufficiently reasoned argument: 246c), although at the same time he believes that the
gods, in fact, do not have souls but just are souls.
The body of the gods as a ‘vehicle’ of rational activity In the case of the
gods, as opposed to human beings and other living creatures, both horses
(that is, the non-rational parts of the soul) are good and come from
good things (Phdr. 246a). This implies that, by contrast with the souls
of human beings, there is no conflict within the souls of the gods and, in
consequence, they are perfectly ‘winged’ (Phdr. 246c). All their parts or
elements pull in the same direction, towards what really exists and is col-
ourless, uniform and intangible, and so they regularly achieve complete
and authentic knowledge of this goal (Phdr. 247e–248a). Also, in terms of
the myth, after each vision of this place ‘outside the heavens’ (Phdr. 247c),
the souls of the gods keep revolving and travelling across the heavens in
circular motions while they order and take care of everything under the
direction of the administrator of the cosmos, that is, Zeus (Phdr. 246e). In
other words, they execute without interruption rational activity according
to their complete knowledge of the truth.
As noted earlier, a body is something that must always be moved in one
way or other. Accordingly, in the mythical image, the chariot, which is
drawn by the horses and the charioteer (that is, by the soul) must repre-
sent the body. Thus, at Phdr. 246e5 we find reference to the ‘chariot’ driven
by Zeus, the leader of the ‘army of gods’: ‘First in the heavens travels Zeus,
the great leader, driving a winged chariot (πτηνὸν ἅρμα), putting all things
in order and caring for all’ (trans. Rowe 1988, but my italics). The Greek
word harma may mean: (a) just the chariot (especially a war chariot,
which is driven by horses); or (b) the chariot and the horses; or (c) just the
team of chariot-horses.59 Insofar as the harma is here described as ‘winged’
and the wings belong to the soul,60 and especially to the horses (which
should be imagined as winged horses such as Pegasus and the horses of
Pelops),61 it can be argued that the meaning in question here is (c) and
so no ‘chariot’ is referred to.62 Besides, in principle, only the soul can be
winged, since it can move itself upwards; the wings represent, I take it, the
59
See LSJ s.v. ἅρμα. Similarly ζεύγος at Phdr. 246a7 may mean both (a) ‘a team, not necessarily a
pair, of beasts’ (see Hackforth 1952 ad loc.) or (b) ‘the carriage drawn by the beasts’. See also LSJ s.v.
But in this case, (a) is more likely as ‘Socrates goes on in the next sentence to talk about the horses’
(Griswold 1986: 262 n. 28). The term συνωρίς (246b2) designates a pair of horses (but also with or
without a carriage).
60
Actually ‘to every part of the soul’, as Griswold (1986: 93) points out. This agrees with what it is said
at Phdr. 251bc: ‘originally all the soul was winged’.
61
See Hackforth (1952: 77) who refers to Pindar, Olympian Odes 1.87.
62
See again Hackforth 1952: 77 (‘the chariot has no symbolic value’), and Griswold 1986: 93 (‘a char-
iot is not actually mentioned in the myth’).
erotic strength that drives the soul ideally towards the truth. However, it is
also clear that the wings start to grow from the body, as if they derive their
initial ‘fuel’ from it.63 Thus, it makes sense to think that harma means
here (b), the whole winged set of chariot and horses. The image seems to
be in this passage, a ‘body-chariot’ which is moved by the strength of the
irrational parts of the soul, the wings of the horses, and kept in the right
direction by the intelligent leadership of Zeus.
Also, at Phdr. 246e–247b, it is clearly asserted that gods (and daemons),
as well as human beings deprived of their mortal bodies, travel in an
ochēma or ‘vehicle’. The term ochēma, which derives from the verb ocheō,
‘to go in a vehicle’, and ocheomai, ‘to transport, to support’, has as its more
general meaning ‘anything that bears or supports’, that is, a vehicle and,
more specifically, a ‘carriage or chariot’.64 Thus, although, as in the case
of harma, ochēma might include in its meaning the animals ridden, or
even refer only to them, the most natural image seems again to be the
chariot that gives support and transportation to the soul, being pulled by
the horses and guided by the charioteer.
In the case of the gods, their ochēmata are moved effortlessly in the way
described above, that is, towards intelligible reality, and in circular move-
ments across the universe, because they are kept stable and fully under
control. The supreme example of this is Zeus who simultaneously drives
his chariot and maintains his leading position, while ordering and car-
ing for everything. Thus we read at Phdr. 247b: ‘The chariots of the gods
travel easily, being well balanced and easily controlled.’ Although in this
passage εὐήνια, ‘easily controlled’ or ‘obedient to the rein’, is suitable for
the horses, ἰσορρόπως (‘well balanced’) alludes rather to the effect of their
action, in combination with the directions of the charioteer, on the whole
vehicle. Besides, even if the obedience or lack of obedience of the horse is
responsible for the equilibrium of the chariots or the lack of it, the kind
of horse present in a soul is connected with the kind of body that the soul
has inhabited.65 This is clear from the link between the black, insubordin-
ate horse in human souls and the soul’s inhabiting mortal bodies (Phdr.
246c, 248cd). Thus, the reason why the horses of the gods are docile to the
rein is related to the kind of body in which they have always lived. Thus,
it can be seen that ochēma in this passage refers mainly to the body of
the gods which, if it functions as a suitable vehicle for their soul, actually
helps the soul to devote itself exclusively to contemplation of the forms
63
See Phdr. 250c; also 246d. 64 See Chantraine 1968 s.v. ὀχέω; LSJ s.v. ὄχημα.
65
See contra, Griswold 1986: 262 n. 28.
and to the expression of this wisdom ‘in practical terms’ through their
administration of the whole universe.
The gods of the Phaedrus as the heavenly bodies Although the description
of the ‘life of the gods’ in the Phaedrus cannot be translated directly into
the account of the movements of the heavenly bodies in the Timaeus,
or indeed into any other astronomical or astrological explanation of the
time,66 many features of this account invite some kind of identification
between the gods in the Phaedrus and the astral gods of the Timaeus and,
consequently, between the nature of their respective bodies. First, the souls
of the gods, being perfectly winged, not only care for all that is soulless,67
but also ‘travel above the earth’ and ‘administer the entire world’ (Phdr.
246bc). This suggests that somehow their souls account for the orderly
movements of the universe. Secondly, Zeus is presented as ‘the great leader
in heaven’ who ‘goes first, ordering all things and caring for all of them,
and is followed by an army of gods and daemons’ (Phdr. 246e–247a). This
description seems to evoke either something like the soul of the world,
responsible for the movement of the whole universe as well as for those
movements that derive from it, or the movement of the outer circle of the
fixed stars which also ‘carries away’ a large part of the astral movements.68
Thirdly, their circumvolutions within the heavens,69 and the circularity of
these movements,70 evoke the orbital courses of the heavenly bodies in the
Timaeus. Fourthly, Hestia, who does not travel with the rest of the gods
but remains at home, seems to allude to the fixed position that the Earth
was at that time believed to have in relation to the circling starry heaven.71
As for the disconcerting fact that, in the Phaedrus, the gods are described
as ‘travelling outside’ (Phdr. 247c), ‘climbing the top of the world’ (Phdr.
247b) and ‘raising their heads and gazing at the things outside the heaven’
66
See Hackforth (1952: 71–4). As he says, the number twelve, which corresponds to the traditional
gods of the Athenian pantheon such as Zeus, Hestia, Ares, Hera and Apollo (Phdr. 246e–247a,
252c, 253b), does not fit into any astronomical scheme or planetary system. See further Poratti
(2010: 357–64).
67
That is, it has the care of a body as any soul does.
68
See Robin 1954 on Phdr. 247a; also Ostenfeld 1982: 234. By contrast, for Eggers Lan 1992, Zeus is
just a moral paradigm and does not stand for the world-soul. See also Griswold 1986: 84, and 259
nn. 13 and 15.
69
Phdr. 247a4: διέξοδοι often refers to the paths of the heavenly bodies. See Rowe 1988 ad loc. and
Hackforth 1952: 73 n. 3. At the same time this also alludes in mythical terms to the ‘excursion’ of
the gods (see LSJ s.v. διέξοδος III.2.)
70
According to Phdr. 247c, ‘the revolution carries them round’ (περιάγει ἡ περιφορά). See also Phdr.
247d, where it is also said that ‘the revolution brings them back again to the same place in circles’.
71
See Hackforth 1952: 73.
(Phdr. 247c), we can see all these features as an anthropomorphic represen-
tation of the full eidetic knowledge that the gods possess in a unmediated
way according to which each of them ‘fulfils his own task’ (Phdr. 247a) –
that is, executes orderly astronomical movements.72 But since, even if this
cosmological dimension of the myth is left aside, in both dialogues the
gods stand for a perfect psychosomatic condition, similar characteristics
of an ideal kind of body can be inferred from both dialogues. Since it is
immortal and self-sufficient, the divine body is able to function solely as
support or as a ‘vehicle’ for the understanding and execution of intelligent
action.73 A body with these features represents the ideal corporeal condi-
tion which only the philosopher, among human beings, could ever envis-
age achieving.
The possible acquisition of a god-like body by the philosopher In spite of

these positive views of the mortal sōma, the Phaedrus also suggests, like the
Timaeus, that only the possession of a body that works as a divine ‘carriage’
or harma would guarantee the attainment of full, permanent knowledge
as well as action that was infallibly in accordance with it. However, this is
an aspiration that only the philosopher could ever dream of fulfilling.
In the case of the Phaedrus, this seems to be suggested through the
complex dynamics of the mythical imagery. In the state previous to their
first incarnation, human-souls-to-be did not inhabit a terrestrial body, but
were transported in a circular path across the universe, like the gods, in
a chariot or harma, while they attempted at the same time to attain con-
templation of the plain of truth. However, unlike the gods and maybe
other inferior divinities, only a few of them were successful at doing this,
and then only partially, or at least not easily – and in fact most of them
never succeeded at all.74 The reason for their partial or complete failure lay
in the force exercised by the horses (Phdr. 248a), especially in the ‘heavi-
ness’ of the dark aspect of their souls, the black horse, which represents, as
seen above, a tendency in human souls that to various degrees pulls them
72
I say it is a ‘disconcerting fact’ because, if we accept that the circular revolutions of the gods in the
Phaedrus represent somehow the movements of the heavenly bodies, outside the last circle of the
fixed stars there was nothing at all and so the forms would be nowhere. But, as I explain, the unim-
peded gaze of divine souls outside the heavens can be taken as an anthropomorphic representation
of the full eidetic knowledge that only the gods really possess. The forms occupy a different dimen-
sion from perceptible reality – but this should not be taken to mean that they occupy any kind of
physical place.
73
On the way that intelligent understanding combines contemplation and action, see Nightingale
2004: 127–31, and also Armstrong 2004.
74
Note μόγις (‘with difficulty’ or ‘partially’), and μὴ καλῶς (‘not perfectly’) at Phdr. 247b. On this
point cf. De Vries 1969 ad loc., and Steinthal 1993.
in the opposite direction to reason’s desire for wisdom, and inclines them
towards what is mortal and earthly.
Thus, all these souls, which have in their first incarnation an exclusively
human kind of existence, adopt different types of terrestrial lives depend-
ing on how far they have been from having a ‘peep’ into the eidetic realm.
After they die and abandon their mortal forms of embodiment, every
human soul is considered responsible for its way of living and is then
judged according to the degree of fairness in the performance of its first
incarnated life. They acquire, correspondingly, different kinds of physical
support: while some of them are sent to ‘prisons’ under the earth, others
are raised up to some place in the heavens (249b). After this first millen-
nium, human souls are able to choose which kind of mortal embodiment
they are going to have and some of them might even prefer a bestial kind
of existence to a human one. All those souls who opt again for a human
life will be able to exercise, as they should have done in their first exist-
ence, the capacity for perceiving the common universal pattern in mul-
tiple instances (249c). But only some of them become aware of this special
ability and are reminded of the eidetic realm. In other words, they lead
a philosophical life (Phdr. 249cd),75 and are able to ‘fly’ from the beauti-
ful young lad they love and worship to Beauty itself, and develop ‘wings’
(249c): that is, they will develop and strengthen their desire for wisdom.
After three thousand years, if they choose to be philosophers in three suc-
cessive phases of mortal embodied existence, they acquire wings steadily
(that is, they acquire a stable god-like configuration in their souls) and
they leave behind any kind of mortal embodiment. This must imply that,
as they acquire a divine kind of existence, their erōs is not distracted any
more by the forces coming from the dark, appetitive aspect of the soul and
its concerns for the mortal body, but flows without impediment towards
the truth, as their souls are transported in a god-like body which works as
an appropriate vehicle for them.76 The myth also suggests that they might
keep this condition at least until the next cycle of ten thousand years,77
75
See further Scott 1987 and Rowe 1988 on Phdr. 249c1–2.
76
Although nothing is specifically said about what happens to human souls which have been reincar-
nated and, afterwards, manage to go to a place in the heavens where the gods dwell, their acquisi-
tion of a godlike condition should imply that they travel in some kind of harma or ochēma as the
gods do. If we accept that the myth can be read with the Timaeus in mind, this could mean that
these souls travel mounted somehow on the heavenly bodies, with which they now share their per-
fect, intelligent movement (cf. Ti. 41e). See above, pp. 37–8.
77
If human souls could actually transform their black horse into a white horse and so metamorphose
into ‘gods’, they would remain forever in this psychosomatic condition. However, it might be the
but perhaps not even the philosopher could manage to escape forever from
mortal embodiment. All the same, even though expressed in a vague form
and in mythical terms, the idea of the possibility for human beings of
acquiring a body that would not be exposed to disease, aging or even death
and would be perfectly designed to guarantee the most intelligent kind of
behaviour seems to find a first formulation here and in the Timaeus.78
Conclusion
We have seen, then, that the two different conceptions of the body pre-
sent in these Platonic dialogues are not contradictory but complemen-
tary, and that the differences in approach and emphasis can be explained
by locating these ideas more firmly in their dialectical context. We have
also seen that the Phaedrus, especially Socrates’ palinode, combines these
two different views creatively and brings them together in a new form.
According to this dialogue, our mortal body constitutes a type of con-
finement which restricts and disrupts the acquisition of full knowledge.
However, at the same time, the wings of desire (that is, erōs for the truth)
are aroused in us only through a mortal body. It is, then, open to us to
manage the restraints of the body and transform them instead into the
opportunity to start our search for wisdom: in other words, to become
philosophers. Those who always succeed in living a philosophical way of
life in their mortal, embodied existence may also dream of the acquisi-
tion of some kind of invulnerable and godlike kind of body which, like
the heavenly bodies, guarantees immortality together with the exercise of
exclusively orderly, intelligent activity. However, the more secure conclu-
sion of the Phaedrus seems to be that the choice of the best kind of life
constitutes a contribution to the development of a rational universe by
leaving behind a trace of ourselves,79 as well as conferring the hope of
becoming, somehow, after we die, part of the heavenly bodies.80
case that the black horse, which yearns for what is earthly, can only be trained but not converted
into something of a different nature. In that case, in a new cycle of ten thousand years, even the
philosopher might return to an existence of mortal embodiment. Cf. Rowe 2007a: 140 n. 55.
78
This has been considered as an actual possibility for biogenetic engineering since the last decades of
the twentieth century. See Harris 2000 and Selinger 2009.
79
On the link between the development of erōs in human souls and the development of cosmic erōs,
see Fierro 2010.
80
In fact nowadays there are scientific speculations which consider that living organisms on earth
ultimately derive from atoms of dying stars. See Reeves et al. 2008: 33. Why not imagine, along the
lines of Plato’s suggestions, a way back to that original state?
ch apter th ree
Socrates in the Phaedo

Noburu Notomi
Re-examining Socrates in Plato

In most of the Platonic dialogues, Socrates is the main speaker while Plato
remains absent or silent. This basic fact generates interpretative difficulties
for modern scholars. Is it possible to separate Socrates from Plato and,
if so, how? In his recent work Christopher Rowe re-examines this issue
and proposes a new approach, namely reading Plato’s dialogues as essen-
tially ‘Socratic’ even after what are usually called the ‘Socratic dialogues’.1 I
explore this issue from a different angle by focusing on the Phaedo.
In a volume of essays which addresses directly the question of the
Socrates–Plato relationship and the periodisation of Plato’s work, Julia
Annas and Christopher Rowe point out a ‘growing disaffection’ with tradi-
tional ways of reading Plato’s dialogues, including the widespread assump-
tion that we are able ‘to isolate a “Socratic” phase of Plato’s thought’.2
In his contribution to this volume, Rowe comments critically on Terry
Penner’s attempt, based on Aristotle’s testimony, to separate the ‘real, his-
torical Socrates’ from Plato.3 In Plato and the Art of Philosophical Writing
(Rowe 2007a), Rowe challenges more explicitly the use of Aristotle as an
authority for separating Plato’s theory of forms from Socrates’ thought.4
I agree with him that we rely too much on Aristotle when we assume
that Plato departs from his master on specific doctrines, for instance in
maintaining that the objects of philosophical inquiry, that is, forms, are
separated from sensible things. I think we should read Plato’s dialogues
without taking Aristotle’s views on such points for granted. We should
read the dialogues without these preconceptions in order to determine
1
See in particular Rowe 2006b and 2007a with Notomi et al. 2009: Section 4 (which reviews his
arguments).
2
Annas and Rowe 2002: ix. 3 See Rowe 2002: 216–20, on moral psychology.
4
See Rowe 2007a: 43, 48.
51
52 Noburu Notomi
how ideas such as the theory of forms arise within the dialogues and how
they are actually presented there.
The Phaedo is a very interesting test case for this question. On the one
hand, this dialogue is usually regarded, along with the Symposium, as
marking the emergence of distinctively ‘Platonic’ doctrines, a process that
culminates in the Republic, and as moving away from the Socratic, and
mostly aporetic, mode of inquiry represented in the so-called ‘Socratic’
or ‘early’ dialogues. It is normally assumed that, when Socrates tries to
prove the immortality of the soul based on the theory of forms, Plato uses
his master’s voice to present his own theory. In this sense, the Phaedo is
regarded as among the first of the ‘Platonic’ dialogues. On the other hand,
when we discuss the historical Socrates, the dialogue is normally cited as a
key source, since it depicts his last day, in particular how he drank the poi-
son, so vividly that we cannot imagine Socrates’ death without thinking of
this scene.5 The fact that we approach the dialogue in these two, divergent,
ways should make us uneasy about characterising one aspect of the dia-
logue, its doctrines on the soul and forms, as straightforwardly ‘Platonic’.
Rowe (2007a) argues that the Phaedo can, in many ways, be read as
a continuation of the representation of Socrates’ philosophical project in
the Apology.6 He focuses on two main themes, based on the usual assump-
tion that Socrates in the Phaedo, unlike Socrates in the Apology, (i) firmly
believes the transmigration and afterlife of the soul, and (ii) proposes a
metaphysical theory of transcendent forms. Against these widely held
views, Rowe first argues (i) that the ideas about, and attitudes toward,
the fate of the soul after death are consistent in the two works, and also
suggests (ii) that the theory of forms presented in the Phaedo is not essen-
tially different from the ideas presupposed by Socrates in the Apology. He
concludes that ‘this dialogue does nothing to alter the picture we seem to
get from other parts of the corpus: that access to the forms, and to knowl-
edge, for human beings, is by philosophy – dialectic – alone’.7
Following his line of thought, I propose to re-examine the question of
the continuity of the two works and to bring out how Plato takes over
Socrates’ core message. I aim to show that the simple dichotomy between
‘Socratic’ and ‘Platonic’ elements is inadequate for interpretation of the
Phaedo because it misses Plato’s intention. I think we can find in this
dialogue essentially the same, Socratic, conception of philosophy as we
5
For differing accounts of the historicity of the account of Socrates’ death, see Bloch 2002 and Gill
2002b.
6
Rowe 2007a: ch. 2 (‘The Phaedo: Socrates’ defence continued’).
7
Rowe 2007a: 116. For philosophy as dialectic, see also Rowe 2001: 43–4.
Socrates in the Phaedo 53
find in the Apology, even if this conception is deepened in various ways.8
It is not enough, in my view, to say that, on the one hand, we have a
quasi-historical report of Socrates’ last days and that, on the other, Plato
offers his own doctrines in Socrates’ mouth. Rather, we can read the whole
dialogue as exploring a single, but rich and complex theme, what it means
to be a philosopher, as Plato understands Socrates to be.
My argument is presented in two stages. First, I examine three aspects
of the dramatic framing of the dialogue: the opening conversation, the
dramatic context of the narrative (Phlius), the significance of the inter-
locutors. Then, I go on to examine key themes within the argumentation
of the dialogue: practice for death and care of the soul, and – what I call –
metaphysics as the awakening of the soul. My aim in doing so is, first, to
bring out how the different aspects of the dialogue add up to an explora-
tion of what it means to be a philosopher, in the Socratic sense. Secondly,
I seek to show how what we might call the psychology and metaphysics
of the argument develops organically out of the Socratic conception of
philosophy and the notions of practice for death and care for oneself that
form a crucial part of this conception. In this way I aim to bring out how
the dialogue as a whole constitutes an invitation to the reader to try to live
the philosophical life.
This discussion reflects certain broader interpretative directions which
it may be useful to make explicit here. First, as already indicated, I follow
Rowe in challenging the idea that we can clearly and usefully subdivide
‘Socratic’ and ‘Platonic’ dialogues – or indeed, ‘Socratic’ and ‘Platonic’
elements within any one dialogue. Secondly, I aim to illustrate the close
and integral links between aspects of the dialogues sometimes taken
separately – between ‘form’ and ‘argument’ as it is put in one collection of
essays focused on this theme.9 On the question of the relationship between
ethics and psychology or metaphysics in the Phaedo, I am not aiming to
make the substantive claim that, in Plato, psychology and metaphysics
can be reduced to ethics or that they constitute, in essence, the same phil-
osophical practice. (This claim is one that is made, or at least implied, by
Gadamer in his readings of Plato.)10 Mine is the more modest claim that
8
In this paper, I use the word ‘Socratic’ to characterise the life and thought of Socrates as shown
in the Apology, leaving aside the question whether this matches the historical Socrates. I strongly
doubt if we can extract a purely historical picture from the extant sources.
9
See Gill and McCabe 1996, discussing the late Platonic dialogues; see also, on the relationship
between literary form and philosophical content, Griswold 1988 and Klagge and Smith 1992.
10
See further the Introduction to Gill and Renaud 2010, which brings out the influence of
Heideggerian existentialism on Gadamer’s interpretation of Plato; also Renaud 1999. Christopher
Gill points out to me this significant difference between Gadamer’s approach and mine.
54 Noburu Notomi
we cannot make sense of the psychology and metaphysics of the Phaedo
without understanding how they derive from, and express, the Socratic
ethical ideal of what it means to be a philosopher. It is also perhaps worth
reminding ourselves that these categories (ethics, psychology, metaphys-
ics) are themselves Aristotelian inventions, and that our use of them to
analyse the Platonic dialogues betrays, in another way, the powerful influ-
ence of the pupil on our readings of his master.11 In all these respects, I aim
to show that a Platonic dialogue such as the Phaedo forms a seamless web
which is, crucially, a web of ‘Socratic’ and ‘Platonic’ strands.12
Socrates the philosopher
A narrative about Socrates

At the opening of the Phaedo, Echecrates and Phaedo hold the following
conversation (Phd. 57a):13
‘Were you there with Socrates yourself, Phaedo, on the day he drank the
poison in the prison, or did you hear of it from someone else?’
‘I was there myself, Echecrates.’
‘Then what was it that he said before his death? And how did he meet
his end?’
Already in this first exchange, two hints on how to read the dialogue are
given. First, the initial word in the Greek text, autos, indicates the focus
of the whole dialogue.14 It is essential that Phaedo himself, that is, his soul,
was present when Socrates conversed with friends on his last day. As we
see later in this discussion, both the soul and the forms are presented using
this term, as ‘alone by itself ’ or ‘itself by itself ’ (auto kath’ hauto). Secondly,
Phaedo is asked to tell what Socrates said and what he did. The conjunc-
tion of word and deed is emphasised as the key object of the report in the
first outer frame, and we cannot tell which is wanted more eagerly.15 This
is natural because all Socrates’ ‘deeds’ are also told (not shown visibly);
11
That is, in addition to Aristotle’s much noted comments on the Socrates/Plato distinction, which
play a prominent role, for instance in Vlastos’ influential account of the ‘Socratic’ element in Plato
(1991: chs. 2–3).
12
See also Gill in this volume, stressing the ‘chemical’ fusion of Socratic and Platonic elements in the
dialogues and the importance of reading Platonic dialogues as organic wholes.
13
Translations of the Phaedo are from Gallop 1975, sometimes slightly modified.
14
Burnyeat 1997: 9–11, points out the importance of the first word autos in the Phaedo. I consider
shortly the important role played by this term in the dialogue. When this reflexive pronoun is
added to any noun or pronoun, it means ‘self, oneself; by oneself, alone’: cf. LSJ s.v. αὐτός.
15
See also 57c5–6 (τὰ λεχθέντα καὶ πραχθέντα), e4 (τοῦ τρόπου καὶ τῶν λόγων).
everything is given in words. So the act and speech make a single topic in
the narrative.
The narrator of the dialogue plays a dual role. In the outer frame con-
versation, Phaedo reports Socrates’ last day to an audience, probably a
few months after the event. He shares with Echecrates and the others the
emotional responses of Socrates and his interlocutors to the course of the
argument, such as surprise, depression and conviction.16 On the other
hand, within the narrated conversation, Phaedo himself saw and heard
what Socrates said and did in prison on that day (57ab). He was not just a
listener, but he himself had a short but memorable exchange with Socrates
in the middle of the dialogue (89a–90d). When Socrates’ friends were
totally depressed by problematic counter-arguments, Socrates began to
stroke his head and squeeze the hair on his neck, and said, ‘So tomorrow
perhaps, Phaedo, you’ll cut off those lovely locks’ (89b). Then he encour-
aged the pupil to fight together with him against hatred for arguments
(misologia). This touching episode forms a turning-point as the dialogue
moves towards the final argument. Thus, Phaedo plays a role both inside
and outside the reported dialogue. This dramatic scheme invites readers
into the dialogue, as if we were present both at Phlius and in prison.
Phaedo’s account of the facts of the situation, especially the list of the
family and friends who gathered around Socrates, is usually taken as a
kind of historical record. But this also contains the brief but significant
comment, ‘I think Plato was ill’ (59b). This observation about the absence
of the author seems to underline the fictional – or at least indirect – nature
of Plato’s report of the whole event.17
Phlius
Phlius (Φλειοῦς) is chosen as the place where the death of Socrates is nar-
rated. It is located in the north-eastern Peloponnese above Sicyon and
near Nemea: it was one of the small but independent cities in the classical
period. Phaedo, a loyal pupil of Socrates, seems to have left Athens after
the execution of his master to return to his home town, Elis. He seems to
have taken a longer route to visit his friends in order to report the event.
The members of the audience who listen to Phaedo’s report are thought
to have Pythagorean backgrounds, since their community was located in
16
Cf. 57a–59c, 88c–89a, 102a: each intervention is of dramatic significance.
17
I believe that this statement is a literary device of the author; for a different view, see Hackforth
1955: 29 n. 2. See also Rowe 1993: 14–15, 17.
56 Noburu Notomi
Phlius. In particular, Echecrates was a famous Pythagorean and a pupil of
Eurytus and Philolaus: the latter is described by Cebes as staying in Thebes
(61de). After the main residence at Croton was destroyed around 450 BC,
many Pythagoreans were forced to emigrate to the mainland of Greece,
and established centres at Thebes and Phlius.18 Aristoxenus, a pupil of
Aristotle, reported that he was personally acquainted with the last genera-
tion of the Pythagoreans, including Echecrates (fr. 19 Wehrli = DL 8.46):
The last of the Pythagoreans, whom Aristoxenus in his time saw, were
Xenophilus from the Thracian Chalcidice, Phanton of Phlius, and
Echecrates, Diocles and Polymnastus, also of Phlius, who were pupils of
Philolaus and Eurytus of Tarentum.
It is thus possible that Phaedo’s audience includes some other Pythagoreans,
such as Phanton, Diocles and Polymnastus.19
I suggest that the reason why Pythagoreans settled here is linked to the
legend that Phlius was, in two senses, the birthplace of the philosopher
Pythagoras. First, the city was said to be the homeland of Pythagoras’
family, and second, he was said to have called himself ‘philosopher’ for the
first time in this place.
Pausanias in his Description of Greece reports that Pythagoras’ ancestor
was a citizen of Phlius who left his native land for Samos for political rea-
sons. When the Dorians attacked the city, Hippasus, the great-grandfather
of Pythagoras,20 insisted that the citizens should defend themselves, but
since they rejected his advice and accepted the Dorian offer, Hippasus
and his colleagues fled to Samos (Paus. 2.13.1–2). Whether this is histori-
cally accurate or not, it is important that this legend was transmitted and
believed by the Phliasians (2.13.2). In particular, the story of Hippasus is
noteworthy since his firm attitude against invaders resembles Pythagoras’
stance against tyranny.
Also, Cicero tells us that Pythagoras himself visited Phlius, where he
is said to have called himself ‘philosopher’ (philosophus) for the first time
(Tusculan Disputations 5.8–9). I think this must have happened when he
left Samos, opposing the tyranny of Polycrates, and moved to southern
18
See Guthrie 1962: 179. In Thebes, Lysis later became a teacher of Epaminondas.
19
Iamblichus, On the Pythagorean Life 35.251 (= Aristoxenus fr. 18 Wehrli, part) and 36.267 (list of the
Pythagoreans). See also Guthrie 1962: 169.
20
Pausanias reports that ‘Pythagoras was the son of Mnesarchus, the son of Euphranor, the son
of Hippasus’ (2.13.2). Another genealogy is recorded in DL 8.1: ‘Some indeed say that he was
descended through Euthyphro, Hippasus and Marmacus from Cleonymus, who was exiled from
Phlius.’ The family origin of Pythagoras was already controversial in antiquity; see Burkert 1972:
206 n. 77.
Italy to found his own community, around 538 BC. The story is also
recorded in Diogenes Laertius in a slightly different form, and both reports
are said to be based on the dialogue of Heraclides of Pontus entitled On
the Woman without Breath. Diogenes situates the conversation with the
local lord Leon at Sicyon (a neighbouring city), who was ruler of Phlius as
well (DL 1.12; 8.8).
According to these sources, in the conversation with Leon, Pythagoras
gave himself a new name, philosophos (lover of wisdom): he is not wise
(sophos), for this description is only suitable for gods. On the other hand,
since this anecdote came from a dialogue of Heraclides, who was a mem-
ber of the Academy with strong Pythagorean sympathies, some scholars
suspect that it might be a projection of the Platonic ideal onto Pythagoras.
However, since the first extant example of this word is found in Heraclitus
of Ephesus, who criticised Pythagoras as a pseudo-sage,21 it is very likely
that it was first used by Pythagoras himself and was already well known by
the beginning of the fifth century BC. It is now generally agreed that the
terms ‘philosophy’ and ‘philosopher’ were especially associated with the
Pythagorean tradition.22
Phlius was associated with the origin of the word ‘philosopher’ probably
because of its special links with Pythagoras. This must be the reason why
Plato sets the narration of this important dialogue in such a small city,
away from Athens. However, since several important Pythagoreans actu-
ally lived there in the early fourth century BC, its fame was certainly not
invented by Plato or Heraclides. Thus, by placing his narration at Phlius,
Plato establishes in this way the idea of ‘the philosopher’ as the leitmotif of
the dialogue.
The ideal ‘philosopher’

The people around Socrates in prison were different from the Pythagoreans
in Phlius. In the Phaedo, Simmias and Cebes said that they had studied
with Philolaus in Thebes (61de). Based on this passage, modern schol-
ars tend to include them among the Pythagoreans, but they were not
Pythagoreans in the strict sense.23 This is obvious from the later tradition,
in particular Iamblichus’ On the Pythagorean Life, which does not include
21
22 B35 DK. I suggest that the word philosophos in this fragment be read in a negative sense (cf.
B22), when taken in association with the critical references to Pythagoras in B40, 129.
22
See Guthrie 1962: 204–5.
23
For the traditional view, see Burnet 1930: 309; Burkert 1972: 92, 198; and Huffman 1993: 326–7,
408–10. Against this, see Rowe 1993: 7, 115–16. On Simmias and Cebes, see also Sedley 1995.
58 Noburu Notomi
these two names in the list of Pythagoreans (which totals 235 members).
On the other hand, Simmias and Cebes were often mentioned as close
associates of Socrates,24 and they may have written some Socratic dia-
logues.25 Therefore, these two were Socratics, and not Pythagoreans.
The immortality (or transmigration) of the soul was a main tenet of
Pythagoreanism, and the ‘harmony’ theory of the soul, which Simmias
proposes in 85e–86d, may well come from Philolaus.26 In the dialogue,
Socrates offers a philosophical proof of the former idea and refutes the lat-
ter one (91c–95a). By contrast, those who listen to the arguments in Phlius
are genuine Pythagoreans. Hence, as David Sedley comments, ‘what we
meet in the Phaedo is the paradoxical spectacle of Socrates having to per-
suade the Pythagoreans of the truth of their own doctrine’.27
The topic of the dialogue was also of crucial importance for Phaedo, the
narrator. It is reported that he was enslaved and made to work as a prosti-
tute after the war with Sparta, but that when he was freed by Socrates, he
devoted himself to Socratic philosophy, which taught him the liberation
of the soul from the body (DL 2.105). He later wrote Socratic dialogues,
notably Zopyrus and Simon.28 One may imagine that, by transmitting to
the people of Phlius the report of Socrates on his last day, Phaedo is set-
ting out on his new life as a philosopher, back in his native city, Elis,
where he is reported to have founded his own school of philosophy. Thus
Plato’s decision to set the frame conversation in Phlius and to use Phaedo
as its narrator symbolises the idea that the dialogue as a whole presents the
ideal of the philosophical life as a paradigm for imitation.
The author Plato (whom Phaedo says was ill)29 uses the dialogue to link
Socrates with the originally Pythagorean notion of the philosopher. He
also depicts his master as the ideal philosopher, who died fearlessly after
trying to demonstrate (the originally Pythagorean idea of ) the immortal-
ity of the soul. Talking about the philosopher is not only a pleasure, but
also a great encouragement for people, in the first instance, the narrator
and his listeners, and then readers of the dialogue, to pursue philosophy
(cf. 58d). Plato’s Socrates in the Phaedo thus shows us what a philoso-
pher is and how he should live and die. It is significant, as I have pointed
out elsewhere, that depicting Socrates as a ‘philosopher’, as opposed to a
24
See Cri. 45b; Xenophon, Memorabilia 1.2.48, 3.11.17.
25
DL 2.124–5 attributes twenty-three dialogues to Simmias and three to Cebes.
26
Cf. Huffman 1993: 323–32, and Sedley 1995: 11–12, 22–6.
27
Sedley 1995: 11, 13.
28
See Rossetti 1973 and 1980; Sedley 1995: 8–9; Kahn 1996: 9–12.
29
Cf. n. 17 above; the comment probably underlines the fictionality of the dialogue.
‘sophist’ (and non-philosopher), was an original feature of Plato’s presen-
tation, whereas the other Socratics cared little about the distinction.30
Therefore, any attempt to isolate elements relating to the historical
Socrates from the dialogue as a whole destroys the cohesive unity of the
dialogue. It seems to me highly implausible that Plato intended his read-
ers to take one part of the dialogue as ‘Socratic’ and another as ‘Platonic’.
Moreover, the moving scene of his master’s death would not have the same
significance without the discussion presented as being held on his last day.
Indeed, Cebes suggests that, without proof of the indestructibility and
immortality of our souls, any courage shown by a philosopher in the face
of death will amount to ‘a senseless and foolish confidence’ (88b).31 The
main arguments for the immortal soul and the transcendent forms are
thus inseparable from the portrayal of Socrates as the ideal philosopher.
Accordingly, it becomes crucial to ask whether we can find a single notion
of ‘philosopher’ which links the representation of Socrates in the Phaedo
with that of the Apology.
Philosophy of soul and form
Practice for death

On the assumption that Plato’s main purpose in the Phaedo is to present
Socrates as the ideal philosopher, let us examine how this idea is formu-
lated there. A philosopher is, firstly and crucially, said to be someone who
is willing to die (62c, 63e–64b; cf. 61c, d), indeed, someone who engages
in ‘practice for death’ (meletē thanatou, 67de, 80e–81a). Socrates defends
this claim against common opinions about the philosopher. In charac-
terising his argument as a ‘defence’ (apologia, 63b, d, 69de), he signals a
link with the presentation of the Socratic philosophical life in the Apology.
This part of the dialogue, sometimes called ‘Socrates’ defence’ (63b–69e),
rarely receives serious philosophical examination, but I believe it provides
vital hints on how to understand the main theme. I focus on three topics
in this part: death (especially 64a–c, 67c–e), care (64c–65a) and wisdom
(65a–67b).
First, let us consider the notion of philosophy as ‘practice for death’.
Practice is a continuous preparation for a certain state; and the state of
30
See Notomi 2010. See also Narcy in this volume, who shows that the ‘sophist-philosopher’ distinc-
tion deployed by Plato was not a merely descriptive but a polemical one.
31
Socrates refers back to Cebes’ statement in 95bc.
60 Noburu Notomi
death (that is, being dead) lies just ahead, as an inevitable fact for Socrates.
How he is to spend his last few hours and conclude his life reveals the way
in which he has lived and is living. Socrates chooses to hold a dialogue
(dialegesthai) with his close friends, that is, to be occupied with philoso-
phy (philosophia, 59a), which is his customary business in life. The topic
chosen for the last dialogue is the philosopher’s willingness to die. This
shows in word and deed how Socrates has prepared himself for death as a
true philosopher.
Socrates says that the truth passes unnoticed by the many: namely,
that a person who engages properly in philosophy pursues nothing other
than dying and being dead (64a). On hearing this, Simmias breaks into
laughter, and claims that the public will agree with this statement, since
it is quite obvious that a philosopher lives a half-life and deserves to die
(64ab). Simmias’ reaction seems quite natural to us. Ordinary people,
today as well as in antiquity, picture the typical philosopher as a hermit or
a pale ghost, in contrast to someone who throws himself into life – thus in
Aristophanes’ Clouds Strepsiades calls such a person ‘half-dead’ (hēmithnēs,
504). But in this common usage ‘death’ is a mere metaphor.
Socrates strongly opposes this popular view. He insists that it is false to
say that this idea is obvious to them, because they do not understand the
true meaning of what he is saying. Ordinary people do not really under-
stand in what sense true philosophers wish to die and deserve to die or,
above all, what death is. We should remember that, in the Apology (29ab),
Socrates warns us that our fear of death is an example of the worst kind
of ignorance (amathia), namely that of thinking we know what we do
not know. Contrary to our supposition, death may be the greatest good
for human beings. I think that Socrates presents death as unknown to us
in response to Simmias, not because we have not experienced it yet but
because the truth about life is not being recognised. The true nature of life
(zōē) will eventually be revealed as an essential quality of the immortal soul
(Phd. 105c–e, 106d). Simmias’ laughter comes from his adoption of the
popular, and metaphorical, understanding of what it means to be ‘dead’
or ‘half-dead’ in life. In rejecting this commonplace kind of ignorance,
Socrates sets out to elucidate the true meaning of ‘practice for death’: he is
willing to die, and dies shortly afterwards with perfect composure.
What, then, is death (thanatos)?
Is it anything else than the separation of the soul from the body (τὴν τῆς
ψυχῆς ἀπὸ τοῦ σώματος ἀπαλλαγήν)? Do we believe that being dead
(τὸ τεθνάναι) is this, namely, that the body comes to be separated from
the soul, alone by itself (χωρὶς μὲν ἀπὸ τῆς ψυχῆς ἀπαλλαγὲν αὐτὸ καθ’
αὑτὸ τὸ σῶμα γεγονέναι), and the soul is, alone by itself, separated from
the body (χωρὶς δὲ τὴν ψυχὴν τοῦ σώματος ἀπαλλαγεῖσαν αὐτὴν καθ’
αὑτὴν εἶναι)? (Phd. 64c, my translation)
Socrates defines death (or being dead) as the soul’s separation from the
body. This definition may at first sight look commonplace and unsurpris-
ing, but we should bear in mind that Socrates presents it as something
unknown to ordinary people.32 It is unknown because we are ignorant of
what ‘separation’ really means.33
When two things are separated, each of them becomes alone by itself
(auto kath’ hauto). The reflexive pronoun auto plus the adverbial phrase
kath’ hauto signifies the isolation of the thing itself from the other factors
or contexts, and therefore it signifies independence, purity and absolute-
ness.34 In the kind of separation that constitutes death, one of the two enti-
ties involved is the body, which remains whole for a while after death but
sooner or later disintegrates; therefore, its state of separation is described in
terms of generation and change (gegonenai, 64c6). The other entity is the
soul: the soul’s being alone by itself implies immortality, as is argued later.
Care for the soul

In the next stage, Socrates explains the idea of practice for death in terms
of two kinds of care. He says that those who love and aspire for wisdom
do not care seriously for their body, nor do they attach great importance
to the pleasures of food, drink and sex. No philosopher regards any serv-
ices to the body (tas peri to sōma therapeias) as being of any value (64d).
Simmias agrees with Socrates on this point (Phd. 64e–65a):
‘Do you think in general, then, that such a man’s concern is not for the
body, but so far as he can stand aside from it, is directed towards the soul
(pros tēn psychēn tetraphthai)?’
‘I do.’
‘Then is it clear that, first, in such matters as these the philosopher dif-
fers from other men in releasing his soul, as far as possible, from its com-
munion with the body (apo tēs tou sōmatos koinōnias)?’
‘Apparently.’
32
Hence it is wrong to assume, as Hackforth does (1955: 44 n.1), that this definition ‘doubtless repre-
sents the normal contemporary view’.
33
The abstract noun chōrismos is used only at 67d4 and 9 in Plato’s corpus; see also the verb chōrizein
in 67c, and the preposition chōris in 67a and 76c.
34
See Broackes 2009: 47–9. The phrase was already used before Plato in non-technical senses in the
fifth century BC, and Plato himself uses this expression or similar ones differently in different con-
texts; see further El Murr forthcoming.
62 Noburu Notomi
Their conversation may well remind us of the ‘care for the soul’ (epimeleia
tēs psuchēs), which Socrates presents as his philosophical mission in the
Apology.35 If we bear in mind the significance of this famous earlier pas-
sage, we may see the sharp contrast between body and soul here not only
as an analysis of our nature or being but also, and crucially, an ethical con-
trast: ‘soul’ and ‘body’ represent two contrasting directions or dimensions
of our life as agents.
To see the link between the two ideas, let us note this famous pas-
sage in Socrates’ defence. He addresses an imaginary fellow-citizen (Ap.
29d7–e2):
Good Sir, you are an Athenian, a citizen of the greatest city with the greatest
reputation for both wisdom and power; are you not ashamed of your care
(epimeloumenos) for money, reputation and honour, so that you can acquire
as much of them as possible, while you do not care for (ouk epimelēi) or
give thought to wisdom (phronēseōs) or truth, or for the soul (tēs psuchēs) so
that it can be as good as possible?
Most people concern themselves only with how to increase their wealth,
fame or belongings. This type of concern Socrates goes on to call ‘care for
the body’ (30ab):
For I go around doing nothing but persuading both young and old among
you not to care (epimeleisthai) for your body or money in preference to, or
as strongly as for your soul (tēs psuchēs) so that it will be as good as possible.
As I say to you: wealth does not bring about virtue, but virtue makes wealth
and everything else good for men, both individually and collectively.36
Socrates here broaches the subject of two types of care, or two ways of life:
the first is care for the body, and the second is care for the soul. These are
presented and contrasted as two fundamental dimensions of our being.
There are five important points in the delineation of this contrast:
(i) The items mentioned in the first type include money (wealth), repu-
tation, honour, and bodily health. The term ‘body’ (sōma, 30a) rep-
resents this group, which is presented as indicating a single focus
in one’s life. By contrast, the second type of life aims at wisdom
(phronēsis), truth (alētheia), and virtue, which are centred on the
soul (psuchē).
35
On the care for the soul in the Apology, see Burnet 1924: 203–4.
36
For 30b2–4, I follow the reading of Burnet 1924: 204, and Burnyeat 2003; see also Rowe 2007a:
66–80.
(ii) The first type of life gives priority to acquiring as much as possible
(pleon echein, pleonektein): quantity is of the utmost importance, in
preference to value. This acquisitiveness is presented as originating in
concern for the body. By contrast, the primary concern in the second
type of life is goodness, and the aim is to make the soul as good as
possible. The goodness of the soul is called ‘virtue’ (aretē). We should
note that ethical terms such as ‘good’ and ‘fine’ are applied only in
the case of the second type.
(iii) Even if one were to get as many material goods as one would like,
one’s soul would not become good. In contrast, all good things,
including the welfare of the state, become good only by virtue of the
goodness of the soul (30b).
(iv) Socrates also uses the expression ‘care for oneself ’: he advises caring
for oneself so as to become as good and wise as possible (36c, 39d).
This expression implies that the soul is our own self. We normally
suppose that our personal identity consists of our physical features
(or how we appear to others), reputation (or how we are assessed by
others), and property (or what we possess). However, these things
are presented as outer ornaments, and thus are called ‘one’s belong-
ings’ (ta heautou), in contrast to ‘one’s true self ’ (heautou, 36c).37 My
belongings cannot constitute the true ‘me’, but provide only an illu-
sory version of my identity and real existence. We should not take
care of those outer things in preference to our inner self, namely the
soul.
(v) Socrates is a person who really cares for virtue (Crito 51a), but what is
more, he urges other people to care for the soul. This activity is both
divine, because it comes from god (Ap. 30a), and political, because
it serves the state by making the citizens good. Therefore, to advise
and persuade each individual to take care of himself (heautou epime-
leisthai) is presented as giving the best service to the state (36cd).
This is the ‘divine mission’ that Apollo inspired through the Delphic
Oracle.
Thus, in these passages, the idea of the separation of the soul from the
body is used as a way of characterising two competing ways of life, or two
possible directions in which we might take our own lives. The distinction
37
In Alcibiades 1 127d–133c, Socrates distinguishes three kinds of objects of care: oneself (hautou),
things of oneself (ta hautou), and things of the body (ta tou sōmatos) (esp. 128a–d). The care for
the first is necessary for the care for the rest. Burnet 1911: 68, 81, attributes this trichotomy to
Pythagoreanism.
64 Noburu Notomi
between these two ways of life also underlies the Phaedo: if someone is dis-
turbed when facing death, he is not a philosopher (lover of wisdom), but a
lover of the body (philosōmatos), that is, a lover of money (philokhrēmatos)
or of fame (philotimos) (68bc, 82c). This dichotomy, or trichotomy, takes
over the core message of the Apology. A philosopher avoids bodily desire
and takes care of his own soul, so as to live a different life from ordi-
nary people who are completely focused on bodily things but do not even
understand how they are actually living (81b–e, 82b–d).
Since we are preoccupied with various bodily concerns and thus already
‘enslaved to the service (therapeia) of the body’ (66d), we are not in a
position to choose freely between the two types of care. A radical shock
is needed to make us want to free ourselves from this stubborn slavery,
in order to realise our true self. For this purpose, Socrates provokes us
by refutation (elenchos) and reveals our ignorance of our own state of
understanding, so as to break our stubborn ignorance or inability to learn
(amathia), as he explains in the Apology. He pursues this goal, not for his
pleasure, but as part of his divine mission to break the strong bonds that
keep ordinary people chained to the body (Phd. 66b–67b, 68cd). Socrates’
philosophical activities aim to turn people’s mental gaze from the body to
the soul. Such a conversion requires a strong mental shock from the out-
side but, in their ignorance, those who are cross-examined and refuted by
Socrates usually feel insulted and become furious with him.
The ultimate aim of the argument about the soul’s immortality is to
promote this ethical conversion. This is made explicit when Socrates, hav-
ing just delivered his proof of the soul’s immortality, stresses the impor-
tance of caring for the soul (Phd. 107c8–d2):
But this much it is fair to keep in mind, friends: if a soul is immortal,
then it needs care (epimeleias), not only for the sake of this time in which
what we call ‘life’ lasts, but for the whole of time … But since, in fact, it
[the soul] is evidently immortal, there would be no other refuge from bad
things or salvation for it, except to become as good and wise as possible (hōs
beltistēn te kai phronimōtatēn genesthai).
Later, in reply to Crito’s request for any final words before his death,
Socrates again says (Phd. 115b5–c1):
What I am always telling you, Crito, and nothing very new: if you take
care of yourselves (humōn autōn epimeloumenoi), your actions will be of
service to me and mine, and to yourselves, too, whatever they may be, even
if you make no promises now. But if you take no care of yourselves (humōn
autōn amelēte), and are unwilling to pursue your lives along the tracks, as it
were, marked by our present and earlier discussions, then even if you make
many firm promises at this time, you will do no good at all.
In these final words Socrates encourages his friends and family to care
for themselves, that is their own souls. This implies that his proofs of the
immortality of the soul in the Phaedo have an underlying ethical motiva-
tion, one associated in the Apology with Socrates’ ‘divine mission’.
The attitude he displays in the Phaedo matches that shown in Socrates’
final request at the end of his speech in court in the Apology. He asks his
enemies to criticise his sons if they seem to care (epimeleisthai) for money
or any such thing, rather than virtue (Ap. 41e–42a).38 The distinction he
draws between the two types of care, for the soul and the body, and the
advice he gives to turn from the one to the other, are fundamental com-
ponents of Socrates’ lifelong message. This explains why he says that there
is ‘nothing rather new’ (ouden kainoteron) in his final words in the Phaedo
(Phd. 115b).
Since the soul is identified with one’s true self, one becomes truly oneself
only when the soul comes to be alone by itself. Separation means purifica-
tion (katharsis, 67a–c, 82d). This is the state in which the soul is separated
from all the bodily senses and desires, which always divert our attention
from the soul, our true self, and from our inquiry into the nature of true
being (82d–84b). It is only through purification that the soul returns to its
original, divine place (69cd, 82bc, 84b, 114c).
Metaphysics as the awakening of the soul

When Socrates explains the nature of a true philosopher (ho alēthēs philos-
ophos) in terms of the care for the soul (64b–65a), he focuses on ‘wisdom’
(phronēsis, 65a). Search for the truth (alētheia) is possible through the soul
alone by itself, but not through bodily senses. We should remember that
‘wisdom’ and ‘truth’ are presented as two fundamental objectives for which
we should care, in Apology 29de (see (i) in the previous section). Therefore,
Socrates’ explanation in Phaedo 64b–65a of how we seek wisdom to attain
the truth forms a way of presenting more fully his conception of what it
means to be a philosopher.
It is only when the soul is separated, alone by itself, that it car-
ries out its proper function, namely, contemplation, free from any
38
Here I interpret the ‘sons’ to mean not just his three sons, but also his philosophical offspring,
including Plato.
66 Noburu Notomi
bodily disturbances. This pure state of contemplation is called ‘wisdom’
(phronēsis), which concerns itself with justice itself (dikaion auto), beauty,
goodness and the other things that are, alone by themselves (65de). ‘Forms’
are thus introduced for the first time in the dialogue as the proper objects
of the soul’s wise understanding. A form is designated by the emphatic
phrase ‘itself ’ or ‘alone by itself ’ (hauto, auto kath’ hauto, 65d–66a, 100b),
which indicates separation. In the theory of forms, the idea of ‘separation’
has two senses: the forms are separated from sensible objects, and also
separated from one another, so that each of them is distinct. Plato’s use
of the same phrase ‘alone by itself ’ (auto kath’ hauto) for the soul and the
form indicates his intention to treat the two in a parallel way and thereby
to correlate the soul with the object of its contemplation. It is no accident
that he uses the same, highly significant, phrase in these two, apparently
different, contexts.
In addition, to admit the existence of forms, as the objects of wise
understanding and philosophical inquiry, means admitting that the soul,
or the true self, really exists. Although ordinary life is contaminated by
various bodily desires and inclinations, the contemplation of forms will
purify and restore our true selves, as is beautifully depicted in the myth
of the Phaedrus. Through our philosophical efforts to view the forms, we
human beings strive to live as good (or ‘divine’) a life as we can.39 The
claims that there is such a thing as a form ‘itself by itself ’ and that there
is such a thing as the soul ‘alone by itself ’, as these are presented in the
Phaedo, are interdependent and mutually supporting. And both claims
emerge as an expression of Plato’s desire to follow through and to explain
further Socrates’ conception of the ideal philosophical life. In that sense,
the psychology and the metaphysics of the Phaedo have an ethical foun-
dation, since they arise out of Plato’s wish to clarify and substantiate
Socrates’ ethical mission. This does not mean, as noted earlier, that, in
my view, psychological and metaphysical claims are reducible to ethical
ones.40 What I wish to stress is the inseparable and mutually supporting
nature of these claims, as formulated in the Phaedo.
Also, I wish to underline the way in which these characteristically
‘Platonic’ doctrines have their origin in the desire to support and convey
the ‘Socratic’ ethical mission. On this point, it is helpful to contrast my
39
Cf. Fierro above, p. 43. We should remember that in the Platonist tradition the aim of human life is
to become like a god as far as is allowed. The digression on philosophers in Tht. 176ab presents this
view most clearly, but see also R. 500b–d, 613ab.
40
See p. 53 above.
account with Aristotle’s as regards the separation of the forms from sen-
sible things. While he explains this separation as the crucial point where
Plato’s philosophical thought departs from that of Socrates,41 it is inter-
preted here as Plato’s way of developing Socrates’ ethical message about
how we should live well as philosophers. In the final form of argument,
Platonic and Socratic elements are no longer separable in a clear and
meaningful way.
The two points just made, the interdependent nature of the claims
about the forms (‘itself by itself ’) and the soul (‘alone by itself ’), and the
integral linkage between ethics and psychology or metaphysics, come out
clearly in two passages during the arguments for the immortality of the
soul. One of the arguments is centred on the notion of learning as recol-
lection (72e–77b). Since we see equal things also appearing to be unequal,
we are aware that they are always deficient in being, compared with the
equal itself (or form of equality). Recollection, mediated through sense
perceptions, is needed to recover our soul’s prenatal knowledge of the
forms. Forms are the ultimate objects of inquiry in the philosophical life.
This argument leads to the crucial conclusion that the soul’s being alone
by itself (after death) and the existence of the forms, which are ‘themselves
by themselves’, stand or fall together (76d–77a). This is put clearly in this
exchange, in which Socrates speaks first (76e5–9):
‘Is it equally necessary that those objects exist, and that our souls existed
before birth, and if the former don’t exist, then neither did the latter?’
‘It’s abundantly clear to me, Socrates,’ said Simmias, ‘that there’s the
same necessity in either case.’
The interdependent status of the claims about the soul (alone by itself )
and the forms (itself by itself ) is here made explicit.
Next, Socrates distinguishes two levels of existence (78b–80b): there
are the true beings that always remain the same, and the changing things
that never remain the same.42 While the latter are perceptible through the
senses, the former are invisible, and can be grasped only through pure rea-
soning. The soul is cognate with unchanging reality, whereas the body is
cognate with the other realm. On the basis of this affinity, Socrates main-
tains that ‘whenever the soul uses the body as a means to study anything’,
41
See Arist. Metaph. M4, 1078b30–2; also A6, 987b6–9. Aristotle’s concern in these passages is with
ontology (explication of the four causes in A and of the status of mathematical entities in M), so
his comments on the forms (by contrast with sensibles) need to be taken in the context of this
concern.
42
I see in this argument a more important role than Rowe 1991.
68 Noburu Notomi
then ‘it is dragged by the body towards objects that are never constant;
and it wanders, and is confused and dizzy, as if drunk, in virtue of contact
with things of a similar kind’ (79c). He goes on to say (79d1–7):
Whereas whenever it [the soul] studies alone by itself, it departs yonder
towards that which is pure and always existent and immortal and unvary-
ing and, in virtue of its kinship with it, enters always into its company,
whenever it has come to be alone by itself, and whenever it may do so.
Then it has ceased from its wandering and, when it is about those objects,
it is always constant and unvarying, because of its contact with things of
a similar kind. And this condition of it is called ‘wisdom’ (phronēsis), is it
not?
Here, a linkage is made between the psychological and ontological claims
and the ethical theme discussed earlier, that is, the two competing ways
of life (that of the body and the soul). Here, Plato’s Socrates marks a con-
nection with his earlier ‘defence’ and with the defence of Socrates’ divine
mission in the Apology.
The metaphysical argument, while supporting the Socratic ethical ideal,
also carries implications about the nature of the ethical ideal that is sup-
ported in this way. The separation of the soul from the body and the gain-
ing of knowledge of forms require a continuous practice which will not be
complete until death. The philosopher’s life is thus lived in the continuing
search for this state of separation. The project of leading this kind of life
is one that changes us fundamentally, as is indicated by the simile of the
Cave in the Republic, which depicts the movement from the cave to the
real world outside (R. 514a–517a). Through this process, we are ourselves
changed so as to live a true life. This change of life consists in the conver-
sion from the sensible to the intelligible, and from the body to the soul.
The metaphysical claims made in dialogues such as the Phaedo and the
Republic thus also carry ethical implications: they make us aware that we
should free ourselves from all evils by caring for the soul. Hence, Plato’s
most distinctive metaphysical claims can be understood as deriving from
the motive of supporting and substantiating Socrates’ ethical ideal – and I
believe that this explains why Plato continues to use Socrates as his main
speaker in these dialogues.
In the Republic, those philosophers who view the forms are compared
to people who have been awakened, in contrast to ‘sight-lovers’, that is,
those who are sleeping but take themselves to be awake (476cd). Plato’s
metaphysics is designed to awaken us, as Socrates did the Athenian peo-
ple, and to urge our soul to be alone by itself to see truth and reality. By
its presentation of the life and death of Socrates, Plato’s Phaedo challenges
and invites us, the readers, to undertake the philosophical journey towards
death. Just as Plato sometimes presents logos as the medium for commu-
nication between souls (Phd. 115c, Alc. 1 130e), so dialogue becomes an
essential medium for philosophy to enable us to realise our selves. It was
Socrates himself, that is, his soul, alone by itself, which held a dialogue with
Phaedo himself on the last day, and about which he talked to his audiences,
including us readers. Socrates is thus made into the ideal philosopher by
the narrator Phaedo, and by the author Plato. By recalling Socrates we are
encouraged to become philosophers like him. Significantly, Phaedo ends
his narrative with this summation of his memory (118a15–17):43
And that, Echecrates, was the end of our companion, a man who, among
those of his time we knew, was – so we should say – the best (aristou), also
the wisest (phronimōtatou) and the most just (dikaiotatou).
43
Part of the argument of the second section, ‘Socrates the philosopher’, was originally published
in Japanese in Notomi 2005: ch. 1. The main part of the third section, ‘Philosophy of soul and
form’, was first read in Italian at the International Plato Society conference Psyche in Platone held
in Como in 2006, and published in Notomi (2007a). Its main claims are located in a wider context
in Notomi 2006. I thank Christopher Gill and Dimitri El Murr for valuable comments on both
arguments and style.
ch apter f ou r
Socratic intellectualism in the Republic’s

central digression
David Sedley
This chapter was originally entitled ‘Socratic intellectualism in Republic

5–7’. I shall start by explaining why I changed the title.
There were two competing divisions of the Republic in antiquity, and
our evidence, the essentials of which were set out by Henri Alline in his
1915 book Histoire du texte de Platon, shows that in the first two centu-
ries AD these two systems existed alongside each other.1 Its redivision into
the canonical number of ten books goes back to Thrasyllus in the early
first century AD, but perhaps no earlier. The alternative division was into
six books. Probably already known to Cicero in the late 50s BC, when he
wrote his own De Republica in six books,2 it was apparently still being
used by Aulus Gellius in the late second century AD,3 and certainly by the
grammarian known as the Antatticist4 around the same time. But mean-
while another near-contemporary, Galen, can be seen using the ten-book
division, which thereafter becomes the standard one.
Even if there is good reason to believe that the six-book division was
the earlier one, does it go all the way back to Plato? The Republic is too
My thanks to audiences at Durham and Paris in April 2010 for helpful comments; to Myles Burnyeat
both for instructive discussion and for showing me his unpublished paper ‘In what sense are the cen-
tral books a digression?’; to Georgia Mouroutsou, Frisbee Sheffield, Melissa Lane and an anonymous
referee for searching written comments; and to Christopher Gill for improvements in the final draft,
as well as for his commentary in this volume. Harold Tarrant was kind enough to show me unpub-
lished work of his own, defending a very different view of the six-book Republic. Above all, my thanks
go to Christopher Rowe, with whom I have had several illuminating discussions of these issues, and
whose work has done much to inspire and shape my own.
1
Alline 1915: 14–19.
2
The force of this evidence is, if lessened, by no means negated by the fact that Cicero had at one
stage envisaged an alternative, nine-book version (Letters to Quintus 3.5.1–2). Replicating the origi-
nal number of books was only one of the available ways of imitating a Platonic model, and he
clearly made no attempt to do so with his three-book On Laws in relation to Plato’s Laws.
3
Gellius 14.3.3 represents Xenophon’s Cyropaedia as his response to the political proposals in ‘more or
less the first two books’ of Plato’s Republic, which could hardly refer to anything less than the first
three books in the familiar division.
4
Bekker 1814: 75–116.
70
The Republic’s central digression 71
long to have been squeezed into a single normal roll, so we must assume
that Plato either published it with his own chosen division into books, or
left it to his copyists to decide where to make the breaks. The more plau-
sible hypothesis of the two is that the original division was his own, but
even if the alternative option were preferred, namely that the decision was
taken by his copyists, we could still infer that Plato at least acquiesced in
it. Should we, then, suppose that original division of the Republic to have
been the six-book one? Or are we to postulate yet another division, pre-
dating both the six- and the ten-book one?
Earlier scholars have been sceptical about the six-book division’s being
traceable back to Plato’s time, but partly, it seems, owing to the mistaken
belief that standardised book divisions were a Hellenistic innovation,
imposed for the first time by Alexandrian scholars.5 Recently, reasons
have emerged for thinking that the majority of the book numbers in the
Aristotelian corpus, at least, do date from Aristotle’s own time, because
they retain the pre-Hellenistic system of numbering.6 This is strong evi-
dence that there were already official book numbers by the second half
of the fourth century BC, and there is no longer any obvious reason to
doubt that Plato had already been following the same practice. Even on
the reasonable hypothesis that there was an Alexandrian renumbering of
the Republic’s books, it very probably retained an older six-book division
and simply retitled the books according to the up-to-date number system.
The six-book schema itself could well go back to Plato’s own day.
The Antatticist’s citations of the Republic name every book from 1
to 6, with Book 6 unmistakably containing the material that our mod-
ern editions put in Book 10.7 They thus suffice to give an approximate
idea of the dialogue’s six-book structure. In a study published in 1897,
whose findings Alline took over more or less intact, Joseph Hirmer had
in fact used these citations to propose a very precise division of the con-
tents into six books.8 Hirmer was perhaps overconfident in his precise
5
See Alline 1915: 16–18; Hirmer 1897: 591.
6
The old system used simply the twenty-four letters of the alphabet, whereas the new one added
three further letters, including digamma for 6. See Burnyeat 2004: 178–9 n. 3, followed by Primavesi
2007. The latter argues plausibly that the full texts of these works survived independently until
the first century BC, possibly through the famous book-cache in Scepsis, thus bypassing the nor-
mal Alexandrian renumbering. The archaic numbering system has also survived in the stichom-
etry of Demosthenes’ letters (Goldstein 1968: 10, 269–71), and in the book numbering of Homer,
Herodotus and Theophrastus.
7
One has no choice but to discount a number (nine, by Hirmer’s count) of deviant citations which
fail to correspond to anything that could possibly be in the books cited. Nevertheless, the remaining
twenty-two or so citations display an unmistakable and self-consistent pattern.
8
See Hirmer 1897: 589–92.
72 David Sedley
divisions, but the evidence cited by him is enough to justify the follow-
ing schema:9
Book 1 probably extended into our Book 2, and brought together the
immoralist statements of Thrasymachus, Glaucon and Adeimantus.
The break may have come at R. 369a, just where the creation of an
ideal city is about to begin.
Book 2 then took the creation of the city down to the end of our Book
3, the point at which the basic social and political structures are in
place.
Book 3 coincided with our Book 4 and part of Book 5. It thus started
with the question whether the newly constructed city is happy, and
continued through the tripartite division of the soul and the identi-
fication of the virtues, down to the end of our Book 5’s discussion of
family arrangements, the first two ‘waves’, at R. 461e.
Book 4 started with discussion of the city’s unity and consequent good-
ness, continued with the third ‘wave’ (the need for philosopher kings),
and ended at the close of our Book 6, with the Sun and Line.
Book 5 will have contained not just our Book 7, with the Cave simile
and the completion of the philosophers’ education, but also, remark-
ably, all or most of Book 8, quite possibly ending as Book 8 does in
the middle of the construction of the tyrannical city and individual.
Book 6, finally, contained some or all of our Book 9, and the whole of
our Book 10.
Although this division yields six books of suitable length, it lacks the
cleanest breaks that the ten-book division can boast. In particular, the
ends of our Books 1, 7 and 9 read as if each might have been designed to
mark a major closure; yet in the six-book schema each of these occurred in
the middle of a book.
This, in turn, suggests that the six-book Republic came across to its read-
ers as less episodic than our version risks doing. The endings of its books
were such as to leave the reader poised for a continuation. For example,
even if its Book 1 formed a natural unit,10 bringing together the critiques
of justice by Thrasymachus, Glaucon and Adeimantus, it left the reader
anxiously awaiting Socrates’ reply to them. This contrasts with our current
9
The evidence is set out in Hirmer’s appendix (Hirmer 1897: 676–8). I performed my own inde-
pendent calculation, based directly on the Antatticist’s citations, and arrived at the same division
as Hirmer. But it has to be admitted that the findings rely more than one would ideally like on the
unproven, if plausible, assumption that the books were all of broadly similar length.
10
Strikingly, almost the same division is independently adopted by Emlyn-Jones 2005.
Book 1, which ends in the kind of apparently terminal failure associated
with the closures of familiar Socratic dialogues.11
The six-book schema, thus understood, is an artful enough construction
to be Plato’s own, and we should consider seriously the possibility that it
was. Alline, assuming Hirmer’s division, supported his contention that the
six-book scheme is unlikely to be Plato’s partly on the ground that it does
not cut the text at the natural joints. But it may be that, on the contrary,
this very fact reflects Plato’s own preference for not marking off discrete
parts. Unlike the ten-book division, the six-book division may have served
to minimise the dialogue’s fragmentation into discrete episodes, and to
maximise its continuity as a single conversation across the unavoidable
endings of scrolls. If ascribed to Plato himself, it indicates that he was
keen for the Republic, despite its great length, to be read as a single con-
tinuous conversation.
The six-book redaction is thus a salutary reminder that we would do
much better to abandon our habitual talk of distinct books, and instead to
stick with Stephanus page references. To pick one example, to the best of
my knowledge the modern debate as to whether Book 1 (of the ten-book
edition) was originally an independent dialogue has taken no account of
this evidence that its marking off as a single book is likely to post-date
Plato’s own redaction of the Republic.12 If instead it originally continued
into what we call Book 2, there is no evident way in which it can have pre-
sented itself as a complete dialogue, capable of publication as an autono-
mous text. Rather, we will have to suppose that it left its reader poised
for more, probably at the point where the triple immoralist challenge of
Thrasymachus, Glaucon and Adeimantus was complete, and Socrates’
reply to it was about to begin.13
Even if the Platonic authorship of the six-book schema were to be
doubted, the probability that it is closer to Plato’s original text than the
ten-book version is compelling reason for trying to defamiliarise the latter.
11
The appeal to the so-called Socratic fallacy in the last sentence of our Book 1 (354c1–3) recalls that
in the closing sentence of the Lysis (223b5–8).
12
There is a useful survey of relevant scholarship in Vegetti 1998: 15–22. As far as I am aware, the divi-
sion between our Books 1 and 2 has been questioned solely by Vlastos (1991: 250–1), and only to the
extent of suggesting that the original Book 1 may have ended a few lines earlier.
13
The Antatticist’s evidence permits us to say with confidence no more than that the division between
Books 1 and 2 of the six-book edition occurred somewhere between 342b and 379d. Given that the
whole of its Books 1–2 coincided more or less with our Books 1–3 (n. 3 above), amounting to
roughly ninety Stephanus pages, it is unlikely that its Book 1 stopped at the end of our Book 1, a
mere twenty-seven pages in. An approximately equal division of its two first books could well put
the end of Book 1 soon after the end of Adimantus’ speech, for example at 368c, where Socrates is
first pressed to answer the challenge (around forty-one Stephanus pages in), or a bit later.
74 David Sedley
Weaning ourselves off the convenient, but misleading, talk of book num-
bers might help our understanding of the Republic in all sorts of ways. We
are exceptionally lucky with all Plato’s dialogues, other than the Republic
and Laws, to have inherited the structurally neutral Stephanus page divi-
sions as our sole mode of citation. Imagine the distortions that our read-
ing of Plato would have risked if we had instead inherited a division of
all the dialogues into chapters and sections, or (even worse) into acts and
scenes.
This at last brings me to the main topic. My title was inspired by the
following observation, which (despite the preceding disclaimers) has to
be based, for the sake of initial clarity, on the traditional book division.
Our Book 4 of the Republic ends by developing Plato’s celebrated tri-
partite psychology, and delineating the cardinal virtues on its basis. Our
Books 8–10 build elaborately on this same psychology, both in describing
a moral decline from the truly just person all the way down to the tyrant
(Books 8–9), and in developing a new critique of poetry (Book 10).14 Yet
in the intervening books, 5–7, the tripartite soul vanishes from view. I do
not think that this is accidental, and my aim is to ask the reasons for its
temporary suppression.
However, with the ten-book division sidelined, the question deserves
to be reframed. My question is not about three particular books, but
about the dialogue’s long central digression. Where does this start and
stop? There is no doubt that at the beginning of our Book 8 the whole
of Books 5–7 are marked as a digression: following the creation of the
ideal city, we learn there, it is time to return to the topic of the inferior
constitutions. That is a topic that was announced at the end of Book 4 but
then postponed so that further social arrangements, corresponding to the
three ‘waves’, could be discussed. Thus the beginning of our Book 8, by
announcing a resumption of a topic postponed at the very beginning of
Book 5, is marking off the whole of our Books 5–7 as a digression.15 It is
quite plausible that this major structural announcement guided whoever
it was (Thrasyllus?) who first marked off Books 5–7 as a distinct part of the
Republic.
However, that three-book section itself includes a smaller and much
more unified passage. This one ends in the same place, at the close of our
Book 7, but begins in the middle of Book 5, at 471c, with Glaucon raising
14
Cf. p. 88 below for the allusion to tripartition in Book 10’s discussion of immortality (611a10–b8).
15
On the significance of digressions in Plato’s dialogues, including the Republic, see Dixsaut in this
volume.
the question whether the ideal city might be realised in practice.16 Socrates
answers with the proposal linked to his third wave, the need for philoso-
phers to rule. This topic, including in particular the education of the phi-
losophers, runs down to the end of Book 7, where it is closed off with
the remark that the city’s creation has now been shown to be possible,
though difficult (540d–541a). Whereas Books 5–7 as a whole are linked, it
might be said, by the sequence of three waves, the shorter passage I have
picked out is unified both by the theme of the city’s practicability and,
more important still, by its strict focus on the nature of philosophy and its
realisation in suitably gifted individuals.
That Plato himself recognised this tighter unity is confirmed in a curi-
ously negative way by the Timaeus. There, it is a well-known anomaly
that Socrates’ summary of the previous day’s discussion on the one hand
reproduces accurately the social and political provisions of the Republic,
but on the other hand totally omits the role assigned there to philoso-
phers in Books 5–7. Whatever Plato’s reasons for this contrived silence, it
is significant that the silence starts in the middle of our Book 5: the first
two waves – that is, the equal status of women and the abolition of the
family – are fully covered as an integral part of the political set-up, and it
is the third wave alone that is set aside. Thus, the passage picked out by
Socrates’ pointed omission in the Timaeus corresponds not to the entire
section Books 5–7, but to the smaller unit that I have indicated, the one
devoted to philosophy and philosophers.
Nothing that I am going on to say would be compromised if one did
take the whole of our Books 5–7 to be the relevant unit. But the first part
of Book 5, which I omit from it, has no significant bearing on moral psy-
chology, and is thus better ignored for present purposes.
While thinking about the issues that I shall be addressing in the remain-
der of this chapter, I have turned frequently to Christopher Rowe’s book
Plato and the Art of Philosophical Writing (2007a), because he more than
any other scholar I know is alive to the issues that I am confronting. If I
shall have occasion to mention points on which we diverge rather than
ones on which we agree, that should not disguise my great debt, and my
admiration for the light he has shed.
Let me start, then, from the introduction of the Good, 502c–506b.
This is the one passage in the whole of the Republic digression that makes
16
Those who have rightly chosen 471c as the beginning of the central digression include Penner
2007a: 23. Averroes, despite having inherited the ten-book division, chose this same point (give or
take a page) in his commentary on the Republic to mark the division between the first two of his
three ‘treatises’: see Rosenthal 1956: 175–6.
76 David Sedley
explicit mention of the tripartite psychology. But it does so in a way calcu-
lated to marginalise it. Socrates has undertaken to sketch the higher stud-
ies that distinguish philosophers. Their intellectual gifts must be brought
into unison by training them right up to the highest branches of learning
(μέγιστα μαθήματα, 503e3), and including the very highest one of all (τὸ
μέγιστον μάθημα, 504d–505a), the study of the Good. ‘You presumably
remember,’ says Socrates (504a, recalling 435cd), ‘how, after distinguishing
three kinds of soul, we embarked on working out what justice, modera-
tion, courage and wisdom each are … Do you also remember our prefa-
tory remark? … We said, didn’t we, that to see them in the finest possible
way would require a lengthier detour (μακροτέρα … περίοδος).’17 A future
guardian of the city, he adds, ‘must take that lengthier detour’ (504cd),
which is, in fact, the highest of all disciplines, the study of the Good itself.
Neither in Book 4 nor here does the text explain fully whether the things
(plural) that will be seen clearly only by the longer route are (a) the three
parts of the soul or (b) what each of the virtues really is. But since the
remainder of the digression will sketch a radical redefinition of the virtues,
while saying nothing more about the tripartite soul, I assume that (b) is
meant.18 It is Book 4’s account of the virtues that is to be in some sense
superseded in the philosophers’ education programme. And this will be
done without the help of tripartition.
The new account is thoroughly intellectualist. The lower soul-parts
make no contribution to it at all. By this, I do not mean to suggest that the
Republic’s earlier division of the soul into three parts has been altogether
discounted. In our passage Socrates several times refers to the intellect, or
at any rate what sounds like the intellect, as if it were just one part of the
soul (490b3–4, 518c4–d1, 527d8, 532c6, 533d1–2), thus reminding us that
the earlier division of the soul still officially stands. Nevertheless, moral
psychology will here be explained in terms of the intellect alone, in a man-
ner much more reminiscent of the Protagoras and Phaedo than of Republic
4. This is the ‘Socratic intellectualism’ of my title. I mean by it, in accord-
ance with convention, the thesis displayed in Plato’s Socratic dialogues,
including for this purpose the Meno and even the Phaedo, according to
which all moral states can be analysed as intellectual ones: every soul aims
17
All translations in this chapter are my own, and follow the text and line-numbering of Slings
2003.
18
I here agree with Adam 1963 ad loc. An anonymous referee points out that 504c9–d8 lends stronger
confirmation to the reading than Adam made explicit. Rowe (2007a: 164–6), if I have understood
him correctly, takes the primary reference to be to the nature of the soul; Penner (2007a: 24–8)
argues for a fusion of (a) and (b).
for the good, and virtuous people are those with the requisite understand-
ing to achieve this universal end. (Whether or not it was also the view of
the historical Socrates is a separate question, on which I make no com-
ment here.)
That the moral psychology of the Republic digression is a resumption of
Socratic themes is made particularly clear by the following consideration.19
When the Good is introduced, two failed candidates for the role are ini-
tially listed: wisdom and pleasure. Pleasure is said to be what the many
identify as the good (505b5), with the further remark that this equation of
good with pleasure is easily refuted by forcing its proponents to agree that
there are in fact bad pleasures (505c6–d1). This neatly encapsulates the
fortunes of hedonism in the Protagoras and Gorgias. In the Protagoras, the
thesis of the identity of pleasure and the good is ascribed to the many, on
the ground that, if pressed on this point, they will not be able to come up
with any other candidate for the nature of the good apart from pleasure.
Socrates, for reasons that remain controversial among interpreters, allies
himself there with the popular identification of pleasant and good. In
the Gorgias, his final argument against any such identification is precisely
the one referred to here in the Republic, that the hedonists can be forced
to admit the existence of bad pleasures (499b6–500a6). Hence it is easy
to feel that the Republic’s discussion of the Good is resuming where the
Socrates of the Protagoras and Gorgias left off.
Much the same applies to the second candidate, ascribed to those who
are ‘more sophisticated’ (κομψοτέροις, 505b6). They hold that the good
is wisdom (phronēsis). And their refutation is reported here to be that,
when asked which wisdom they mean, they are unable to give a satisfac-
tory reply: all they can say is ‘Wisdom about the good’, leaving them with
the fatally circular definition that the good is wisdom about the good
(505b8–c5). In a parallel to the case of pleasure, this too is easily identified
as a position which the Socratic dialogues have shown Socrates both flirt-
ing with and ultimately finding incoherent. The classic text is Euthydemus,
and in particular its second protreptic passage. The first protreptic passage
in that dialogue had already established that knowledge is sufficient for
happiness and that we should therefore pursue philosophy. The further
question raised in the second protreptic passage is which knowledge we
need in order to be happy. After much searching, the discussion finally
turns to the art of kingship (βασιλικὴ τέχνη, 291d7). But unlike medicine
19
And cf. Rowe 2007a: 241: ‘the whole of 505b5–c11 is a kind of patchwork of elements from
pre-Republic dialogues.’
78 David Sedley
or farming, whose product is easily identified, it proves to be unclear what
good product is generated by kingship. The good has already been agreed
to be knowledge (292b1–2), and if that knowledge is the kingly art then
we are left with no idea what knowledge the kingly art is meant to gen-
erate beyond itself (292a4–e5). At the end of the discussion Socrates con-
fesses to being in deep aporia, and asks who is going to rescue them from
the stormy waves, the trikumia, of their argument (292e6–293a6).
The intertextuality between this protreptic passage and the Republic
digression is unmistakable.20 In both passages the question is what the
goodness is that a philosopher king needs to know. In the Republic, the
postulation of philosopher kings marks the third of the three great ‘waves’
of protest that Socrates anticipates; in the Euthydemus, the problem about
what kind of goodness philosopher kings can impart, given the assump-
tion that the good is knowledge, is called a triple wave (trikumia).21 And,
above all, the Republic’s objection to identifying the good with knowledge,
on the grounds of a concealed circularity, formalises the very objection
that Socrates had made against himself in the Euthydemus.
In short, we are given abundant clues that the Republic’s account of
the Good will be resuming a Socratic discussion from the Euthydemus
at exactly the point where Socrates begged for help. When the Republic
takes up and pursues the same issue, we find ourselves in the company
of a speaker who still has much in common with the Socrates of those
aporetic dialogues. I mean by this a Socrates who does not claim to be
able to define the Good (506c), but who nevertheless has come to under-
stand the necessity for there to be some objective good over and above
all good states of mind, one that can serve as the proper object of kingly
knowledge. This is equally the Socrates of the Charmides, who struggled
with the idea that a virtue might consist in knowledge which has noth-
ing but knowledge as its object. In the Republic digression Socratic moral
psychology is being resumed and developed, with the help, as we shall see,
of refinements worked out in the intervening Phaedo. And this time the
Socratic dialogues’ repeated demand that virtuous knowledge have some
object other than knowledge itself will at last be properly addressed.
20
I first learnt of this intertextual correspondence (which is more extensive, and hence more unmis-
takable, than I have conveyed above) from Myles Burnyeat in the 1980s; see now Burnyeat 2002:
63 n. 46, where he himself cites Narcy 1984: 183. One may want to draw conclusions about the
relative chronology of the two dialogues, but my present point need not depend on that subsidiary
question. See also McCabe 2002a; and, on the ‘waves’ of aporia in the Euthydemus, McCabe in this
volume.
21
On the meaning of this term, see Sedley 2005.
The continuation of a Socratic quest need not in itself constitute a
reversion to an earlier Socratic position. But there is plenty in this long
digression to confirm that the moral psychology assumed and explored
there is such a reversion. Take the role of desires. No mention at all is
made of lower soul-parts with their own distinct desires. Thus, when the
innate or natural virtues of the potential philosophers are set out, their
natural moderation (σωφροσύνη) is expressed without any hint of Book
4’s characterisation of moderation as orderliness. In Book 4, moderation
was presented as a ‘control of certain pleasures and appetites’ (430e4–8),
above all those of the appetitive part of the soul. But here in the digres-
sion we are given a quite different account, one thoroughly reminiscent of
the Phaedo, in that it is founded not on soul division but on soul–body
dualism. Desires (called ἐπιθυμίαι at 485d6) are not subdivided between
different soul-parts, but are treated as a single psychic force, one that can
be channelled either into intellectual pursuits or into those of the body.22
Provided that you focus your desires on intellectual pleasure, bodily
pleasure will be neglected by default, and you will ipso facto be moderate
(σώφρων) (485d6–e5):
But when someone’s desires (ἐπιθυμίαι) converge strongly on a single thing,
we presumably know that this person’s desires for other things are weaker,
like a diverted stream … Well, when someone’s desires have flowed towards
the sciences and everything of that kind, they would concern themselves
with the pleasure of the soul in its own right, I think, and would leave aside
the pleasures achieved through the body, provided that the person is a real
philosopher and not a fake one … Such a person is moderate (σώφρων),
and in no way a money-lover, because the things for whose sake a big
cash-flow is desirable are things more appropriately desired by another type
than by this one.
That this reunification of desire under a single heading is by no means
a casual shift of psychological emphasis is confirmed by the section on
the Good. There the Good is described as ‘that which every soul pursues,
doing everything for its sake because it intuits that there is such a thing’
(505e1–2). This account is in manifest tension with Book 4’s psychology.
Can it really be said, for example, that according to the Book 4 account the
soul of Leontius, when he gave in to his appetites and gazed at the corpses
(439e5–440a5), was acting for the sake of the good? The tension between
the two accounts confirms that we have returned for present purposes to
22
This is what Lane 2007 has helpfully discussed as the ‘hydraulic model’ of desire.
80 David Sedley
a monistic psychology more familiar from the Gorgias (467c5–468c8) and
Meno (77b6–78b6), where all desire is explicitly for the good alone.23
The account of moderation that we have encountered is one sign of a
Socratic intellectualist approach to virtue in the Republic’s central digres-
sion. However, the strongest manifestation of this same approach comes
soon after the Cave allegory. This key passage needs to be approached
carefully, in two stages. The first stage reads as follows (518b7–d2):
‘If this is true,’ I said, ‘we should hold a belief about them along the fol-
lowing lines: that education is not the sort of thing some people profess it
to be. What they say is that when knowledge is not in the soul they them-
selves put it there, as if they were putting sight into blind eyes.’
‘Yes, they do say that,’ he replied.
‘But the meaning of our present account,’ I said, ‘is that, just as if it were
impossible to turn an eye from what is dark to what is bright except along
with the whole body, this power which is in everybody’s soul and the organ
which everybody uses for learning have to be turned round along with the
whole soul from what is in a state of becoming, until it can bear to gaze
upon that which is, and the brightest part of that which is, which we say is
the Good. Is that right?’
‘Yes.’
This is one of many places in the Republic digression where we find Plato
developing an elaborate account of intellectual vision illuminated by the
Good, a power systematically analogous to that of ocular vision illumi-
nated by the sun. When the prisoner released from the cave turns his
whole body towards the light, this symbolises the fact that intellectual
enlightenment means turning the mind’s eye, and with it the whole soul,
towards the ultimate source of illumination, the form of the Good. Here,
then, we have one of the digression’s occasional reminders that the divi-
sion of the soul has not actually been abandoned.
The wording leaves it underdetermined whether the intellect could in
theory have turned to the truth while leaving the rest of the soul behind.
But a very natural reading is that when you turn your intellect towards the
Good the rest of your mental powers ipso facto go along with it. There is
no hint that, in the process of philosophical enlightenment that is at issue
here, the non-rational faculties exercise any independent sway. And this
23
At 490a8–b7 Socrates shifts briefly into the psychological language of the Symposium to describe
the trainee philosopher’s passion for learning. This incorporates into the Republic digression yet
another of Plato’s earlier psychological models, again a monistic one according to which all desire is
for the good (more strictly, for everlasting possession of the good).
intellectualist way of reading the passage finds support in what follows
almost immediately (518d9–519b5):
‘Hence the other so-called virtues of soul seem to verge closely upon those
of the body. For the reality is that, having not been present in the soul
previously, they are later created in it by habit and practice. But the virtue
of being wise (tou phronēsai) turns out to belong to something altogether
more divine, it seems: something which never loses its power, but is made
either useful and beneficial or useless and harmful by the way it is turned.
Or haven’t you noticed, in people who are called bad but wise, how keenly
their little soul focuses on, and how sharply it discriminates between, the
things towards which it is turned, showing that it does not have bad vision,
but is compelled to place itself at the service of vice, so that the more
sharply it sees the more harm it does?’
‘Quite so,’ he said.
‘Yet this faculty,’ I said, ‘of people with such natural talent, if it were,
right from birth, beaten into shape and stripped of the lead weights, as it
were, that relate it to becoming, weights which as a result of eating and of
pleasure and overindulgence in such things become part of their nature
and turn their soul’s vision upside down – if this faculty were rid of those
things, and had turned round to look at truths, the same faculty of the
same people would be seeing even those truths with the great sharpness
with which it sees those things to which it is now turned.’
From this key passage we can extract the following points.
The difference between the philosopher and the talented scoundrel lies
exclusively (so far as we can tell) in the way they each direct their intellec-
tual gaze. The metaphysical aspect of this choice is admittedly spelt out in
terms germane to the Republic and with no obvious Socratic background:
the philosopher’s soul fixes its gaze on being, the scoundrel’s on becoming.
Thoroughly Socratic, on the other hand, is the underlying intellectualist
assumption that what morally wrecks a talented soul is simply the way it
directs its intellectual gaze. No room is left here for the power of the two
lower parts to drag the rational part down.
Instead of the lower soul-parts competing with, and potentially pervert-
ing, the intellect, the soul as a whole is seen as somehow pitted against the
body. The talented scoundrel’s intellect has been weighed down – diverted
from the truth – by its focus on bodily indulgences. We are being invited
to read this in conjunction with the immediately preceding reference to
the virtues of the body (518d10). But how?
The key contrast is between the virtue of ‘being wise’ (phronēsai,
518e1) and the other ‘so called’ virtues of the soul. Several questions arise.
Assuming that the ‘so called virtues of the soul’ are so named incorrectly,
82 David Sedley
is this because they are misleadingly called ‘virtues’, or because they are
misleadingly said to belong to the soul? And if the latter, is the truth that
they belong at least in part to the body? Or when they are said to verge
closely upon the virtues of the body, does that closeness consist merely in
a degree of resemblance to bodily virtues? What in any case are the virtues
of the body? Again, is the ‘virtue of being wise’ just a variant name for the
virtue usually called ‘wisdom’ (phronēsis)? And is the ‘more divine thing’
to which this authentic virtue belongs the whole soul, or just the intellect,
both soul and intellect often being credited with a degree of divinity by
Plato?
These questions are so slippery to handle, and the text so unlikely to
yield a clear answer by itself, that a more promising tactic will be to inter-
pret the passage with the help of some intertextual interpretation. The
intertext that seems to me to be clearly indicated here is Phd. 68b8–69d3.
In this Phaedo passage Socrates opens by distinguishing the wisdom-lover
(φιλόσοφος) from the body-lover (φιλοσώματος). But the simple bipar-
tition immediately becomes a tripartition, since body-lovers are further
analysed into the money-lover (φιλοχρήματος) and the honour-lover
(φιλότιμος). Thus, we have a tripartition (philosopher, honour-lover,
money-lover) which anticipates the three psychological types examined
and compared in Republic 9 (580d–581c6), where they are distinguished
from each other according to which part of the soul predominates and
sets their agenda.24 But the tripartition is here mapped onto a Phaedo-style
bipartition, which contrasts the philosopher, devoted to the care of the
soul, with the non-philosopher, whether money-lover or honour-lover,
whose primary concern is for the body. This equivalence offers us a key for
interpreting the Republic passage. We should therefore read on.
Both types of body-lover, we next learn (with a later resumption at
83e5–7, cf. also 114d8–115a3), are credited with virtues which are in reality
no such thing. Their ‘moderation’ and ‘courage’ turn out on inspection
to be skills for the prudential management of bodily desires and pleas-
ures, with a view not to eliminating these but on the contrary to maximis-
ing them. Implicitly, foregoing a bodily pleasure now, or facing a danger
now, is for them simply a device for enjoying greater bodily pleasure or
security in the long run. They measure value in terms of pleasure, pain
and fear, not appreciating, as the philosopher alone does, that the only
true currency is wisdom (φρόνησις). Wisdom alone confers value on other
supposed goods, the virtues included. And wisdom, properly understood,
24
The same tripartition recurs at Phd. 82c2–8.
is the soul’s purification from all that is bodily. The implication of the
Phaedo passage, as Neoplatonist interpreters appreciated, is that real cour-
age and real moderation are attained only when wisdom purifies the soul
of its concern for the body’s welfare, safety and pleasure, and liberates it to
pursue its own intellectual agenda.
It is, I think, this purificatory virtue, as explained in the Phaedo, that
is being recalled in our passage of the Republic. The ‘so-called’ virtues
of the soul are to be identified as those whose link to the soul is tenu-
ous because their ultimate objective is in fact the care of the body. The
‘virtues of the body’ upon which these quasi-virtues of the soul are said
to verge are, no doubt, exemplified by health and strength, and possi-
bly also wealth, since the Phaedo passage classes the money-lover as a
type of body-lover. The point being made about the non-philosophical
quasi-virtues – quasi-moderation and quasi-courage – will be that they
collude with these bodily virtues, all of them alike being geared towards
maximal bodily wellbeing.
But the Republic passage is not simply reiterating the Phaedo’s critique
of popular virtue. What it adds is its own newly formulated criterion for
distinguishing real virtue from quasi-virtue. Already in the Phaedo real
virtue was associated with the soul’s liberation to pursue its own agenda.
In the Republic, this is interpreted in terms of the direction of the soul’s
gaze. Making a soul into a philosophical one, as we saw earlier, simply
requires turning the already-present intellect in the right direction. But
many people, we read here (518b7–c2), think that education is, on the
contrary, conferring on the soul a new capacity, analogous to putting sight
into blind eyes. It is this unexplained contrast that now comes to the fore,
in elucidating what is wrong with the quasi-virtues: ‘For the reality is that,
having not been present in the soul previously, they are later created in it
by habit and practice’ (518d10–11).
What this reference to habit and practice tells us is that quasi-virtues
are demotic virtues. According to both the Phaedo and the Republic (Phd.
82a11–b3, R. 500d5–9), demotic virtues are not the exercise of any intel-
lectual capacity, but simply a propensity for good civic habits, instilled
into the citizens’ souls by training. They are not directly identifiable with
the virtues of Republic 4, which are analysed in terms of soul-parts but,
as their presence in the Phaedo confirms, are simply a de-intellectualised
imitation of real virtues.25 As is indicated by their presence in the Phaedo
25
Traditionally (at least since Plotinus 1.2.3.8, and quite possibly since Aristotle, Nicomachean Ethics
1116a15–21, Eudemian Ethics 1229a13, cf. [Aristotle], Magna Moralia 1191a5–10), the same notion of
84 David Sedley
and confirmed by the Republic, their inferiority is to be understood by
reference to their focus on the body as distinct from the soul. No doubt –
although the point is nowhere developed – citizens habituated into good
behaviour are ultimately motivated by a concern for their own bodily wel-
fare, especially through material acquisitions, rather than for any kind of
intellectual fulfilment. We should remember that in Callipolis the com-
mercial class, in return for their compliance with the power structure, are
allowed to own all the property. But even outside an ideal city, as the
Phaedo (68b8–69d3) has emphasised, quasi-virtues are fuelled by the desire
for bodily fulfilment.
Returning to the virtue of ‘being wise’, phronēsai (518e1), which our
Republic passage says belongs to something more divine than the body,
we can now be confident that the reference is indeed to the virtue of wis-
dom, phronēsis, and that the ‘something more divine’ to which it belongs
is either the whole soul (as the Phaedo would undoubtedly say), or the
intellect insofar as the Republic considers this a distinct soul-part. Since
his early Socratic phase Plato had regularly represented phronēsis as the
principle of all virtue. What we have here, however, is not the familiar
Socratic view that the other virtues have to be informed by wisdom in
order to be genuine virtues at all, but the Phaedo’s extreme version of that
view, according to which the wisdom at issue is not practical wisdom,
guiding the virtues in moral conduct, but pure intellectual wisdom that
altogether transcends the body and its concerns. (This is not, of course, to
deny that once it has risen all the way to a full dialectical understanding
of the Good such wisdom can then inform practical action as well, which
indeed it must do in philosopher kings.)
It is now time to ask exactly what is going on. Why has the central digres-
sion of the Republic shifted into a Phaedo-like version of Socratic psychology,
one which will be replaced once more with the tripartite model the moment
the digression is over? The general problem is one that has been highlighted
‘demotic’ or ‘civic’ virtue has been detected at R. 4, 430c3: πολιτικήν γε (sc. ἀνδρείαν). I doubt this
reading, because the topic at that point is still the courage of the city, and the phrase should there-
fore mean ‘at any rate that of a city’, exactly matching the reference to moderation and justice ‘in
the city’ (ἐν τῇ πόλει) that immediately follows, at 430c8. Personal courage is not addressed until
442b10–c2. But what, in that case, is the relation of demotic virtue to the other two kinds, namely
psychic balance (Book 4) and intellectual purification (Books 5–7)? The answer in relation to intel-
lectual purification is given in the present passage, 518d9–519b5. In relation to psychic balance, it
will be given in passing at 590c7–d6: while true psychic balance is when one’s soul is governed
by its own powers of reason, the next best thing is when it is governed by reason imposed from
outside. The latter is clearly ‘demotic’ virtue, imposed by politicians. How far short it falls of true
psychic balance is brought out in the myth, at 619b–d, where a soul habituated in merely demotic
virtue proves capable of making the worst possible choice of its next life.
in particular by Christopher Rowe’s work, although he places more of the
focus on Book 10, with its expressed doubts about tripartition, than on
Books 5–7. His view is that Plato never meant to abandon Socratic intellec-
tualism and replace it with tripartition, and that his occasional resort to talk
of tripartition represents a potentially misleading but perhaps unavoidable
mode of discourse about the soul’s incarnate life, during which most people
come to function as if they were under the influence of irrational forces,
despite the fact that these are not part of their real nature.26
My own alternative view tries to leave room for a less unified picture of
Plato than this. My starting point here, one that I have articulated else-
where,27 is that, whether or not unitarianism is judged to offer the most
helpful analysis of the Platonic corpus, the original unitarian was none
other than Plato himself. Plato is constantly trying to show that, wher-
ever we might have suspected him of placing some philosophical distance
between himself and Socrates, he is in fact doing no more than continue,
interpret and perfect Socrates’ own lifetime project. An example is the
way that the Republic starts by pointedly recalling the elenctic Socrates of
the early dialogues, but goes on to present the eventually successful def-
inition of justice as a seamless continuation of his work, not as any kind
of new beginning. (The continuity is all the more emphatic if we follow
the six-book division and thus eliminate the break at the end of our Book
1.) The recurrence, at the heart of the Republic, of Socratic intellectualism
reads to me like part of this same unitarian strategy, an affirmation that
the new psychological model has not supplanted the old.28 It nevertheless
leaves us with the delicate task of explaining how Socratic intellectualism
can remain true, at a point in Plato’s work where tripartition gives every
appearance of having superseded it.
In tackling this problem, two preliminary points need to be made. The
first is structural. The Republic is not alone among Plato’s dialogues in
the function played by its digression. The Theaetetus digression, placed
at the exact mathematical centre of that dialogue, plays a significantly
similar role. There the main body of the dialogue investigates the defin-
ition of knowledge from a persistently empiricist set of assumptions, and
the digression is unique in rising far above those assumptions. Like the
Republic digression, this digression too focuses on the nature of philo-
sophical knowledge as radically detached from the world of particulars,
26
Esp. Rowe 2007a: 171. 27 Sedley 2004.
28
Rowe (2007a: 19 n. 52) asks how my interpretative strategy can deal with the Republic, and what I
say here is meant as a partial answer.
86 David Sedley
and presents the true philosopher as so far above his physical and civic
surroundings as to be barely aware that his neighbour is a human being.
To point out this parallel structural anomaly in the Theaetetus may seem
like doubling the problem rather than solving it. But what it does show
at the very least is that the abrupt switch into and out of Socratic intel-
lectualism at the centre of the Republic, whatever its meaning, is an artful
and repeatable device on Plato’s part.
The second preliminary point is thematic. The anomalous juxtaposi-
tion of tripartite psychology and intellectualism is by no means unique
to the Republic. It recurs in the dialogue that is in many ways its sequel,
the Timaeus. And there once again the intellectualist model is the one
associated with the philosopher’s unique attainments. A large part of the
dialogue is devoted to the soul’s tripartition, which is certainly not repre-
sented as a mere ad hoc explanatory or heuristic device, since it is rooted in
the basic construction of the human body. In the light of this exposition,
at Ti. 89e3–90a2 an account of human wellbeing is advanced, based on
the harmonious relation between the three soul-parts, closely recalling the
psychic harmony described in Republic 4. But it is immediately followed
by a climax in which intellectualism takes over (90a2–d7). True happiness
is now located not in the harmony of the whole soul, but in the correct
alignment of the revolutions that occur in the rational mind alone, while
the other two soul parts are ignored. As in the Theaetetus digression, this
state of intellectual fulfilment is presented as a godlike one, which for that
very reason transcends and eclipses the essentially human contentment
enjoyed by the harmoniously balanced tripartite soul. That this juxtaposi-
tion of the two psychological models is integral to Plato’s thought, and
not a mere compromise, is further witnessed by its early antecedents in the
Gorgias,29 and, more strikingly, by its survival into the work of Plato’s lead-
ing pupil. For the classic enigma of Aristotle’s Nicomachean Ethics – how
the lives of moral virtue and intellectual virtue can be successively recom-
mended in a single treatise without the least indication that the author has
changed his mind – is really just a reformulation in Aristotelian terms of
the phenomenon we have witnessed in the Republic.
These clues suggest the following lesson for our understanding of the
Republic. Plato really does believe that the tripartite analysis is an advance
over his inherited Socratic psychology, at least to the extent that it is rooted
in human nature and provides the best possible account of the incarnate
29
Compare Grg. 467b–468c, on the wish for the good as the universal human motivation, with pas-
sages such as Grg. 492d–494a, 507c-508a, on virtue as psychic orderliness.
human soul. However, tripartite psychology is of no particular help for
understanding the moral and intellectual mindset that lifts the philoso-
pher to a higher and more godlike plane. For what marks the real philoso-
pher is not a reasoned control of irrational and semi-rational drives, but
ascent to a level of understanding where the demands of the carnal world
simply fade into the background, lose their relevance, and have no more
than a bare minimum of motivational purchase. Whether this is achieved
because the philosopher’s intellect is naturally dominant, or because his
earlier education in the ideal city is assumed to have made it so by now, is
left unclear. Plato no doubt prefers to leave both options open, since not
all philosophers live in an ideal city.30
One might at this stage pose the following question. Should we say that
the Republic digression offers us a better analysis of the very same virtues
that were previously defined on the basis of the tripartite soul? Or that
it presents an analysis of virtues which are themselves better virtues than
those previously described, since they are those of the philosopher alone?31
There is no need to choose. On the one hand, Socrates certainly does not
say that he has switched to talking about a different set of virtues, but,
on the contrary, implies that the ‘longer route’ taken here is a route to
an enhanced understanding of the very same virtues. On the other hand,
it is a consequence of Plato’s paradigmatism that the best way to define
the essential features of x just is to define the essential features of the best
possible x. For instance, at Timaeus 46e7–47c4, the best way to describe
the function of the eye is to describe the best possible function of the
eye, namely that of facilitating astronomy. Likewise, what we meet in the
Republic digression is the best possible account of the virtues, namely an
account of the highest virtues, those of a philosopher.
One reason for insisting that, in some sense, the same virtues are at issue
is that, in the digression, Socrates speaks as if the body/intellect dualism
he is now employing can be used to characterise non-philosophers as well
as philosophers. The clever scoundrel, as we saw, is defined by his intel-
lectual gaze as effectively as the philosopher is by his. The quasi-virtues
30
Cf. R. 520b.
31
I have used the plural ‘virtues’ here, even though 518d9–e1 may seem to single out wisdom alone as
an authentic virtue (cf. Meno 89a3). The Phaedo allows that courage, moderation and justice, once
informed by wisdom, become authentic virtues, supporting the soul’s detachment from bodily
concerns. And the insistence of the digression in the Republic that trainee philosophers should have
natural courage and moderation, defined as we have seen in a rather similar way, suggests that the
picture offered in the Phaedo is fully assumed here too. Conceivably, calling it ἡ … τοῦ φρονῆσαι
(sc. ἀρετή) (518d11–e1), rather than simply φρόνησις, is designed to allow this broader meaning: all
the virtue that comes with exercise of the intellect.
88 David Sedley
of ordinary people are likewise successfully characterised by their focus
on the body’s goals rather than on those of the soul. Thus, what we have
in the Republic digression aspires to be a complete moral psychology, not
one applicable to the philosopher alone. Nevertheless, its leading merit
in Plato’s eyes is undoubtedly that, unlike the tripartite psychology, it is
one that can correctly capture the intellectualised virtue distinctive of the
philosopher.
In the final book of the Republic Socrates expresses uncertainty about
the correct psychological theory to adopt, now that the soul has been
shown to be immortal (611a10–612a7). The tripartite account, he insists,
has been the right one to use for the incarnate soul, but it does not neces-
sarily capture the truest nature of the soul when it exists independently of
the body. It is one of Christopher Rowe’s merits to have insisted on the
importance of this passage for the overall understanding of the dialogue’s
psychology.32 His remarks tend towards taking the passage as Plato’s admis-
sion that a unified Socratic psychology remains the correct model.
The primary point of the passage concerns the soul’s post mortem nature.
When it is released from the body, Socrates is saying, it may well turn out
to leave the tripartite structure behind as well. And as we know, Plato will
in his late work favour the view that the discarnate soul is in fact unitary.
But he will not do so until he writes the Timaeus, having first, in the
Phaedrus, sought to retain a tripartite structure even for the disembodied
soul. So the aporetic remarks in Republic 10 cannot, I think, in context
represent a clear decision in favour of the monistic psychology over the
tripartite one. What he says there is that it is ‘not easy’ for something
composite to be everlasting (611b5–7), not that this can be ruled out.33 The
tension between the two accounts of psychology is being pointedly left in
place as the dialogue draws to a close.
Nevertheless, Rowe is right that, at any rate in the longer term, this
passage proves to have profound implications for Plato’s comparative
evaluation of the two psychological models. At the close of the Republic
Plato does at least gesture towards the possibility that the soul, when
fully detached from the body, is a unified intellect. If this possibility is
32
Rowe 2007a: esp. 166–9.
33
See also 612a3–4: ‘and then one would see its true nature, whether multiform, uniform, or what-
ever’. I take this to mean: ‘see whether its true nature is multiform, uniform, or whatever,’ and not
(as I think it is understood by Rowe 2007a: 142, 168): ‘see its true nature, regardless of whether it
[the soul? its nature?] is multiform, uniform, or whatever.’ The latter would be referring to features
of the soul that do not depend on its being multiform or uniform, but the preceding discussion has
not focused on any such issues.
confirmed, as it eventually will be, then the same Socratic monism will be
the model that conveys, if not the whole truth about the incarnate soul, at
any rate the irreducible core of its nature.
Thus the Socratic understanding of the soul, showcased in the Republic’s
central digression, and advertised at the dialogue’s close as possibly the
truest analysis, is enabled to retain its place at the very centre of Plato’s
philosophy.
ch apter f i ve
Timaeus in the cave

Thomas Johansen
Introduction
Unitarianism was the norm amongst ancient interpreters of Plato.1 One
strategy they used to maintain the unity of his thinking was to argue that
different works were saying the same things but in different modes. So,
for example, the Republic was saying ethically (ethikōs) what the Timaeus
was saying in the manner of natural philosophy (physikōs).2 In this paper,
I want to offer an interpretation of the Cave image in Republic 7 which
lends support to this division of labour, and so indirectly, at least, to a uni-
tarian understanding of Plato’s thinking across these two works.
The division of labour seems well supported by the Timaeus. The work
starts with Socrates’ resumé of his account of an ideal city much like that of
the Republic. The speakers are to reciprocate with accounts putting Socrates’
ideal citizens into action, Timaeus giving an account of the generation of
the cosmos including that of human beings, and Critias showing human
beings endowed with Timaeus’ nature and Socrates’ education engaging in
war (Ti. 27ab). Listening to Timaeus, we find that the good order of the
soul – tripartite as in the Republic but now with added physiology (69d–
71e) – corresponds closely to the just soul we know from the Republic. It is
an order that Timaeus projects onto the cosmos as a whole, just as Socrates
in the Republic found justice of the soul ‘writ large’ in the city (R. 368c–369a,
442d). For the world-soul is made of the same materials and structured in
I am grateful to audiences at the Universities of Sarajevo and Oxford, and to Despina Fragkoulopoulou,
for criticism. The comments of Christopher Gill and Dimitri El Murr also helped me clarify the argu-
ment. I had the good fortune of having Christopher Rowe as a colleague in my first academic job: one
could have no better mentor.
1
Cf. Rowe 2006a: 15–16. Rowe 2007a shows the relative weakness of current developmental inter-
pretations of Plato compared with a more unitarian approach. For a discussion of Rowe’s argu-
ments, see Sedley, Gill and El Murr in this volume.
2
Cf. Arius Didymus, On the Philosophical Schools 59.1, 25–7 Mullach (= Stobaeus, Ecl. 2. 49.18–22
Wachsmuth).
90
Timaeus in the cave 91
the same kind of way as the rational human soul (41d–42d) and controls the
cosmos in much the same way that human reason controls the virtuous per-
son (44ab, 47bc). Timaeus’ universe looks much like a cosmic counterpiece
to Callipolis; both are fashioned by benevolent craftsmen, a god and a phil-
osopher respectively.3 The politics of the Republic and the cosmology of the
Timaeus seem a joint project, at least from the viewpoint of the latter work.
There are, indeed, points in the Timaeus when the account does not
just join up with that of the Republic but overlaps with it. Thus, in the
discussion of the diseases of the soul (Ti. 86d–87b), Timaeus presents
intemperance as an ailment caused both by the body and by an unedu-
cated upbringing. Cornford, as if echoing the ancient view, is prompted
to comment: ‘This is not the place to pursue further the topic touched
upon in the last sentence [sc. 87a7–b8] – the corrupting influences of an
ill-governed society and the reform in education needed to correct them.
That belongs to the moral and political discourse of the Republic; the
Timaeus is a physical discourse.’4 But even in the account of the bodily
diseases, political notions seep through. For instance, the simple bod-
ies’ wanting to get more than their share (pleonexia, 82e3) causes unrest
(stasis, 82a6) and breeds war (polemos, 83a5, 88e5) within the body. In the
Republic, pleonexia was the key word for the tyrant’s exploitation of his
subjects (344a1). Proper therapy whereby friendly elements are arranged
together (philon para philon, 88e6) will restore proper order (taxis, 88e3),
recalling the friendship obtaining between the parts in the virtuous person
(R. 442c10) and city (Ti. 18a1).5
Timaeus’ physiology is replete with political terms which get their full
explication in the Republic.
However, my focus in this discussion is not primarily the Timaeus
but the Republic. I shall argue that, just as the Timaeus is not innocent
of Republic-style politics, so, conversely, the Republic is informed by
Timaean-style cosmology. I am not making any claims here about the
compositional order of the works; rather my thesis is that there are points
in the Republic where cosmology, of the sort developed in the Timaeus,
3
Cf. Ti. 28a–29a with R. 500d6. 4 Cornford 1937: 349.
5
Compare also the composition of the body of the world at 32c (φιλίαν τε ἔσχεν ἐκ τούτων). Dimitri
El Murr has pointed out to me the Empedoclean background to this use. However, in Empedocles
philia also has a political dimension: see, for example, 31 B128–30 DK. The infusion of cosmology
with moral and political significance is a common feature of Presocratic philosophy adopted by
Plato. Compare the uses of the notion of equality (isotēs, to ison) in Archytas 47 B3 DK, Parmenides
28 B8.49 DK and Empedocles 31 B17.27–9, which form the background to Socrates’ claims about
the power of equality in the cosmos at Gorgias 508a.
92 Thomas Johansen
intrudes on the ethical and political account, in a way that shows the
essential interconnectedness for Plato of the human sphere and the rest of
the cosmos.
Two projects in the Cave?

The Cave image is a particularly interesting case study for this thesis. The
Cave, as it is first presented (514a–521b), is clearly about ethics and politics.
The prisoners are concerned with matters of justice and virtue; they are to
be liberated by the implementation of the ideal city; and their behaviour
towards any potential liberator echoes the Athenians’ treatment of Socrates.
But as we read on (521c–541b), it becomes clear that Socrates thinks that
the liberation is to be brought about by an educational curriculum of ten
years of mathematics, which includes the study of astronomy and harmon-
ics, followed by dialectic, culminating in a vision of the form of the Good.
Here it is less clear what exactly this mathematical programme has to do
with ethics and politics. Malcolm Schofield sees the Cave as engaged in
two different projects. One is a narrative about the abject condition of
ordinary people in the city, particularly the democratic city, and their need
for redemption. This he refers to as the ‘Narrative’. Another is a reinter-
pretation (‘the Commentary’) of the narrative in terms of the philoso-
pher’s ascent through an education in mathematics and dialectic. Myles
Burnyeat had already argued that the philosophical curriculum is a reinter-
pretation of the first narrative: ‘it is only in retrospect that we learn that
the Cave has to do with mathematics as well as cultural values (532bc)’.6
Against this, Schofield protests that: ‘The proposal doesn’t seem to take
the full measure of the incommensurability between Plato’s two projects in
the Cave.’7 ‘Incommensurability’, I take it, implies not, or not primarily,
that the Narrative and the Commentary are inconsistent,8 but that they
are concerned with fundamentally different things in such a way that there
is no single perspective from which both can be understood together. By
contrast, in Burnyeat’s reading, the two are commensurable because the
Commentary integrates the political and ethical message of the Narrative:
it is by learning about unity and harmony in mathematics that we also
learn about the sort of values that will make people and cities good.
6
Burnyeat 1999b: 243; cf. 2000: 45. 7 Schofield 2007: 217 n. 4.
8
However, Schofield 2007: 230 also writes: ‘Narrative and philosophical commentary upon it work
out ideas that are not merely different, but in terminology as well as in substance at odds with each
other.’
There is a question, which I shall address later, about whether, even if
Burnyeat is right about the Republic, this helps my thesis that elements of
cosmology of the sort pursued in the Timaeus are to be found in the Cave.9
For it may seem that the mathematical curriculum itself is hostile to such
a cosmology. However, if Schofield is right, my thesis, as far as Republic
7 is concerned, is a non-starter. If the political narrative is incommensu-
rable with the mathematical commentary, it would by the same token be
incommensurable with a cosmological account, since it is the presence of
harmony and proportion in the cosmos that makes the strongest case for
thinking that the cosmology of the Timaeus and the ethics and politics of
the Republic join up.
However, for politics and mathematics to be commensurable, we do
not have to think that we learn directly about ethics from mathematics.
For example, it may be that there is a higher standard, ultimately the form
of the Good, of which both the human good and mathematical values
are instances. Commensurability would be ensured at a higher level than
both the human good and mathematics. As Christopher Gill has argued
(Gill 2007), the form of the Good, even understood as ‘the one’ or ‘unity’,
may be a conception that crosses different categories of entities that are
good, or one, in different ways. On such a reading, it may still be true that
ascending to the form of the Good requires, or at least is best undertaken
through, studying mathematics, perhaps by virtue of the paradigmatic
way in which mathematical entities exhibit unity and proportion. While
understanding mathematics may, then, represent significant progress
towards understanding the good, mathematics may not tell us about the
good at exactly the level of specificity at which it applies to human affairs.
Put differently, while we may learn something about justice, friendship
and harmony from mathematics in order to apply it to ethics and pol-
itics, we may have to generalise from the specifics of their mathematical
representation.10 This might also help explain why the philosopher cannot
simply, on the basis of having studied mathematics, return to the cave and
produce the just city, but needs to ascend all the way to the form of the
Good. In order to show the relevance of mathematics to politics and eth-
ics, and thus their commensurability, there is therefore no need to find
direct relevance in every detail of the mathematics.
9
With Schofield 2007: 221 n. 12, I use ‘the Cave’ for Plato’s image (unless it is qualified as such), and
‘the cave’ for the cave described by the image.
10
Compare the need at Symposium 211a–212a to survey all the different kinds of beauty, including
beauty in the sciences, in the ascent to the form of Beauty.
94 Thomas Johansen
The Cave follows the images of the Sun and Line. Socrates says that
we are supposed to apply the image of the Cave to the earlier images (R.
517a8–b1). As compared to the Sun, it is clear that the Cave too stipulates
two realms, one inside and one outside the cave, where the inside corres-
ponds to the perceptible world illuminated by the sun and the outside to
the intelligible world made knowable by the form of the Good. It seems
clear also that the Cave posits, at least roughly, four cognitive stages: look-
ing at the images at the back of the wall, looking at the visible objects that
cast those images, looking at images outside the visible world and looking
at the originals of those images. The Line similarly distinguished four cog-
nitive states, corresponding to reflections of perceptible objects (eikasia),
the perceptible objects themselves (pistis), mathematical objects (dianoia)
and forms (noēsis), in ascending order of clarity and truthfulness.
The Cave, then, integrates key features of the previous images. But it
also does something new. Socrates introduces the image as being about
the effect on our nature of education and lack of education. The Sun and
the Line are thus integrated into a story about how we may or may not
advance towards having knowledge. One immediate difficulty with under-
standing the Cave in this way, however, is that it seems to make us start
from too low a level. Socrates says that the prisoners are ‘like us’ (515a5).
If we are confined to the lowest level of education then we are in a per-
manent state of looking at reflections of physical objects. But this seems
absurd. For it tells us that we are not seeing concrete physical objects, but
merely their reflections. It is as if you were merely looking at them in a
mirror.
This problem has been widely noted in the literature on the Cave and
received different kinds of solution.11 One solution is simply to accept that
this is what Plato thinks about perception: when you are perceiving things
you are just seeing images. There were theories of perception in antiquity
according to which this was true. So Democritus, for example, held that
in perception we are seeing objects by their being reflected in our eyes.12
If one held a theory of perception like this, one might also hold that what
we are seeing are not physical objects themselves but merely their reflec-
tions. However, the image of the Line contrasts apprehending a physical
object directly (pistis) and merely seeing its reflection in reflective surfaces
(eikasia). So Socrates cannot very well be saying now that whenever we are
seeing we are merely seeing reflections on reflective surfaces.
11
See Joseph 1948: 34; Malcolm 1962: 42.
12
Cf. Aristotle, On Sense Perception 438a5–7; Theophrastus, On Sense Perception 49–50.
I think the answer is that the Cave is not about perception as such,
but about our state of understanding, the degree of development of our
intellect with respect to knowledge. The Cave, as Socrates said, is about
our state of education or lack thereof, which indicates primarily the state
of our intellect (nous) with respect to knowledge. It is our intellect that
is in the darkness of the cave, not our faculty of vision: indeed, Socrates
is explicit that the ‘eye’ by which we see in the cave is the intellect.13 And
there is after all plenty of visible light in our natural world: the compari-
son between the strong light outside of the cave and the weak, flickering
light inside, is a comparison between the conditions in which our intellect
is made to work.
Compare Socrates’ description of the activities of the prisoners: they
are concerned with discerning (kathoran)14 what the shadows are on the
back of the wall, and they give each other prizes for being able to predict
the future on the basis of which shadows have followed or coincided with
which in the past. That is to say, they are trying to answer the sorts of
questions a philosopher might try to answer by reference to things that
are in fact mere shadows. This sort of activity, no doubt, involves per-
ception: perception is still used – or misused, rather – by the prisoners’
intellect to settle the questions. The shift from looking at reflections to
beginning the ascent out of the cave is therefore not a simple matter of
switching from using one kind of faculty, perception, to using another,
the intellect. Rather, as Socrates puts it (518c), it is a matter of using one’s
intellect – one’s ‘eye’ – in a different way by turning it towards the sort
of object that can truly answer the questions that occupied the prisoners.
This object is, ultimately, the form of the Good, since it is the cause of the
being and coming into being of all things, including ultimately the being
and coming into being of the images inside the cave (517c3). But looking
at the statues is a move in that direction since these are more real than
their reflections (515d3).
Once we appreciate that our status as prisoners relates to our under-
standing of the world, we can begin to appreciate also why it might be
appropriate to think of us as being in a permanent state correspond-
ing to seeing reflections of physical objects. Our understanding of what
13
Socrates talks of the conversion of the whole soul, corresponding to turning around the whole
body (518c4–d1), but it is the intellect, corresponding to the eye of the body, that leads and the rest
of the soul that follows. See also the discussion by Sedley in this volume, pp. 80–2.
14
R. 516b8–d2. The verb was used earlier at 476d1 for those who could distinguish reality from appear-
ances; γνωματεύοντα, ‘distinguish’ at 516e8, is, similarly, a term describing cognitive success.
96 Thomas Johansen
constitutes human beings,15 animals or other such physical objects is at the
level of somebody who is trying to understand them merely on the basis
of their reflections.16 This is because we use what appears to us in percep-
tion to judge what the perceptual objects are, but what appears to us in
perception is no more indicative of what the perceptual objects are than
a reflection of a cow in water will show what a cow is. Imagine a biolo-
gist developing a taxonomy of animal species merely on the basis of out-
ward appearance or a comparative linguist providing etymology merely on
the basis of phonetic resemblance. Or, to use Plato’s example in Book 7,
an astronomer who studies the planetary motions merely on the basis of
what he sees.
This might still seem an overly pessimistic diagnosis of our epistemic
state. Surely we do have a basic grasp of physical objects. Everybody
grasps something about water even if they do not perhaps know it is H2O.
For example, everybody knows that water is liquid and transparent. To
call what we grasp about water a mere shadow of H2O seems unduly to
distance the properties we grasp from their underlying constitution. For
while being liquid and transparent may not be essential features of water
they are properties that follow from the essence of water, H2O, given a
certain temperature. I think we would say this even without taking into
account that H2O would still only be an image of the real Water, the form
of Water, for Plato.
At this point, we may think it appropriate to narrow the scope of the
reflections in the cave. So it may be that Socrates is not saying that we are
in an epistemic state of eikasia with respect to all objects, including water,
for example, but that he has in mind ethical and political values, notions
such as ‘justice’ and the other virtues. After all, the Cave presents an ascent
which will end in grasping the form of the Good and which will enable
the philosopher to implement justice and virtue in the cave. Socrates
talks, accordingly, of how the philosopher on his return to the cave will
be confused and unable to discuss with the prisoners about the shadows
and statues of justice (517d9). But once accustomed to the darkness of
15
Burnyeat 1999b: 238–41 emphasises the way in which the reflections inform the prisoners’ view of
what a human being is and so provide their self-image.
16
We might think that there is a further downgrading of our cognitive state implied by the Cave
compared to eikasia in the Line: reflections in water or mirrors (even allowing for the relative pov-
erty of ancient mirrors) can after all be quite precise with a fair amount of content represented –
otherwise Narcissus would hardly have fallen in love – while the flickering shadows of an object
represent at best the outline of an object with no internal content. We might say, therefore, that
the Cave puts the prisoners in a worse cognitive situation than somebody who has access to objects
merely through reflections in water and mirrors.
the cave, the philosopher will, in virtue of what he has learnt about the
form of the Good, be infinitely better at discerning the images of the fine,
the just and the good (520c1–7). There seem to be good grounds, then, for
thinking that Socrates’ concern in the Cave is specifically with our grasp
of ethical notions. It is our ethical understanding, then, which is at the
very low level of eikasia. And this seems not such an implausible claim,
given that justice, goodness and fineness are extremely difficult notions to
become clear about.17
There is, however, a narrower and a wider way of taking the claim that
the Cave is about our understanding of values. We may take the claim to
mean either that it is only about ethical and political values to the exclu-
sion of other kinds of entities. And this seemed to be the attraction of the
claim when we were considering the problem of saying that our under-
standing of objects such as water was just that of eikasia, given that we
surely grasp some real features of water, such as its transparency or liquid-
ity. For now we can forget about these cases and just focus on values,
since there seems no good reason why the prisoners should have grasped
anything about what goodness is, corresponding to the transparency or
liquidity of water. The difference between the two cases, one could say, is
that, while our everyday interaction with ordinary physical objects such as
water will give us at least some rudimentary grasp of some of its real prop-
erties – transparency, liquidity and so on – if one grows up in a corrupt
society such as late fifth-century Athens, there is no guarantee that one
will grasp any of the real features of justice and the other virtues.
However, there is also a wider way of taking the claim that the Cave
is concerned with values, which does not exclude the possibility that
it is also concerned with a range of other objects, including perceptual
objects such as water. This is because values are represented also by other
perceptual objects and because, in order to grasp these objects, in what-
ever way is appropriate to such objects, one needs to recognise the way
in which they represent those values. On this reading, values for Plato
extend beyond the human and social realm to the cosmos. If grasping
the goodness or beauty of a perceptual object is a necessary condition of
grasping what it is, we can therefore also see why the prisoners are ignor-
ant of standard perceptual objects. To use the previous example of water,
the prisoners may grasp that water is wet and transparent, but they have
no grasp of what it is about water that makes it good or beautiful. For this
reason, the lowly comparison of their state of understanding with eikasia
17
Cf. Euthyphro 7c–e.
98 Thomas Johansen
is appropriate. The wider reading of value shares, then, the same kind of
response to the problem of eikasia as the narrow reading of value.
We can sum up the three positions that have been mentioned here as
follows:
(a) wide value-neutral reading: the appearances in the cave are of all per-
ceptual objects;18
(b) narrow value-positive reading: the prisoners observe appearances
only of political or ethical entities, for instance justice, but not, for
instance, horses, tables and chairs;19
(c) wide value-positive reading: the prisoners are concerned with appear-
ances of all perceptual objects which fall under the same kind of
explanation as political and ethical entities.20
We have seen that (a) presents difficulties in terms of explaining the pris-
oners’ inferior epistemic state; (b) seems to fare better on that score; while
(c), preserving the concern with values, keeps this advantage. I want to
give some reasons now for preferring (c) over (b).
Consider, for a start, how Socrates described the form of the Good
as a cause of knowledge in the first two images. The image of the sun
told us that the Form of the Good is the cause of knowledge and truth,
without restriction (508e3–4). Moreover, the image made the form of the
Good the model for the sun and its cause,21 while the sun was the cause of
vision and becoming in the visible world: ‘the sun not only provides vis-
ible things with the power to be seen but also with coming to be, growth,
and nourishment’ (509b).22 But if the form of the Good is the model and
cause of the sun, and the sun is the cause of coming into being in the
natural world, then we may well infer that the form of the Good, at least
indirectly, will help to explain the natural world. Socrates seems then to
envisage an explanatory role for the form of the Good in relation to the
natural world, albeit a role mediated through the sun. In the image of
the Line, the first principle – not named as such, but surely the form of
18
Cf. White 1979: 186: ‘appearances presented to brute sensation.’ White’s proper interpretation is
more sophisticated than (a).
19
Cf. Malcolm 1962: 43: ‘The ordinary man is at L2 (the ability to recognise particulars) with respect
to ordinary objects but not with respect to value-particulars;’ cf. Irwin 1995: 276.
20
This is, broadly, the view of Burnyeat 1987: 227–34 and 2000.
21
507a3–4: τοῦτον δὲ δὴ οὖν τὸν τόκον τε καὶ ἔκγονον αὐτοῦ τοῦ ἀγαθοῦ κομίσασθε. Cf. 506e3–
4: ὃς δὲ ἔκγονός τε τοῦ ἀγαθοῦ φαίνεται καὶ ὁμοιότατος ἐκείνῳ. ‘Being the child of ’ clearly
means more than ‘being similar to’. The claim seems to be that the sun is similar to the Good itself
because the Good is its parent.
22
Translations of Republic throughout are by Grube as revised by Reeve, in Cooper 1997.
the Good – was presented as the principle of all knowledge. It is also, in
the first instance, the principle of the other forms, then of mathematical
truths, and finally the visible world insofar as it can be grasped by belief
(pistis) as an image of the intelligible world.
For the Cave to focus on ethical notions to the exclusion of all other
objects would, then, be surprising, at least if we think that what gives
knowledge of such ethical notions is ultimately the very same thing that
is the basis of all knowledge, the form of the Good. Socrates’ point seems
in the other two images to have been that if you really want to understand
things in general then you have to start with the Good itself. This should
already make us prepared to think that Socrates in the Cave could well have
the broader scope of appearances in mind while maintaining the emphasis
on value, because of the pivotal role of goodness in all knowledge.
The scope of the appearances inside the cave

What about the evidence of the Narrative of the Cave itself? Socrates
describes the objects whose shadows the prisoners see: ‘all kinds of arte-
facts, both statues of people and other animals’ (514c1–515a1). It is easy to
make sense of the reference to people on the narrow reading: what we are
concerned with in ethics are, after all, human values, the excellences that
tell us what a human being ought to be like. However, statues of human
beings (andriantes) are just one example, albeit a prominent one, of the
sorts of statues being carried. The other statues are referred to as ‘other
zōia’, which may refer to figures in general, not just figures of animals.23
However, given the distinction drawn with andriantes, it seems likely that
we should think primarily of representations of other animals, as indeed
translators commonly do.24 The scope of what the statues represent seems
deliberately wide, the figures being of all sorts (pantodapa, 514b9), but
with a focus on humans and animals.25 There seems, therefore, to be no
restriction on the kinds of figures that cast shadows on the back of the
cave wall; and specifically there seems to be no restriction compared to the
objects of eikasia in the image of the Line, where Socrates spoke of reflec-
tions of ‘the animals about us and all plants and the whole class of objects
23
LSJ quotes this passage (s.v. ζῷον II) as a case of zōion meaning figures in general, ‘not necessarily of
animals’.
24
So Jowett (‘and figures of animals’), Shorey (‘shapes of animals as well’, Grube/Reeve (‘statues … of
other animals’) and Waterfield (‘and animal models’).
25
Compare Socrates’ description of the materials that the statues are made of: wood, stone and all
other sorts of material (pantoia).
100 Thomas Johansen
made by man’ (510a). The generality of the scope seems a problem for the
‘narrow scope value reading’, which takes the reflections to be specifically
about human values.
One might retort that all the figures in the cave are skeuasta (515c2,
cf. skeuē, 514b), that is, artefacts, and so ought to belong specifically to
the human sphere. But neither man nor the other animals are artefacts,
or at least not human artefacts. The thought may be, then, that human
beings and animals are divine artefacts, a thought presaged perhaps by the
image of the Sun (507c–508b), with its talk of the sun as a god and cause
of living beings, and fully developed in the Timaeus.26 Alternatively, the
implication of skeuasta may be that, while more real than their reflections,
the statues are still less real compared to the real things outside the cave,
which would be phuteuta (cf. 597c), just as a statue of Napoleon is less real
than the living man himself.
Here is another detail that I think fits badly with the narrow scope
reading: the role of the fire inside the cave. It is one of the clearest points
of contact with the image of the Sun. When a prisoner is released, he sees
not just the statues but also the fire itself (auto to phōs, 515e1), which we
are told explicitly corresponds to the sun in the heavens (517b3). When he
leaves the cave, he sees reflections of human beings, other animals (516a)
and the planets (516b): as inside the cave, the reflections extend beyond
the human sphere to other living beings and the heavens. Finally, the phi-
losopher turns to the sun, that is, the form of the Good itself, at which
point he reasons about how this ‘provides the seasons and the years, gov-
erns everything in the visible world, and is in some way the cause of all the
things that he used to see’ (516b9–c2).
We need to look at the visible sun as a necessary part of the ascent.
Somebody might object that this claim takes literally what is a symbol
within a story. While in the image of the Sun it was clearly the sun in the
sky that was the analogue to the form of the Good, in the image of the
Cave notions such as a fire and the statues are not to be taken literally but
have to be decoded in view of the overall educational message. The point
about decoding is clearly right and the image of the Cave differs from
that of the Sun in this regard. However, as we have just seen (517b3), it is
26
If so, one could think with Sedley 2007a: 245 that the statue-handlers represent the other gods,
corresponding, perhaps, to the lesser gods in the Timaeus. Two issues to be resolved on this reading,
however, are first that we are told nothing in the Cave about the statue-handlers having made the
figures they carry, and secondly that the comparison of them with thaumatopoioi (514b5) may be
taken to imply their intent to deceive the prisoners. So, as Dimitri El Murr points out to me, the
term thaumatopoios is used in the Sophist (268d2) to refer to the sophist as such.
Socrates who offers us the decoding of the fire in the cave as its standing
for the sun. In relation to the image of the Sun, the Cave performs a kind
of analogical displacement: within the image of the Cave, what was in the
earlier image the sun in the sky is now the form of the Good, while the
fire in the cave has now taken the place of the sun in the sky. Similarly,
the animals inside the cave, which were living beings according to the Sun
and the Line, are now artefacts (skeuasta) compared to the living beings
outside of the cave.
Shortly afterwards, Socrates says (517c1–5):
Once one has seen [the form of the Good] … one must conclude that it is
the cause of all that is correct and beautiful in anything, that it produces
both light and its source in the visible realm, and that, in the intelligible
realm, it controls and provides truth and understanding, so that anyone
who is to act sensibly in private or public must see it.
What is striking about this passage is not just that Socrates offers light and
the sun as examples of the work of the form of the Good in the visible
realm, though this in itself shows that Socrates’ interest in the manifesta-
tion of goodness extends beyond the human sphere to the cosmos. What
is also remarkable is that Socrates in the same breath as he says that the
form of the Good produces the sun also says that the person who wants
to act well in private or politically must know the form of the Good. The
cosmos and the human sphere are thus set up as parallel expressions of the
power of the form of the Good. There could be no clearer indication that
we need to think of the human good alongside the cosmic good as both
manifestations in the visible world of the form of the Good.27
The fact that the Cave puts the sun alongside human goodness as a
manifestation of the form of the Good should be sufficient to refute the
narrow value-positive reading. But it does not show that understanding
the sun or its light or the seasons or any such thing is directly relevant to
our understanding of the human good. It could be that both are manifes-
tations of the form of the Good but that, in order to learn about either
of them, one has to turn to the form of the Good and reason about each
independently with reference to that form. The fact that both human vir-
tue and cosmic beauty are manifestations of the form of the Good might
appear only once we have grasped the form of the Good. So seeing the
connection between the two would not form part of the ascent from
the cave, but would rather serve as a confirmation that one had reached
27
Contrast Strang 1986: 29, who describes the fire in the cave as symbolising ‘a false and counterfeit
good’.
102 Thomas Johansen
the highest level of understanding of the good, from which all kinds of
goodness are seen as interconnected. In that case, we might maintain that
Socrates brings in the sun in order to illustrate the wide range of good
things produced by the form of the Good, and the way that it brings
about all beauty and correctness ‘for all’. Even so, we might further say,
the focus is still firmly on the human dimension, which can be explained
with reference to the form of the Good, independently of the cosmic
dimension. On this view, the concession made to the wide value-positive
reading would be relatively uninformative since the references to the cos-
mos would be detachable from the ethico-political account.
However, this suggestion divorces the role of the sun as a manifestation
of the good from its role as the source of light which illuminates the cave.
When Socrates says that the form of the Good brings about beauty and
correctness, he mentions light first of all and refers to the sun obliquely as
the source of light (517c3). Since the primary role of light in the image is
to illuminate the cave so as to create reflections and to make things visible,
we are surely, first and foremost, to think of the sun with reference to the
way it illuminates the cave so that we can see. But then, bearing in mind
that seeing in the Cave stands for understanding, it must also be the case
that we understand things in the visible world, including human affairs, in
a way that is somehow mediated by the sun’s light. This mediation could
be construed in a number of different ways; but I take it that fundamental
to any of them must be the idea that the sun somehow makes things in
the visible world intelligible by reference to what is good. It is as good that
the form of the Good spreads light in the intelligible world and the light
in the cave is somehow a weaker version of this light.
The sun as a source of insight

Can we, then, make sense of the idea of the sun’s light helping us to grasp
visible things insofar as they are good? The Timaeus offers a model. The
sun plays a key role in our cognition of the cosmos. So (Ti. 39bc):
God lit a light, which we now call ‘the sun’, to provide a clear measure of
the relative speeds of the eight revolutions,28 to shine throughout the whole
heaven, and to enable the appropriate living creatures to gain a knowledge
of number from the uniform movements of the same. In this way and for
this reason there came into being night and day, the period of the single
and most intelligent revolution; the month, complete when the moon has
28
Reading καθ’ ἃ in 39b3.
been round her orbit and caught up the sun again; the year, complete when
the sun has been round his orbit.29
Not only does the passing of night and day give us the notion of the
number one, the sun also by illuminating the other planets reveals to us
the numerical ratios of their revolutions. Moreover, we have been given
eyes, made of the same kind of fire as the sun – recall the description of
the eyes as ‘sun-like’ in the Republic – so that we can observe the heavenly
motions, learn about number and order, and regulate the periods of our
souls accordingly (Ti. 47ab). Observing the sun’s light is, then, the key
first move in grasping the order of the visible world.
If this model is applied to the Cave, it helps to solve a number of inter-
pretative problems. First, it helps to overcome a tension between two
aspects of the account. One of these is Socrates’ description of the cave
as the visible world,30 and his presentation of the sun in its capacity of
spreading light, which we know is a condition of visibility. The other is
his use of sight in the image of the Cave to stand for understanding, that
is, for exercising the capacity of the eye of the soul which only gets fully
realised outside the cave. The key point is that we are supposed to use our
eyes intelligently. The sun prompts us through its light to grasp by our
intellect the basic numerical proportions that order the visible world. The
purpose of the eyes themselves is fulfilled by identifying what the light
shows about the world, that is, its order.
Secondly, the model of the Timaeus gives us a grip on the difference
between the shadows and the statues in the cave. The statues are those
aspects of visible objects that we come to understand when we see them
properly in the light of the sun, that is to say, the way in which they rep-
resent goodness and beauty. The difference between the chained and the
released prisoner lies in the extent to which they grasp these values. In the
Timaeus, this is a matter of grasping the mathematical structure underly-
ing the visible object. The statues, then, would be the mathematical struc-
tures of the various objects, which constitute their underlying natures and
bring about their various appearances.31 Thus, Timaeus refers to several
cases where people observe perceptual phenomena but fail to understand
their mathematical structure.32 For example, he describes how we observe
29
All translations of the Timaeus are by Lee as revised by Johansen in Lee and Johansen 2008.
30
See, in particular, R. 532cd.
31
Timaeus suggests that the principles of these mathematical structures constitute the subject of a
higher study than cosmology (48cd, 53d) so, in identifying the mathematical structure of water, for
instance, there is no need to think that we have grasped the form of Water. We are still in the cave.
32
Karasmanis 1988 makes a similar point about the prisoners in the cave.
104 Thomas Johansen
the four ‘elements’ apparently turning into each other (49bc), in a way
that turns out to be misleading once we understand their mathematical
structure (54c). Another example is Timaeus’ discussion of the relative
movements of the planets (Ti. 40cd), which contrasts those who observe
the phenomena as expressions of mathematical regularities with those who
do not and who therefore fail to comprehend how and when they appear.
The latter are the people who Timaeus thinks will be reincarnated as ani-
mals (Ti. 91d–92c). Like the prisoners in the cave, they fail to analyse by
reason what it is they observe, and they too try, like prophets, to predict
the future (see 40d, with R. 516c) without knowing its causes.
Finally, the Timaeus model offers us a way of explaining the ethical
relevance of cosmology. We become better human beings, more orderly,
more temperate, by imitating in our souls the motions of the heavenly
bodies (Ti. 47c). While Timaeus falls short of saying that cosmology is
the only route by which we can become good, he is explicit in saying that
we should study cosmology to this end (Ti. 90d). The ethical relevance of
the heavens is, of course, familiar not just from the Timaeus but also from
Gorgias 508a, where Socrates criticises Callicles for ignoring the values rep-
resented by the kosmos in his advocacy of trying to get more than one’s
share (pleonexia). The message is clear: there are direct lessons to be learnt
about how to live one’s life from studying the cosmos.
In the Republic, apart from the Cave, we find this thought most explic-
itly expressed when Socrates says in Book 9 (592b) that there is a paradigm
(paradeigma) of the just city laid up in heaven for the wise man to imitate.
Socrates is here referring to the way in which the heavenly bodies display
the same kind of order (politeia) that defined justice in the city and in the
individual soul.33
However, one reason why one might still resist the assimilation of the
Cave to the Timaeus – and to the Gorgias, which presents the cosmic order
as a subject of geometry (508a) – is, again, the absence of any explicit math-
ematics in the Narrative (514a–521b). The Timaeus relies on a specifically
mathematical conception of the cosmic order which is to be reproduced in
the human soul, the sun’s light having a special role because it teaches us
about number. But there is no indication within the Narrative that we have
to learn mathematics. There is a reference to ‘the seasons and the years’, as a
paradigm of the good work of the form of the Good, and we saw Timaeus
refer to ‘day and night, the months and returning years, the equinoxes and
solstices’ as the phenomena from which we learn about number. Although
33
See Burnyeat 1999a: 781.
what matters for Socrates about these phenomena is, clearly, that they illus-
trate beauty and correctness,34 there is no indication that these values must
be conceived mathematically. So the crucial detail in the Timaeus account
that links the human sphere with the cosmos seems missing from the Cave.
It is, of course, true that there is no mathematics in the Narrative.
Rather, as Burnyeat said, it is only in retrospect, in the Commentary
(521c–541b), that the relevance of mathematics becomes explicit. However,
there is a reason for this absence, namely that the prisoner’s release and
ascent from the cave is not described in the Narrative from the point of
view of Callipolis. Thus, the perspective of the prisoner is not that of some-
one being educated through the mathematical curriculum of Callipolis. So
the violent pulling (helkoi … biai, 515e4) of the prisoner in the cave con-
trasts with the gentle pulling (hērema helkei, 533d2) of the student being
taught mathematics and dialectic in Callipolis. Notice also that the release
of the prisoner happens suddenly (exaiphnēs, 515c6, cf. 516a4), while the
student in Callipolis enjoys a ten-year education in higher mathematics,
on top of the introductory mathematics he learnt as a child (536d). The
release of the prisoner is much more reminiscent of the abrupt shock of
somebody bumping into Socrates in the market place and being exposed
as ignorant,35 than it is of systematic schooling. The prisoner’s ascent is
thus characterised by painful resistance (515d–516a). Part of the reason for
this resistance is, no doubt, that the prisoner in the unjust city is weighed
down by recalcitrant irrational desires, while the aspiring philosopher in
Callipolis has already been emotionally primed for the ascent.36 But there
is the further crucial difference that there is no educational support in the
unjust city whereby the conversion can best be effected. The difference in
the modes of ascent is also what calls for different arguments to be used
to persuade the prisoner, who owes nothing to the unjust city, and the
philosopher, who owes his education to Callipolis, to return to the cave
(520bc). But the difference in mode of ascent should not overshadow the
point that the stages of the ascent are the same: whoever we are, prisoner
in the unjust city or budding philosopher in Callipolis, we have to liberate
ourselves from perceptually based judgements about the nature of things,
and turn to a broad rational inquiry into the representations of value,
including the cosmos – an inquiry that can only be satisfied fully through
an understanding of the form of the Good. Schofield is then right to
say that there is a fundamental difference between the Narrative and the
34
Cf. ὀρθότητος at Ti. 47c2. 35 See in particular R. 516d.
36
See the pertinent remarks by Gill, p. 113 below.
106 Thomas Johansen
Commentary in how the ascent is portrayed. But the difference lies not
in the incommensurability of the ethical and the mathematical. Rather, it
lies in the absence of the mathematical curriculum from the unjust city; it
is only that curriculum that shows how the same ascent is best conducted.
Hence, this ascent will only happen spontaneously (automatoi, 520b) or
by some divine chance in the unjust city, whereas in Callipolis it will hap-
pen systematically and deliberately.
Cosmology in the philosopher’s curriculum

This brings us to the objection, raised earlier, that the Timaeus’ concep-
tion of cosmology is at odds with what Socrates says about the proper use
of astronomy in the Commentary. In the Timaeus, we are supposed to
study the motions of the observable planets, but Socrates tells us to leave
the things in the sky alone in favour of the study of the purely intelligible
motions of geometrical solids (530bc). Astronomy is meant to draw us
away from coming-into-being towards changeless being, that is, to get us
out of the cave: so is Timaeus-style attention to the visible heavens not
ruled out as a part of the ascent within the cave? It is, of course, true
that Timaeus, as we have seen, wants to calculate the invisible motions
of the heavenly bodies and wants us to assimilate our intellects to these.
Nonetheless, the motions are still those of the visible planets and so they
fall foul of Socrates’ injunction.
The first thing to say is that Socrates wants mathematics not just to get
us out of the cave, but also to perform the first stages of the ascent within
the cave. So at 532b6–d1, Socrates says:
Then the release from bonds and the turning around from shadows to statues
and the light of the fire and, then, the way up out of the cave to the sunlight
and, there, the continuing inability to look at the animals, the plants, and
the light of the sun, but the newly acquired ability to look at divine images
in water and shadows of the things that are, rather than, as before, merely
shadows of statues thrown by another source of light that is itself a shadow
in relation to the sun – all this business of the crafts we’ve mentioned has the
power to awaken the best part of the soul and lead it upwards to the study
of the best among the things that are, just as before, the clearest thing in the
body was led to the bright thing in the bodily and visible realm.
As Burnyeat has pointed out,37 the passage strongly suggests that the
mathematical sciences, as Socrates has described them, have the power not
37
Burnyeat 1987: 227 n. 37.
just to lead the soul out of the cave but also to release the soul from its
chains in the first place, for the two are presented as a continuous process
brought about by ‘the crafts’. There is additional evidence at 524c. Here,
Socrates distinguishes between perceptions that provoke us to think about
what the thing perceived is – say, what it is to be one or two – and those
perceptions that do not. The example of the former is the perception of
something as one, and the example of the latter is the perception of a fin-
ger (524d10–525a5). Since this inquiry is supposed to arise out of percep-
tual experience and force us to think about what the numbers are, it makes
sense to say that this happens at the lowest level of the cave. Also, the lan-
guage of forcing and pulling the soul to turn around in the direction of
being in a way that induces aporia is strongly reminiscent of the descrip-
tion in the Narrative of the prisoner’s release from his chains, although, as
I noted earlier, the prisoner is dragged violently and the mathematician
gently (R. 515c6–d7). The turning around of the soul is clearly meant to
begin at the first stage within the cave but, if so, it must be the case that
the mathematical thinking begins already at this stage.
Is there, however, any indication within the mathematical curricu-
lum that attention should be paid to the visible world? Socrates clearly
deplores the kind of empirical astronomy that simply describes the specta-
cle of the heavens (529b). Even so, he finds a use for observing the heavens
(529d7–530b4):
Therefore, we should use the embroidery in the sky as models (para-
deigmata) in the study of these other things [i.e. the eternal motions]. If
someone experienced in geometry were to come upon plans very carefully
drawn and worked out by Daedalus or some other craftsman or artist, he’d
consider them very finely executed, but he’d think it ridiculous to examine
them seriously in order to find the truth in them about the equal, the dou-
ble, any other ratio … Then don’t you think that a real (tōi onti) astronomer
will feel the same when he looks at the motions of the stars? He’ll believe
that the craftsman of the heavens arranged them and all that’s in them in
the finest way possible for such things. But as for the ratio of night to day,
of days to a month, of a month to a year, or of the motions of the stars to
any of them or to each other, don’t you think he’ll consider it strange to
believe that they’re always the same and never deviate anywhere at all or
to try in any sort of way to grasp the truth about them, since they’re con-
nected to body and visible?
Socrates, crucially, says that the admirer of the heavens is a geometer;

that is, his admiration is not based simply on visible properties, as in the
case of the star-gazer, nor, for that matter, is it based simply on general
108 Thomas Johansen
considerations of good order. Rather, it is an appreciation of the degree
to which the heavenly bodies illustrate the intelligible ratios on which the
‘real’ astronomer focuses.38 We should admire the way that the heavens
exemplify mathematical order, while realising that what we see falls short
of instantiating it fully.39 The visible heavens offer paradigms by which to
think about good order. The passage fits well with the final lines of Book
9 (592b), where Socrates, as we saw, recommends using the heavens as a
paradigm (paradeigma, as here) of good order. But it also fits the details of
the Cave, where the one who ascends has to look at the sun.
One might think that what is valuable about looking at the planets here
is just that it reminds us of what we know anyway from real astronomy.
This does not do justice to the term paradeigma, where the term, as often,
has a normative dimension: a paradeigma tells us how to think correctly
about something, by offering a particular case by which to grasp the gen-
eral.40 But one further sign that Socrates takes the role of the senses seri-
ously in appreciating the beauty of the heavens is his remark that the eyes
have been designed for astronomy, just as the ears have been designed for
harmony (530d). If there was no positive use for observation in astronomy,
the teleological function of the eyes, which the image of the sun called
‘god’s work’ (508a), would go unfulfilled. The Timaeus (47ab), as we saw,
has much to say on this point.
These are modest concessions to the study of the visible heavens, and
we should expect nothing else, given Socrates’ concern with how we can
get out of the cave. Nonetheless, this marks the appreciation of the math-
ematical order of the heavens as a stage in the ascent. In the Timaeus, on
the other hand, the ascent stops there: we are told nothing about pure
astronomy and nothing about how to use astronomy to ascend to the
study of being. However, despite their different emphases, in making the
claim that humans can and should learn about being good from the heav-
ens, the two works are of a piece.
There is one final worry about the assimilation of the two works. Even
if the ascent from the cave shows the ethical relevance of mathematics,
does the Timaeus not represent an increased mathematisation of ethics,
compared with the Republic? In the Timaeus, it seems, one can become
good without ascending to the forms by reproducing in one’s soul the
38
Another instance of Plato’s extravagant punning on to on and its cognates in the Republic: a real
astronomer is an astronomer of what is (to on), not of what comes to be.
39
There is no tension between these two points: an illustration may well make you realise what some-
thing is without its being an exact copy.
40
Cf. the use of paradeigma in historiography, e.g. Thucydides 2.37.1.
mathematical order of the visible heavens. But the Republic may, as I have
suggested, only be committed to the view that one has to learn about
mathematics in order to learn about the form of the Good, after which
one can become good. Also, as Gill has argued, it is not clear that the form
of the Good is itself a mathematical concept. The paradigmatic role of the
heavens guides us towards an understanding of the form of the Good, but
we may also have to abstract from the specifically mathematical character
of this paradigm. No doubt, talk of ‘becoming good’ is too simple here.
In the Timaeus, while cosmology may make us just (42a8), sensible (44b7)
and measured (59d), the higher intellectual virtue that belongs to nous
requires the study of forms (51e), which constitutes an exercise distinct
from cosmology (59cd). It is also clear that grasping the mathematical
structure of visible objects is not nearly enough to enable us to understand
why they are so structured (48b–d). There are, then, indications that cos-
mology has to ascend to the study of the forms for us to attain the sort of
knowledge of explanatory principles characteristic of nous. While math-
ematical cosmology may, in both works, get us some of the way towards
being good, it remains an open question whether the knowledge ultim-
ately required to be fully virtuous is mathematically based.
ch apter Si x
Reflective commentary (1)

‘Socratic’ psychology in Plato’s Republic
Christopher Gill
In this discussion, and in the companion piece by Dimitri El Murr, the

objective is to explore interpretative and philosophical issues raised by
other chapters in the volume and by Christopher Rowe’s recent work.
Our aim is to give the volume a more dialectical character, by respond-
ing to claims made in the volume and in related work by Rowe, and by
underlining their larger implications for current scholarship on Plato.1
The claim considered here is that we can find in the Republic the continu-
ing presence of ‘Socratic’ psychology. A related claim is that the ‘Socratic’
strand in the psychology of the Republic is, in some sense, more philo-
sophically profound or fundamental than the other strand or strands in
the Republic. These claims are made in David Sedley’s chapter in this vol-
ume, and by Christopher Rowe in Rowe (2007a) (especially ch. 5) and
(2013).2
In broad terms, their claims form part of a reaction against a very wide-
spread feature in Platonic scholarship, especially English-language schol-
arship, in recent decades. This feature is the assumption that we can trace
within the early and middle Platonic dialogues the gradual replacement of
‘Socratic’ ideas on psychology, ethics and epistemology or metaphysics with
‘Platonic’ ones. A linked assumption is that Plato sometimes signals to the
reader the stages of this process of replacement. In the interests of brevity,
I will refer to scholarship shaped by these assumptions as expressing the
I am most grateful to Christopher Rowe and David Sedley for reading an earlier draft of this essay and
clarifying their positions (though, of course, I am responsible for any remaining misinterpretation of
their views) and for their willingness to have their discussions probed in this context. I have also taken
account of the comments of an anonymous reader for Cambridge University Press.
1
Cf., to some extent, Gill 1996a, an ‘afterword’ to Gill and McCabe 1996, which both discusses some
of the claims made in the volume and offers its own view of the core theme of that volume.
2
Although the views of Sedley and Rowe on these topics are not identical, they are, I think, suffi-
ciently close to make it useful to take them together in this way. Where my comments apply more
specifically to the work of one or other scholar, this is made explicit in the course of the discussion.
110
‘Socratic’ psychology in Plato’s Republic 111
Standard Development View (SDV).3 There have also been a number of
challenges to the SDV, which aim either to revise previous accounts of the
developmental process or, more fundamentally, to question whether this
process of the replacement of ‘Socratic’ with ‘Platonic’ ideas can actually
be found in the dialogues.4 Rowe (2007a) is centrally concerned to mount
this challenge, and this is continued in Sedley’s chapter in this volume and
Rowe (2013). I am in sympathy with their challenge to the SDV, which,
I think, rests on a number of highly questionable assumptions about the
form and philosophical objectives of Plato’s dialogues. I am concerned
here to probe two aspects of the discussions by Rowe and Sedley. One is
the specification of what should count as a ‘Socratic’ psychological idea.
The other is the question how we should understand the relationship
between ‘Socratic’ and ‘Platonic’ strands within the psychological thought
of the Republic. On both topics, as in some other recent treatments, I will
stress the importance of locating Platonic ideas firmly in the dialectical
context of the specific dialogue in which they occur.5
First, I review the main aspects of the Republic taken by Rowe and
Sedley to express ‘Socratic’, as distinct from ‘Platonic’, thinking. One is
the passage in 611b–612a, which raises the question whether the ‘true’
psyche is the tripartite, complex one presented in Books 4, 8–9, or a uni-
fied one, characterised solely by its love of wisdom. Rowe sees here at
least a strong suggestion that the unitary (Socratic) psyche is the true
one, while Sedley sees it as highly significant that the question is raised in
such a prominent way near the end of the dialogue.6 This passage is taken
along with several in the central digression7 which bear on the psycho-
logical state of the philosopher. One passage (485de) presents desire as a
motivational ‘flow’ (rheuma), which is wholly modifiable or ‘plastic’, and
which may be directed equally towards philosophical learning or bodily
pleasures. Another passage (518b–519b) presents education – at least edu-
cation towards wisdom – as a matter of turning round the mind, or ‘eye’
of the psyche, and with it the psyche as a whole, towards philosophy and
3
A hugely influential work in promoting the SDV is Vlastos 1991, esp. chs. 2–3. A developmental
approach is often adopted in the study of Socratic–Platonic psychology, e.g. by Lorenz 2006: ch. 3,
esp. 28–30, Brickhouse and Smith 2010: ch. 7.
4
See e.g. Kahn 1996 (which accepts the idea of development but aims to revise the analysis of the
developmental process), and Annas and Rowe 2002, which examines the underpinning assumptions
of the developmental approach, and offers a number of alternative approaches.
5
For this view, see Gill 2000a, 2002a: 153–61, 2006 and 2013.
6
See further Rowe 2007a: 140–1, 165, 170–2; Sedley, pp. 88–9 above.
7
The digression consists of Books 5–7 in the familiar ten-book division, though the part of this sec-
tion most relevant for the present topic begins in 471c (the third of the three ‘waves’ of paradox in
Book 5). In the six-book division, as Sedley notes, Book 4 begins at 471c.
112 Christopher Gill
away from other objectives, including sensual pleasures. These passages
are seen as implying a more unified, or more purely ‘intellectualist’, con-
ception of human psychology than is presented in the tripartite accounts
of the psyche in Books 4 and 8–9.8 In a related point, Rowe has argued
that Socrates’ argument with Thrasymachus in Book 1 implies a strongly
‘intellectualist’ conception of virtue, which he also takes as a further indi-
cation of a persistent ‘Socratic’ strand in the Republic.9 These passages
are taken as indications that Plato still adheres to ‘Socratic intellectual-
ism’, that is, as Sedley puts it, ‘the thesis … that all moral states can be
analysed as intellectual ones: every soul aims for the good, and virtuous
people are those with the requisite understanding to achieve this univer-
sal end’ (pp. 76–7 above).
The first question I want to raise about these suggestions by Rowe and
Sedley is this: do the passages they cite really mark a reversion to ‘Socratic’
psychology? Of course, the answer depends on what sense we give to that
notion. But I think that on one, rather common, view about what this
means, these passages do not express Socratic psychology – at least not
wholly or straightforwardly. Let me take what is often seen as illustrating
Socratic psychology, the final argument in Plato’s Protagoras (349b–360e).10
The argument, very broadly, is designed to show that courage, like the
other virtues, is a form of knowledge, specifically knowledge about what
is and is not to be feared. In the course of this argument, the idea of being
‘overcome’ by pleasure or pain is re-analysed in terms of a failure of cal-
culation, or a failure to acquire the art of measurement (352d–357e). Does
this argument reflect the kind of ‘Socratic’ position that is seen by Rowe
and Sedley as reaffirmed in the features of the Republic noted earlier? To
some extent, the answer is, clearly, ‘yes’. The Protagoras argument purports
to show that all people, cowards and courageous alike, aim at what they
think is good, and that the difference in their mode of behaviour reflects
8
See Sedley, pp. 79–81, 83–4 above; also Rowe 2013.
9
See Rowe 2007a: 186–97, referring esp. to 334b–335e, 349a–350c, 352d–353e, also Rowe 2010. Rowe
2013 argues against the view sometimes taken that Book 1 of the Republic is a distinct (or earlier)
work because of its ‘Socratic’ character. I will not pursue the question of Book 1 here (but cf. the
doubts about the separateness of Book 1 expressed also by Sedley, p.73 above).
10
My brief comments on the Protagoras argument are meant to reflect obvious and widely agreed
features; for a recent reading, summarising current interpretative debate, see Brickhouse and Smith
2010: 70–88. The psychological side of this argument is generally taken as Socratic, even though
there is no agreement about whether the hedonism is Socratic or not (353c–354e). Penner and Rowe
2005 take the Lysis, rather than this argument in the Protagoras, as the prime example of Socratic
intellectualist psychology. For a full re-examination of the nature of Socratic intellectualism in
psychology, including a critique of Brickhouse and Smith 2010, see Rowe 2012b.
their failure or success in gaining this kind of understanding.11 To this
degree, the argument matches the criteria of ‘Socratic intellectualism’ in
the comment just cited by Sedley.
On the other hand, the Republic passages fail to display certain other fea-
tures that are strikingly present in the Protagoras argument, which can be
characterised as universality and holism. The Protagoras argument is expli-
citly presented as applicable to all people: for instance, as just noted, both
cowards and courageous go towards things they are confident about (which
do not inspire fear), although these are opposite things. Similarly, Socrates’
re-analysis of what it means to be ‘overcome’ by pleasure or pain, in terms
of the failure of the art of measurement, only works if the account of
human psychology assumed applies universally (359c–360d, cf. 356b–357a).
On the other hand, most of the passages in the Republic cited earlier refer
only to the psychological state of an ethical or intellectual elite, defined by
contrast with the condition of most people. The central digression in the
Republic (at least after 471c) is focused on the characterisation of the natural
qualities and advanced education of this elite, and the passages noted illus-
trate their special psychological state. Although the idea of the plasticity of
desire (R. 485de) could, in principle, apply universally, it is, in fact, only
applied to the kind of redirection of desire displayed by the ‘true philoso-
pher’ (485e1). The idea that education represents the turning around of the
psyche is, similarly, only applied to the redirection of the mind, the ‘eye’ of
the psyche, and the kind of (intellectual) education involved is explicitly
contrasted with the more habituative kind that produces other qualities.12
Sedley, admittedly, does see in this passage a universal model of ethical edu-
cation, based on a body–intellect dualism and intended to replace the tri-
partite psychology and the type of education linked with this. But I think
the focus here is narrower, and one that continues the theme (dominant in
484a–504a) of finding the right kind of education to ensure the philosoph-
ical nature is not corrupted, a view discussed more fully later.13
The Book 10 passage (611b–612a) raises rather complex questions in this
connection; my comments here are supplemented by points made later
in this essay. Socrates presents the ‘true’ nature of the psyche (most easily
11
See esp. Prt. 358b–360d, including the point that all people go towards what they see as good;
hence cowards and brave alike go towards what they are confident about (what they see as good),
even though these are different things, and they do so because of their knowledge or ignorance of
what is good.
12
R. 518b–519b. The contrast with habituative education (presumably, the first phase of education, for
auxiliaries and potential guardians) is marked in 518d9–e2.
13
See Sedley, pp. 80–1, 83–4 above, referring to 519ab; see also ‘in the psyche of each of us’ (518c5). For
my view, see further p, 115 below.
recognisable if considered in abstraction from the body) as unified and
purely rational or intellectual. He contrasts this nature with the heavily
encrusted, multiform (tripartite) condition of the psyche as it manifests
itself in its embodied form. Does this passage make – what we might call –
a metaphysical statement, about the nature of psyche as such (for every-
one) or an ethical one, about the ideal psychic condition which we should
take as our goal, though we may not achieve it? As often in Plato’s writings,
this distinction cannot be drawn sharply. But, in part at least, the passage
offers a new formulation of an ethical ideal, that of psychic unification (or
harmonisation), which is fundamental to the argument of the dialogue as a
whole.14 Although this ideal is often presented, as here, as applying in prin-
ciple to everyone, the argument as a whole also strongly suggests that it will
only be realised, if at all, under very restricted and exceptional conditions.
The philosopher-ruler, by this stage in the argument, has come to represent
this ethical ideal of psychic unity;15 and the philosopher-ruler’s achievement
of this psychic state is presented earlier as dependent on a complex set of
factors, including the combination of exceptional natural ability and an
elaborate two-stage educational programme.16 Hence, although this psy-
chic state is presented as the ‘true’ one for everyone, in practice it seems
inevitable that in life – and maybe also after death – it is only those who
come closest to this ideal philosophical condition who have much hope
of achieving the ‘true’ state.17 This is a very different kind of ‘universal-
ism’ from that implied in the Protagoras argument, according to which all
people, whatever their ethical character, act ‘rationally’ in the sense that
they act and feel according to their beliefs and understanding. Hence, if
both passages express a ‘Socratic’ approach to moral psychology, the notion
of ‘Socratic’ is being used very differently in each case.
A similar point can be made if we consider a second striking feature
of the Protagoras argument, namely the unified or holistic character of its
psychological assumptions. The argument purports to explain the overall
psychological state of the people concerned, including their fears or con-
fidence in situations of danger. And it does so by postulating a close link-
age between beliefs and reasoning and emotions; hence, one’s effectiveness
14
For other formulations, see e.g. R. 443d–444a, 591b–d. For an ‘ethical’ reading of R. 611b–612a, see
Rowe (2007a): 170–2, summarised in text to n. 21 below.
15
See e.g. the climactic contrast between the happiness of the ‘king’ (i.e. ‘philosopher-ruler’) and the
tyrant in R. 580b–c, supplemented by arguments based on the superior pleasurability of a psychic
state dominated by the rational (philosophical) part (586e, cited in text to n. 27 below). For com-
ments highlighting the psychic disunity of other types of person, see R. 554c–e, 561e, 575a.
16
See further text to nn. 23–6 below.
17
The following myth of Er suggests that psychic disunity (injustice) in life persists after death (e.g.
618b–620a).
in using ‘the art of measurement’ to challenge ‘the power of appearances’
determines whether or not one feels fear or confidence in situations of
danger (353a–360e). Although the passage is not formulated as a general
theory of human psychology, it implies a view which is similar to Stoic
thinking or modern ‘cognitive’ theories of emotion, in assuming that
emotions and desire are shaped by beliefs and reasoning, both overall and
in specific situations.18
Is the same view also presented or implied in the passages in which
Rowe and Sedley find a distinctively ‘Socratic’ approach in the Republic?
There is a holistic dimension in two of the passages: 485de, which presents
desire as a uniform flow (rheuma), which is modifiable in its objects,
and 518c–519b, which presents education as a matter of turning around
the mind along with ‘the whole psyche’ towards what is ultimately real,
namely the good (518c). However, as just noted, this process is seen as
operating not universally but only for the ethical-intellectual elite (poten-
tial philosopher-rulers). Also, the stress is solely on one aspect of psycho-
logical experience, intellectual activity or philosophising, described in 518c
as the function of the ‘eye’ of the psyche, the mind or intellect. Similarly,
in 611b–612a, philosophy or ‘love of wisdom’ (philosophia, 611e1) is pre-
sented as the key factor that enables us to understand the ‘true’ nature of
the psyche and what underpins its essential unity. In the Protagoras argu-
ment, by contrast, what is at issue is a different, and in some ways broader,
range of psychological phenomena (the emotions of all types of people
in situations of danger). In both cases, arguably, the theory assumed is
‘rationalist’ or ‘intellectualist’, in that it presents reason shaping emotion
or desire, rather than emotion or desire shaping reason, or each function
operating separately. But the Republic passages seem to be ‘intellectualist’
in a rather stronger sense than the Protagoras argument, since they are
focused solely on intellectuals and on the workings of the intellect.19
If we take the Protagoras argument as a paradigmatic, or at least exem-
plary, case of ‘Socratic’ psychology, how does this affect the Rowe–Sedley
view that there is a strongly ‘Socratic’ strand within the Republic? The
outcome of the comparison is ambiguous. You might conclude that the
Republic shows the same Socratic assumptions (those summarised by
Sedley) but directed towards a different objective, namely characterising
18
On Stoic cognitivism regarding emotions, see Brennan 2003; on modern cognitivist theories, see
e.g. Deigh 2010: 26–32; Helm 2010: 304–6.
19
At least, this is how I read these passages, although (as noted at p. 113 above) Sedley sees at least
some of them as implying a more universal psychology. Rowe also suggests that the psychology
is universal, in the sense that the ideal philosopher is, for Plato, a paradigm to which all human
beings should aspire (see 2007a: 273–6 and refs. in n. 6 above on R. 611b–612c).
certain aspects of the psychological life of an ethical or intellectual elite
rather than offering a picture of human psychology that applies univer-
sally. Or you might question whether there is enough philosophical con-
tent in common to justify classifying them both as ‘Socratic’ – especially if
this is taken to imply a contrast with a distinctively ‘Platonic’ psychology.
The comparison also raises more general questions, which are significant
both for the Rowe–Sedley approach and the SDV, which they challenge.
What criteria do we use for determining what counts as ‘Socratic’ ideas
and, in applying these criteria to specific Platonic passages, how narrowly
or how broadly do we define what is supposedly ‘Socratic’ in these pas-
sages? A yet more fundamental question is this. What is the evidence that
Plato (as distinct from modern scholars) is interested in the question of
what counts as ‘Socratic’, rather than ‘Platonic’, in his dialogues?20 On the
face of it, the distinctive form of the Platonic dialogues, in which the fig-
ure of Socrates is used as the vehicle for a whole series of sharply different
ideas and dialectical methods, seems calculated to frustrate, rather than
satisfy, the desire to answer this question in any definitive way. A second
general question bears more specifically on the Rowe–Sedley approach.
The main thrust of their interpretation is to question or nuance the sharp
distinction between ‘Socratic’ and ‘Platonic’ sometimes drawn by schol-
ars influenced by the SDV. But in that case, is it still useful to deploy
this distinction at all? Should we not accept that the arguments in Plato’s
dialogues represent a chemical fusion of ideas and methods, to which the
historical Socrates was surely a key contributor, but one whose precise con-
tribution can never be defined exactly, in part because Plato himself was
not fundamentally interested in that question? Perhaps it would be better
to abandon the use of the ‘Socratic–Platonic’ distinction in our readings
of Plato, and to focus on the only solid distinctions that Plato himself pro-
vides, that is, between different dialogues, such as the Protagoras and the
Republic, and the arguments these dialogues offer.
I return to these general questions later. First, however, I take up the
second topic on which I want to probe the Rowe-Sedley approach, the ques-
tion of the relationship between (what they see as) Socratic and Platonic
strands within the overall thinking of the Republic on ethical psychology.
As Rowe (2013) stresses, it is crucial that there should be an intelligible rela-
tionship between these strands if the core argument of the dialogue that
20
In modern discussions, massive weight is placed on some passing comments by Aristotle on this
distinction (see e.g. Vlastos 1991: ch. 3); but indications in Plato are far more open to dispute. R.
438a is standardly taken as an outright rejection of ‘Socratic’ psychology; but this reading is conjec-
tural and open to challenge (see some pertinent comments in Ferrari 2007: 168–9).
justice constitutes happiness is to be a coherent one. Rowe and Sedley dis-
cuss various possible forms of resolution; the common element is that the
‘Socratic’ strand is seen as providing the more profound or fundamental
insights, and the ‘Platonic’ one is in some way subordinate or secondary.
For instance, in his book Rowe suggests that R. 611b–612a indicates
that the unitary (purely intellectual) psyche is our ‘true’ one, whereas the
accounts of the tripartite psyche show the result of people making them-
selves divided by giving way to competing motives and thus treating their
psyche as if it were fundamentally divided. At first sight, the passage seems
to be focused solely on the contrast between the discarnate (unified) and
embodied (tripartite) condition of the psyche. But Rowe also sees the tri-
partite psychological state as one that is presented as a mark of personal
failure, or as a reflection of the social pressures that sway our judgements.21
His most recent discussion (Rowe 2013) goes rather further in seeing com-
patibility between the two strands. Here he focuses on the question of the
relationship between two accounts of justice in the dialogue, a ‘Socratic’
formulation in terms of knowledge, and a ‘Platonic’ one in terms of the
harmony or cohesion between the parts of the psyche or polis that enables
them to do their own job. Rowe sees these conceptions as appropriately
applied to different types of people in the ideal polis: the first only to
(fully educated) philosophers and the second to well-trained auxiliaries.
The two definitions are compatible to the degree that knowledge informs
the psychological state of both types of figure, even if only the philosopher
actually possesses knowledge for herself. Rowe gives special attention to a
passage in Book 9 (590cd) which stresses that it is better for people to be
ruled by properly constituted reason, even if this is imposed from outside,
than by their own mistaken judgements. This suggests to Rowe that the
two kinds of justice can have a partly common structure or that there can
be degrees of (the same kind of ) justice. Both forms of justice constitute
expressions of knowledge; but in the inferior or secondary type, the know-
ledge is supplied from outside rather than grasped directly by the person’s
mind. The harmonisation of the parts of the psyche and of the political
structure of the ideal state (with which the definitions of justice are closely
associated in Book 4) form the necessary preconditions of this secondary
kind of justice. The partial similarity between the two forms of justice
enables us to say that there is a single subject under debate in the dia-
logue. But, contrary to what is widely supposed, the argument does not
21
See Rowe (2007a): 170–2. On this passage in R. 10 see also text to nn. 14–17 above.
depend on the replacement of Socratic claims about the virtues and about
psychology with Platonic ones.
Sedley makes some partly parallel suggestions about how to reconcile
the two distinct strands that he sees as present in the Republic. Like Rowe
in his most recent treatment, Sedley thinks that Plato operates with two
kinds or levels of virtue, and that the two psychological theories (intellec-
tualist and tripartite) are linked with one or other of them. Although the
tripartite theory provides the best possible analysis of most human psy-
chological and ethical states:
[it] is of no particular help for understanding the moral and intellectual
mindset which lifts the philosopher to a higher and more godlike plane.
For what marks the real philosopher is not a reasoned control of irrational
and semi-rational drives, but ascent to a level of understanding where the
demands of the carnal world simply fade into the background … and have
no more than the bare minimum of motivational purchase. (p. 87 above)
Although Sedley, in this respect, sees both psychological accounts as
presented by Plato as valid, and as compatible, he sees certain ways in
which the intellectualist or Socratic view is given greater weight. This is
partly because the Socratic theory is linked with the ideal philosopher,
who functions as the paradigm of the virtues in the Republic. It is also
because the argument of the Republic ends with the passage in 611b–612a
in which Plato at least suggests the possibility that that the unitary view
of the psyche, rather than the tripartite one, shows its true nature (611b10,
612a4). To this extent, Sedley thinks that Plato presents himself as the ‘ori-
ginal unitarian’ (p. 85 above) in showing his adherence, at a fundamen-
tal level, to Socratic ideas even when he also introduces theories which
diverge from these.22
I see the force of the proposals made by Rowe and Sedley about how to
integrate the two strands of thinking whose distinctness they also stress.
However, I would favour an interpretation that sees both these dimen-
sions as aspects of a single, integrated theory or argument, which runs
through the Republic as a whole (and perhaps only there). In broad out-
line, the core argument that justice constitutes happiness is formulated
in terms of a complex theory combining, in an exceptional way, psycho-
logical and political strands and centred on an interconnected two-stage
educational programme. The first stage is designed to produce, through
habituation in social norms, certain patterns of belief, aspiration and
22
Sedley sees the same process at work in the Theaetetus and in two, contrasting, accounts of embod-
ied psychology in the Timaeus (69c–72d, 90a–d).
desire; in this respect it aims to develop the functions of all three psychic
parts, including certain rational functions. The second stage is designed
to achieve, through purely intellectual means (mathematics and dia-
lectic), independent and objective understanding of ethical and commu-
nal norms, culminating in knowledge of goodness. This stage is directed
solely at the rational part of the psyche; but, if successful, it brings with it
a transformation of aspirations and desires, which harmonises and unifies
all three functions in a way that goes beyond the condition produced by
the first stage. Justice, in the full sense, both at the psychological and pol-
itical level, can only be achieved by the combination and co-ordination
of the two stages. Thus, for instance, the norms implanted in the first,
habituative, stage of education are based on the understanding of the
fully educated philosopher. The philosopher is entitled by this under-
standing to ‘mould’ the patterns of the community to which she is ‘equal-
ised’ or ‘assimilated’ in her virtue.23 But the philosopher cannot achieve
this understanding without passing successfully through the first stage of
education and showing that its patterns have become deeply embedded
in her character.24
Against the background of this interpretation of the Republic, I would
attach a rather different significance to some of the passages presented by
Rowe and Sedley as expressing a distinctively Socratic psychology. For
instance, the passage on (advanced) education as the ‘turning around’ of
the psyche (518b–519b) is regarded by Sedley (pp. 80–1 above) as present-
ing an alternative (‘intellectualist’) conception of ethical development to
the habituative shaping of the virtues which is also recalled in this pas-
sage (518d9–e2). But we can also, and, I think, more plausibly, see the
two forms of education as meant to be co-ordinate and interdependent.
In the preceding discussion in the Republic, it is repeatedly stressed that
those who possess the natural qualities of the ‘philosophical nature’ need
also to have their characters shaped by the right kind of social environ-
ment if they are not to be corrupted (485a–504d). This explains why the
privilege of moving from the auxiliary class and its mode of education to
the advanced, intellectual, stage of education is reserved for those who
pass the tests, which show that their character as a whole (including emo-
tions and desires) has become thoroughly shaped by the norms implanted
23
See R. 498d8–499a2, esp. 498e3; see also 401e–402a and 500c–501b.
24
See further on this point below. For the interpretation outlined in the preceding paragraph, see Gill
1996b: 267–75, 279–87, 291–300 (esp. 267, 274–5, 295–7); also Gill 1998, 2004 and 2013.
by habituative education.25 Thus, I think it is implied that the ‘turning
round’ of the eye of the psyche depends on the prior harmonisation of the
parts of the psyche: it is not an independent process.26
Also, as cited earlier, Sedley suggests (p. 87 above; cf. 79–80) that the
psychological state of the (fully educated) philosopher can only be effect-
ively conveyed by replacing the tripartite psychic analysis with an alter-
native, intellectualist one, for instance, in the passage on the plasticity of
desire (485d–e). But Plato also expresses that psychological state in, for
instance, the arguments at the end of Book 9, about the superior pleasures
of the philosopher – arguments that are crucial for the core claim of the
Republic that justice constitutes happiness – which deploy the idea of the
tripartite psyche. One important theme in these arguments is that only
the philosopher, and not the other two types of person, can recognise that
a more ‘real’ pleasure comes from knowledge of what is more real (‘that
which is always the same and immortal and true’).27 Thus, for instance, we
are told (586e4–587a1, my translation):
When the psyche as a whole follows the lead of the philosophical [that is,
rational] part and does not cause internal strife, it is open to each part to
do its job and be just in other respects, and also for each part to enjoy the
pleasures that are most appropriate to it and that are the best and, as far as
possible, the truest.
These ideas are very close to those in the central digression which are
taken by Rowe and Sedley to mark a recurrence of Socratic intellectual-
ism. But they are firmly linked in this context with the tripartite psychol-
ogy, suggesting that for Plato himself this feature of psychological life can
be expressed effectively in those terms. Indeed, it is possible to read this
passage as conveying, in more fully articulated form, what is meant in the
earlier, more impressionistic, passage about the plasticity of the ‘flow’ of
desire from one object to another.28
What follows from these points, and my alternative reading of the
Republic, for the broader interpretative questions raised here? In outlining
this alternative reading, I am not aiming to re-assert the SDV, accord-
ing to which, in the Republic, Plato replaces one psychological theory,
the Socratic, with another (tripartite) Platonic one. I accept that Rowe
25
R. 502 –503a7, 503e1–2, referring back to the earlier criteria of selection of guardians from the larger
auxiliary class (412d–e, 413c–414b); also 535a–536a, 537c9–d2. See Gill 1996b: 274–5.
26
By implication, habituative education (518d9–e2) could have helped to ‘knock off’ the ‘lead weights’
of sensuality that prevent some naturally talented people from seeing ethical truth clearly (519ab,
esp. a9–b1).
27
R. 581b, 582b–d, 585c1–2 (cited in text), 585d–586b. 28 See Gill 1996b: 295–7.
and Sedley have identified features in the dialogue which seem, at least
by some criteria, Socratic (even if they only partly match the supposedly
Socratic argument that closes the Protagoras). But what seems to me much
more open to question is whether Plato himself is as conscious of the dis-
tinction between these two strands as Rowe and Sedley are. His readiness
to re-describe the same phenomenon (the psychological state of the phil-
osopher) in ‘Socratic’ or ‘Platonic’ terms suggests otherwise. It also seems
to me open to question whether Plato himself – as distinct from modern
scholars – is strongly aware of, or interested in, this distinction.29 What
the discussions by Rowe and Sedley indicate to me is that the argument of
the Republic represents a chemical fusion of – what we call – Socratic and
Platonic strands. Plato’s aim, as I see it, is not to provide clues to help us
unpick this fusion, but to use these strands to construct a compelling and
coherent argument within the Republic as a whole and to challenge us to
engage philosophically with this argument.
29
See n. 20 above.
ch apter seven
Reflective commentary (2)

Appearance, reality and the desire for the good
Dimitri El Murr
Among current Plato scholars, Christopher Rowe is ‘notable for the atten-
tion he has devoted to the topic of the Socratic/Platonic conception of the
good.1 My aim in this discussion is to offer some reflections based on his
seminal work, while also considering the contributions of David Sedley
and Thomas Johansen in this volume. In particular, I focus on a crucial
problem raised by Socrates’ account of the Good in Republic 6.
The issue I address is this: when, in Republic 6, Socrates claims that
every soul desires the Good and then, a few pages later, makes the Good
nothing less than a grand ontological and cosmic principle analogous to
the sun in the sensible realm, is he using the same notion?2 Building on
Aristotle’s famous objection that Plato’s idea of the good is not practically
realisable (prakton),3 most commentators have either held that the good
has no ethical or political role to play in the Republic (most famously,
Martin Heidegger),4 or that Plato is simply confused and the metaphysical
notion of the good is only homonymous with its ethical counterpart, our
human good.5 In my view, both options are misleading, and the work of
Rowe (and others) will help me to argue that it is the understanding of
the way human desire works, and its specific relation to what is good, that
gives Plato a firm basis for his broader ontological approach.
My thanks to George Boys-Stones, Christopher Gill, Christopher Rowe and David Sedley for reading
an earlier draft of this paper, and to an anonymous reader for Cambridge University Press.
1
See Rowe 2003a, 2003b, 2005 and 2007a: 239–65.
2
That the idea of the good is endowed with a significant cosmic role may not be self-evident, but
seems to be implied by such claims as Socrates’ at e.g. 517bc: ‘The form of the good … is cause for
everything of everything (πᾶσι πάντων) right and beautiful, as both progenitor of light and of
the source of light in the sphere of the seen, and the source itself of truth and intelligibility in the
intelligible sphere’ (trans. Rowe). Throughout this paper I quote the Republic from Rowe’s Penguin
translation (2012a).
3
See EN 1096b32–4: ‘For even if the good that is predicated in common of things is some one thing,
something separate “itself by itself ”, it is clear that it will not be anything doable or capable of being
acquired by a human being’ (trans. Rowe, in Broadie and Rowe 2002).
4
See Heidegger 2001: 192: ‘ἀγαθός, gut, hat ursprünglich keine moralische Bedeutung.’
5
For this view, see Cooper 1999: 148–9.
122
Appearance, reality and the desire for the good 123
At R. 506d, after Socrates has claimed that the study of the idea of the
good is the object of the highest learning (megiston mathēma), Glaucon
insists that Socrates should define the nature of the Good. No one else,
after all, Adeimantus noted earlier, is better equipped than Socrates to ful-
fil this task.6 But Socrates famously refuses to give in to his interlocutors’
demands and argues for a different approach to the problem: he is willing
to talk about the offspring (ekgonos) of the Good, which bears a very close
resemblance to it. Given that, like Plato’s brothers, we must content our-
selves with mere images, let us focus on what Socrates actually says about
the closest thing to the Good, its offspring: the sun.
To put it in a nutshell, the analogy used by Socrates develops the dis-
tinction, made at the end of Book 5, between the plurality of sensible
particulars and the uniqueness of the form that cannot be perceived by
the senses. Hence Socrates makes use of a specific type of analogy, of the
form A : B :: C : D, one that may be labelled, using Aristotle’s termin-
ology, ‘discontinuous’:7 the intelligible sphere, where the Good stands, is
radically distinct from the sensible one, where the sun rules. The set of
entities belonging to the first sphere (that is, forms) are grasped by the
intellect, and not seen, whereas sensible things are seen and not grasped
by the intellect (507b). Moreover, the analogy shows that, just as the sun
produces light and makes vision possible, the power of the Good is to
produce truth and being. Indeed, the possibility of vision is only indir-
ectly dependent upon the sun, since the sun produces light which makes
vision possible. Similarly, the Good makes knowledge possible insofar as
it creates the condition under which intelligence can not only grasp its
specific objects but can also recognise their status as essences, that is, as
eternally stable entities. The ‘condition of the Good’ (hexis tou agathou,
509a5) is therefore of a causal nature: what the Good does is to grant being
(ousia) to a certain set of entities accessible to intellect alone, thus confer-
ring on them the power to escape from becoming (genesis). The Good is
thus responsible for the distinction between the sensible and the intelli-
gible realms.
Can one find in Socrates’ mouth an explicit argument for this claim? It
seems that the Good is just said to be the principle that makes the forms
real and intelligible, and unaffected by becoming, just as the sun makes
sensible things visible: the analogy merely illustrates this point but does
6
R. 506bc: ‘No,’ he said, ‘and it doesn’t seem to me to be right, either, Socrates, for somebody to be
willing to talk about other people’s views but not about his own, especially when he’s been busying
himself for so long with it all.’
7
On this distinction, see Aristotle, EN 1131a31–b6.
124 Dimitri El Murr
not prove it. So Julia Annas seems right in claiming that ‘Plato is putting
forward [these] thoughts, though he leaves them deliberately schematic’
because ‘presumably he believes that they are true, but has no idea how to
argue for them, and perhaps thinks that they are not the kind of truth that
can be argued for’ (1981: 246).
To understand why Socrates gives this extraordinary prominence to this
particular form, and why he holds that ‘even while the Good is not itself
being, [it] is even beyond being, superior to it in dignity and in power’
(509b), let us consider how the study of the megiston mathēma is justified a
few pages earlier (R. 504e–505b):
Anyway, it’s something you heard about on more than a few occasions.
Either you don’t remember, or you’re deliberately attacking me just to make
trouble. I think the second is the more likely, because you’ve heard often
enough that it’s the form of goodness that is the most important subject,
since it is what brings about the goodness and usefulness both of just things
and of everything else. And you pretty much know that’s what I’m going
to say now – as you know I’ll go on to say that we don’t have sufficient
knowledge of it; and if we don’t, even if we had the greatest knowledge pos-
sible of everything else, and not of this, you know that it’s of no more use
to us than possessing anything if good doesn’t come of it. Or do you think
it takes us any further on, to have made any acquisition you like, but never
a good one? Or to be wise about anything and everything else, but with the
good left out, and have no wisdom about anything beautiful and good?
The highest object of knowledge is the idea of the good. It is not diffi-
cult to see why this is so: every good thing for us derives its goodness and
usefulness from this idea, so that if one knew everything except the idea
of the good, this knowledge would be of no use to him. As Adeimantus
will point out later, and as Socrates states explicitly in this passage, there
is nothing new in these claims, as far as Socrates is concerned. In the
Charmides, for instance, a long section (172c–174d) is devoted to show-
ing that it is necessary to postulate the existence of a knowledge of good
and bad in addition to the particular sciences. The reason why the know-
ledge of good and bad is at the top of the scale of sciences is that it gives
its usefulness to every particular science. This is only one example among
many other possible ones. As Rowe has argued, Socrates’ explicit reference
to previous conversations on the same topic seems to underline that this
passage and the following one, on the account of the Good offered by the
many and the more subtle people (505bc),8 constitute ‘a kind of patchwork
8
See the pre-Republic parallels to 505bc examined by Sedley, pp. 77–8 above.
of elements from pre-Republic dialogues’ (2007a: 241). Surely, this is also
true of the passage immediately following the rejection of the definition of
the Good as pleasure or knowledge, where Socrates deals with the desire
for the Good (R. 505de):9
And isn’t it also plain that whereas many would choose to do, or possess, or
think things that seemed to them just and beautiful, even when they were
not so in fact, they draw a line when it comes to good things? They won’t be
satisfied with getting things that merely seem good; they’ll insist on seeking
out what really is good. This is one sphere in which nobody needs to be
told to scorn mere appearances.
Note the universality of Socrates’ claim here. By considering the par-
ticular status of the good from the standpoint of human motivation, and
by comparing the desire for beautiful things (ta kala) or just things (ta
dikaia) and the desire for good things (ta agatha), Socrates takes the ana-
lysis of 504e–505b one step further. As far as ordinary virtues are con-
cerned (justice, piety, temperance and so on), most people believe that
the mere appearance of such virtues is often perfectly sufficient. The aim
of the Republic as a whole is to show that this opinion is wrong-headed
and that desiring the appearance of justice rather than desiring to be really
just comes from complete ignorance about what is really good for our-
selves. Even so, this is how most people behave. Yet the same ignorant
people would never admit that they are striving for the appearance of the
good: when they go for the appearance of justice, they do so because they
think that the appearance of justice is a real good. What grounds Socrates’
claim here is thus the idea that there is an essential difference between two
orders of desire. Given that what I desire is always my own good,10 then,
when I desire to appear just or pious, I do so because I think the appear-
ance of justice is really good for me. So even desire for the appearance of
justice is motivated by desire for a real good. Socrates then makes his next
move, the crucial one, within the same line of thought (R. 505d–506a):
What every soul pursues, then (ho dē diōkei), the very thing for the sake of
which it does everything it does – divining that there is such a thing, but
puzzled and unable to get an adequate grasp on what exactly it is, or come
9
For parallel passages on the universal desire for the good, see e.g. Meno 77b–78b, Gorgias
467b–468c.
10
It has been argued that Socrates is dealing here with some sort of impersonal or ‘non-self-referential’
good: see White 1979: 35 and 2002: 198–211. I cannot make sense, at least in a Platonic context, of a
notion such as a ‘non-self-referential’ good. On the issue of egoism and self-interest in Plato’s ethics
(and the effort made by scholars to discharge Plato from egoism) see Penner 2006a: 156–60. See
also Penner 2006b: 246–62, and Rowe’s arguments against this interpretation at 2007a: 252–3.
126 Dimitri El Murr
to any stable conviction about it as it can about everything else, and so
missing any benefit there might have been in anything else – are we going
to say that the best of the citizens ought to be similarly in the dark on a
subject like that, and a subject of such importance, when we’re going to put
everything in their hands?
The whole passage is a single sentence depending on the first relative
pronoun (ho: ‘what’): the reference of this pronoun, which is picked up
by the following demonstratives ‘a subject like that, and a subject of such
importance’ (to toiouton kai tosouton), cannot be different from that of
the demonstrative pronoun ‘this’ (touto) that one finds six lines below at
506a6.11 This pronoun refers unambiguously to the object whose know-
ledge the guards must acquire, namely the ‘highest learning’ (megiston
mathēma) which is, of course, the idea of the good.
Socrates then adds that most people fail to understand what the idea
of the good is, but that everyone divines that it is. What does every soul
divine about the good without being able to grasp its true nature? What
every soul divines stems from what Socrates has noted a few lines above,
namely that desiring just things (ta dikaia) and desiring good things (ta
agatha) are not similar activities. I can desire to appear just because I think
that this appearance of justice is good for me, but I cannot desire the
appearance of the good because what I desire is always my own good.
Thus, one can desire to possess something which is only apparent because
it seems profitable, but one cannot desire that profit to be only apparent.
The way desire works in each and every one of us reveals that the dis-
tinction between appearance and reality is deeply implanted in our souls.
What the uneducated soul divines, but what it is incapable of grasping, is
that the Good can only exist as a reality, that is, as an eternal essence, or
a form.
The presentation of the prisoners inside the cave, which, as Socrates
makes clear, integrates key features of the two previous images, illustrates
this point by depicting the role of the good in relation to the epistemo-
logical and ethical dimensions of our human nature. As Johansen puts it
(p. 95 above): ‘the sort of object that can truly answer the questions that
occupied the prisoners … is, ultimately, the form of the Good, since it
is the cause of the being and coming into being of all things, including
ultimately the being and coming into being of the images inside the cave
11
In addition, notice how closely 505d–506a parallels 505ab, a passage explicitly devoted to the idea
of the good. The clause ἀποροῦσα δὲ καὶ οὐκ ἔχουσα λαβεῖν ἱκανῶς τί ποτ’ ἐστὶν echoes αὐτὴν
οὐχ ἱκανῶς ἴσμεν at 505a5–6; similarly, διὰ τοῦτο δὲ ἀποτυγχάνει καὶ τῶν ἄλλων εἴ τι ὄφελος ἦν
at 505e3–4 echoes οἶσθ‘ ὅτι οὐδὲν ἡμῖν ὄφελος at 505a7.
(517c3)’. Even though only the philosopher will turn out to be capable of
moving out of the cave and understanding what the good really is, that is,
what is best for each thing and for all of them, the allegory illustrates the
consequence of Socrates’ claim about the universal desire for the good: no
one, not even the least educated individual, is so confused as to have no
glimpse at all of true reality.
It is now possible to clarify Socrates’ hyperbolic and somewhat mysteri-
ous claim that the form of the good is ‘not itself being, [it] is even beyond
being, superior to it in dignity and in power’(509b). If the idea of the
good is specifically endowed with the extraordinary power to distinguish
appearance and reality it is because the good is the only thing among the
many things one desires that one will see as implying this distinction. If
every soul desires the good, as something truly real, perhaps Plato is not
so confused in placing the idea of the good at the top of his ontological
hierarchy of forms.
I conclude with a few remarks on how the interpretation sketched
above fits into the overall programme of the Republic.
As Sedley shows (pp. 74–5 above), the section of the Republic I have
examined belongs to a long digression which Plato has deliberately placed
at the centre of his dialogue. This digression, extending from 471c, in the
middle of Book 5 in the familiar ten-book division, to 541a at the close of
Book 7, is unified by Glaucon’s question about the practicability of the
ideal city and Socrates’ paradoxical argument (paradoxos logos, 472a) in
response to that challenge. This is the ‘third wave’, showing the need for
philosopher-rulers and therefore the definition of the true nature of phi-
losophers. What stands at the exact centre of this central digression is the
passage I have been considering in detail on the relation between human
desire and the idea of the good (505d–506a). In my opinion, this location
is no accident.
First, if I am right in claiming that the difference between appearance
and reality is at the forefront of Socrates’ argument here, this passage can
be read as a partial response to Glaucon’s challenge, formulated in Book
2 as a follow-up to Thrasymachus’ defence of a merely conventional exist-
ence of values. Glaucon famously argued that it is only the appearance,
or reputation, of justice that is universally favoured, and that it remains
to be shown that justice as such is better than injustice. Indeed, since the
appearance of justice or piety entails the social gains that come with good
reputation, why bother with striving for the real thing? We desire many
things: wealth, honours, success, but also justice, piety and all the other
virtues whose possession are advantageous for us. But we desire these
128 Dimitri El Murr
virtues because we consider that the reputation of having them is already
beneficial. Socrates’ point, in the passage I have been considering, is sim-
ply that, when we desire to benefit from the appearance of justice or piety,
we do so because we consider that this appearance is really good, since
desiring the appearance of any value entails that we consider the value of
this appearance to be really good. So anyone is bound to recognise that
one thing at least seems to exist independently of any opinion, and this is
the idea of the good. This central claim will lead, ultimately, to Socrates’
final response to Glaucon’s challenge in Book 9, where he will show that
injustice is always practised unwillingly and that it is worse to seem just
but be unjust.
Secondly, and more broadly, the central position within the Republic
of the passage on the desire for the good, which is embedded in a context
which Sedley sees as expressing the main tenets of Socratic moral psych-
ology (p. 78 above), has a further significance. It indicates that Plato sees
no contradiction between what we tend to see as a purely Socratic claim
about human action (the universal desire for the good) and what we typ-
ically describe as genuinely Platonic (the tripartite soul). However, there
is room for debate about how we should explain this juxtaposition of
seemingly divergent theories. On Sedley’s account of the central digres-
sion of the Republic, Plato chooses to revert to an intellectualistic account
of moral psychology because it does capture, more accurately than the
tripartite psychology, the kind of intellectualised virtue that is distinctive
of a true philosopher.12 On Rowe’s account, tripartition is never a correct
explanation of human motivation and action (the soul’s desires are always
and only for the good), but something Socrates resorts to because, in the
Republic, he is led to develop a large-scale, political solution to the prob-
lem of human ignorance. (Tripartition is how non-philosophical people
choose to think of themselves and is therefore a mode of discourse access-
ible to them.)13 Although there are significant differences between these
interpretations, both of them offer material for clarifying the issue on
which I have been focusing here.
How should one account for the relationship between the univer-
sal desire for the good, which is by and large the core idea of Socrates’
intellectualist position throughout the dialogues,14 and the idea of good
to which Plato gives an exceptionally grand role in the Republic? One
solution, in line with Sedley’s views, would be to argue that ‘the original
12
See pp. 86–7 above. 13 See Rowe 2007a: 170–5.
14
See e.g. references in n. 9 above.
unitarian was none other that Plato himself … constantly trying to show
that, wherever we might have suspected him of placing some philosoph-
ical distance between himself and Socrates, he is in fact doing no more
than continue, interpret and perfect Socrates’ own lifetime project’ (p. 85
above). From this perspective, it is Plato’s reflections on the crucial claim
of Socratic intellectualism that led him to confer on the Good that extra-
ordinary preeminence. An alternative account, following Rowe, would
be to hold that if we read Socrates’ intellectual autobiography at Phaedo
97c–99c alongside Republic 6, we understand that ‘the alleged “Platonic”
(cosmic) view of the good is in essence an application or extension of the
“Socratic” (human, or humanist) view to a broader context (in fact, the
broadest context possible)’ (2007a: 248). So ‘Plato wants us to associate
that more ample view of the good with Socrates’ (2007a: 248). In other
words, what we see as a typically Platonic view and a specific contribution
of the Republic is nothing more than a position Plato constantly associates
with Socrates.
It is not my purpose here to decide which of these two options should
be preferred. What I have tried to bring out is that each of these two inter-
pretations, in their own specific way, helps us to understand that Plato
was never so confused as to make the form of the good an entity discon-
nected from human practice.
ch apter ei ght
Waving or drowning? Socrates and the

sophists on self-knowledge in the Euthydemus
M. M. McCabe
The structure of the Euthydemus

The Euthydemus gives the first impression of being one of the most well
organised of Plato’s dialogues. Compared, for example, with the lopsid-
edness of the Parmenides, the sagging centre of the Cratylus, or the sheer
immensity of the Laws, its clear structure promises the reader philosophi-
cal clarity as well. And yet its careful construction has worked, throughout
its history, to its disadvantage.
This conversation between Socrates and Crito falls into seven separate
parts. In the frame, Crito complains that yesterday there was a discussion
between Socrates and some strangers which he had been unable to hear
because there was such a crowd around them.1 They were sophists, Crito
presumes: please would Socrates explain to him their wisdom? Socrates
apparently acquiesces: at any rate, he recounts to Crito the conversation
that took place, which he describes as trying to turn young Clinias to
philosophy. There follow five distinct argumentative chunks, carefully
arranged: three sophistic episodes, interleaved with two Socratic episodes,2
It is a pleasure to offer this essay to Christopher Rowe, as a mark of a long friendship, argumentative
in all the right ways and none of the wrong ones. I should also like to thank audiences at Berkeley,
Cambridge, London and Toronto for extremely helpful discussions and comments, and, in particular,
Dom Bailey, Nick Denyer, Christopher Gill, Verity Harte, Mark Textor, an anonymous discussant in
Cambridge and the referees for Cambridge University Press.
1
There is a curious dispute about how many strangers there were: Crito thinks there is one; Socrates
corrects him and insists that there were two. This counting and individuating of the sophistic broth-
ers runs through the dialogue and here prefigures a central theme. I have argued elsewhere that
the Euthydemus is interested in general metaphysical principles of individuation and identity (e.g.
McCabe 1998); here I consider how it may also focus on central questions in epistemology, notably
a contrast between a version of internalism and externalism.
2
I eschew the usual label ‘protreptic’ for the Socratic episodes, since that seems to me to privilege
them over other parts of the dialogue. For not only do the Socratic episodes run into serious dif-
ficulties, as I shall argue, but also the conceit that the sophists are actually offering something clever
may make theirs protreptic, too, just because they force the reader to wonder why.
130
Self-knowledge in the Euthydemus 131
culminating in complete victory (and a self-refutation) for the sophists.
The dialogue closes by returning to the frame, as Crito recounts his own
encounter on the previous day with an anonymous character – who just
might be Isocrates – who complains that philosophy is worthless and
shameful. Does the dialogue counter that challenge?
The five central episodes are held together by their dramatic frame;
this enclosure gives the dialogue an illusion of unity. But this may just
be an illusion. For the arrangement of the episodes is paratactic: they are
aligned, but not apparently connected, at least at the level of philosophi-
cal content.3 This parataxis might offer two separate dialogues: a (very
short) Socratic dialogue on wisdom, running into an aporia at 292; and
a tour de force on sophistry. The Socratic episodes found an after-life in
the Stoa (the Stoics got their terminology of promoted indifferents from
Euthydemus).4 The sophistic episodes, by contrast, were grist to Aristotle’s
mill: the Sophistici Elenchi either uses the Euthydemus or has a common
source for many of its sophistries. But it is, surely, a peculiar way to think
of the dialogue – as two separate dialogues. Otherwise, however, why are
these two projects (wisdom and sophistry) interwoven in this way? In par-
ticular, why did Plato think that these logical puzzles should be intercon-
nected with this ethical material? And in general what might we say about
the relation between logic and ethics thus presented so starkly?
Vertical and lateral order

While it is true that the episodes seem paratactically arranged, their rela-
tion to the enclosing frame is not a trivial one. For example, at the very
centre of the dialogue the conversation between Socrates and Clinias
(who has apparently by now read the Republic)5 is suddenly interrupted by
the frame discussion, which both continues and criticises what has been
said so far. Here the role of the frame is not merely formal; it brings the
direct conversation of the dialogue itself into reflective scrutiny. This has
the effect of focusing the critical attention of the reader on the methods
and principles of that direct argument. And this reflective feature of the
3
For example, the transition between the first sophistic episode and the first Socratic episode seems
abruptly to change the subject from a set of puzzles about knowing and learning to a characteristi-
cally Socratic inquiry into the nature of happiness. I have benefited here from discussion with Philip
Krinks (see Krinks 2011).
4
As far as I can see, the terminology comes precisely from the imagery here; see McCabe 2002b; and
see also Annas 1995, Long 1996, Striker 1996.
5
See below, n. 13.
132 M. M. McCabe
dialogue runs through the sophistic episodes, too, which regularly report
on the terms of argumentative engagement.6
Therefore, the dialogue is not merely carefully structured laterally (from
one episode to the next), but vertically, too: a major part of the discussion
in the frame and across the episodes focuses on the methods, reasons and
assumptions of the first-order discussion itself. This should discourage the
‘two dialogues’ thought; at least, in the frame there is a unified and sys-
tematic reflection on the ways of Socrates in contrast to the ways of the
sophists.7 In what follows I shall argue that this amounts to a deep discus-
sion of philosophical principle, to be seen in the continuity across the last
two episodes as the dialogue draws towards its apparently chaotic close.
A handbook of fallacy?
The ‘two-dialogues’ worry is more urgent, however, at the level of content:
it is sometimes hard to see how the content of the sophistic episodes can
be thought to have any bearing on the content of the Socratic ones. After
all, it is a common view that we have, in the sophistic sections of the dia-
logue, a handbook of fallacies, whose purpose is to allow us to spot and
diagnose sophistical argument. That the arguments in question are falla-
cies is, on this view, just taken to be obvious, something we can just see –
in the same way, it is often supposed, as we can just ‘see’ validity.8
But things may not be so simple. We may easily recognise that the con-
clusion of a sophistic argument is false; Socrates’ dog is not, as a matter
of his own biography, his father. Equally, we may recognise that Achilles
will in fact overtake the tortoise. But we take Zeno’s arguments to be para-
doxes and the sophists’ to be fallacies: why? Perhaps the thought is that
paradoxes rely on assumptions whose credibility – or otherwise – needs
to be uncovered, whereas fallacies work by ignoring principles which just
are obvious, or recognised in advance, or common sense. Hence, to diag-
nose a paradox we may have to think about all sorts of complicated issues
in logic, or metaphysics, or even ethics, very hard indeed; to diagnose a
fallacy, we just need to see where the trick is turned. This makes falla-
cies, if you like, theory-laden: the mistake is made by the contravention of
6
In the first episode, Dionysodorus remarks that Clinias will be refuted, ‘whichever answer he might
give’, 275e5; in the second, the discussion of contradiction and consistency is echoed in the accusa-
tions between the protagonists, e.g. 285e, 287b3. See further on the third episode below.
7
Consider the running theme of the defence and the importance of consistency, e.g. 287ab; or the
puzzles about what counts as a λόγος in the second sophistic episode, especially at 285e–287a.
8
See Lear 1980: 1–15.
assumptions whose credit is already sound. The arguments of the sophists
in the Euthydemus – it is commonly supposed – are of this very sort; and
they are perpetrated by characters who couldn’t give a fig for logical prin-
ciple anyway.
Of course, we had better not think that the difference between a good
and a bad argument, or between a paradox and a fallacy, rests on the moral
character of the person who puts it forward. But we had better not think,
either, that access to the principles violated by fallacies is immediate, so
that they can just be spotted and diagnosed readily enough. At least, this
thought may be anachronistic: although Plato is clearly interested in logi-
cal principles, he may not have a fixed account of what those principles
might be, or of how we might defend them. Indeed, as I shall suggest, one
of the purposes of the Euthydemus is to take a crack at getting hold of just
such principles – and how we might defend them against attack. Only
after that (and with the addition of more theory about what it is to ‘see’
validity and the like) can we confidently say that fallacy is just something
we ‘see’.
Principles of non-contradiction
The development of the third sophistic episode of the Euthydemus
(293a–303a) is dramatic: rapid shifts of interlocutor; a striking squabble
between the sophists where they seem to disagree9 – despite their earlier
claim that disagreement is impossible; and an increasingly dilapidated
sequence of arguments. In this section of the dialogue the impression of a
ragbag of ready-made sophistries becomes more and more acute.
However, the opening two sections of the third sophistic episode –
the discussion of knowledge, 293b–296d, and the ‘dog’ arguments,
297b–299a – do have some internal continuity. For they turn on an
extreme principle of non-contradiction (PNC) which insists that noth-
ing can be both thus and so and not thus and so, simpliciter (293b, 293d,
298a–c):10
Extreme PNC: It is not the case that anything, if it is thus and so, is also not
thus and so.
9
Especially in the context of my argument here, ‘disagreement’ is a more felicitous rendering of
ἀντιλέγειν than ‘contradict’.
10
I leave on one side for now the question whether this is a linguistic or a metaphysical principle. I
think that contrast is of interest for the Euthydemus, but to elaborate it would make the puzzles I
am attending to here too loaded. I call this a Gross LNC in McCabe 2006.
134 M. M. McCabe
This principle is extreme just because it does not allow any qualification of
the properties it describes, and this ostentatiously leads to counterintuitive
conclusions, such as that Socrates is a puppy.11 But the exchanges between
Socrates and the sophists reveal that each of the interlocutors subscribes to
some version of PNC. If these conclusions are to be avoided, what may be
needed is a principled defence of a qualified PNC, showing in particular
which qualifications might be thought to matter, and why:
Qualified PNC: It is not the case that anything, if it is thus and so, is also not
thus and so, in the same respects, at the same time, etc.
This defence – at least within the limits of the individual arguments –
seems to be unavailable to Socrates, with the result that he must concede.
Consider the dog.12 The sophists’ conclusion that Socrates’ dog is his
father relies both on some version of PNC and on ignoring any of the
qualifications on the predicate terms that he tries to make (that his dog is
the father of puppies; that this is his dog, not merely his, etc.); so it relies on
Extreme PNC. But of course Socrates cannot merely enter any old qualifi-
cation on his dog to evade the conclusion: he cannot just stipulate that the
only father he has is Sophroniscus. Rather, some of what is required here
is a series of standing orders, for example, about the persistence of individ-
uals through relational change, an issue that has been running since 283,
and that turns up again, for instance, at 296d and 298bc. But a collection
of assumptions about identity will not do the trick in every case where
PNC needs qualification. The arguments – to protect against the false-
hoods that can be derived from Extreme PNC – need to show that we
should prefer Qualified PNC to its Extreme counterpart. And they need
to show how and why: how a variety of principles is needed and why this
variety is defensible. This project, I shall suggest, is at issue at the end of
the second Socratic episode. The topic is self-knowledge.
The third wave

That this is an important moment in the dialogue seems obvious, and the
vertical axis of the dialogue is now at its most dense. For suddenly (298c)
the meek Clinias seems to find his voice, rejecting what Socrates suggests.
11
One could set this up differently so that any PNC is Extreme, and the qualifications are entered
within the predicate term: predicates, on this account, are always contextually complex, PNC sim-
ple. This is not, I suggest below, how Plato envisages it nor how Aristotle later explicates PNC in
Metaphysics Γ.
12
I have discussed this in more detail at McCabe 2006.
Crito interrupts (290e), incredulous that Clinias might indeed have said
such a thing, and the discussion moves into the frame, where it is exposed
to some cutting criticism. At the end of all of that, Socrates declares him-
self to be drowning in the third wave of puzzlement (293a), and he calls on
the sophists for help. This time, for the first time, they seem to oblige.13
Socrates has reason to fear drowning. For the argument that has gone
before is dense and its point is difficult to establish. The wave in which
he risks drowning rises just when Socrates wonders whether what he has
shown so far about knowledge is either regressive or circular or merely
tautological. The sophists’ lifebelt is the suggestion that the knowledge
Socrates is interested in has been his all along, even though he did not
notice it: and they imply that this rescues him from regress, circularity
and tautology. Or this, at least, is so if the sophists so decide (296d). Can
Socrates show that knowledge is not merely a matter of what some sophist
decides? Are there any principles here?
The sophists to the rescue

At 293a Socrates appears to be drowning. He has been trying to give an
account of the knowledge that he had earlier claimed to be the only good
‘itself by itself ’ (281de); but all of his attempts reduce to regress or circu-
larity or vain repetition. Euthydemus comes to the rescue with alacrity,
and he offers to show Socrates that he has had the knowledge that he seeks
all along. Socrates’ surprising response is that he would be delighted to be
shown that he had the knowledge all along, since that would save him the
bother of having to learn it.
Here for the first time, and surprisingly, Socrates’ interests seem to coin-
cide with those of the sophists. However, as I shall argue, they cannot save
him from the drowning he fears. For the account of knowledge the soph-
ists give precludes what Socrates was looking for when he risked drowning.
Indeed, as it is developed in a sequence of arguments, sophistic knowl-
edge provides a detailed alternative to Socrates’ quarry. As a consequence,
the juxtaposition of the two episodes offers not merely a surface contrast
13
This whole sequence is not only dramatically striking but redolent of the language and the theories
of the Republic (e.g. the description of the relation between the geometers and the dialecticians, to
which I shall return). Now while I shall not here defend the claim that Plato’s thought developed,
especially by means of critical engagement with his own work, the peculiar allusiveness of this
passage is not adequately explained (i.e. not given a philosophical explanation) by the idea that it is
proleptic of what will come later, in the Republic (see Kahn 1996). Rather, if you like, it is ‘meta-
leptic’: it reflects critically backwards on the Republic, and the telegraphic nature of the allusions is
evidence that it does so; see McCabe 2002a.
136 M. M. McCabe
between a Socratic way of thinking about knowledge and something that
sophists might play with, but rather two detailed and contrasted accounts
of what we should say about knowledge. Those accounts, as I shall sug-
gest, are contextualised, in the arguments that follow, in a discussion
about how we might qualify the application of PNC.
Euthydemus opens by asking Socrates to subscribe to Extreme PNC
(293b8–c1):14 ‘So do you think that it is possible for one of the things that
are, not to be the very thing which it happens to be?’ ‘I do not indeed, by
Zeus.’15 There follow three brisk arguments to show that Socrates knows
everything and that he knows everything all along, always, just so long as
he concedes – as he does – that he knows something. By the conclusion,
Socrates seems forced to concede his own eternal omniscience. But surely
this is not something the old curmudgeon of the opening of the dialogue
would say (272c). What has gone wrong?
The arguments proceed by making a great deal of fuss about qualifica-
tion. So in the first argument (293b–d): if (as he agrees) there is something
that he knows, then Socrates is a knower. Socrates qualifies: of the thing
that he knows. It makes no difference, says the sophist: if Socrates is a
knower, then he knows everything. No, no, says Socrates, there are many
things that he does not know. But then he is not a knower. Socrates quali-
fies again: of the things which he does not know. No matter, insists the
sophist, he is still not a knower, whereas before he conceded that he was
a knower. So (293c8–d1): ‘Thus you happen to be at once both this same
person as you are and then again are not [sc. this same person] in the same
respects.’16 Socrates is shaken. It is impossible that something can be the
14
Translations are my own, following the text printed in Burnet 1909.
15
ἆρ’ οὖν δοκεῖς οἷόν τέ τι τῶν ὄντων τοῦτο ὃ τυγχάνει ὄν, αὐτὸ τοῦτο μὴ εἶναι; The exact pars-
ing of this sentence matters. Euthydemus asks whether it is possible that τι τῶν ὄντων (the subject
term: ‘one of the things that are’) is not αὐτὸ τοῦτο (the predicate term: ‘that very thing’) which is
explicated by τοῦτο ὃ τυγχάνει ὄν ‘which it happens to be’. What is denied is Poss [∃x] (Fx & ~Fx);
so Nec ~[∃x] (Fx & ~Fx) where ‘F’ is tightly specified (what is asserted of something is the very same
thing as what is denied). The phrasing could be taken to present a general rule about identity (‘that
very thing’) or about predication (‘which it happens to be’). The passage taken as a whole turns
on a problem about predication and the properties of things; so this is how I take this formula-
tion, and its instantiation at 293b8–c1. But it is possible that the sophists’ position rests on some
connected metaphysical assumptions which require them to collapse predication into identity; see
McCabe 1998, Bailey 2012.
16
καὶ οὕτως τυγχάνεις ὢν αὐτὸς οὗτος ὃς εἶ, καὶ αὖ πάλιν οὐκ εἶ, κατὰ ταὐτὰ ἅμα. This echoes
the phrasing of 293b9–c1, but now expressed in the masculine singular (second person: τυγχάνεις
ὤν, ‘you happen to be’) so that αὐτὸς οὗτος ὃς εἶ is the predicate term (‘this same person as you
are’). In the second conjunct πάλιν οὐκ εἶ picks up the negative copula, and the predicate term
is implied (‘this same person’). So this may be taken to be an application of the PNC enunci-
ated there, emphasised by the logical modifier. What of κατὰ ταὐτά? At first glance, it concedes
Socrates’ qualifications: the negative disjunct will turn out true when just one respect fails, and
same thing and not (293d4–5).17 So he cannot be both a knower and not
a knower; and this, he supposes, implies that if he knows something –
which he concedes – he knows everything (293d6).18 So he already has the
knowledge he seeks.
That the argument concludes thus has Socrates at the mercy of an
Extreme PNC, and that is confirmed by the sophists’ refusal to allow him
any qualifications. But is this merely skulduggery on their part? Or do
they have grounds to suppose that Socrates has agreed to Extreme PNC?
I suggest that they do: Socrates agrees to an unqualified version at the
outset; fails to insist on the importance of his qualifications during the
argument; and agrees to a conclusion which appears to follow only if we
disallow qualification. The conclusion itself may highlight the fact that
qualification is at issue; for it apparently invites Socrates to admit that the
qualifications he sought have collapsed into a contradiction that asserts
both that he is the same person as he allowed himself to be, and is not that
person at all.
In the second and third arguments, likewise, the argument turns on
how we might qualify the claim that someone (if they know one thing) is a
knower.19 In the second argument (293e–294e), where Socrates is the ques-
tioner, the conclusion is that anyone knows everything always. Socrates
expresses his admiration at this move (295a6–9): ‘For if I have escaped my
own notice being wise,20 and you demonstrate that I know everything and
always did, what greater piece of luck could I have for my whole life?’
the argument that Socrates is either a knower or not a knower at all will be unsound. From the
sophists’ point of view, however (and ignoring the egregious editorial comma after εἶ), κατὰ ταὐτὰ
may simply recall the missing predicate term (‘this same person’), and maintain the validity of the
argument on the basis of Extreme PNC. From the reader’s point of view, things are more complex:
this move allows the sophists’ argument to be valid, but it emphasises the very point of difference
between Socrates and the sophists, and invites us to attend to the PNC in question (compare the
terminology of the cognate principle of opposites at Republic 436c–d).
17
Here taking τοῦτο as the subject and τὸ αὐτὸ as predicate.
18
The sequence at 293d4–6 makes clear the inferential relations between three claims: Extreme PNC;
its instance that Socrates cannot be a knower and a non-knower at once; and the inference that if
he knows one thing he knows everything. The role of the unqualified instance of PNC is central.
19
Dionysodorus here repeats the insistence that no one can know some things and not others, on the
basis that no one can be both a knower and a non-knower, 293e–294a. Ctesippus, in opposition,
wants evidence that there is nothing that the sophists do not know. Ctesippus’ position is grounded
in particular qualifications, unlike Dionysodorus’ position (294b–d); and he insists that this is a
problem for the sophists – they deny it.
20
The shift of terminology, from what I have translated as ‘knowledge’, ἐπιστήμη (with reservations:
better for many of these contexts would be ‘understanding’), to ‘wisdom’, σοφία, is revealing; I shall
return below to the connection between these discussions and an account of virtue. The sophistic
version is found in their ‘piece of wisdom’ at 293d8.
138 M. M. McCabe
In the third argument (296a–d), where the sophists are once again the
questioners,21 Socrates agrees that when he is a knower, he knows with
something (his soul), and always with that same something; so since he
knows something, he always knows something, and always everything.
Repeatedly Socrates enters a qualification; and repeatedly the sophists dis-
allow it, and take themselves to be allowed the astonishing conclusion that
Socrates knew everything before even the heavens came into being: ‘And,
by Zeus, you yourself will always know, and know everything, if I please,’
says Euthydemus (296d3–4, my italics).
Knowledge and its conditions

Notice, first, what Socrates is thus committed to: omniscience through-
out time, but an omniscience of which he himself appears to be igno-
rant. Socrates, here at least, seems to find this startling; and we – if we
think that somehow the sophists are the villains of this piece – might do
so too, since they seem to have come up with something that elsewhere
we cheerfully attribute to Plato himself: the conditions for the theory of
recollection.22
Notice, secondly, that this commitment is reached just by virtue of the
sophists’ having refused to allow Socrates any qualifications on what he
knows. In particular:
(a) Socrates is not allowed to say that he knows something but not eve-
rything (so his knowledge cannot be qualified in scope: this works
twice – whether one can ever know just one thing and whether one
can know just one field). (Argument 1: knowing one thing implies
‘being a knower’ simpliciter.)
(b) Socrates is not allowed to insist that some kind of condition on
knowledge is that he knows that he knows it (so knowledge does
not have a higher-order condition, his knowledge is not qualified by
order). (Argument 2: escaping one’s own notice knowing.)
21
They focus the argument on the predicate ‘knower’ (ἐπιστήμων, 295b2), starting with the conces-
sion to Socrates that he is a knower of something (which has not been in dispute), and then argu-
ing that that with which he is a knower is that with which he knows. As in the first argument, the
sophists both rely on Socrates’ concession that he knows something and on the apparently simple
predicate ‘knower’. Socrates, conversely, focuses on what he knows, and so on the qualification of
knowledge, e.g. at 295a8, rather than on a general description of his state of mind. I shall suggest,
below, that this focus may be what causes him such trouble, because of the model of knowledge
into which it buys; but it is, of course, consistent with his escaping his own notice knowing.
22
See McCabe 2009.
(c) Socrates is not allowed to say that he came to know something
(so his knowledge cannot be qualified by time); this precludes, of
course, his learning anything, and rules out the protreptic which
is the explicit object of the dialogue’s search (276–8). (Argument 3:
eternal omniscience.)
Thus, the sophists deny any conditions on knowledge which might
qualify PNC. Socrates’ qualifications are represented by the sophists
as being somehow extraneous, even arbitrary – and on this basis he is
overcome. It remains a question, then, whether he may have grounds
to insist on qualification (especially, here, in terms of the field of knowl-
edge, its order and its temporal qualification) and what the nature of such
grounds would be.
Notice, thirdly, that Socrates’ knowledge or otherwise seems to depend
on what the sophists say – not only is it something he himself might be
completely ignorant of, it is also something that they determine: ‘And, by
Zeus, you yourself will always know, and know everything, if I please.’ It is
this knowledge that Socrates describes as a windfall.
Well, perhaps knowledge is as the sophists portray it?
First, structure: knowledge might be, as some would say, an all-or-nothing
affair – Socrates either knows, or not; and if he knows then he knows all
there is to know. This knowledge has no claims to internal articulation
(there is no logical relation between the one thing Socrates claims to know
and all the other things he ends up knowing, as such);23 and it is neither
ordered (Socrates knows, whether he knows that he knows or not) nor
bounded (there is no restriction on field, no specification of what should
be the content of the things Socrates knows).
Secondly, causation: whether or not Socrates knows depends on some-
thing other than himself. His own engagement is something of which
he may be completely unaware. What is more, whether he knows or not
seems to be caused merely by the decisions of the sophists, so the causal
structure of Socrates’ knowledge is external to, even extraneous to, that
knowledge itself.
Thirdly, value: the value of knowledge of this sort appears to be entirely
contingent – it is a windfall. This, too, presses on the external features of
the knowledge Socrates would thus have – its value is not something con-
veyed, for example, by his own engagement in it.
23
The relation between one piece of knowledge and all of it, that is to say, is based on the logic of the
property of being a knower, and not on anything about the knowledge itself or its structure.
140 M. M. McCabe
One might be reluctant to suppose that the sophists have any such
thing as a theory of knowledge (I suspend judgement, for here, on whether
Euthydemus and Dionysodorus are refuted by being silenced, in the sense
that they can take no part at all in the dialectical exchange and could have
no theory of anything). But the account of knowledge that they seem here
to foist on Socrates is especially externalist – the knowledge that Socrates
would have is determined completely by circumstances outside himself,
by the way things are and by the decisions of others (it is as the sophists
please). On such a theory of knowledge, it seems, the value that it has
depends entirely on external circumstance: it would indeed be a windfall
that as a matter of fact (and despite one’s own ignorance of it) one knows
everything, is right all the time. (Cassandra might not have thought of it
in quite this way …)
Suppose, then, that we find the sophists pressing on their interlocu-
tors two general principles: an Extreme PNC and an extreme externalist
account of knowledge. Neither principle implies the other (for example,
we could maintain extreme PNC and be relativists, or be externalists
but have a more refined version of PNC). But they might be mutually
supportive: this epistemology is consistent with an extreme PNC and
gives a principled way to account for apparent counter-examples to it.
So what the sophists press on Socrates poses a complex threat both to
any more qualified account of contradiction, and to any more subtle
epistemology – a threat to which Socrates (if we take seriously the inter-
textuality of this passage with both Meno and Republic) may be espe-
cially vulnerable.
The role (or lack of a role) of self-knowledge in an account of knowl-
edge is, I propose, central to understanding the difference between
Socrates’ view of knowledge and that of the sophists, as well as his account
of why knowledge may be both valuable in itself and not merely a wind-
fall. This turns on how Socrates understands what is involved if his knowl-
edge escapes his notice – or, conversely, on how he might suppose one can
know either oneself or one’s own knowledge itself. In particular, I shall
argue, he asks us to attend to the question of self-knowledge as knowing
what one does and does not know (we do not escape our own notice), and
to the question of self-knowledge as a central commitment of the knower
(not merely at the will of others nor merely as a windfall). This, in turn,
may give him some kind of resource for rebutting Extreme PNC by show-
ing that there are principled reasons to refine it – if he can explain how
self-knowledge might work.
Making and using

If the sophists force the conclusion that Socrates is omniscient, in the
sequel Socrates himself has no direct rebuttal. Instead, any response he
might have in response to the sophists’ view is offered in the preceding,
Socratic, episode (288d–293a). That this episode comes at an important
moment in the dialogue is emphasised by its framing. For suddenly (289c)
the meek Clinias seems to find his voice, rejecting what Socrates suggests.
Crito interrupts (290e), incredulous that Clinias might indeed have said
such a thing, and the discussion moves into the frame, where it is exposed
to some cutting criticism. It is only at the end of all this that Socrates
declares himself to be drowning in the third wave of puzzlement (293a).
The second Socratic episode (288d–293a) had started by recalling the
conclusion of the first: that wisdom alone is good ‘itself by itself ’ (281de).
That is, wisdom or knowledge24 is both good in itself and causes or pro-
motes the goodness of other things (it does not merely facilitate our
acquisition of existing goods).25 So we should philosophise (288d), because
philosophising will allow us to acquire that very knowledge.
Now that account of philosophising – as the coming to possess knowl-
edge – may be inadequate, Socrates suggests. For – as they already agreed –
unless we know how to use something, the possession of it is of no value
to us. Equally, no knowledge of how to make something is any use to us
unless we know how to use what we have made (289a4–6). So merely hav-
ing the knowledge that is good itself by itself is no good without knowing
how to use it. Why not?
I set aside the complexities of Socrates’ various discussions (here and
elsewhere)26 of the relations between possessing, using and making.27
24
In what follows, I cash in the argument as being about knowledge; but the conditions for knowl-
edge that follow bring this conception of knowledge closer to something we might understand as a
state of soul, even a virtue, hence wisdom.
25
I sidestep the question whether this also claims that wisdom/knowledge is either necessary or suf-
ficient for happiness. Wisdom/knowledge is what I call a ‘transformative’ good, responsible for the
goodness of other goods (281de); hence the repetitive causal language of the first Socratic episode.
At least we shall be better off with it than without it. On the argument see McCabe 2002b, 2005.
26
E.g. at Meno 87–8.
27
It might be thought that here Socrates conflates two models: one of knowledge of how to use a
possession one already has, the other of knowledge how to make (some such possession) and how
to use it. But the possession in question is itself knowledge, so the acquisition of the possession
and the making of it are here treated as analogous – in a dialogue which trades on puzzles about
how we come to know (e.g. in the sophists’ hands at 275d–277c, in Socrates’ hands at 282a–d and
repeatedly in the frame). The sequel makes clear that Socrates has in mind a model in which the
142 M. M. McCabe
Socrates moves directly from thinking about knowledge as a possession to
considering the skill of using it, and comparing that to the skill of making
(by 289b). In so doing, he insists that skills – of making and of using –
may be distinct: just as a possession is no good without the skill to use
it (no good my owning an Aston Martin if I cannot drive), so the skill
of making the possession and the skill of using it are distinct (the skill of
making fast cars is distinct from the skill of driving them). There is, then,
a preliminary point – skills can be distinguished by their fields, or their
objects. And further: skills may be related, and when they are, they may be
ranked: the maker of the Aston Martin is subordinate to the user because
the user sets the ends. This places, then, epistemological demands on both
of the related skills; while the content of each may be distinct, they can-
not be completely opaque to each other – the maker needs to understand
what the user wants and the user to understand what the maker is able to
make. So between ranked skills, or ranked knowledges,28 there needs to be
some shared content, even if the distinct skills are possessed by different
people.
But Socrates and Clinias are seeking that knowledge which is good in
itself, the possession of which is philosophy. Suppose that this knowledge
were such as to make us immortal: this immortality would be no good to
us unless we knew also how to use it. So whatever the knowledge that is
philosophy makes (whether immortality, or something else) this know-
ledge needs also the knowledge how to use it. And if this is to be the
good we seek, then: ‘“We need some such knowledge, my beautiful young
man,” I said, “in which the making and the knowing how to use what is
made coincide”’ (289b4–6).What sort of coincidence does Socrates have in
mind? Not that the knowing how to make this knowledge and the know-
ing how to use it should be the same in content – this is what he has
been denying. Nor, either, the earlier, weakish claim that the two fields
are somehow related, and accessible to each other, since that too is here
taken to be inadequate. So what is it for the knowledge of making and the
knowledge of using to coincide?
If knowledge is my valued possession, it is not something that I can
hand over to someone else to use, at least if ‘handing over’ is meant in the
same sense as handing over the keys to the Aston. Knowledge that I pos-
sess and do not know how to use is not alienable. It is no good seeking out
relation between the various sorts of knowledge is explained in terms of what we should say about
their owner.
28
My rebarbative English here reflects the way that Plato encourages us to reflect on just what would
individuate one ‘knowledge’ from another.
someone else who can use it: it needs to be me who does the hard work.
In that case, the knowledge how to use the knowledge that I have must
be mine, too. The argument that runs up to Crito’s interruption suggests
that we should think about who has the knowledge. There may well be
two fields of knowledge which contribute to whatever makes knowledge
good in itself, but the knower could not have one without the other – the
knowledges he has may be distinct but inseparable. So the knowledge in
question needs to belong to the same person: the beneficiary of what is
good itself by itself. How is that going to work?
The interruption
In the sequel, this condition of coincidence becomes clearer: the point
is not that we are searching here for a single knowledge which we could
possess (or fail to possess) all at once; nor merely for two interconnected
fields, even if they are ranked; but rather for some account of who has
the knowledge they seek, a knowledge that is itself articulated in complex
ways. Could the knowledge they seek be the skill of the speechmaker?
Clinias vigorously rejects the suggestion: for speechmakers have to hand
their speeches over to someone else. We should say the same for generals,
who are unable to do anything with their booty, but must hand it over to
the politicians to use. We should say the same, too, for geometers, astron-
omers and calculators: these people do not make what they hunt (their
‘diagrams’, diagrammata),29 but they find their prey out there in reality
and hand them over – to the dialecticians.30
This sudden and shocking shift, from the well-worn example to some-
thing highly specialised (and lifted from the Republic: compare 522–5), is
marked by Crito’s interruption. What would it be for the geometers to
hand over their diagrams to the dialecticians? Is the point that geometers
do one thing and dialecticians another, so that dialectic is separable from
geometry? Or is the point the one with which Socrates began – that the
knowledge they seek must be such that its possessor both has it and knows
how to use it? The focus of Socrates’ attention, that is, is on the person
who knows; but it includes an account of the complexity of knowledge
itself.
29
This echoes the original contrast between having a possession, making something and using it:
hunting what is already out there vs. making vs. using. Throughout, the focus of Socrates’ attention
is on the knower.
30
The examples resound with the metaphors of the dialogue: in what follows the bird-catcher is por-
trayed as vainly pursuing his catch.
144 M. M. McCabe
If Clinias has read the Republic, perhaps we should have done so, too.
Consider the training of the philosopher king: he has done all this geom-
etry and calculation and astronomy: and only after that does he reach dia-
lectic. But dialectic is synoptic (Republic 537): it is able to deploy its skills
across the subject matter of geometry (mathematical objects somehow
conceived) as well as across astronomy and the rest. So, while dialectic is
not the same as mathematics, and while someone could be a mathemati-
cian without being a dialectician, the converse is not true. So the relation
between dialectic and mathematics is not merely one of rank (where dia-
lectic, like the skill to drive fast cars, sets the ends) but also one of order:
dialectic has a view over the other skills, and somehow includes them.
Likewise, here the higher knowledge ‘will itself know how to use what it
got by making or hunting, and which will be such as to make us blessed’
(290d5–7).31 In that case, the relation between the two is not only one of
rank, but one of order: the higher knowledge includes the lower in its
content.
Regresses, tautologies and circles – the toils of self-knowledge

After Crito’s interruption, the higher-order features of the knowledge
they seek are mirrored in the higher-order features of the discussion itself.
But they seem to be getting into trouble. Perhaps, Socrates suggests, this
account of knowledge simply postpones the account of its value, because
we do not have a determinate account of its content? Like someone vainly
trying to catch birds, what we are trying to find is indefinitely postponed
(290d, 292b–d). Or perhaps the knowledge they seek produces happiness,
which is constituted by the knowledge they seek? In that case they come
round in a circle to where they began (291b). How will they avoid regress
or circularity?
They now propose some kind of compromise. Suppose the knowledge
they seek transmits itself, and is knowledge of itself (292d) – is some kind
of self-knowledge. This suggestion might be thought to meet the coin-
cidence condition by having this knowledge somehow be indexed to
itself, and so trivially belonging to the same person. And thus it may ter-
minate the regress – but at the risk of something worse. For it is hard
to see just what it means to say of knowledge that it is knowledge of
31
This point about the happiness peculiar to gods surely recalls the earlier point about the knowledge
that makes us immortal, 289b1–3.
itself. Is this reflexivity not merely a circle, but some kind of tautology? Is
self-knowledge a dead-end (the ‘Corinthian’ move of 292e)?
Recall the model of the interruption (290e–293a), which allowed the
interlocutors to reflect on what has already been said. Strikingly, the inter-
ruption illustrates how cognitive conditions may indeed be reflective or
higher-order without becoming trivial. If knowledge is qualified by: what
someone knows (the field), who knows it and how (the order), when they
know (the time) – then self-knowledge can be qualified by a higher-order
dimension. After all, we can see that saying that I know that I know is
not the same as saying that I know; saying that I know what I know and
do not know makes a different claim from merely claiming that I do or
do not know a given thing. And, what is more, we might think that these
higher-order features of knowledge are important. If I know, maybe I
know that I know? Maybe that is somehow or other what it is for me to
know?
Socrates is still troubled, and there rises the third wave. Is Socrates’
problem that self-knowledge is impossible or absurd? Perhaps it is, at least
on these conditions. If it is a condition on knowledge that I know that
I know, when I know, and if each use of ‘know’ has the same terms and
conditions,32 how can this not be vulnerable to a regress? If it is a condi-
tion on knowledge that the knowledge that I know is itself the knowledge
that knows it, how can this not be circular? Otherwise, if I insist that
when I know, I know what I know, how does that add anything to what
I claimed to know; how can the claim to know that I know be other than
tautological? But all of these difficulties arise, of course, if I try to insist
that knowledge has an important second-order dimension and that this
focuses attention primarily on the internal dispositions of its owner. Why
not just suppose, against the thrust of Socrates’ argument, that knowledge
just turns up in me, and that when I know, it is just true that I do, what-
ever I think about it?
Consider thus two different things (among many) that have gone on
in this Socratic episode. The first is about how knowledge needs to be
qualified – in field, in order, in time, in value. The second is about how
knowledge needs to ‘coincide’, where this coincidence is explained in
terms of the person who is the beneficiary of the knowing. Now notice
what – in the threats of the third wave – Socrates is here committed to
32
Notice, for the moment, what this second-order dimension does not involve: the arguments of this
section of the Euthydemus are not about subjective consciousness, nor about insisting that when I
know I must be aware that I know. Reflective knowledge may not be at all the same thing as what
we understand by consciousness.
146 M. M. McCabe
saying about knowledge: that we understand it, not merely in terms of its
external conditions (not just in terms of what it is about), but its internal
ones: its articulation, its ordering and its value to its possessor (in terms
of what it is to know). For Socrates, the knowledge they seek needs to
have a reflective dimension – so that the knower knows what they know
(and what they do not) and knows how what they know is articulated and
deployed. And, for Socrates, this reflectiveness is a necessary condition of
knowledge’s having the value they want to give it: for this knowledge to be
good itself in itself, it has to be fully accessible to its owner, and this acces-
sibility needs to be itself a part of its value. Socrates’ epistemology is thus
thoroughly internalist; and this reflects in the importance for it of some
role for self-knowledge.
This, in turn, renders it vulnerable to regress, tautology and circularity.
For knowledge of this kind has to be qualified repeatedly, and without
legitimate qualification regress, tautology and circularity threaten. This
could happen in two quite different ways, of course – on the one hand
by arguing that any qualifications on knowledge themselves violate some
PNC (Extreme PNC); on the other by arguing that there is an alternative
epistemology which does not require such elaborate qualification (such
as the externalist epistemology urged by the sophists). So Socrates’ view
requires both a complex epistemology and a refined PNC. As with the
epistemology of the sophists, refined PNC does not imply this epistemol-
ogy; but the epistemology requires the PNC, on pain of the third wave.
Socrates versus the sophists

But this, then, shows just why Socrates is vulnerable to the sophists’ argu-
ments. For once they claim, via Extreme PNC, that anyone who knows
one thing is a knower tout court and cannot then be a non-knower, the
force of the qualifications on this application of PNC needs to be defended
by more than sheer obstinate refusal to concede. Yet Socrates’ attempt to
produce a defence is stymied by his failure to give a theoretical account, in
the previous episode, of the conditions that would provide the qualifica-
tions needed. A qualified PNC, at this stage, still needs a justification.
Now notice where that justification needs to come. In the second
Socratic episode, the qualifications he seeks to impose bring in the third
wave. In particular, the higher-order features which push the argument
towards self-knowledge provoke regress or circularity or tautology. It would
be an obvious move to reject those features and insist on an account of
knowledge in which it is determined by factors external to the knower
(perhaps by the body of truths, if not by the determinations of the soph-
ists themselves). But then – we might ask in response to the sophistic puz-
zles – just what sorts of constraints could we say there are on someone’s
claim to know? Are the constraints limited to what can be known – in
which case, why not everything, or nothing at all? Or are there interesting
things to say about the condition of the knower? Answering these ques-
tions, after all, will be central to deciding whether Socrates was right to
claim that knowledge, or wisdom, is the only good itself by itself.
And we are inclined to reject the sophists’ conclusion as false. What is
more, the structure of the dialogue itself encourages us to see that con-
siderations of the orders of knowledge may well be important – for the
second-order dimension of the dialogue, prominent at this central point,
relies on the possibility that we are able to reflect critically on what we
think, even if we do not know anything. The dialogue itself invites us
to think that there is something important about what Socrates wants to
say about knowledge, just as it invites us to see that the self-refutation
Socrates urges at the end (that the sophists sew up everyone’s mouths and
their own) describes the disintegration of the last sophistic episode, which
began with an externalist epistemology. All of that suggests that the inter-
nal conditions on knowledge that Socrates urges and the dismal external-
ism of the sophists are still dialectically engaged by the end.
Coda: Socrates … and Aristotle

The sophists deny Socrates an appeal to common sense in blocking their
arguments – he cannot just say, of the odd outcomes they produce, that
they are false: not least because they deny, anyway, that there is such a thing
as falsehood. He needs, instead, to produce a theory that countervails; and
the theory he wants is in trouble from the start. However, that theory has
elements which – were it possible to explain how self-knowledge of that
kind were possible – could do the blocking he needs. He may not need
rescuing by the sophists, if he is able to breast the wave himself.
This suggests that what Socrates needs is not agreement to PNC plus
some matters of fact or observations to support the qualifications he
needs – since the merit of those qualifications is itself up for grabs in
the sophists’ account. Instead, he needs to find a theory about knowl-
edge that complements the logical principle, PNC. This suggests, in turn,
that where Socrates needs to go in protecting a modified PNC is to other
theories (theories of identity, theories of knowledge) to defend whatever
qualifications he needs. My proposal is that deep in the structure of the
148 M. M. McCabe
Euthydemus is a discussion of just how PNC can be modified; and in
particular an (unresolved) inquiry in the passages I have just discussed
into how some of those modifications might be achieved by a theory of
knowledge.
Notice, first, that the omniscience argument shares an assumption
about knowledge that Socrates would approve of: that knowledge, when
acquired, is something that denies qualification: being a knower is know-
ing everything. This is the good that Socrates seeks at 281; and the windfall
he may have without knowing it. But unless this knowledge can be quali-
fied over time, it will be impossible to learn (to come to know something;
or to know one thing and then another: this is the focus of attention in
the discussions of protreptic that surround the first set of sophistic argu-
ments against the possibility of learning, 276a–278e, and that surround
the puzzles of change throughout the third sophistic episode). So Socrates’
account requires knowledge to be both holistic and learnable. But unless
such knowledge is structured, then there will be no distinction between
knowledge by use and knowledge for making (required for the arguments
at 288–291); and any such distinction relies on the qualification of knowl-
edge (by use, for example, or for making). Furthermore, it is a failing of
any account of knowledge – or so Socrates suggests by his puzzlement
at 290–1 – that we might be unaware that we know, when we know. So
his account, therefore, requires that knowledge be qualified also in terms
of order: we can (or perhaps must) know that we know when we know.
Socrates’ account of knowledge – as his resistance of the omniscience
argument suggests – is internalist; and such an account demands qualifi-
cations. It relies, therefore, on a qualified PNC.
The deep structure of the dialogue lies in the connections between the
arguments that are explicitly advanced in the framed dialogue and the
frame in which they are advanced. In the Euthydemus in particular, as I
have argued, these interconnections are especially complex and baroque:
especially in the way that one, apparently discrete, episode of the dialogue
is reflected on by its sequel, and thence in the frame itself. This technique
in the dialogue then invites the reader to see just where issues of principle
need to be understood before the standoff between Socrates and the soph-
ists can be resolved.33
In closing, let us notice the difference between Plato’s approach to PNC
and Aristotle’s account of the most important principle of all (Metaphysics
1005b19–23): ‘It is impossible for the same thing to belong and not to
33
On this compare and contrast Rowe 2007a, esp. 1–51, 266–72.
belong at the same time to the same thing and in the same respect (adding
whatever else should be added to meet the logical difficulties).’ There is an
argument, I think, for finding the Euthydemus in the near background of
Metaphysics Γ, and, in consequence, a real contrast between the dialecti-
cal position of the Euthydemus and Aristotle. For Aristotle it seems that
we have the authoritative principles – the basic principles of logic, PNC
and LEM, as well as some other thoughts about relativism – and then the
qualifications supplied ad hoc (supplying the things needed to meet the
‘logical difficulties’). For Aristotle, some principles of logic are of a dif-
ferent order from other philosophical principles. These principles of logic
are necessary for any articulation of other principles, and so they are prior
to other important philosophical principles about, for example, identity
or cognition.34 For Plato, however, the same may not be true. Perhaps for
Plato – as the complex discussion of PNC in the Euthydemus exemplifies –
general metaphysical principles are not subordinate to principles of logic,
but are needed to defend them, and vice versa. So Plato’s view may be that
all these principles are interconnected and interdependent. Might he be
right?
34
Plato and Aristotle do agree, as the Euthydemus shows, that PNC cannot be directly demonstrated,
but must be shown by refutation.
ch apter Ni ne
Why was the Theaetetus written by Euclides?

Michel Narcy
Introduction
Part of the plot of Plato’s Theaetetus is that it was written not by Plato,
but by Euclides; indeed, it was virtually dictated to Euclides by Socrates
himself! At the beginning of the dialogue, Euclides reminds his friend
Terpsion of the time when he went back and forth from Megara to Athens
to visit Socrates, using his encounters with him to complete his transcript
of the account by Socrates of his dialogue with the young Theaetetus. As is
well known, the Theaetetus is the only Platonic dialogue in which, instead
of being criticised (as in the Phaedrus), writing is presented with empha-
sis as a guarantee of truthfulness in reporting the words or thoughts of
an absent speaker. In the prologue, the absent speaker is Socrates; and
Euclides is proud of having been able to write down the exact words of
Socrates’ arguments with Theaetetus. In the course of the reported argu-
ment, it is Socrates himself who stresses the necessity of sticking to what
was written by his main philosophical opponent, Protagoras, since the lat-
ter is no longer alive and thus not in a position to explain his own thesis
further.
This warning recurs three times in Socrates’ mouth, but couched in two
different ways. On the first two occasions, Socrates represents Protagoras
as blaming him for not taking into account what he has written (162d),1
or for interpreting it in an unfair or unintelligent way (166c).2 In these
two cases, in order to defend himself against Socrates’ attacks, Protagoras
1
Tht. 162d6–e2: ‘gods, whose existence or nonexistence I exclude from all discussion, written or spo-
ken’ (italics mine). Unless otherwise stated, the translations of Platonic dialogues are taken from
Cooper 1997.
2
Tht. 166c7–9: talking about pigs and baboons, ‘you show the mentality of a pig yourself, in the way
you deal with my writings (εἰς τὰ συγγράμματά μου)’ (my italics). See Burnyeat 1990, n. 19 ad loc.,
suggesting that the connotations of ‘the Greek pig’ are ‘intellectual as much as social or moral’. It
should be noted that the reproach spoken by Protagoras is immediately followed by his reassertion:
‘I take my stand on the truth being as I have written it’ (166d1; my italics again).
150
Why was the Theaetetus written by Euclides? 151
asks him to recall exactly what he wrote. But on the third occasion, it is
Socrates, by contrast, who maintains that, since it is impossible for him to
ask Protagoras for explanations of his famous man-measure principle, the
only way of analysing it rigorously is to stick to the exact wording of this
principle in the sophist’s writings (169e, 171b).3
Perhaps we can see in this insistence on the necessity of reporting the
words of the opponent accurately, and thus, when he is absent, resorting
to his writings, the reason why it was so important for Euclides to write
down the account by Socrates of his conversation with Theaetetus, and
why he was so anxious to obtain from Socrates the exact wording of the
arguments. It is plausible to suggest that this concern with exact word-
ing has something to do, as we shall see later, with Socratic definitions of
eristic;4 and I shall suggest that this is the reason why Euclides was chosen
to be Socrates’ secretary. It is very unlikely that Plato cast Euclides in the
role of Socrates’ secretary and set the scene in Megara without any specific
intention. We do not know the exact dramatic date of the reading we
are invited to attend at the house of Euclides or whether or not Euclides
was still alive when Plato wrote the Theaetetus (some time after 370 BC).
But we do know from later tradition that Euclides was at this time well
known as the founder of the Megarian school, otherwise called the ‘Eristic’
school;5 and Plato must have known that his readers were aware of that
fact. Nevertheless, as presented in the prologue of the Theaetetus, the most
3
Tht. 171b7–8: ‘But Protagoras again admits this judgment [sc. the opinion of those who take his own
as untrue] to be true, according to his written doctrine?’ (final italics mine). At 169e7, ‘an appeal to
his own statement (ἐκ τοῦ ἐκείνου λόγου)’ necessarily refers to the sentence ‘things are for every
man what they seem to him to be’ (170a3) as to a written sentence: hence the present ‘says’ used by
Socrates (170a3) and in turn by Theodorus (170a5), the subject of which is Protagoras as the peren-
nial author of this written and thus immutable sentence.
4
Cf. R. 5, 454a7: ‘To pursue mere verbal (κατ’ αὐτὸ τὸ ὄνομα) contradictions of what has been
said’; 454b5–6: ‘we’re bravely, but in a quarrelsome and a merely verbal fashion (ἀνδρείως τε καὶ
ἐριστικῶς κατὰ τὸ ὄνομα), pursuing the principle’; Tht. 168b8–c1: ‘base [one’s] argument upon the
use and custom of language … drag words this way and that anyhow’ (trans. modified). I will come
back later to the last passage, where this critical definition of eristic and of how it differs from dia-
lectic is put into Protagoras’ mouth.
5
See Diogenes Laertius 2.106 and Suda s.v. Εὐκλείδης (= SSR II A22); [Galen], History of Philosophy 7
(= SSR II A27). I take ‘Megarian school’ in a broad sense. About the actual existence of a Megarian
school, most scholars are currently doubtful (see Giannantoni 1990: iv. 33–66, referring to Fritz 1931
and Döring 1972). What Giannantoni discusses is whether Euclides was the founder of an institu-
tionalised school such as the Academy of Plato or the Peripatos of Aristotle. But not being the head
of an institution such as the Academy did not prevent Euclides from being in Megara the leader of a
more or less informal, yet quite recognisable, circle of ‘companions’, just like Socrates in Athens. On
the other hand, nobody contests that Euclides developed a variant form of Socraticism ‘through the
eristic components already present in Socrates’ method’ (Giannantoni 1990: iv. 56). See also Zilioli
in this volume, p. 171, for the idea that Plato’s use of Euclides establishes a link with the Megarian
school.
152 Michel Narcy
significant feature in the characterisation of Euclides is his enthusiasm for
recording meticulously his master’s voice. Thus, even if we are right to
think that Plato is alluding through him to the Megarian or Eristic school,
Euclides is presented straightforwardly in the Theaetetus as a Socratic.
Is it possible to be both a Socratic and an Eristic? In the scholarly lit-
erature, ‘eristic’ is ordinarily used as a synonym for ‘sophist’, and we are
used to thinking of a clear-cut opposition between the Platonic (or indeed
Xenophontic) Socrates and the sophists. Indeed, from the Meno to the
Philebus, every time Socrates mentions eristic, he opposes it to dialectic or
what he describes as a genuinely philosophical conversation.6 Every time,
except one: in the Euthydemus (272b10), Socrates not only mentions eris-
tic without making any allusion to dialectic but, instead of maintaining as
usual the superiority of dialectic over eristic, he declares that he himself is
longing for eristic and tries to persuade his friend Crito to come with him
and learn it!
Socrates’ desire for eristic: the Euthydemus

The Euthydemus is the narration by Socrates, intended for his friend Crito,
of the encounter he had, the day before, with Euthydemus, the epony-
mous character of the dialogue, and his elder brother Dionysodorus, both
of them being presented at the beginning as sophists. The conversation
rapidly turned into a competition between Socrates and the sophists
about who was more effective in teaching virtue and, in the first instance,
in motivating a boy to want to acquire it, namely Clinias, a young fol-
lower of Socrates, who was available for this experiment. Socrates did not
emerge as the winner of this contest; and that is the reason why he con-
ceived the idea of being taught eristic by the two sophists. By telling Crito
about this misfortune, he hopes to persuade his friend to join him. After
presenting this proposal as the reason for telling Crito the whole story, at
the end of his account, after reporting how he asked the two brothers to
6
See Men. 75cd: ‘One of those clever and disputatious debaters (τῶν σοφῶν τις εἴη καὶ ἐριστικῶν
τε καὶ ἀγωνιστικῶν)’ vs. ‘friends [who] want to discuss (dialegesthai) with each other’; at 80e–81a,
the ‘debater’s argument’ (eristikon logon) is opposed to ‘the priests and priestesses whose care is to
be able to give an account (logon didonai) of their practices’. At R. 454a, erizein and dialegesthai are
opposed as ‘pursuing mere verbal (kat’ auto to onoma) contradictions of what has been said’ and
‘examining what has been said by dividing it up according to forms (kat’ eidē)’. At 499a, ta eristika
are contrasted with ‘arguments (logoi) which search out the truth … for the sake of knowledge’, i.e.
philosophical discussions. In the Philebus, ‘dialectic’ indicates a method of division rather than a
way of discussing, but here too (at 17a) we get an expression of the difference between ‘dialectical
and eristic discourse’ (dialektikōs vs. eristikōs poieisthai tous logous).
take Clinias and himself as their pupils (304b5), he urges Crito again to
come with him and study with them (304b7–c1).
The usual scholarly response is to take as ironical Socrates’ declaration
that he wants to learn eristic, and, similarly, his insistence on involving
Crito in this process. In my book on the Euthydemus, I have brought out
the problems in this line of interpretation.7 Certainly, we can suppose that
Socrates was ironical in his treatment of the two sophists, but why should
we imagine that he is also ironical in his treatment of Crito? The request to
the two brothers to be taught their expertise, namely eristic, may well be
ironical; but, on the following day and out of their presence, why should
Socrates tell Crito he has the same desire, if it is not true? Also, during
the discussion on the day before, Socrates might not have been ironical
throughout. Even assuming that he was only pretending to be overcome
by the silly sophisms of the two brothers, he had no reason to be ironi-
cal when he was talking to Clinias in order to give them an example of
how to train a young man in virtue. But he has ended in an aporia about
the nature of the supreme science (292e): again, if he was ironical in his
attitude to the sophists, why should he be ironical to Clinias? If he was
not ironical, and the aporia was real, then he really failed to show that he
was a better teacher than the sophists.8 So it seems that we are bound to
take seriously Socrates’ desire to learn eristic and to involve in this process
some of his friends and students. After all, this desire is the reason offered
by Socrates for telling Crito what happened the day before; and what else
could Socrates’ reason for telling this tale be? More precisely, why other-
wise should Plato show him expressing this desire?
7
Narcy 1984: ch. 2.
8
The most common way of making sense of the Euthydemus is to take it as discrediting eristic by con-
trasting it with the ‘Socratic method’, and in this view Socrates’ incapacity to display his superiority
to the two sophists is obviously ironical. I am not prepared to endorse this view without reserva-
tions, because the ‘Socratic method’, as defined by other ‘Socratic’ dialogues, is clearly misrepre-
sented in the Euthydemus. To anyone who was brought up in classic Platonic scholarship, it is clearly
shocking to hear Socrates say that he is longing for eristic, and this seems to be the main reason why
the ironical interpretation of his saying it is so compelling. But perhaps it is because we are used not
to translating ἐριστική so much as to transliterating it, as ‘eristic’ (or equivalents in other modern
languages). Should ἐριστικὴ τέχνη be translated, I think that ‘art of debating’ would be a quite
good translation. And who will maintain that the ‘Socratic method’ is alien to the art of debating?
Certainly, Socrates is eager to distinguish himself from those who debate about names instead of
things; but it is worthy of notice that to debate about names does not imply to be unfair or to lack
seriousness, precisely because there are cases in which seriousness or honesty in discussion requires
names to be taken seriously – my claim is that the discussion of Protagoras in the Theaetetus is such
a case. In other words, there are cases in which one cannot help but debate about names. On this
view, the fact that Socrates had wished to learn the art of debating – even of debating about names –
is not at all shocking. For another reading of the relationship between Socratic dialectic and that of
the brothers, see McCabe in this volume.
154 Michel Narcy
One could retort: what reason could Plato have for showing Socrates
seriously expressing a desire to learn eristic? To this question the response
is that Plato probably considered eristic to be a more serious business than
most of his interpreters. Let us remember that, in the Sophist, sophistic is
presented as one of two kinds of eristic, the other one being called ‘chat-
ter’, a name that Proclus interprets as referring to dialectic.9
So, which type do Euthydemus and Dionysodorus belong to – the
sophistic or the chattering one? Near the start of the dialogue, the two
brothers are characterised (once and only once in the whole dialogue) as
sophists. In 271b9–c1 we read: ‘They look like some new sophists’ (καινοί
τινες αὖ οὗτοι, ὡς ἔοικε, σοφισταί). Interestingly enough, we do not
know whether Socrates or Crito utters these words. All modern schol-
ars, it seems, attribute them to Crito, who would thus be saying: ‘I don’t
know either of them, Socrates. They look like some new sophists. Where
are they from? And what is their wisdom?’ But up until Stallbaum (1836),
there were editors and translators, beginning with Ficino, who divided up
the sequence of words in another way:
Crito: I don’t know either of them, Socrates.
Socrates: They look like some new sophists.
Crito: Where are they from? And what is their wisdom?10
In my opinion this older arrangement makes more sense than the more
recent one. At the end of the dialogue, Crito reports a conversation
between himself and an anonymous interlocutor who criticises Socrates
sharply for wasting time debating with the two brothers, and to this inter-
locutor Crito says in defence of Socrates: ‘But surely philosophy (philos-
ophia) is a charming thing (charien ge ti pragma)’ (304e6–7). The word
‘philosophy’ has already been used in the course of the dialogue, that is,
the conversation reported by Socrates. In particular, Socrates says that he
obtained from Dionysodorus the declaration that he and his brother are
‘the men of the present day best able to exhort a man to philosophy (eis
9
Proclus, On the Parmenides 658.22–3: ὑφ’ ὃ [sc. τὸ χρηματοφθορικὸν] δηλονότι τάξομεν τὸν
διαλεκτικόν (τὸν codd.: τὸ Luna, Steel). According to Proclus, the Eleatic Visitor does not use
the term ‘eristic’ in a pejorative sense, but only as ‘the fact of opposing arguments to one another
and of raising objections (τὴν ἀντίθεσιν μόνην τῶν λόγων καὶ τὴν ἐνστατικὴν ἐνέργειαν), for it
is possible to contradict rightly and to wrangle in a good as well as in a bad way (ἔστι γὰρ καὶ
ἀντιλέγειν ὀρθῶς, καὶ ἐρίζειν εὖ καὶ κακῶς [codd.: εὖ καὶ καλῶς Luna, Steel]), since eris is said to
be double’ (658.9–15).
10
My translation. In fact the attribution of these lines has been controversial since the first editions of
Plato: Ficino divides them between Crito and Socrates, but Stephanus attributes them all to Crito,
and so does Cornarius. See Bekker 1826 ad loc., according to which there is already a discrepancy in
the manuscripts.
philosophian) and the practice of virtue’ (274e9–275a2). But we know from
the very beginning of the exchange (271a3) that Crito did not hear any of
this conversation, and that is the reason why he asks Socrates to tell him
about it. So if, without having heard anything, or at least anything clear or
definite (ouden saphes), he thinks that Socrates was doing philosophy with
the two brothers, why should he call them sophists on the following day?
Now, if it is Socrates who calls them sophists, what does he mean by
saying that ‘they look like some new sophists (kainoi sophistai)’? Kainoi is
ambiguous: it can mean ‘newly arrived’,11 and in this case the two broth-
ers may be viewed as representatives of the historical sophistic movement,
perhaps Protagoras’ followers. But kainoi can mean also ‘sophists of a new
kind’,12 and in this case we may think, together with Louis-André Dorion
(2000: 37), of ‘philosophers who rather look, in the eyes of Plato at least,
like sophists of a new kind’.
We all know that Plato is quite able to present as sophists certain char-
acters in his dialogues who have been regarded, instead, as philosophers
by some thinkers and readers, whether contemporary or not, as well as
by modern scholars. Dorion suggests that this applies to Euthydemus
and Dionysodorus. In his view, the fact they are called ‘new sophists’
does not mean we are being asked to identify them as representatives of
the so-called fifth-century ‘sophistic movement’. Rather, they should be
seen as certain contemporaries of Plato with whom he is competing in
his appropriation of the Socratic legacy, namely the Megarians. Indeed,
Dorion argues, as dialecticians, the Megarians were especially suited to be
the target of the attack carried out in the Euthydemus.13 Thus, Euthydemus
and Dionysodorus, characterised by Socrates as sophists, constitute in fact
some sort of caricature of the branch of Socraticism that the tradition has
called the Megarian school.
This raises another question. If Euthydemus and Dionysodorus repre-
sent (even in caricature) the Megarian school in the Euthydemus, there is
necessarily some connection between them and Euclides. Dorion does not
11
Jowett 1892 translates: ‘A new importation of Sophists’.
12
Sprague in Cooper 1997 translates: ‘another new kind of sophist’.
13
Dorion takes seriously the possibility of a connection between the sophisms elaborated by the two
brothers in the Platonic dialogue and several Megarian arguments of which we have testimonies.
This connection would be confirmed, according to him, by external data, such as the mention of
Euthydemus and Dionysodorus in SE, M. 7.13 (= SSR II B12 = Eubulides fr. 63 Döring), together
with four other philosophers known as Megarians. Dorion also notes that Diogenes Laertius, after
reporting the invective addressed by Socrates to Euclides because of his interest in eristic (‘You will
be able to get on with sophists, but with men not at all’), adds: ‘For he [Socrates] thought there was
no use in this sort of hair-splitting, as Plato shows in the Euthydemus’ (DL 2.30, trans. Hicks 1925 =
SSR II A3 = Euclides fr. 9 Döring).
156 Michel Narcy
address this point nor, as far as I know, do any of the scholars he mentions
who also see the two brothers as Megarians.14 But it is easy to make objec-
tions to his hypothesis on the basis of the contrast between the presenta-
tion of the two ‘sophists’ in the Euthydemus and that of Euclides in the
Theaetetus. First, Euclides is in no way presented as a sophist, nor as an
eristic. Secondly, while Euthydemus and his brother, whether Megarians
or not, are clearly caricatures, there is nothing of this kind in the charac-
terisation of Euclides. Last but not least, Euclides appears to be as faithful
to Socrates as possible, whereas in the Euthydemus the two brothers are
openly disdainful of him.
The only way to escape this objection, and to my mind the way to
make the best sense of both the Euthydemus and Euclides’ ‘authorship’
of the Theaetetus,15 is to question Dorion’s claim that in the Euthydemus
Plato intends to challenge the claim of the Megarians to be legitimate
heirs to Socrates.16 If I am right in claiming that we have to take seriously
Socrates’ desire for eristic, it follows that Socrates’ role in the Euthydemus
is not to condemn eristic and those, possibly the Megarians, who practise
it, but rather to warn his friends, in the person of Crito, or Plato’s own
disciples, represented by the character Clinias, that they would be wrong
to despise eristic even when it amounts to quite silly jokes. Perhaps this
is the meaning of the last words spoken by Socrates to Crito: the same
Socrates who has opened the conversation with Crito by describing the
two brothers as sophists closes it by saying: ‘Pay no attention to the practi
tioners of philosophy, whether good or bad. Rather give serious considera
tion to the thing itself ’ (307b6–8). This amounts to saying that these ‘new
kind of sophists’ are also some kind of philosophers, maybe ‘good or bad’,
but philosophers in any case; and this consequently casts doubt on the
14
The consensus about the identification of Euthydemus and Dionysodorus as Megarians is less
widespread than Dorion’s claims in 1995: 49 n. 4, and 2000: 36 n. 9. According to Thompson 1901:
278–81, and Hawtrey 1981: 30, Euthydemus and Dionysodorus represent Antisthenes as well as the
Megarian school.
15
Many readers will naturally think of two alternatives: either Plato changed his mind about Megarian
eristic between the Euthydemus and the Theaetetus, or he criticises in the Euthydemus a degenerate
version of the Megarian school. Both explanations imply that Socrates is ironical throughout the
Euthydemus, and are thus entangled in the difficulties discussed above.
16
See Dorion 2000: 41: ‘Since these [sc. the Megarians] pretended, as heirs to Socrates, to practise a
true dialectic, it became essential for Plato to show clearly that Socratic dialectic, as he viewed it,
had nothing in common with eristic, apart from a purely formal similarity … It is thus quite plaus-
ible that Plato has transferred to the Euthydemus, where the protagonist is Socrates, the debates
which, at this moment, were bringing him into conflict with the Megarians … This device is very
clever, for it amounts to saying that Socrates himself repudiates some of his followers, in this case
the Megarians’ (my trans.).
impression we may form from other dialogues that Socrates assumes there
is a clear-cut opposition between sophists and philosophers.
Socrates on what a sophist is

We can obtain from another dialogue confirmation that Socrates is not
so sure of the difference between sophist and philosopher. At the begin-
ning of the Sophist, in order to draw the Eleatic Visitor into discussion,
Socrates tells him that: ‘the genuine philosophers (οἱ μὴ πλαστῶς ἀλλ’
ὄντως φιλόσοφοι) … sometimes take on the appearance of statesmen, and
sometimes of sophists. Sometimes, too, they might give the impression
that they’re completely insane’ (216cd). So here, according to Socrates,
‘sophist’ is one of the appearances that the philosopher is likely to take on.
One is tempted to conclude that philosopher is the genus of which states-
man, sophist and madman constitute the species, so that the sophist, as
I have just suggested in the light of the end of the Euthydemus, would be
some kind of philosopher. But then Socrates reformulates his statement
in question form: do the Visitor’s fellow citizens take as one, two or three
different kinds (genē) the people referred to by these three names? And
surprisingly, in the reformulation, the three names – and so the potential
species – are no longer statesman, sophist and madman, but statesman,
sophist and philosopher (217a4). If we set aside the replacement of mad-
man by philosopher, this wording implies that philosopher is no longer
the common genus, but one of the potential species, side by side with, but
different from, sophist.
We have here only one of several hints, in Plato, that the difference
between philosopher and sophist – and between Socrates and sophists – is
not as clear-cut as many commentators maintain. Certainly, I shall not
pretend that Socrates could be confused with one of the fifth-century
sophists he is shown debating with in Plato’s dialogues. Plato’s Socrates is,
undoubtedly, an opponent of Gorgias and his pupil Polus, Hippias and,
above all, Protagoras, whose doctrines, political as well as physical and
epistemological, he refutes at length in no fewer than two dialogues. But
the term ‘sophist’ is not always, in Plato’s writing, a term describing a
determinate historical group; in my opinion, it is not a historical term at
all. When Protagoras, in the dialogue named after him, applies the term
to himself (Prt. 317b), it is not in order to link himself to contemporar-
ies such as Hippias and Prodicus, who are both present in Callias’ house,
or Gorgias or Thrasymachus. He does not do so to claim membership in
some sort of new movement, in the famous ‘sophistic movement’ which
158 Michel Narcy
was the subject of Kerferd (1981). On the contrary, Protagoras uses this
term to link himself to older poets such as Homer, Hesiod and Simonides,
or hierophants and prophets such as Orpheus and Musaeus. Significantly,
he does not name any of those among his contemporaries that we call
‘the sophists’; the only contemporaries he names are largely or completely
unknown to us. What, then, is the meaning of ‘sophist’ in Protagoras’
mouth? Plato makes him quite explicit on this point (317b4–5, my trans.):
‘I acknowledge myself to be a sophist, that is, to teach people’ (καὶ
παιδεύειν ἀνθρώπους: the καὶ is epexegetical).
A sophist is a teacher. That is the reason why Protagoras can claim com-
munity with such prestigious or revered figures as Homer or Orpheus:
they were, and still are, in the time of Protagoras and Plato, the teachers
of Greece, and sophist is the new name – according to Protagoras, the true
name – of a teacher, whatever he teaches.17 It is clear that in this sense, if
a philosopher, such as Socrates himself, is a teacher, one can rightly name
him a sophist. Socrates’ ambivalence over the relationship between the
terms philosopher and sophist at the beginning of the Sophist thus appears
to reflect common linguistic usage.
On the other hand, the prologue of this same dialogue, the Protagoras, is
celebrated for the quite unequivocal contrast that Socrates draws between
sophist and philosopher. As is well known, this prologue consists of the
conversation Socrates manages to have with the young Hippocrates before
going to Callias’ house to meet the sophist Protagoras. Like the Sophist,
the conversation of this dialogue centres on the question of the nature of
a sophist. Strikingly enough, Socrates’ answer to this question is given in
the same terms as those used by the Eleatic Visitor in his second and third
definitions of a sophist. Insofar as he delivers teaching for money, says
Socrates to Hippocrates, a sophist is a kind of merchant (emporos tis) or
retailer (kapēlos, 313c5) – which inevitably recalls Sophist 223d6–11, where
the business of exchange (metablētikē) is subdivided into retail (kapēlikē)
and trade between cities (emporikē). But let us pay some attention to the
difference between Socrates’ method of definition and the Visitor’s. Unlike
the Eleatic Visitor, Socrates does not proceed by the method of division:
his way of characterising the sophist is by means of a comparison. As
soon as the sophist is said to appear like a salesman, he is compared to a
salesman in the most ordinary sense, that is, as someone who sells goods
17
Probably we have the same neutral use of the term ‘sophist’ at the beginning of the Lysis (204a6–7)
where Socrates calls Miccus, the one who teaches (didaskei) in the new palaestra, hikanos sophistēs:
‘a good teacher’ (‘a very eminent professor’: Jowett 1892).
designed for the body, not for the soul, and especially food. But, by con-
trast with the neutral characterisation in the Visitor’s division, it rapidly
becomes clear that Socrates intends by this comparison to depreciate the
sophist, or at least to make Hippocrates mistrustful of Protagoras. In fact,
the only point that the comparison is used to make is that, unlike the doc-
tor, the salesman does not know whether what he puts on sale and adver-
tises is good for the health of the body or not. Similarly, Socrates claims,
the sophist does not know whether or not the types of knowledge that he
offers and advertises are good for the soul of his pupils (or customers).
The implication is that, in Socrates’ view, trade always has some touch of
dishonesty: it appears that the aim of Socrates’ assimilation of the sophist
to a salesman is to transfer to the former the seemingly inevitable dis-
honesty of the latter. So, before Hippocrates meets Protagoras, and before
Protagoras claims to be a sophist, Socrates has inculcated in the former’s
mind a pejorative meaning of the word ‘sophist’.
Thus, in the Protagoras, neither in Protagoras’ mouth nor in Socrates’
mind is the word ‘sophist’ a historical term. Although, according to
Protagoras, ‘sophist’ is the proper term to designate an activity as ancient
as teaching, the antiquity and prestige of which is testified by such great
figures as Homer, Orpheus and so on, in Socrates’ eyes it seems the very
word ‘sophist’ is a value judgement, and one that is immediately negative.
It is clear that, insofar as we think of Socrates as philosophising, here,
unlike in the Sophist, we are bound to think of him as different from,
even opposed to, a sophist. Because he lacks knowledge, Socrates does not
teach, and a fortiori does not do so for money; unlike a salesman trying to
sell goods designed for the soul, he is concerned not with the success of
his commerce, but the soul of his addressees.
So this Socrates who is speaking to Hippocrates does not seem to be at
all ambivalent about the respective value of sophist and philosopher. The
same Socrates, in many other passages of Plato,18 appears to be anxious
to mark the boundary between eristic and dialectic, and presents him-
self as the champion of dialectic against eristic. But he would not be able
to mark this boundary without knowing what eristic is as well as dialec-
tic; and it will become clear shortly that he is quite able to go beyond
this boundary and practise eristic when he thinks it necessary. In fact,
the Euthydemus seems to be the only dialogue in which he appears to be
unable to do that.
18
See n. 6 above.
160 Michel Narcy
The disappearance of the distinction between

dialectic and eristic
Let us go back to the Theaetetus and to the celebrated paradox it con-
tains: in the peroration of the so-called ‘Apology of Protagoras’, Socrates
represents the latter as drawing a distinction between ‘controversy’ and
‘discussion’ (revised Levett translation in Cooper 1997), that is between
agonizesthai and dialegesthai (167e4–5). Agonizesthai consists in ‘playing
about and tripping up one’s opponent as often as one can’ (167e5–6),
which reminds us of Socrates’ remarks on the first demonstration of their
skill by the two sophists of the Euthydemus: ‘I call these things “frivol-
ity” (paidian) because even if a man were to learn many or even all such
things, [he] would only be able to make fun of people, tripping them up
(hyposkelizōn) and overturning them (anatrepōn)’ (Euthd. 278b3–7). On
the other hand, dialegesthai consists in ‘helping one’s opponent to his feet
again (epanorthoun) and pointing out to him only those of his slips that
are due to himself ’ (167e7–168a1). Thus, the difference between agonis-
tic (a synonym for eristic) and dialectic amounts to a difference between
a controversial and somewhat hostile attitude, and a critical but friendly
one. Also this link with friendship is to be connected with the way in
which Socrates, in the Meno, makes friendship a condition of dialectic. By
contrast with eristic, the dialectical way of answering, which is distinctive
of philoi, consists not only in answering the truth, but also doing so ‘in
terms admittedly known to the questioner’.19 Supplementary references
could easily be found in Plato to show that the distinction that Protagoras
is advocating is unequivocally Socratic in spirit.
This is no surprise, one could observe, since it is Socrates who is speak-
ing, suggesting how Protagoras might defend himself against Socrates’
previous attacks. True; but the question is this: why does Socrates lend his
own conception of dialectic to Protagoras? To understand this, we have to
remember the reason why Socrates decides to speak on Protagoras’ behalf.
As he tells us at the beginning of his supposedly Protagorean speech,
Theaetetus, young and inexperienced as he is, was unable to answer
Socrates’ objections effectively. But these objections, Socrates himself
has previously confessed, were similar to those of some ‘mercenary skir-
misher of debate’ (πελταστικὸς ἀνὴρ μισθοφόρος ἐν λόγοις, 165d6); in
other words, Socrates has acted towards Theaetetus as if he were himself a
19
Men. 75d3–7. Contrary to what I wrote in Narcy 1994: 88 n. 2, I adopt at d7 the emendation of
ἐρωτώμενος in place of ἐρωτῶν (cf. Narcy 2007: 28 n. 26).
practitioner of eristic. Hence the final plea of ‘Protagoras’: when Socrates
represents the sophist as urging him to prefer dialectic to eristic, he does
not allege that the historical Protagoras shared his own conception of dia-
lectic and condemnation of eristic; he simply keeps beating his breast. So
we hear from Socrates himself not only that he is perfectly able to practise
eristic, but also that he is actually doing so. However, up to this point
he has expressed this in a self-deprecating way, so that we are justified in
expecting him to make amends and to assume from now on a properly
dialectical way of refuting Protagoras’ thesis.
This change of attitude would consist in no longer trying to ‘trip up and
overturn’ Protagoras and being content with ‘pointing out to him only
those of his slips that are due to himself ’. Indeed, as soon as he starts to
speak for himself again, Socrates offers a concession made in his defence
by ‘Protagoras’, namely that ‘some men are superior to others in questions
of better and worse, these being “the wise”’ (169d6–8). Here he refers to
what is the core of Protagoras’ defence, namely that the wise man is not
the one who knows the truth better than others (since ‘each one of us is
the measure of what is and of what is not’), but the man who can ‘change
a worse state into a better state’ (167a4). As Socrates will show later at
length, it follows from this concession that ‘some men are wiser than their
fellows and others more ignorant’ (171d6–7).20 But this concession, which
will turn out to be ruinous for the man-measure principle, was in fact
nothing but an interpretation – the most charitable one that Socrates was
able to propose – of what Protagoras claimed when alive. So, is the ‘slip’,
if it is an actual slip, ‘due to himself ’? Is not Socrates once again refuting
Protagoras, not on the basis of answers similar to those that the sophist
would himself have given, but through concessions other than those he
would have agreed with?21
It is true that, in order to respond to the first point made by ‘Protagoras’
in his defence, namely that so far it is not the sophist himself but
Theaetetus who has been overturned, Socrates has replaced Theaetetus
with Theodorus. He has replaced a teenager with the only adult in the
audience and, what is more, a former companion of Protagoras, and
thus, apparently, an authoritative witness to the sophist’s doctrines. But
Theodorus has already told Socrates that he ‘could not consent to have
[Protagoras] refuted through [his] admissions’ (162a5–6), before confessing
20
The examples will be first doctors and legislators (171e–172b), but then also farmers, musicians,
cooks and orators (178de).
21
Cf. 166a7–b1.
162 Michel Narcy
that he has turned his attention from dialectic to geometry (165a2), so
that he is not really any more suitable than Theaetetus to face Socrates
on Protagoras’ behalf. Thus, the request put by Socrates into Protagoras’
mouth (a) to refute him only through answers he could have given him-
self, (b) without confining his attack to the letter of the sophist’s doctrine
(166e1) but trying to find out its real meaning (168b4), appears to be some
sort of double bind. Socrates cannot meet condition (b) without probably
disobeying condition (a), while the better way to meet condition (a) is to
disobey condition (b). The latter is the way out of this aporia that Socrates
chooses: he will concentrate his attack on quotations of Protagoras, justi-
fying his procedure of interpreting them in his own way by the fact that
Protagoras cannot do it himself.
As soon as he has obtained from Theodorus an agreement on the con-
cession recalled above (169d9), he observes that each of them could be
refused authority for concluding an agreement on behalf of Protagoras, so
that the only solution is to appeal to ‘his own statement’ (ἐκ τοῦ ἐκείνου
λόγου, 169e7). Immediately after, he goes on to say: ‘He [sc. Protagoras]
says (phēsi) that things are for every man what they seem to him to be’
(170a3–4). This sentence echoes the one uttered by Socrates a few lines
above (168b5–6) where, speaking on Protagoras’ behalf, he challenged
himself to refute this statement properly. Since it was Socrates who was
speaking, it was then possible to think that he was paraphrasing or perhaps
even interpreting in his own way the doctrine of Protagoras. But now the
logic of Socrates’ argument implies that this sentence is a literal quotation
and, in spite of the hesitation of translators, we can get confirmation of
this from a comparison between the four occurrences of this statement in
the dialogue (152a6–7, 161c3, 168b5–6 and here, 170a3–4). Apart from the
third instance (168b5–6), given as direct quotation of Protagoras, in each
case the statement is introduced by one of the Greek verbs that mean ‘say’:
successively legei (152a6), eirēken (161c2) and finally phēsi (170a3). Through
this successive order, the accuracy of the statement – that is, the fact that
it is an actual quotation – is increasingly stressed. From legei, which is
ambiguous between ‘say’ and ‘mean’,22 Socrates switches to eirēken, which
unequivocally means ‘say’, and then to phēsi, which refers rather emphati-
cally to the affirmative way in which the statement was actually said.23
22
See Cornford 1935, note to 152a6. Apparently avoiding deciding between the two senses, at 152a6
and 9 Cornford translates legei as ‘he puts it’; but in the same footnote he writes that: ‘since Crat.
386a repeats the formula in almost the same words, it may well be a quotation’. For a more quali-
fied position, see McDowell 1973, note ad loc., translating the word as ‘means’.
23
See Fournier 1946.
So if I am right and, above all, if Plato’s Socrates is consistent, the state-
ment he gives from the very beginning of the discussion as the core of the
doctrine of Protagoras is taken from a writing of the sophist, probably
from his work entitled Truth, which began, we know from 161c3–5, with
the statement of the man-measure principle.24 Indeed, unauthorised as he
is to speak for the late sophist, and consequently obliged to limit the dis-
cussion to answers that could be given by Protagoras himself, Socrates has
no other way of doing so than to quote from the sophist’s writings. But
‘Protagoras’ in his apology urged Socrates not only to refute him only by
reference to his own words, but also to ‘try to find out what [he] means’
when he maintains that everything is in motion and that for each person
things are what they seem to him.25 In short, ‘Protagoras’ asks Socrates
both to respect the letter of his sayings and to interpret them. How is
this possible? Certainly, Socrates says, if Protagoras were present, he could
refute at length any erroneous interpretation of what he said or wrote, and
thus help to find out what he really meant; but this is not the case, so that
‘we have got to take ourselves as we are, and go on saying the things which
seem to us to be’ (171d3–5).26 However closely they stick to the formulae
used by Protagoras, Socrates and Theodorus cannot but interpret them
in the light of their own understanding. This amounts to doing exactly
what ‘Protagoras’ told them not to do in his apology: that is, to interpret
phrases and words used by the sophist on the basis of their common sense
(ἐκ συνηθείας ῥημάτων τε καὶ ὀνομάτων), as the many do who drag
words this way anyhow (ἃ οἱ πολλοὶ ὅπῃ ἂν τύχωσιν ἕλκοντες) so as to
involve each other in all sorts of difficulty (168b8–c2).
24
According to Cambiano 2007: 118–19, the proposition ‘things are for every man what they seem
to him to be’ is a consequence inferred by Socrates from the man-measure principle. ‘The whole
argument,’ he writes, ‘is constructed on a series of consequences that can be deduced from the
premise stated in Protagoras’ writing, namely the man-measure principle, and that are admitted by
Theodorus, a champion of Protagoras more “authoritative” than Theaetetus’ (my trans.). In fact, as
argued above, Socrates points out, and Theodorus admits, that the latter is not an ‘authoritative’
witness to Protagoras, and that is the reason why they have to obtain an agreement, not with each
other, but with what Protagoras himself said (i.e. wrote): ‘from his own statement (ἐκ τοῦ ἐκείνου
λόγου)’. In the latter phrase, ἐκ (‘from’), according to Cambiano, might point to an inferential pro-
cedure but, to my mind, if this were the case, one would expect ἀπό rather than ἐκ.
25
168b4: τί ποτε λέγομεν. The usual sense of λέγειν in this interrogative form is ‘mean’, and here this
sense is confirmed by the context, since ‘Protagoras’ states immediately the two theses that Socrates
has to examine.
26
It should be noticed that Socrates does not base the validity of his argument on its logical correct-
ness but on its consistency with the statement ‘things are for every man what they seem to him
to be’, which implies that Protagoras, not Socrates as his interpreter, is actually the author of this
statement.
164 Michel Narcy
The practice attributed here to the many, that is, of making use of the
respondent’s answers regardless of how he means them, is the core of the
lesson given to Socrates by the two professional teachers of eristic in the
Euthydemus (295c4–7):
‘Well then,’ I [Socrates] said, ‘if you ask a question with one thing in mind
and I understand it with another and then answer in terms of the latter,
will you be satisfied if I answer nothing to the purpose?’ ‘I shall be satisfied,
he [Euthydemus] said, although I don’t suppose you will.’
Thus, to forbid the respondent to add anything in his answer to the
terms in which the question was asked appears to be the rule of eristic.
According to this rule, in the Euthydemus Socrates’ qualifications of his
own answers, intended to avoid any misunderstanding, are rejected as
superfluous utterances (paraphthegmata). In the Theaetetus, it is Protagoras
who is prevented from giving superfluous answers, paraphthengesthai, by
death and not by a rule, though the result is the same. The dialogue with
him turns out to be a discussion of the phrases and words he employed,
to which he can no longer add any qualification. The point is not that the
questioner, Socrates, has a free hand as regards Protagoras’ phrases and
words (rhēmata kai onomata), as the two brothers in the Euthydemus do
towards him, but that, fair though Socrates may be, he cannot help deal-
ing with the sayings, or rather the writings, of Protagoras as if he, Socrates,
were an eristic. This means that eristic turns out to be the only way to be
fair with Protagoras or, more exactly, with what he has written. In short,
after blaming himself in the ‘apology of Protagoras’ for behaving like an
Eristic, Socrates now justifies himself for behaving in this way, alleging
that to behave in any other way (either accepting Theodorus as spokes-
man for Protagoras or speculating on what the latter could mean rather
than sticking to what he actually wrote) would be unfair.
In his apology, ‘Protagoras’ is presented as telling Socrates: ‘Don’t be
unfair in your questioning’ (μὴ ἀδίκει ἐν τῷ ἐρωτᾶν, 167e1). In my
translation of the Theaetetus, I compared this sentence with Aristotle,
Sophistical Refutations 171b23: ‘In contradiction, eristic is an unfair way of
fighting (en antilogia(i) adikomachia hē eristikē estin).’27 In fact, as Dorion
has recently pointed out to me,28 the comparison is inaccurate. Whereas
for Aristotle lack of fairness is intrinsic to eristic, what becomes clear in
Socrates’ response to his own apology for Protagoras is that, under these
27
Narcy 1994: 94 n. 3 and 338 n. 206. 28 By letter: 9 September 2011.
circumstances, the best possible way of treating Protagoras fairly is to
argue with or about him in a manner that closely resembles eristic. It is
unfair, ‘Protagoras’ said, not to keep controversy (agonizesthai, a synonym
for erizein) distinct from dialogue (dialegesthai), eristic from dialectic. As
noted earlier, agonizesthai is ‘to play about and trip up one’s opponent as
often as one can’, whereas dialegesthai is ‘to be serious and keep on helping
one’s opponent to his feet again, and point out to him only those of his
slips which are due to himself or to the intellectual society which he has
previously frequented’ (167e6–168a2). To what extent does this distinction
apply to the case of Socrates against Protagoras? First, it is too late to help
Protagoras to his feet again. Besides, it is not Protagoras’ slips, but those
of Theaetetus that Socrates aims to point out, in order to counteract the
influence of the intellectual with whom ‘he has previously frequented’,
namely Protagoras. Consequently, what can Socrates do to achieve this
aim other than overturn and ‘trip up’ Protagoras? Thus, in order to act as
a dialectician towards Theaetetus, Socrates has to engage in controversy
(agonizesthai) with Protagoras. Of course, Socrates’ efforts to overturn
Protagoras have nothing playful about them; but this is the only differ-
ence between his mode of agonizesthai and the eristic of a Euthydemus.
If we leave on one side the playful component of eristic, the distinction
between eristic and dialectic that Socrates has lent to Protagoras is, for his
present purpose, non-significant.
Perhaps this is the best clue to the special interest shown by Euclides in
Socrates’ account of his discussion with Theaetetus. After all, Euclides had
certainly listened to many other discussions of Socrates; but from what
he says in the prologue he does not seem to have cared about reporting
any other, so we can guess that this one was indeed of special interest
to him. I suggest that the reason why he was so interested in Socrates’
account, to the point of questioning Socrates again and again on the exact
terms of the arguments with Theaetetus and Theodorus, has something
to do with the fact that, when Plato wrote the Theaetetus, Euclides and
his school had made themselves famous for their eristic. Viewed in this
light, the Theaetetus may appear as a defence of Euclides: the intention of
Plato might well have been to show that what his contemporaries called
eristic was nothing else than Socratic elenchus, and that what had earned
Euclides the name of eristic was nothing other than his faithfulness to
Socrates’ teaching in the field of argumentation.
So to have the Theaetetus written by Euclides, including a refutation of
Protagoras that illustrates what may be a serious kind of eristic, is a way
166 Michel Narcy
of underlining that Socrates himself did not despise eristic provided it is
fair, and that he was even a master of it. In this sense, the Theaetetus and
Euthydemus are two faces of the same coin: the one showing Socrates com-
ically interested in eristic, and the other giving a serious demonstration
that he had nothing yet to learn about this subject.
ch apter ten
The Wooden Horse

The Cyrenaics in the Theaetetus
Ugo Zilioli
Introduction
In this contribution, I aim to show how locating the Platonic dialogues
in the intellectual context of their own time can illuminate their phil-
osophical content.1 I seek to show, with reference to a specific dialogue
(the Theaetetus), how Plato responds to other thinkers of his time, and
also to bring out how, by reconstructing Plato’s response, we can gain
deeper insight into the way that Plato shapes the structure and form
of his argument in the dialogue. In particular, I argue that the subtler
thinkers (hoi kompsoteroi) discussed by Plato’s Socrates at Tht. 156a3 are
Aristippus and the early Cyrenaics. (Recent scholars, such as Giannantoni
and Tsouna, have rejected this identification, which was earlier defended
by Schleiermacher, Grote, Zeller and Mondolfo.)2 Further, I claim that,
once we recognise that the subtler thinkers are most likely to be the early
Cyrenaics, we can make better sense of the scope and content of the argu-
ments Plato puts forward at Tht. 156a–160e (especially 156a–157c). Also,
I suggest that this identification helps us to understand a crucial part of
Tht. 184b–186e. Here Plato, in exploring the account of perception offered
at 156a–157c, uses the metaphor of the Wooden Horse to illustrate the
conception of perception that he attributes to thinkers such as Protagoras
and, in my view, the early Cyrenaics, who maintain that knowledge is a
form of perception.
1
My discussion aims to follow the approach of Christopher Rowe (e.g. in 2007a), both in exploring
the philosophical significance of the setting and characterisation of the Platonic dialogues and in
seeking to locate the dialogues in the intellectual context of their own time. This essay is also offered
as a tribute to his perceptive and illuminating readings of Plato.
2
See Giannantoni 1996, who reviews the debate, with full bibliographical references. See also Tsouna
1998: 124–37.
167
168 Ugo Zilioli
The Theaetetus as a peirastic dialogue

On my interpretation, the Theaetetus is a dialogue ad homines. The dia-
logue, I think, offers answers to the question ‘what is knowledge?’ by con-
structing ideas that are not characteristic of (the historical) Socrates or
of Plato. Rather, the answers provided are the products of Plato’s – and
perhaps Socrates’ – engagement with the theories of knowledge that were
dominant during their lifetimes. I think here of Protagoras’ relativism and
the various epistemological theories formulated, more or less completely,
by Aristippus, Euclides and Antisthenes, who were the supposed founders
of Socratic ‘schools’: the Cyrenaics, the Megarians and the Cynics, respec-
tively.3 I take it that Plato’s aim in the Theaetetus is not purely exegetical:
rather, he believes that such views were worth investigating philosophically
because of their widespread currency and theoretical appeal. In particular,
Plato’s Socrates criticises these views because they present knowledge as
based exclusively on perception.4
On this reading, the Theaetetus offers material that is relevant for a his-
torical reconstruction of the ideas of those thinkers, even though Plato
presents this evidence in an oblique way, shaped by the overall needs of
the argument. The obliqueness with which Plato presents the views of
his opponents should not be taken as meaning that these views are only
vaguely based on their ideas or that Plato is making them up. I believe
that the ideas addressed by Plato’s Socrates in the dialogue have a firm his-
torical basis and that they would have been recognised by Plato’s readers
as belonging to specific thinkers. It is true that Plato’s Socrates draws con-
ceptual connections between such views not drawn by the original think-
ers; even so, the core of those views is, I claim, historically accurate. This
is part of what made a dialogue such as the Theaetetus historically credible
to its initial readers.
On this interpretation, in the Theaetetus Plato’s Socrates is arguing ad
homines, namely against those thinkers who maintained perception-based
accounts of knowledge. This explains why the dialogue is aporetic: it
3
This reading of the Theaetetus presupposes that, at the time Plato wrote the dialogue, some of the
main tenets of the Socratic schools were already formed, at least in outline (see further on the his-
tory and intellectual status of these schools, Zilioli 2012: 3–15). An alternative reading, which takes
Protagoras’ secret doctrine to be reflecting the views of Democritus, has been recently adopted by
Ademollo (2011: 226–8). I thank an anonymous reader for Cambridge University Press for pointing
this out to me.
4
When this paper was already written, I became aware of a very recent attempt to interpret the first
part of the Theaetetus as aimed to deal critically with accounts of perception in Plato’s own time:
Balansard 2012. For an interpretation that takes even the last part of the dialogue to be concerned
with perception, see Brancacci 2010.
The Wooden Horse: the Cyrenaics in the Theaetetus 169
provides no answer to the question what knowledge is because its main
aim is to refute a certain type of view.5 This reading of the dialogue goes
back a long way in the history of Platonism and appears from time to
time in current Platonic scholarship. The Theaetetus was understood in
antiquity as a ‘peirastic’ dialogue, namely one written specifically to refute
someone’s theories.6 In his commentary on the Parmenides, Proclus indir-
ectly suggests the view that the whole of the Theaetetus is designed to
refute Protagoras’ doctrines, understood as including the theory of per-
ception developed at Tht. 152a2–157c3, which is only partly Protagorean.7
In the third book of the First Principles, Damascius alludes to an interpret-
ation of Socrates’ dream (the philosophical core of the third part of the
Theaetetus) that understands it as philosophically centred on Protagorean
elements, thus extending the critique of Protagoras right up to the final
section of the dialogue.8
Alcinous, Didaskalikos 4 offers us insight into a different interpretation
of the dialogue, according to which the Theaetetus shows what knowl-
edge is not: the dialogue addresses the epistemology of the sensible world,
whereas the Sophist is concerned with the knowledge of what is not sensi-
ble, namely the forms. Although this interpretation is different from the
sceptical one adopted by the Academics, according to which the dialogue
aims to show that knowledge is unattainable, both accounts insist on the
refutative character of the dialogue and recognise that there is a serious
confrontation in the dialogue between perception-based theories of per-
ception and the approach of Socrates/Plato, a recognition that seems to
me crucial for any credible (and historically grounded) reading of the dia-
logue. Closer to us in time, another interpreter of the Theaetetus who takes
it as a battlefield of contrasting ideas about knowledge is Lewis Campbell.
In his commentary on the Theaetetus, alongside Protagoras he carefully
5
I accept Burnyeat’s view (1990: 2 and all of Part III) that there is an epistemological progression
from the first to the third definition of knowledge, in that the latter is more comprehensive than
the former and, to this degree, closer to the truth. Yet all three definitions of knowledge in the
Theaetetus are refuted by Socrates. The impasse with which the dialogue ends is, I think, real not
apparent; nor does it pave the way for a better (and true) definition of knowledge, ready to be dis-
covered by a perceptive reader, as is maintained by Sedley 2004: 8–13, 30–7.
6
‘Peirastic’ comes from πειράω, which is a term used also for the kind of confrontation typical of
agonistic races or war battles: see LSJ, s.v. The dialogue is not a maieutic one, according to the clas-
sification reported at DL 3.49–51. On the principles of classification of the Theaetetus by Thrasyllus
(and before Diogenes Laertius), see Sedley 2009.
7
Proclus, in Prm. 657.5–10 Steel (against modern commentators, Proclus takes the Wax Block section
of the Theaetetus to be part of Plato’s own reply to Protagoras); 654.15–26. See Sedley 1996: 81.
8
Damascius, On First Principles 3.169.5–22 Westerink. These interpretations of the general strategy
of the dialogue may have their ultimate source in Academic interpretation of the Theaetetus, as sug-
gested by Sedley 1996: 84–9.
170 Ugo Zilioli
identifies traces of Megarian, Cyrenaic and Cynic thinking throughout
the whole dialogue (1883: xxviii–xli). Recently, Timothy Chappell’s run-
ning commentary on the Theaetetus (2004: 62–4, 67–85, 204–12) defends
the view that Plato is arguing against empiricist theories of knowledge
maintained in his own time.
If the Theaetetus is best understood as a peirastic dialogue, where Plato
seriously confronts his rivals’ views about perception and knowledge, it
will make good sense to try to locate in the dialogue his references to the
philosophers whose views he is examining. It is in this context that I place
my attempt to give reasons for identifying the subtler philosophers of Tht.
156a3 with Aristippus and the early Cyrenaics.
The choice of characters: Euclides

Some initial reasons for supporting the reading of the Theaetetus as a dia-
logue ad homines are provided by Plato’s choice of characters. The dialogue
is narrated by a slave of Euclides of Megara to Terpsion, another citizen
of Megara, Euclides himself being present at the narration. According to
the Phaedo, both Euclides and Terpsion were present on Socrates’ final
day, together with Antisthenes. (On that occasion, Aristippus is said to be
in Aegina: Phd. 59bc.) The narration of the Theaetetus is also imagined as
taking take place in Megara, where Plato seems to have fled after Socrates’
death (DL 2.106). The anonymous commentary on the dialogue remarks
that Plato wrote this prologue in order to dedicate a ‘mighty dialogue to
a weighty man’ (sc. Euclides: coll. III.50–IV.6). In addition, at least in one
point, the Theaetetus seems to be aimed at an audience from Megara (and
Cyrene), or at least one that knew these places well. When pressed by
Socrates to defend his friend Protagoras, Theodorus alludes to the story of
Sciron, who used to sit on the rocks on the road from Megara to Corinth.
Sciron compelled travellers to wash his feet and, while they were doing so,
he kicked them over the cliffs into the sea, where an enormous tortoise
was waiting to tear their bodies to pieces (Tht. 169a9). The brief allusion
seems to point to Socrates’ own fatal end. The fact that the passage refers
to a legend of Megara might have made the significance of the passage
more obvious to a Megarian reader.9
9
The second legend briefly referred to by Theodorus in the same passage (169b2–3) is that of Antaeus,
a giant, son of Poseidon and Gaia, who lived in a cave and forced all passers-by to wrestle with him.
After defeating those unfortunate passers-by, he killed them and put their bodies in the temple of his
father. In this case, the legend comes from Cyrene, the city of Theodorus and Aristippus. Socrates’
brief reference to it would not have required any further explanation for a Cyrenaic reader.
The Theaetetus is thus, in an important sense, the dialogue of Euclides
the Megarian, as much as the Phaedo, for instance, is the dialogue of
Phaedo.10 More crucially, the Theaetetus is the dialogue of Euclides, the
Megarian, in that there are traces of Megarian thinking throughout the
dialogue. Even if one does not accept my view that the Theaetetus is a dia-
logue aimed at refuting Protagorean and Socratic theories of knowledge, it
remains the case that most of the dialogue is taken up by refutation of the
definitions of knowledge that Theaetetus proposes. The refutation of these
definitions is made mostly through a reductio ad absurdum, namely by ini-
tially assuming the validity of what Theaetetus says and then by showing
that his definitions lead to absurdity. Thus, for instance, it is shown that
man is not the measure because if so, every man would be self-sufficient
in knowledge and would not need to be taught; or that movement is not
the only principle of reality, because, if so, language would be impossi-
ble. Both the dialectical exercise of refuting others’ theories and the logi-
cal method through which these refutations are conducted remind us of
Euclides’ eristic art and its negative character.11 Socrates’ references at 164d1
to those clever men (deinoi andres) and to those ‘lovers of contradiction’
(antilogikoi) at 164c10 can also be understood as suggesting Euclides and
his followers.
Moreover, in the Theaetetus at more than one point we seem to be
reminded of what are usually described as Megarian paradoxes. Diogenes
Laertius (2.108) ascribes to Eubulides ‘many dialectical arguments’, which
he cites by name without giving the precise content of each argument.12
This is perhaps because such arguments were so notorious that they did
not need any further specification; it is also very probable, as Döring
has argued,13 that Eubulides did not himself invent them. The Theaetetus
shows that at least one of these arguments was, undoubtedly, circulating
10
In connection with the Phaedo, Boys-Stones 1999: 1–8 brings out the linkage between the narrator
chosen by Plato and the message of the dialogue. For a different view on the significance of the
choice of Euclides as the object of the narrative of the Theaetetus, see Narcy in this volume.
11
See DL 2.107 (SSR II A34 = Euclides fr. 29 Döring [1972]) on the reductio ad absurdum, and DL
2.30 (SSR II A3) and 107 (respectively frr. 9 and 8 Döring) on the eristic art. On Euclides’ rejection
of the analogical method, see again DL 2.107. On Eubulides as Euclides’ ‘eristic’ successor, see DL
2.108 (SSR II B1). Campbell (1883: xxxv) believes that the Theaetetus as a whole has a Megarian tone,
since it is uniformly centred on refutation of theories. (It might be objected that the refutation of
others’ theories is the most prominent feature of Socrates’ philosophical activity. I do not want
to deny this, but just to stress that Euclides, and his immediate followers, took that approach of
Socrates and developed it more fully: cf. also Muller 1985: 106; Narcy in this volume.)
12
This is Diogenes’ list of the arguments of Eubulides: the Liar (pseudomenos); the Veiled Man
(enkekalummenos or dialanthanōn); the Electra (a variant of the enkekalummenos); the Sorites; the
Horned Man (keratinēs); the Bald Man (phalakros).
13
See Döring 1972: 108 with Giannantoni 1990: iv. 83–8.
172 Ugo Zilioli
before Eubulides. I am referring to the enkekalummenos, the Veiled Man.
This paradox goes roughly like this. You know your father; you are placed
in front of a person covered by a veil: you do not know him. But this per-
son is your father, therefore you both know your father and do not know
him.14 At 165b2–4 we seem to be reminded of this paradox, when Socrates
asks Theaetetus the ‘most alarming’ question, namely how it is possible ‘for
a man who knows something not to know this thing which he knows’.15
The example Socrates chooses to illustrate this is that of someone who,
‘clapping his hand over one of your eyes … asks whether you see his cloak
with the eye that is covered’ (165b9–c2). The analogy between Socrates’
words and the Veiled Man cannot be closer. The same analogy seems also
to surface later in the dialogue, when Socrates attempts to explain the
phenomenon of false judgement by treating knowing as an alternative to
not knowing (188a–c). In Socrates’ accounts of the problem of false judge-
ment, one can find the following statements: ‘If [a man] knows a thing,
it is impossible that he should not know it; or if he does not know it, he
cannot know it’ (188b1); and: ‘a man certainly doesn’t think that things he
knows are things he does not know, or again that things he doesn’t know
are things he knows’ (188c2–3). A reference to the Megarian paradox is
clearly on display here.
The subtler thinkers

A second choice of character that is illuminating for the ad homines read-
ing of the Theaetetus that I am recommending is, obviously enough,
Protagoras. More than half the dialogue is devoted to the discussion of
his views, or of views indirectly springing from his maxim that man is the
measure of all things. We start off at 151e3–152a4 with the identification
between Protagoras’ maxim and Theaetetus’ first definition of knowledge
as perception. From this identification, a long list of other philosophi-
cal theories and ideas follows: Protagoras’ secret doctrine (152d1–e1),
Heraclitus’ theory of movement (152e2–153d5), the subtler thinkers’
account of perception (156a2–157c3), the idea that wisdom is impossible
in Protagoras’ world (161c1–162a3), a further defence of Protagoras’ views
(166a1–168c5), the dichotomy between the philosopher and the person in
14
This formulation of the Veiled Man paradox can be found, together with its variant, the Electra,
in Lucian, Auction of Lives 22. Aristotle treats the paradox as a case of paralogism ‘upon accident’
(παρὰ τὸ συμβεβηκός, SE 179a33–b4).
15
See also 166b4–5. Translations of the Theaetetus are from Burnyeat 1990, emended where noted. For
other passages and authors translations are my own unless otherwise specified.
the law court (172c3–177c5). Thus, we really do seem to be overwhelmed
(as Socrates points out at 172b9–c1) by a discussion that becomes more
intricate as it goes on.
In the Theaetetus, Protagoras is clearly a symbol for a whole range of
approaches to knowledge that Plato’s Socrates wants to identify as being
perception-based. Yet the core of Protagoras’ secret doctrine is a fascinat-
ing theory of perception that Socrates ascribes to some ‘subtler thinkers’
mentioned at 156a3, who are most likely to be the same as the students
of Protagoras identified at 152c10. After giving his initial exposition of
Protagoras’ relativism at 151e–152c, Socrates suggests that Protagoras taught
a secret doctrine to some of his pupils (152c10). The content of Protagoras’
secret doctrine is stated briefly at 152d2–e1, expanded at 153d8–154b8 and
finally presented fully at 156a2–157c3, the passage that is most relevant
here. The subtler thinkers are contrasted with some unnamed materialists
and said to have their own mysteries that Socrates reveals. In accordance
with Protagoras’ secret doctrine, these thinkers maintain that everything
is in movement (kinēsis) and that there is nothing beyond movement
(156a5). They believe that there are two kinds (eidē) of movement, each
infinite in extension, one with an active power (to men poiein), and the
other with a passive one (to de paschein) (156a6–7). From the intercourse
of these two kinds of movement, ‘there come to be offspring, infinite in
number, but always twins: one is the perceived thing, the other is the cor-
responding perception, which is on every occasion generated and brought
to birth together with the perceived thing’ (156a7–b2). These perceptions
have names such as ‘seeings, hearings’ and so on (sensible perceptions) and
‘pleasures, pains’ and so on (emotions) (156b3–4).
Socrates then explains how, according to this theory of perception,
objects cease being ontologically indeterminate and become determinate
(such as coloured) for a perceiver. For instance, a stone does not possess
whiteness in itself but it becomes white once the perceiver and the per-
ceived thing have encountered each other so that the perception eventu-
ally arises (156c6–157a5). It is clear that in this picture there is no room
at all for perceptual error: the perceiver is infallibly aware that the stone
is white for her. As Socrates put it earlier when discussing the views of
Protagoras, ‘perception is always of what is, and free from falsehood, and
it is knowledge’ (152c5–6). This is a reason why the subtler thinkers may
have been also the disciples of Protagoras: they restate the kind of incor-
rigibility that his relativism attaches to the perceptions of the individual.
Socrates eventually brings in a point about language, when he illustrates
further the philosophical implications of the subtler thinkers’ theory of
174 Ugo Zilioli
perception: a stone does not possess the whiteness in itself but it becomes
white once the perceiver and the perceived object have come across each
other and perception eventually arises (156c6–157a5). According to the the-
ory, objects are ontologically indeterminate prior to perception. Socrates
implies a link with the initial formulation of Protagoras’ secret doctrine
when he says that, according to the subtler thinkers, it is clear that ‘noth-
ing is one thing just by itself, but things are always coming to be for some-
one’ (157a8–157b1).16 In the light of this claim, the verb ‘to be’ has to be
abolished, as ‘those wise people say’ (157b3–4), and a whole range of new
expressions needs to be coined. Such new expressions include ‘coming to
be’, ‘undergoing production’, ‘ceasing to be’, ‘altering’, instead of terms
like ‘something’, ‘someone’s’, ‘my’, ‘this’ (157b4–7). Even the relation of
simple denotation is at risk: when we say ‘stone’ or ‘man’, we are using
a convention (otherwise our words will not be meaningful); but there is
nothing out there in the world, such as a man or a stone. We had better
describe such objects simply as aggregates or collections (athroismata) of
parts (157b9–c2).
The point of the statement about the need for a new language is that,
instead of a metaphysics of objects, in the theory of perception endorsed
by the subtler thinkers there is a metaphysics of process. This kind of
metaphysics is the only one that is able to make good sense of the kind
of indeterminacy that lies at the heart of the subtler thinkers’ theory.
The only way that things can display a certain ontological character (for
example, being white) and can emerge from their intrinsic indetermin-
acy is by coming into contact with a perceiver at a given time. But this
interaction can only be momentary, since it only lasts for a brief period of
time; this interaction is also private both to the perceived thing and the
perceiver (as Socrates says at 154a2).17 Since perceivers are many, there are
also many ways in which things are perceived. The best way to account for
this vast array of conflicting perceptions experienced by different perceiv-
ers is to replace the notion of objects with that of processes. This, how-
ever, forces the subtler thinkers who endorse this replacement to call for
a new language: in place of an object-centred language, there has to be a
language based on processes, in which names such as ‘stone’ or ‘man’ are
empty terms. These names do not really indicate any ontologically stable
and determinate object out there; however, we keep those names as a kind
16
This view about indeterminacy is originally stated at 152d6 and restated at 182b3–4.
17
‘What we say a given colour is will be neither the thing which collides, nor the thing it collides
with, but something which has come into being between them; something peculiar to each of
them’ (153e7–154a2, trans. modified).
of useful convention in the framework of a new language of processes –
otherwise, we would be trapped into a wholly solipsistic world, that is,
one private to each perceiver.
The key ideas in the theory of perception endorsed by the subtler think-
ers are, therefore, the following: everything is (in) movement; perceptions
are incorrigible and private to the individual perceiver; there is a need for
a new language (required by the fact that things are indeterminate). Let
us now see whether these ideas are also present in the richest available
source for the epistemological thought of the Cyrenaics, namely Sextus
Empiricus, Against the Professors 7.190–200.
Sextus Empiricus on the Cyrenaics

Sextus’ report appears in his survey of the criterion of truth of the dogmatic
philosophers. With regard to the Cyrenaics, Sextus says (M 7.191–2):
The Cyrenaics claim that the affections (pathē) are the criteria [of truth]
and that they alone are apprehended and are infallible (adiapseustos). None
of the things that have caused (tōn pepoiēkotōn) the affections is, conversely,
apprehensible or infallible. They say that it is possible to state infallibly
and truly and firmly and incorrigibly that we are being whitened (leukai-
nometha) or sweetened (glukazometha), but it is impossible to say that the
thing productive of the affection in us is white or sweet, because one may
be disposed whitely even by something not-white or may be sweetened by
something not-sweet.
Slightly later (M 7.194–5) he adds:
We must therefore say either that the affections are the phainomena or
that the things productive of the affections are the phainomena. If we say
that the affections are the phainomena, we will have to maintain that all
phainomena are true and apprehensible. If, on the contrary, we say that the
things productive of the affections are the phainomena, all phainomena will
be false and not apprehensible. The affection occurring in us tells us noth-
ing more than itself. If one has to speak but the truth, only the affection
is therefore actually a phainomenon for us. What is external (to d’ektos) and
productive of the affection perhaps is a being but it is not a phainomenon
for us. We are all infallible with regard to our own affection, but we are all
uncertain with regard to the external object (to ektos hupokeimenon).
Two of the three ideas noted earlier in the theory of perception ascribed
to the subtler thinkers of the Theaetetus are obviously present in Sextus’
account: the privacy and incorrigibility of one’s own perception, and the
use of a different kind of language, more in accordance with how the world
176 Ugo Zilioli
actually is. On the first point, Sextus insists that, for the Cyrenaics, each
perception is infallible and incorrigible; indeed, every affection (pathos) is,
on the same principle, incorrigibly private to the perceiver undergoing the
affection. Again, as in the Theaetetus, the first example that Sextus reports
is of an affection of whiteness. The expression that the Cyrenaics were
reported to use is ‘to be whitened’ (leukainomai).18
With the expression ‘I am being whitened’ the Cyrenaics invented a
neologism for capturing the philosophical innovation that was already
fully present in the theory of perception ascribed by Plato to the sub-
tler thinkers in the Theaetetus. In the latter theory, as we have seen, much
emphasis is put both on the movement that makes every perceptual act
momentary and on the dissolution of reality into an indeterminate sub-
stratum, where processes replace objects. The tense of the expression ‘I am
being whitened’ may refer to the instantaneousness of the perceptual act:
every perception (using the term, aisthēsis, employed in the Theaetetus) or
affection (to use the canonical Cyrenaic term, pathos, referred to by Sextus)
is limited to the very moment when it takes place. Secondly, and most
crucially, the expression ‘I am being whitened’ avoids any reference to the
(reality of the) objects causing the perception or affection. Despite the
fact that Sextus often uses Sceptical terms to characterise Cyrenaic views,
he also describes the objects as ‘causing’ the affections. But I believe that
he is just expounding the doctrines of the Cyrenaics by adopting the dis-
tinction between appearance and reality. The Cyrenaics may indeed have
adopted the same distinction (which was central throughout Greek phil-
osophy), but they ended up by adopting a rather new approach to real-
ity. Despite the fact that Sextus often uses Sceptical terms to characterise
Cyrenaic views, he also talks as if the Cyrenaics thought that there really
are objects ‘out there’ that ‘cause’ the affections they experience. This, it
seems to me, is strictly inaccurate. No doubt the Cyrenaics thought there
was some state of affairs external to the individual (as all Greek thinkers
seem to have assumed); but if it is true, as the subtler thinkers tell us, that
this state of affairs is utterly indeterminate and we have only a metaphys-
ics of processes, then the Cyrenaics would not have wanted to use expres-
sions such as ‘object’ which imply a stable or determinate reality.19
18
The use of these neologisms is attested also in Plutarch, Against Colotes 1120E.
19
By relying on ancient sources other than Sextus and Plato’s Theaetetus (such as e.g. Philodemus and
the anonymous commentator on the Theaetetus), I have recently argued that the Cyrenaics may be
well understood as endorsing a metaphysics of indeterminacy, denying the existence of objects: see
Zilioli 2012: ch. 4. For the Cyrenaics as not questioning the existence (over time) of both objects
and subjects (and, hence, as rejecting indeterminacy), see Tsouna 1998: 131–7.
The need to avoid any reference to objects seems to be lurking behind
the Cyrenaic view about the conventionality of language, which Sextus
briefly reports. Since we are all infallible in our affections but we all
make mistakes about the object causing the affections, the Cyrenaics say
(M 7. 195–6):
No criterion is common to human beings, common names are assigned to

objects. All in common in fact call something white or sweet, but they do
not have something common that is white or sweet. Each human being is
aware of his own private affection. One cannot say, however, whether this
affection occurs in oneself and in one’s neighbour from a white object.
According to Cyrenaic epistemology, we are incorrigibly aware of our own

affections, but we can never know whether the object we perceive as white
is white in itself. Neither can we know whether another person perceives
the object as white. We do not have either any common affection or access
to a common world of objects from which our affections arise. If we had
that one world, our affections could still be private to us, but we could
compare them with those of others on the basis of an objectively shared
element, namely that very world. Since we do not have any common
world or common affections, we have to revert to common names, to be
assigned conventionally to those qualities, such as white or sweet, that we,
human beings, all experience, though with reference to different objects
and in different circumstances.
Other important analogies can be drawn in this respect too with the
Theaetetus. Socrates says there with reference to the theory of the subtler
thinkers that ‘nothing is one thing just by itself, but things are always
coming to be for someone’ (157a8–b1). Socrates also remarks that, accord-
ing to the latter theory, we had better not speak of things such as stones or
men: there is really nothing in the world out there as such. (What we call
‘man’ or stone’ are aggregates emerging, so to speak, from perceptual proc-
esses: 157b9–c2.) Similarly, Sextus points out that, on the Cyrenaic view:
‘what is external and productive of the affection perhaps is a being, but it
is not a phainomenon for us’ (M 7.194). On this basis, the Cyrenaics may
have maintained that there is something out there but that this is no more
than an indeterminate substratum (or hupokeimenon). This substratum,
however, is not, on this view, either discrete or made up of objects per-
fectly identifiable by an essence (in the strong sense of ‘essence’ intended
by, for example, Aristotle). There is no ‘essence’ there to be grasped.
That is why it makes no sense to say ‘I perceive the stone as white’; it
makes much better sense to say ‘I am being whitened’ when I happen
178 Ugo Zilioli
to see what we conventionally call a stone. In the Theaetetus, the sub-
tler thinkers are still referring to objects, although their theory implies
that there are none and although these thinkers are also made to invoke
the invention of a new language, which replaces the language of being
with that of becoming and which avoids direct reference to objects as
determinate entities. Despite the way that Sextus presents the position,
it was probably the later Cyrenaics, rather than Aristippus and the early
Cyrenaics, who coined a new language to express both the indeterminacy
of the world and the privacy of the individual’s affections.20
The detailed comparison I have offered between the theory of the sub-
tler thinkers in the Theaetetus and Sextus’ account of Cyrenaic doctrines
has brought out the analogies between the two texts, in particular as
regards perceptual incorrigibility, indeterminacy and the need for a new
language. The final element in the theory of the subtler thinkers that is
worth investigating is the idea that everything is (in) movement. From the
intercourse of the two kinds of movement, active and passive, there come
to be twin offspring, the perceptions and the corresponding perceived
objects. These perceptions have names such as hearings, seeings and pleas-
ures and pains (Tht. 156b4–6). Thus, perceptions are seen as the results of
movement and as movements themselves, being themselves subject, as any
other thing is, to the law that everything is movement. The reference to
pleasures and pains is, in this context, particularly illuminating. According
to the subtler thinkers’ theory, in the Theaetetus there has been no refer-
ence at all to pleasures and pains as kinds of perception. Why does Plato
present these quite naturally as perceptions in the context of the subtler
thinkers’ theory? The answer is that, by including pleasures and pains in
the category of perceptions, he is here signalling, indirectly, a reference to
the Cyrenaics.
One of the fundamental tenets of the Cyrenaics, crucial to their ethics,
was in fact that the two essential affections characteristic of human beings,
namely pleasure and pain, are two movements. As Diogenes Laertius puts
it (2.86):21
The philosophers who followed the teaching of Aristippus and were called
Cyrenaics had the following beliefs. There are two affections (pathē), pain
20
I argue for this historical attribution in Zilioli 2012 ch. 5, section 8.
21
Diogenes is clearly advancing the view that these definitions of pleasure and pain can be traced back
to Aristippus. Aristocles seems to be inclined to ascribe this view to Aristippus the Younger (T5 =
Eus., PE 14.18.32). Other sources explicitly ascribe the view to Aristippus, the supposed founder of
the school: SSR IV A174 (= Athenaeus, Deipnosophistae 12, 544ab) and 181 (= Cic. Fin. 1.26).
and pleasure, pleasure being a smooth motion (leia kinēsis) and pain a
rough motion (tracheia kinēsis).
On this basis, we find a striking resemblance between the concepts of
pleasure and pain as defined by the Cyrenaics and what the subtler thinkers
of the Theaetetus hold about perceptions, including those, such as pleasure
and pain, which were not mentioned in the dialogue before the report of
the subtler thinkers’ theory of perception. In both theories, pleasure, pain
and all other perceptions, or affections, are best understood as (the results
of ) movements. Perhaps we do not want to go as far as Dümmler, who
claimed that the theory of movement that is such an important ingredient
in the subtler thinkers’ theory of perception does not have a Heraclitean
basis but is, ultimately, Aristippean.22 Still, the analogies between the ideas
ascribed to the subtler thinkers in the Theaetetus and those we recognise
as authentically Cyrenaic are so many that it can hardly be a matter of
coincidence.
The Wooden Horse

Thus, there are good grounds for identifying the subtler thinkers of Tht.
156a3 with the early Cyrenaics. In the Theaetetus, Plato cannot name the
Cyrenaics directly as having elaborated a set of views about perception of
which Protagoras was the founder because at the time the dialogue is set
(399 BC) Aristippus and his early followers had not developed their doc-
trines fully. But the identification of them with the subtler thinkers is, I
believe, crucial for understanding Plato’s final critique of the thesis that
perception is knowledge (Tht. 184b–186e).
Almost at the end of his lengthy treatment of Protagoras, Heraclitus
and the subtler thinkers, after refuting the view that all is movement
(179d–183b), Socrates puts forward his final argument against the iden-
tification of knowledge with perception (184b–186e). Very briefly, the
argument goes like this: we perceive a thing as white by means of our
sense organs; by means of these sense organs, however, we are unable to
say whether the object we see as white is identical with something else,
whether it is really something, or whether it is similar to something else.
In short, by means of the sense organs, we are unable even to conceive
of ‘being and not being, likeness and unlikeness, the same and different’
(185c9–10). As Theaetetus puts it: ‘The mind itself (autē … hē psuchē), by
22
Dümmler 1901: 56–67. He also underlines the importance of the Hippias Major (287d–289e) for a
correct understanding of the relationship between Heraclitus and Aristippus.
180 Ugo Zilioli
means of itself (di’ hautēs), considers the things which apply in common to
everything’ (185e1–2). There is a clear distinction, Socrates adds, between,
on the one hand, perceptions and, on the other, the cognitive operations
that the mind makes by itself: truth and being can be grasped by the mind
and not by the senses (186c1–10). Since the operations the mind makes by
itself in its search for truth and being cannot be performed by perception,
the latter has no share in the grasping of being and hence it has no share
in the grasping of truth either. Perception therefore cannot be knowledge.
What is at the root of this epistemological picture is, plainly, the organ-
ising activity that the mind has to perform to account for those properties
that are not perceptible, such as identity and similarity. More crucially for
us, the mind itself is responsible, on Plato’s account, for organising the
various pieces of information that perception, through our sense organs,
provides us with. For this purpose, perception has to be conceptualised.
He conveys this view eloquently through the metaphor of the Wooden
Horse (184d1–5):
It would surely be strange if we had several perceptions sitting inside us, as
if in wooden horses, and it wasn’t the case that all those perceptions con-
verged on some one kind of thing, a mind or whatever one ought to call it:
something with which we perceive the perceptible things by means of the
senses, as if by means of instruments.
As the scholia recognise, Plato is here referring to the well-known
wooden horse of Troy: no Greek would have missed the reference.23 It is
clear that, in using this metaphor, Plato is not only expounding his view
about the relationship between perceptions and mind, but is also indi-
rectly criticising the view that does not recognise fully the fundamental
role of the mind in the cognitive process. If we do not recognise this role,
we will conceive of ourselves as a kind of wooden horse, with a plurality
of perceptions sitting inside us without any organising centre. But, in this
case, these perceptions would be (according to Plato’s masterfully evoca-
tive suggestion) deceptive ones, as the wooden horse of Troy was deceptive
for the Trojans. Their deceptiveness does not consist (as I interpret Plato’s
point) in their being false perceptions, but in their being uni-sensorial
perceptions, that is, isolated perceptions which simply register an affection
of the body without being able to interact epistemologically with other
23
See Greene 1938 ad loc. At Odyssey 4.271–89, Menelaus praises the ability of Odysseus by remem-
bering the episode when, by imitating the voices of their wives, Helen calls by name all the men
sitting inside the wooden horse. Odysseus saves his fellows on that occasion by persuading them
to remain silent.
similar perceptions. On this view, perception is simply a unidimensional
affection of the body, achieved through the stimulation of the appropri-
ate sense organ. It can hardly be denied that the affection that someone
undergoes is true for them; yet, on Plato’s view, the affection is deceptive
insofar as it focuses on just one perceptual property of the object. We can
see a stone as white; we can sense a stone as hard, but there is no way,
according to this theory, to put these two isolated affections into the same
epistemological picture (see Tht. 185a1–6).
By stressing again the epistemological distance between mind and per-
ception, Socrates refers implicitly to this view in two other passages. First,
186b11–c2: ‘There are some things that both men and animals are able
to perceive by nature from the moment they are born: namely all those
affections (pathēmata) that, by means of the body, converge on the mind.’
Slightly later, he returns to this point (186d2–5): ‘Knowledge is located not
in our affections (en tois pathēmasin) but in our reasoning upon these.’
In these two passages, the term Plato uses to refer to the simple altera-
tion of the body in the perceptual process is, strikingly enough, pathos,
‘affection’, and not the more common aisthēsis, by which he has so far
described the sensations in the Theaetetus. There are two other passages
in the dialogue where Plato uses the term pathos. The first is at 179c3,
where Socrates says: ‘if we focus on the momentary affection (pathos) each
of us has, from which there come to be his perceptions and the judge-
ments which conform to them – well, it is harder to refute these latter as
not being true’ (179c2–4). In another passage, Protagoras is made to ask
whether it is possible that ‘a man’s present memory of an affection (pathos)
which he has experienced in the past but is no longer experiencing is the
same sort of affection as he then had’ (166b2–4).
In the latter passage, the point made by Socrates forms part of an
argument designed to show how teaching will not be possible if one is
incorrigibly correct in one’s affections. (This is seen as one of the main
upshots of the subtler thinkers’ theory of perception.) The relationship
between the affection and its memory is the core theme of the passage.
Socrates denies the possibility that the memory of the affection is the
same affection as was experienced earlier. This point sounds undoubtedly
Cyrenaic: Aristippus, famously, denied any value to the memory of past
pleasures on the grounds that what counts is the pleasurable affection of
the moment (however long this may be) and that we cannot recreate the
affection of past pleasures as this was originally experienced.24 Curiously
24
See SSR IV A174 = Athenaeus, Deipnosophists 12, 544ab.
182 Ugo Zilioli
enough, another passage of the Theaetetus (184b–186e) may be relevant
here. Theaetetus remarks that, as far as good and bad are concerned, the
mind considers their existence ‘in relation to one another, calculating in
itself things past and present in relation to future’ (186b1). Thus, the mind
is able to provide the kind of linkage between what is good and bad in the
past, present and future that Aristippus, by grounding his ethics on the
pleasure of the moment (namely on the transitory affection of the body),
was not able to provide.
These analogies between the Theaetetus and other sources on the
Cyrenaics, together with the use of the key term pathos in the Wooden
Horse section (in place of the more common aisthēsis) may, once again,
point towards the identification of the subtler thinkers and the early
Cyrenaics. In his critique of the thesis identifying perception with knowl-
edge, Plato is, in my interpretation, in effect arguing specifically against
a position stated and defended earlier in the dialogue (156a2–157c3),
namely the full-scale identification of perception and knowledge. This
passage is, without doubt, the one where the subtler thinkers, who were
the front-runners of perceptual subjectivism and relativism in Plato’s time,
propose their theory of perception. In showing the unavoidable role of the
mind in cognitive processes, including perceptual processes, Plato is at the
same time showing the main weakness of the account of perception he has
so far been confronting. On this account, namely the one defended by the
subtler thinkers at 156a2–157c3, a perception is simply an affection; we can
grant to such affections (to use the term that appears most in 184b–186e)
or perceptions (to use the term that appears most in 156b–157c) their
incorrigibility and we cannot properly say they are false. Yet these affec-
tions or perceptions are unable to tell us anything about properties such
as identity, similarity and being that only the mind is able to grasp. Even
worse, as Plato argues, affections and perceptions, conceived in Cyrenaic
terms, are uni-sensorial events, with no residual capacity to be elaborated
into a perceptual picture that pulls together the unidimensional informa-
tion provided by the different sense organs into a coherent whole.
In addition, the Cyrenaic doctrine of ‘internal touch’ (tactus interior or
intumus) can be interpreted as a direct, though later, response to Plato’s
challenge based on the organising role of the mind in perceptual proc-
esses. According to Cicero’s testimony, the Cyrenaics maintained that ‘the
only things they perceive are those which they sense by internal touch, for
instance, pain and pleasure, and they do not know whether something
has a particular colour or sound, but only sense that they are themselves
affected in a certain way’ (Ac. 2.76). In another passage, Cicero says: ‘What
about touch, of that touch philosophers call interior, of either pleasure
or pain, in which the Cyrenaics believe that only there is the criterion
of truth (iudicium), because it is perceived by means of the senses?’ (Ac.
2.20). According to this evidence, the Cyrenaics seem to replace Plato’s
idea of the organising mind (expressed in the Wooden Horse metaphor)
with that of internal touch, which appears to be a mental equivalent of
physical sense organs. It would seem that the cognitive role that Plato
attributes to the mind, understood as an organising principle, is seen as
fully explained by reference to the idea of internal touch.25
The Siege
That Plato’s main target in the Wooden Horse section is the Cyrenaic
account of affection, which earlier in the dialogue (156a2–157c3) is ascribed
to the subtler thinkers, is clear also from two other important sources on
Cyrenaic thought. In his critique of Cyrenaic epistemology, Aristocles first
criticises the idea, which forms part of the subtler thinkers’ theory of per-
ception, that subjects and objects of perception do not exist independ-
ently of each other by remarking, in Aristotelian fashion (Aristocles T5
Chiesara, at Eus., PE 14.19.3–4):26
These three things must necessarily coexist, the affection itself, what causes
it, and what undergoes it. He who apprehends the affection must necessar-
ily perceive also what undergoes it. For he will not know that something,
for example, is warm, without knowing whether it is himself or a neigh-
bour, now or last year, in Athens or Egypt, someone alive or dead, a man
or a stone.
Secondly, he adduces an argument on the self-consciousness of affections
that is very close to Plato’s argument in the Wooden Horse metaphor. In
the same passage, Aristocles asks (T5 Chiesara at Eus., PE 14.19.5):
How will he [the man with an affection] be able to say that this is pleasure
and that pain? Or that he had the affection by tasting, by seeing, or by
25
On the possible organising role of the internal touch for the Cyrenaics, see Zilioli (2012), 125–7,
where I defend the idea that they admitted of extra-affective activities in their epistemology.
26
Compare Tht. 156a–157c, and in particular 156a7–b2: ‘From their intercourse [sc. the intercourse
of passive and active elements], and their friction against one another, there come to be offspring,
unlimited in number but coming in pairs of twins, of which one is a perceived thing and the other
a perception, which is on every occasion generated and brought to birth together with the perceived
thing.’ See also 156b7–c3 and 156d4–5. Note the pairing of man and stone, which occurs both at
Tht. 157c1 and in Aristocles. For Aristocles’ On Philosophy, I follow the translation by Chiesara 2001
unless otherwise stated.
184 Ugo Zilioli
hearing? And by tasting with his tongue, seeing with his eyes and hearing
with his ears?
With such questions, Aristocles suggests, I believe, that simple Cyrenaic
affections are impossible to have without a proper organ that is ultimately
responsible for organising the data provided by the sense organs.27 This
is Plato’s main point in the Wooden Horse metaphor. In providing two
counter-arguments which can be read as directed against the arguments
contained in two sections of the Theaetetus (156b–157c and 184b–184e),
Aristocles confirms the philosophical linkage between the two sections of
the dialogue, sections on which we have been so far concentrating and
which, even on Aristocles’ reading, expound Cyrenaic views.
A last passage strengthening the suggestion that in Tht. 184b–186e Plato
is, among other things, elaborating his critique of a Cyrenaic theory of
perception is the earliest source on the thought of the Cyrenaics, namely
the account of the Epicurean Colotes, as preserved by Plutarch. Plutarch
reports Colotes’ view on the epistemology of those philosophers (Against
Colotes 1120CD):
[The Cyrenaics], placing all affections and all sense-impressions in them-
selves, believed that the evidence coming from them was not sufficient, as
far as assertions on external objects are concerned. Distancing themselves
from external objects, they shut themselves up within their affections as
in a siege (eis ta pathē katekleisan hautous en poliorkia). In doing so, they
adopted the locution ‘it appears’ but refused to say in addition that ‘it is’
with regard to external objects.
In these few words, Colotes is made to report the kernel of Cyrenaic
epistemology: only affections are knowable, and incorrigibly true. What
is striking, however, is how he describes such a position: the Cyrenaics
are said to ‘shut themselves up within their affections as in a siege’. This,
undoubtedly, reminds us of Plato’s Wooden Horse, which in turn would
have reminded any Greek, including Colotes, of the siege par excellence in
Greek legend, that of Troy. I suspect that what is behind Colotes’ image of
the Siege is, in fact, Plato’s Wooden Horse metaphor.
In this metaphor, Plato speaks of a plurality of perceptions sitting inside
us as though in wooden horses; in using this metaphor, I have argued,
he implies that perceptions are deceptive insofar as they are partial and
need to be elaborated further by the mind. These deceptive perceptions
are the same affections that Colotes speaks of in the passage just reported.
27
On the Platonic background for Aristocles’ argument, see Chiesara 2001: 139 and Grote 1865: 335.
By accepting only the evidence of the affections of the individual and by
refusing to make any statement about how the world out there really is,
the Cyrenaics, Colotes argues, are deceived by themselves, insofar as they
believe as incorrigibly true what they sense: these affections are deceptive,
since they do not say anything about the real objects (and their actual
properties) causing the affections. The best image able to capture this state
is that of a siege; this metaphor is as evocative as the Wooden Horse image.
The latter draws its inspiration from the former and points in the same
direction: in both cases, the ultimate responsibility for the deceit is the
fact that the perceiving subject adopts a Cyrenaic type of epistemology.
This is plainly evident in the passage by Colotes and also in the Theaetetus
passage (Tht. 184d1–5): by comparing individuals to Trojan wooden horses,
Plato makes clear that the siege for the individual is brought out by the
individual himself.
If the metaphor of the Siege takes its inspiration from the metaphor
of the Wooden Horse, it is reasonable to say that the earliest source for
Cyrenaic doctrines, namely Colotes, understands the critique that Plato
elaborates at Tht. 184b–186e as directed against the Cyrenaics. If this is
so, we will be in a better position than before to understand more fully
the critical arguments that Plato advances against perception at Tht.
184b–186e. In particular, we will be in a better position to grasp how
these arguments are directed against a model of perceptual knowledge
that Plato has depicted vividly in a preceding section of the dialogue,
namely at 156a–157e, where the subtler thinkers make their appearance. I
am not aware of any previous attempt to link so closely the two sections
of the Theaetetus considered here and also to explain the thought of Tht.
184b–186e in the light of what is said at 156a–157e. This linkage has been
made possible by dealing with a question, namely that of the identity of
the subtler thinkers, which seemed at first sight a purely exegetical ques-
tion but which has proved to offer us better insight into the philosophical
arguments of the Theaetetus.28
28
This chapter derives from a revision of a paper I read at the Durham 2010 conference and has bene-
fited from the discussion there. I thank especially David Sedley and Terry Penner for their helpful
comments. George Boys-Stones has been kind enough to discuss the main ideas of this paper both
at the conference and by email. As is characteristic of him, Christopher Gill offered me very useful
comments on the paper, both at the conference and a few months later, on a walk we together had
in Dublin. He also went through the written version of the paper more than once, providing sug-
gestions for improving my English when needed. The research that has made this paper possible
has been funded by the Irish Research Council for the Humanities and Social Sciences (IRCHSS),
whose generous support I am happy to record. Lastly, I thank an anonymous reader for Cambridge
University Press for useful comments on an earlier draft.
ch apter el eve n
The Wax Tablet, logic and Protagoreanism

Terry Penner
Introduction
The Problem of False Identity Beliefs (hereinafter FIB) is one of three
acute philosophical problems Plato thinks he needs to face up to in the
Theaetetus and Sophist. The other two I shall call the Problem of Non-Being
(hereinafter NB) at Tht. 188c–189b, Sph. 236e–239c and passim, and the
Problem of False Statement (hereinafter FS) at Sph.261c–263d. While false
identity beliefs are my main focus in this paper, I shall argue that it is
Plato’s view that this problem cannot be handled without dealing with
the other two problems as well. As for the other two problems, they are,
for Plato, so closely related that Plato appears to treat NB, the problem of
non-being, as solved merely by solving FS, the problem of false statement.
Why Plato should see these problems as so closely related tends to remain
obscure, I think, even within the tradition of analytical philosophy, which
has produced some of our very best Plato interpreters, and which is the
tradition in which I grew up and to which I owe most.
FIB is the only one of Plato’s three problems that modern analytical
philosophers think call for more than cursory attention.1 It is broached
Many of the themes of this paper are close relatives of themes Christopher Rowe and I have been
developing over the years, most recently in Penner and Rowe 2005. I have learned hugely from him
in the many intense discussions we had in the years we were co-workers (along with the odd moment
of a no doubt justified impatience which he will have felt for me after what might be three straight
days of constant discussion, when we would inevitably get a little fratchy). Equally importantly, I also
learned from the friendship that bloomed in that partnership. One of the things we claim in our 2005
book is that friendship is a joint striving for the truth – something I have found, and been grateful for,
over and over again within the community of students of ancient philosophy. Thanks also to George
Boys-Stones, Dimitri El Murr and, for his ever-kind – and comprehending – help to me, Christopher
Gill.
1
Modern analytical philosophy dispenses with NB – at the outset – via the existential quantifier
and the associated theory that existence is not a predicate. At the same time, it dispenses with FS
via the introduction of the proposition; and with FIB via a Fregean theory of psychological contexts. I
believe Plato would have regarded all three moves as well motivated, but excessive – as throwing out
the baby with the bathwater. For a much more detailed and improved discussion of these issues,
186
The Wax Tablet, logic and Protagoreanism 187
in the False Belief section of the Theaetetus, the second of the three main
philosophical parts of that dialogue. The other two parts are the long sec-
tion on the Protagorean view that Knowledge is Perception (151d–186a),
and the concluding section on the claim that Knowledge is True Belief
with an Account (201c–210d). The second part, which concludes with the
easy refutation of the suggestion that Knowledge is True Belief, is mostly
taken up with a question that may seem irrelevant to the rest of the dia-
logue: the question how False Belief is possible. There are five principal
arguments concerning False Belief: the Argument from Knowing and
Not-Knowing, the Argument from Being and Not-Being, the Argument
from Interchange (allodoxia), the Wax Tablet model, and the Aviary
model. All but the second are explicitly directed at FIB. Of the remaining
four arguments, only that based on the Wax Tablet model (WT) has been
of much philosophical interest to the modern analytical philosopher. To
see why this is so, I shall begin by explaining FIB – both in Plato’s version,
and in the modern Fregean version – and then look at how moderns plan
to solve the problem.
Suppose that when Socrates sees Theodorus coming in the distance, it
happens that:
TH1: Socrates believes that the one coming is Theaetetus.
What is this a belief about? What does it refer to? It will seem natu-
ral to many to assume that it is about the people Socrates thinks it is
about, namely, the person who is the one coming and the person who is
Theaetetus. On that assumption, we should it have that:
TH2: Socrates believes that Theodorus is Theaetetus.
But suppose it is also the case that:
TH3: Socrates knows both Theaetetus and his teacher Theodorus – and knows
them very well.
In that case, it would seem that:
TH4: Socrates does not believe that Theodorus is Theaetetus
– contradicting TH2! But TH4 seems quite solid. At least it does, given
the Argument from Interchange’s plausible-looking account of believing
(or thinking or judging) something in terms of saying something to oneself
juxtaposing more explicitly the modern handling of these three problems with Plato’s, see Penner
forthcoming, tentatively entitled ‘Plato takes aim at Intentionality’.
188 Terry Penner
(Tht. 189e): would Socrates say to himself, knowing both Theaetetus and
Theodorus as he does, that Theodorus is Theaetetus? Yet he does clearly say
to himself that the one coming is Theaetetus.
The problem here appears to flow from the assumption made above
about the seemingly natural belief in TH1, that:
ABT: What Socrates’ beliefs are about is what he thinks they are about.
On this assumption, the belief in TH1 must be about the person who is
the one coming. And that person is Theodorus. But then we can surely
substitute ‘Theodorus’ for ‘the one coming’ and that will get us from the
true statement TH1 to the false statement TH4, thus apparently violat-
ing the principle of the substitutivity of identicals salva veritate. This
substitutivity principle is central to ‘extensional contexts’. In the present
discussion, we may take extensional contexts to be those that involve no
reference from within a psychological context.2 I shall call this problem
the ‘Theaetetus problem’. The differences between beliefs produced by
substitutivity show us why moderns suppose that this is a problem about
reference from within ‘psychological contexts’. By modern lights, there is
one ‘logic of extensional contexts’ and one ‘logic of psychological con-
texts’: substitutivity works in extensional contexts, but not in psychologi-
cal contexts.
Consider now the classic Morning Star/Evening star case in Frege (1892)
of such failures of substitutivity.3 Suppose that:
FR1: Gilgamesh believes that the Morning Star is not the Evening Star (when,
unbeknownst to Gilgamesh, actually it is, since as later astronomers discov-
ered, both the Morning Star and the Evening Star turn out to be the planet
Venus).
Yet surely it is false that:
FR2: Gilgamesh believes that the Morning Star is not the Morning Star.
(Frege says the falsity of this last claim is analytic – an instance of the law
of logic that says that nothing can fail to be self-identical.) Once more,
2
Consider the perfectly correct argument – in an extensional context:
Theodorus is Theaetetus’ only teacher;
Theodorus is the one coming;
therefore, the one coming is Theaetetus’ only teacher.
Such an argument must surely be granted – even by those who do not know that Theodorus is
Theaetetus’ only teacher.
3
Translation at Beaney 1997: 151–6.
ABT would make Gilgamesh’s first belief be about the object that is the
Evening Star, which is, as we now know, the object that is the Morning
Star. So if FR1 is true, so should FR2 be true. But it apparently is not. So
once more the principle of the substitutivity of identicals salva veritate
is violated in this psychological context. (I shall call this the ‘Gilgamesh
problem’.)
The Fregean solution to the Gilgamesh problem is to suppose that ABT
is false, and, therefore, to ‘change the subject’4 – in the Gilgamesh prob-
lem, from the Evening Star, which Frege describes as the reference of the
linguistic expression ‘the Evening Star’, to the Fregean sense, or meaning,
of the linguistic expression ‘the Evening Star’. Put in a variant form, ‘the
Evening Star’ in the sentence ‘Gilgamesh believes that the Morning Star
is not the Evening Star’ does not refer to – is not about – the object which
is the Evening Star, that is, the planet Venus – that is, the Morning Star.
It is about the object that is the Evening Star under the description ‘The
Evening Star’, which is not at all the same thing as the object that is the
Evening Star under the description ‘the Morning Star’, let alone the object
that is the Evening Star tout court.
Here is a different way to make the same point: the difference between
the first belief and the second belief lies not in the sameness of reference
for the linguistic expressions ‘Morning Star’ and ‘Evening Star’ – for then
these would not be different beliefs – but in the difference of the mean-
ings of the two expressions, that is, in the two different ways in which
the expressions involved take one from the expression to the same refer-
ence or, if you prefer, in two different ways of approaching the same refer-
ence, or two different conceptions of the same object, or two different sets
of directions for determining, or picking out, the reference. I shall call all
these variant solutions ‘Frege-style’ solutions.
For Frege-style theorists and, therefore, for many interpretations of
Plato, TH4 is not about the reference of the expression ‘Theodorus’ or
the reference of ‘Theaetetus’. It is about the meanings of these expressions,
thus allowing TH1 to say something different from TH4, so that TH1
may be true, and TH4 false. (Or, in the other variant, the first is about not
Theodorus, but about Theodorus under the description ‘the one coming’,
while the second is about not Theodorus but Theodorus under the descrip-
tion ‘Theodorus’.) Now, it is because this is just the kind of thing we see
4
See Furth 1968 on non-existence (outside of belief contexts). If the method of ‘changing the subject’
shows us the way to handle belief contexts themselves (Furth does not suggest this), shouldn’t the root
problem be the same: not something special about one context or another, but, say, about falsity?
See further Penner forthcoming.
190 Terry Penner
in the modern solution to the Gilgamesh problem that modern analytical
philosophers are so taken with the fourth of the arguments Plato gives in
the False Belief section concerning false identity beliefs – the attempt at
a solution of FIB by means of WT. For WT is mouth-wateringly close to
modern Frege-style treatments of ‘identity within psychological contexts’
and similarly close to something we may call the ‘logic of psychological
contexts’.
Let’s see how WT’s solution (190e–195b) goes. Plato begins by model-
ling Socrates’ memory bank as a wax tablet, in which memory traces are
engraved. Then we change the subject from these two people, Theaetetus
and the one coming (that is, Theodorus), to the following three objects:
(1) the the-one-coming visual percept (opsis, aisthēsis: 193c–194b),
(2) the Theaetetus memory trace (mnēmeion) in Socrates’ Wax Tablet,
and
(3) the Theodorus memory trace (mnēmeion) in Socrates’ Wax Tablet.
With the change of subject to these three entities, we must also of course
‘change the predicate’. (The point is not that Socrates thinks that the
the-one-coming visual percept is identical with either memory trace.) So
‘change the predicate’ from ‘Socrates thinks falsely that x is y’ (the iden-
tity predicate) to ‘Socrates mismatches u to v instead of matching it to z’.
Then the problem that apparently cannot be solved using only two enti-
ties (Theaetetus and the one coming, that is, Theodorus) is solved with
these three entities (an opsis or aisthēsis and two mnēmeia) and a different
relation between the three. As a result we can say that:
TH5: Socrates mismatches his the-one-coming visual percept to the Theaetetus
memory trace in the Wax Tablet, instead of matching it to his Theodorus
memory trace.
The problem seems to have disappeared! It is easy to see why this awak-
ens the interest of interpreters who are analytical philosophers. Socrates
no longer believes that Theodorus is Theaetetus. Is WT’s solution not
simply a special case of the modern solution, using variants of the enti-
ties to which Frege changes the subject? In Frege’s theory we would also
‘change the subject’ from two objects (Theaetetus and Theodorus, that is,
the one coming) to three objects – the respective senses of ‘Theaetetus’,
‘Theodorus’ and ‘the one coming’. (And there will be a corresponding
‘change of predicate’.)
I should note, for purposes of clarity, that Plato also allows for a fur-
ther way besides simple mismatching to get a false belief involving one’s
memory bank. This is that the wax of the model may in some cases be
too soft, or too hard or too shaggy to allow the memory imprint faith-
fully to record a clear imprint of the person who is being remembered
(194c–195b). This shows us that the mnēmeia are not just dummy entities
standing in place of the people being remembered. The solution of WT is
accordingly not compatible with saying that Socrates mismatches his opsis
or aisthēsis to the person Theaetetus (when that person is not the person
who is the one coming), instead of correctly matching it to the person
Theodorus (when that person is the person who is the one coming). One
does not want the Wax Tablet solution saying that Socrates mismatches
his ‘the one coming’ visual impression to the person who is not coming.
This is why there is so much ‘changing of the subject’ in TH5. The job
would appear to be adequately done only by ‘changing the subject’ from
‘the one coming’ to the opsis or aisthēsis and to the mnēmeion and only
then observing ABT. As it is, any mismatching or correct matching must
be to these other entities called mnēmeia. Thus, in clear violation of any
attempt to maintain the principle ABT, no person at all will occur in WT.
Socrates, so to speak, does not have any person in mind here. Rather what
he is doing is, inaccurately, shuffling around visual percepts and memory
traces! (The same applies, mutatis mutandis, when we are dealing with
Fregean senses in the modern solution to the Gilgamesh problem.)
In the method of ‘changing the subject’, we have a suspiciously simple
general recipe for solving philosophical problems. ‘You can’t solve it with
two objects (say, two persons)? Easy. Produce at least three objects (say, a
visual percept and two memory traces) and then restate the problem in
terms of these three objects!’
Plato then (195e–196c) raises a difficulty for WT’s solution to FIB. It is
that if we go beyond cases involving perceptual misidentification to errors
in pure mathematics, such as Little Johnnie believing that 5 + 7 = 11, there
will be no opsis to which one can change the subject. So WT will fail as
applied to mathematical mistakes.
When, early on in my first, Geach-induced, student flush of Fregean
enthusiasm, I contemplated the ‘Theaetetus problem’ and thought about
the comparison with Frege’s Gilgamesh problem, it seemed to me appal-
ling that Plato should have rejected WT because of this mathematical
example, instead of opting for one of two other possible reactions which
seemed perfectly open to him – reactions that would have enabled him
to fend off this problem of mathematical mistakes in a clearly Frege-style
manner. One such reaction would have been to construct a mathemat-
ical analogue to opseis or aisthēmata, appealing to, say, mathematical
192 Terry Penner
constructions of numbers – multiple modes of constructing the same
numbers – instead of to numbers themselves. We could have a [5+7] con-
struction which matches the number 12 in our memory bank, a [3x4] con-
struction which matches the number 12 in that memory bank, an [11+1]
construction of the number 12, and so on. (I shall speak also of an [11]
identity construction of the number 11 and a similar [12] construction,
neither of which is identical with the numbers 11 and 12, just as neither
the Theaetetus-mnēmeion nor the Theodorus-mnēmeion are identical with
the persons Theaetetus and Theodorus.) Then we could speak of Little
Johnnie mismatching his [5+7] construction to his [11] construction, and
the problem would be solved without having Little Johnnie think that 12
was 11. (But he wouldn’t be thinking of any numbers: just experiencing
various ways of constructing numbers.) Once we have at least these three
constructions, instead of just the two numbers 11 and 12 (that is, 5 + 7), the
problem will be solved.5
Even better – and even more Fregean – would be to ‘change the subject’
not just to an analogue of opseis, but to a generalisation of all such ana-
logues of opseis and mnēmeia. That is, Plato could have changed the sub-
ject to our more general approach to the one coming (or to the number 12),
so that we could have a the-one-coming approach (to Theodorus), as well
as our Theodorus approach (to Theodorus) and our Theaetetus approach to
Theaetetus. (And similarly for our different constructions of the numbers
12 and 11.) Then we just speak of mismatching one approach to another
approach. And the same trick could be done with Little Johnnie’s false
belief that [5+7] = 11. (Instead of speaking of two different approaches to the
same object, we could have spoken of two different conceptions of the same
object. These entities could all perfectly well function as non-linguistic
analogues of Frege’s meanings or Fregean senses of words, or of the modern
descriptions under which.)
(I pause here to emphasise the point, which will be of signal impor-
tance later in this discussion, that with these Frege-style solutions we
will have dropped all (direct) reference to people or to numbers, and will
speak only of opseis, mnēmeia, mathematical constructions, approaches or
conceptions.)6
5
I shall discuss the evident need that the mnēmeia be not the numbers 11 and 12, but the respective
identity constructions in Penner (forthcoming). It is enough for my present purposes to concentrate
simply on analogues to aisthēseis or opseis.
6
There will still be indirect reference to people or numbers, but only via a designation relation
between sense and reference that holds outside the believer’s mind. See the early sections of Kaplan
1968.
If Plato had reacted in either one of these ways, he would have been giving
essentially a Frege-style solution. So he had the Fregean solution in the palm
of his hand – and turned away from it! Was it really, one may ask, beyond the
wit of Plato to opt for one of these alternative reactions? In any event, why
doesn’t he at least keep WT as a partial solution for perceptual cases?
This all raises the question, ‘What is Plato thinking of in rejecting this
Frege-style solution?’ Put otherwise, why does Plato turn away from WT
entirely, instead of seeking some fairly easy modification of the model?’
The ‘linguistic turn’

The linguistic-semantical cast Frege and analytical philosophers give to
their solutions to FIB will have been noticed above in the frequent use
of the terms ‘linguistic expressions’. Things referred to are references of
linguistic expressions; and meanings are meanings of expressions. Given
the extraordinary importance such treatments assign to meanings – and
other things with the same identity conditions, such as descriptions under
which – it will be an essential preliminary to my deploying my answer
to the main question that I address the question just articulated about
meanings, which are so central to Frege-style solutions. Would Plato have
thought that it is linguistic expressions that refer to existing things via their
meanings? Or would he have thought that it is we who refer to them via
the linguistic descriptions we give of them?
It is a consequence of the modern ‘linguistic turn’, that for moderns:
M1: In general, it is words and expressions that have references via their mean-
ings, and it is the meanings that (given a particular way the world is) deter-
mine what that reference is.7
To see how all this applies to sentences, I turn to a supposed entity first
posited by Aristotle in the interests of his entirely new discipline of Logic:
M2: It is sentences that say something and that have truth values; they do so via
the propositions that the sentences express; and it is the propositions that
(given a particular way the world is) determine what that truth value is.
I note further here that if the reference of a sentence (supposing sentences
do have references) is determined by the references of its parts then, for
Frege, the truth value of a sentence – surprisingly, but quite correctly – is
7
As the meaning of an expression (its ‘reference conditions’) stands to the ‘semantic value’ of the
expression (its reference), so the proposition expressed by a sentence (its truth conditions) stands to
the semantic value of the sentence (its truth value). See PROP4, p. 217 below.
194 Terry Penner
its reference. In addition, the meaning of a sentence (= the proposition the
sentence expresses) will turn out to be determined by the meanings of its
parts. (There will be more about propositions later when we come to FS.)
Notice that I am not saying that Plato would have held this view of the
reference of sentences. For he does not think words or sentences refer to
things or say things at all. He thinks people refer to things and say things.
What someone says by means of a sentence is accordingly, in Plato, not
‘inscrutable’ in the way it is in Frege and in the standard modern seman-
tics for first-order logic. What a person is saying by means of a sentence
(which I shall interpret below as saying the same thing as what the person
is referring to in what they say) always requires, in simple subject-predicate
cases, the singling out both of the object referred to in the subject expres-
sion, and the property referred to by means of the predicate expression.
In the Cratylus, we see how Plato would have reacted to the modern
‘linguistic turn’. In this dialogue, he rejects the doctrine of linguistic con-
ventionalism – a doctrine that he thinks of (surprisingly!) as a Protagorean
view. The Protagorean position on language (taken as a special case of ‘as
things appear to us, so they are for us’), is that:
PR1: what seems to us (δόκει) to be the name of a thing is the thing’s name (in our
language or idiolect).
As I read this ‘linguistic Protagoreanism’, its intention is to say that our
conventions for the word ‘cutting’ determine (given the actual state of
the world) whether cutting is going on. Looked at in this way, the lin-
guistic Protagorean’s position begins to look like the sophisticated mod-
ern Fregean doctrine we have just been noting, that meaning determines
reference (given the actual state of the world: see Cra. 384c10–385e3 with
385e4–386a4, 386d8–387c4).
If in this particular linguistic Protagoreanism meanings determine ref-
erence, it is not at all that way for Plato. At Cra. 387a, Plato arrestingly
raises the question whether, if we’re trying to cut something, we are to
cut in whatever way we wish and with whatever instrument we wish. Are
we to cut in accordance with what we believe cutting to be? Or in accord-
ance with what our conventions for the word ‘cutting’ determine?8 Or
8
If in the conventional rule ‘name n shall name object o’, what n and o stand for are matters for our
conventions to decide (see the important passage Cra. 385e4–386a4), then we do not have to find
out first what are real objects and what are not, before assigning names – a point Plato just slips in at
Cra. 433e4–5. For Plato, the problem with linguistic conventionalism is its failure to see that objects
come first, names and conventions come later. Since they will frequently misrepresent the objects,
Plato will certainly have denied that names, meanings or conventions determine reference.
do we not want rather to cut in accordance with the real nature of cutting
even if – I take him to be saying – that real nature differs from what our
beliefs and the conventions of our language pick out as what cutting is?
(Plato raises the same questions, mutatis mutandis, for the real nature of
burning and – exquisitely – naming: can we simply declare by convention
that what makes names correct is their being declared correct by conven-
tion?) To change examples, is a researcher who is trying to discover the
real nature of cancer trying to find out what the reference is that is deter-
mined by the present meaning of ‘cancer’? Surely not. Our researcher may
well find the meaning assigned to the word, thanks to the consensus of
contemporary scientists, to be well off the mark. No, our researcher will
say, I do not want to talk about that. What I want to talk about is the real
nature of cancer, however that may differ from what others think it is or
what they define it as, and even from what I currently think it is – and
even if we are destined never to get a fully satisfactory answer. Such is
Socratic-Platonic Ultra-Realism.9
We see here how thoroughly Plato rejects the modern ‘linguistic turn’
in which:
LingT: a statement (that is, what a person says by means of a sentence) becomes
(reduces to) what the person’s sentence says, understood in terms of the prop-
osition the sentence expresses.10 Similarly, what a person is referring to by
means of a given linguistic expression (if anything) becomes what the per-
son’s referring expression refers to (if anything).
Meanings are meanings of words, propositions are propositions expressed
by sentences. And in the Linguistic Turn, propositions determine truth
value, while meanings determine reference.11 Christopher Rowe and I say,
by contrast, that Socratic-Platonic Ultra-Realism would have resisted both
of these reductions from what we are saying, and what we are referring to,
to what our sentences and words express or refer to.12
9
‘Ultra’ because this it goes well beyond the realism that says that some expressions have real refer-
ences, with such references being determined by their meanings. See Penner 2005b: 157–60, 168,
171–2, 176, and 2011: 282–91, esp. 286–8, as well as Penner and Rowe 2005: 205–10 on the ‘Principle
of Real Reference’.
10
This is also clear in Aristotle’s syllogistic. When actual predicates replace the predicate letters A, B,
etc., in such syllogistic formulas as ‘Every A is B’, the result is a sentence. (The proposition, we shall
see below, turns out to be the meaning of the sentence.)
11
A few modern philosophers (Donnellan, Kripke, Kaplan, Burge and others) are Ultra-Realist in
some areas – about proper names, referential definite descriptions, demonstratives and natural kind
terms. Elsewhere, their approach remains thoroughly Fregean.
12
See Penner and Rowe 2005, index s.vv. ‘what speakers intend to refer to’ and ‘what speakers say’.
See also Penner 2007b.
196 Terry Penner
With so much by way of foreshadowing the quite different approach I
believe Plato takes to psychological contexts from the modern Frege-style
approach, let us turn to the question why Plato rejects WT entirely. One
possibility is that he would be unhappy for the most part about abandon-
ing the principle ABT.13 In favour of considering this possibility is the fol-
lowing consideration. Surely, on at least some occasions, as when:
TH6: Socrates believes truly that Theaetetus is Theodorus’ best student, Socrates’
belief really is about the person who is Theaetetus and about the person who
is Theodorus’ best student.
Why should any doctrine about psychological contexts disturb our intui-
tions concerning aboutness at all, except when there is falsity involved?14 It is
true that Socrates wouldn’t say that:
TH7: he (Socrates) believes that Theaetetus is the gifted young mathematician
whose death is reported at the beginning of the Theaetetus …
… (since Socrates was himself long dead before Theaetetus’ death).
Nevertheless, it is surely still arguable that:
TH8: Socrates believes that the gifted young mathematician whose death is dis-
cussed at the beginning of the Theaetetus is Theodorus’ best student.
What problem – analogous to FIB – arises here if we are supposing
that Socrates’ belief in each of (TH6)–(TH8) is about the person who is
Theaetetus?
What I am proposing here is not without parallels in analytical phi-
losophy – if we leave aside belief contexts. Just consider how analytical
philosophers treat statements a person makes in a so-called ‘extensional
context’, such as the statement someone might make that:
FR3: Theaetetus is Theodorus’ best pupil.
Within an extensional context, this will be true if and only if:
FR4: the gifted young mathematician whose death is discussed at the beginning
of the Theaetetus is Theodorus’ best pupil.
Each of the last two statements is equally about the person who is
Theaetetus. That is all Plato would need to propose holds also with true
identities in belief contexts.
13
At least provided that what people think they are referring to is, say, the real nature of cancer, and
not what the present meaning of ‘cancer’ determines. See also n. 15 below.
14
See n. 1 above and Penner forthcoming as cited there.
This possibility suggests that the problem with false identity beliefs may
not arise so much because of belief contexts as because falsity is involved.
This is exactly what I shall argue in the sequel. In so arguing, I shall not be
affirming that Frege’s decision to ‘change the subject’ for true beliefs as well
is entirely unmotivated. Indeed, I believe it is very well motivated – for any-
one whose primary concern is, like Aristotle’s, to provide a logic that will
apply in a simple and clear way not only to belief statements we know to
be true and belief statements we know to be false, but also to belief state-
ments about which we do not yet know (and may never know) whether
they are true or false. How is a logician to do this in advance? By treating
belief statements as involving ‘beliefs’ that are the same thing whether they
are true or false. This, we shall see, is an extension of Aristotle’s motive in
the area of extensional logic for adding, irreversibly, to the ontology of
Western philosophy meanings and propositions. By contrast, I shall claim,
Plato would have believed in neither. Plato might instead have asked,
‘What if logic is a friend, but a greater friend is truth and real existence?’
I am myself with Plato here. When I think about my children, or my
grandchildren, and believe things about them, I thereby place them in a
psychological context. Does my mind then at most only (directly) encoun-
ter the meanings of the words I use in referring to them (or what those
meanings determine)? Or even only encounter what the meanings of cer-
tain descriptions (under which I think of them) determine? Am I not,
at least quite often, thinking about them – the children and grandchil-
dren – even where my conceptions or descriptions of them are wrong and
misleading (as they of course all too often are)? The remarks above about
the Cratylus would surely have us say here: ‘Yes, it is them I am thinking
about, even if I am wrong about them in numerous ways. At least often.’
(See Socrates’ not wholly successful attempt to break down Cratylus on
just this point at Cra. 430a–433b: the name is, at best, merely a more or
less adequate imitation or picture of the thing named.) This impulse I
take to be both true and profoundly Socratic-Platonic. It is not always
true, of course. Who is Socrates thinking about when he says ‘Hi there,
Hermogenes!’ to Cratylus (Cra. 429e)? The person who is Hermogenes,
i.e. Cratylus? There is no such person.15 But perhaps – as I have been
suggesting – cases of false identifications and non-existents can become
problematic because too much falsehood is involved, not solely because of
15
Some beliefs, say about my children, although in various ways false, remain true enough that it is
still them I am thinking about. (See nn. 34–7 below, with text thereto.) Other false beliefs cannot
be in this way overridden. (See below, p. 210.)
198 Terry Penner
psychological contexts. Perhaps the problem is not one that requires that
we entirely abandon ABT.
But can I really maintain that these are the kinds of consideration that
lie behind Plato’s complete rejection of WT?
Why does Plato reject the Wax Tablet model?

Let us examine the evidence on the question why Plato rejects WT, and
see whether Plato’s reasons have anything to do with what I am suggest-
ing: his reluctance to abandon the principle ABT. Does Plato think he
has an alternative to trying to fix up WT along Fregean lines? I think he
does – and that he opposes any such Fregean lines of thought. He puts it
in the following way in confronting the failure of WT:
CH: [So, given that false belief cannot be the misfitting of thought to percep-
tion,] either
(A) there is no such thing as false belief, or
(B) it is the case that what one knows one may also not know.
‘Which do you choose?’ Socrates asks; and Theaetetus replies that this is
an impossible choice, and Socrates apparently agrees that neither option is
possible (196c4–d2).
But if Plato is rejecting (A) – as surely he is – then he may well be tell-
ing us here that it is (B) he wants us to accept. He may be accepting (B)
in spite of its inadequate prima facie credentials, given that its denial, that
you can’t both know and not know the same thing, is presupposed in
all four of the arguments concerning false identity beliefs, including the
Wax Tablet argument. (See 188a10–b1 with b3–5; 190d5–6; 192a1–b2; and
196b8–10 with c1–2.) Hence the opting for (B) would be the rejecting of
the Wax Tablet model.
This suggests that what we have here is a familiar Socratic mode of apo-
retic argument, as at the end of the Hippias Minor (376b4–6), daring us,
as it were, to recognise for ourselves where the problems lie and only just
barely suggesting how they are to be dealt with. All right, but what has the
False Belief section been saying all along that leads to (B) that you can both
know and not know the same thing? The answer is in the immediately pre-
ceding passage – at 196b8–10 with c1–2: that Socrates won’t say to himself
that this person he knows (think of Theaetetus, whom Socrates knows ex
hypothesi) is someone else he knows (think of Theodorus, that is, the one
coming, where Socrates also knows Theodorus ex hypothesi). And where
does this claim come from? It comes from the dreadful-looking argument
(187e–188c) with which the False Belief section of the Theaetetus begins,
the so-called ‘Argument from Knowing and Not Knowing’ (AKNK).
Now, AKNK is an argument that commentators – other than those
earnest seekers of Plato’s ‘logical fallacies’ – tend to rush over in some
embarrassment. But perhaps we need to re-examine this argument. The
central part of this argument – which I call the ‘Terrible Trichotomy’,
TT – goes like this (188b3–c9):
TT: For any two objects, either (1) you know both, in which case you would
not confuse them, or (2) you know neither, in which case you could not so
much as bring either of them before your mind, or (3) you know only one
(in which case you also could not confuse it with one you do not know: for
you can’t think things you know are things you do not know, and vice versa).
Now what is said to be impossible in TT is precisely what WT seems to
show is eminently possible, namely, that:
TH9: Socrates sees someone in the distance, and acquires the false identity belief
that the one coming is Theaetetus, when actually (unbeknownst to Socrates) it
is Theodorus.
But let us not despair of TT. Let us not rush to a diagnosis of the ‘fal-
lacy’ – or look for special senses of the word ‘know’. On what assumptions
does Plato lay out this trichotomy? Plato tells us that it is based on the
assumptions that:
A1: For any given thing one thinks about, one must either know it or not know
it (188a7–8);
and:
A2: It is impossible if you know a thing, that you also not know it (188a10–b1).
But (A2) is not only the reason Plato gives for rejecting option (3) of TT.
It is also – arguably – his reason given for rejecting option (1), namely
confusing something you know with something else you know. (Suppose,
in option (1), where you know a and you know b, that you confuse them.
Then you would both know a and – since you confuse it with b – you
would not know a.) Now, if – surprisingly – we deny (A2), we are affirm-
ing (B). But, given the failure of WT, (B) is our only alternative to accept-
ing that (A) there is no such thing as false belief (p. 198). And yet – how
could (B) be true?
The idea here seems to be: with Theaetetus, that is, the one coming,
we never come by complete knowledge – an agreeably Socratic idea. We,
200 Terry Penner
as it were, both know this person and we do not know him. And once
more, there is an analogue in extensional logic: see the following remarks
in Frege (1892) (trans. Beaney 1997: 153):
The sense of a proper name is grasped by everyone who is sufficiently famil-
iar with the language or totality of designations to which it belongs; but
this serves to illuminate only a single aspect of the [reference], supposing
it to have one. Comprehensive knowledge of the [reference] would require
us to be able to say immediately whether any given sense attaches to it. To
such knowledge we never attain.
Comprehensive knowledge of the reference: this is what Plato is talking
about when – to much modern scorn – Plato says that there is no knowl-
edge of individuals.16
So what is this line of thought that not only underlies both options
(1) and (3), but also leads to the second alternative, (B), in reacting to
the failure of WT? Is there some more general view that underwrites the
entire idea behind the trichotomy, some – surprising, but hard to refute –
assumption which, if correct, would render TT an excellent argument,
instead of the perfectly dreadful argument it is normally taken to be? If
so, then might it not be that general view that Plato wants us to see as the
source of the aporia with which WT concludes?
Plato’s rejection of Protagoreanism

I suggest there is such a – surprising – general view that underlies TT,
one that is not at all easy to refute and that would render TT an excel-
lent argument. To give up this view would justify giving up TT and the
assumption (B) that you can’t both know and not know the same thing!
So here is my question. What if the opseis and asthēseis in the Wax Tablet
are the aisthēseis of the ‘Knowledge is Perception’ part of the dialogue –
Protagorean appearances? If, as is generally granted (and as I shall pres-
ently argue independently), Plato rejects the Protagoreanism of this part
of the Theaetetus, then the use of TT throughout the False Belief section
of the Theaetetus, instead of showing Plato (rather foolishly) arguing in
accordance with TT, would show him giving an additional reductio ad
absurdum of Protagoreanism.
16
It was Christopher Rowe who brought home to me the importance of the (Socratic!) point that so
far is the Republic from thinking anyone has knowledge of the individual which is the form of the
Good that it can do no better than offer similes and allegories when speaking of the form of the
Good. See also R. 505a4–b1, 505e1–5.
On the Protagorean phenomenon/noumenon view of knowledge,17 the only
things of which we have any knowledge whatever – or to which we have any
epistemological access at all – are these Protagorean appearances (incorrigibly
known Protagorean appearances, as we shall see). Thus, the Theaetetus spells
out the Protagorean dictum ‘Man is the measure of all things, of things that
are that they are and of things that are not that they are not’ as:
PR2: As things appear to us, so they are for us.
The ‘for us’ is, in this case, to be explained as follows: what we start off
thinking of as the wind outside us appears cold to me, and appears not
cold to you; and if it appears cold to me, it is cold – for me. That is, we are
both right – having the kritērion in ourselves – as to what is so in our inner
appearance.18 Hence each of our appearances is incorrigible – and incor-
rigibly true. (They are not just neither true nor false à la Wittgenstein, Ryle,
Sellars or Rorty.) They are true – in our inner appearance. Our appear-
ances are all incorrigibly true.19 The only things that it is intelligible to
suppose are so – the only ‘real facts’, as Peirce (1878: 301) might have put
it – are things of the sort the wind is cold for me and the wind is not cold for
you. For thanks to the constant Heraclitean flux which the Theaetetus takes
to be part of Protagoreanism, there is no stable truth out there involving
the wind.20 (The wind is of course the perfect symbol of Heraclitean flux,
as it is of the inaccessible noumenon for Kant, and for Hume – so long as
he does not leave his study.) In fact, these truths about our inner appear-
ances are the only truths it ‘makes sense’ to suppose we have.21
17
Sextus Empiricus, Outlines of Pyrrhonism 1.31–4, a somewhat inferior rehearsal of the material of
the Theaetetus.
18
‘For us’ here = ‘in our inner appearances’, there being no truth outside of any perceiver definite
enough ever to make appearance false. (See below, n. 20, on the relevance of Heracliteanism to
keeping any things there may be outside of us indefinite enough to block any impediment to
Protagoreanism.) Just so with conventionalism in PR1 above (pp. 194–5), ‘for us’ = ‘in our lan-
guage’, there being no absolute truth outside the given language about what the name of some-
thing is. Similarly for the relevant analogues to ‘for us’ in the Protagorean accounts of legality and
morality.
19
Compare B55–6, 60 of Kant’s First Critique.
20
If there should be any stable truths outside of us – either about perceivers or about things per-
ceived – that are in any way connected with our inner appearances, the appearances would be
giving hostages to the future, which could result in their being falsified. See the way in which
Heracliteanism is used – à la strict empiricism – to ensure that personal identity over time is not
allowed to become a hostage to the objective future – by its being the same person who is first
appeared to in one way, then in another, e.g. at 157ab, 159a–c. See below (pp. 207–8) on the failure
of this attempted use of Heracliteanism.
21
But Berkeley found it perfectly easy to argue this on the basis of incorrigibly known ideas without
Heracliteanism. It is just that Hume and Kant – like the Protagorean in the Theaetetus – were
(commendably) less willing to abandon the world we all accept with a kind of animal faith.
202 Terry Penner
I now return to TT. I have suggested that TT should be a perfectly
good argument if, as Protagoras supposes, the only entities it ‘makes sense’
to suppose there are, are incorrigibly known Protagorean appearances. (And
notice that this makes TT a perfectly good argument without the bringing
in of any special senses of ‘know’. All the work is done by the Protagorean
thesis as to what things there really are.)22 Let’s see how this works in TT:
(1) If you have incorrigible knowledge of both Theaetetus and Theodorus,
you could hardly think the one is the other.23 For, as just argued, any
such belief would show you did not have incorrigible knowledge of
either: you would, as (B) (p. 198 above) says, both know Theaetetus
and – because you do not know that he is not the one coming – not
know him.
(2) You couldn’t think anything at all about two things you do not
know – for there are no such objects (only incorrigibly known
Protagorean appearances).
(3) And you couldn’t think that something you know is the same thing
as something you do not know, since there is not anything you can
think of at all that you do not incorrigibly know. Alternatively, once
more as (B) says, you would then be both knowing something and
not knowing it.
‘But meanings are not Protagorean appearances!’

‘Yes, but what has all this about Protagorean entities to do with a modern’s
use of meanings and propositions? Surely moderns, in their commitment
to meanings and propositions, aren’t committed to Protagorean entities!’
I am not so sure about this. When I think about such modern entities as
meanings of expressions and propositions expressed by sentences, or about
my conceptions of things, or the descriptions under which I think about
things or my approaches to things, they seem to me to be represented as
entities that we embrace (if we do) because while we may be mistaken
22
See above, p. 190–2, on problems that are insoluble with n objects becoming soluble if you posit at
least one additional object (multiple appearances vs. just the one wind). This is exactly what hap-
pens whenever an ambiguity is posited. In my experience, the positing of ambiguities and multiple
senses of a word almost always has all the advantages of theft over honest toil.
23
Compare a Protagorean being asked to accept that this one sense datum to the left of his visual
field is that other sense datum to the right of the visual field: see Cornford 1935 on 193b6–7. It is an
implication of WT that one memory trace cannot be confused with another (at least in the case of
two things known but not perceived): a different sort of Protagorean appearance! (We might call
these ‘neo-Protagorean appearances’: constructs which Plato adds to Protagoreanism for purposes
of the Wax Tablet model.)
about the things our meanings or propositions determine, or the things
our conceptions or descriptions are of, we think we know at least what we
mean in using the words or descriptions, or what we believe or what our
conceptions or descriptions are. (‘I may not get right what the thing is
that I am thinking about, and I may not even be describing it rightly. But,
look, I know what I mean.’ Similarly, ‘I may not know whether or not
what I believe is true. But I do know what it is that I believe.’)
The question at issue here seems to me very similar to something
Plato asks about in the Argument from Interchange or allodoxia (AI) at
189b–190e. Here Plato asks the following question about beautiful and ugly.
Would you ever say to yourself that beautiful is ugly, or that odd is even or
that two is one? When Theaetetus answers ‘No’, Socrates proceeds to work
TT against this answer, in order to show that no false belief would then
be possible that depends upon such predicate pairs as beautiful and ugly.
A natural gloss here would surely be: ‘I may be wrong about what things
I call beautiful or ugly; but I can’t mistake what I mean by “beautiful” and
“ugly”.’ The incorrigible knowledge of meanings here (with the italicised
mean of the Moorean ‘what we mean’)24 I have called conceptual incor-
rigibility – by contrast with the perceptual incorrigibility of Protagorean
appearances currently under discussion. But just as, so far as Protagorean
appearances are concerned, their incorrigibility ensures that TT will show
false belief about such appearances to be impossible, so too, so far as incor-
rigibly known meanings are concerned, there is no false belief possible from
our confusing one of them with the other. This leads to the conclusion that
a false belief is not possible in the cases considered in AI either. For you
could never so confuse beautiful with ugly, in supposing that Theaetetus is
ugly (when really, what he is, in truth, is beautiful).25
In these things I mean that I have suggested are present in AI, we have
close cousins to Protagorean appearances. Indeed, in a previous discussion
(Penner 1987: 57–62 with 20–40) I argued that very early on (in the recol-
lection argument at Phaedo 74a10–c5) Plato was prepared to argue that he
could use a claim of conceptual incorrigibility to argue for the existence of
the forms.26 If the ‘Knowledge is Perception’ part of the Theaetetus rejects
24
See Penner 1987: 27–39, 80–6 and passim.
25
So much for the view that Plato does not consider false predicational beliefs in the Theaetetus.
26
In parts of the Gorgias and Meno, and in the Phaedo as a whole, I see a Plato who has become – for-
tunately briefly – more Pythagorean than Socratic (Penner 1987: 335 n. 26). At Phaedo 74a–c, esp.
b1–2, Plato has Socrates ask: ‘Has equality ever appeared to you to be inequality?’ – to which the
expected answer is: ‘Never’ (Penner 1987: 57–62; 69–86 with 26–40). This is in direct contradiction
to what we find in Plato’s rejection of AI: see Penner 1987: 314–17.
204 Terry Penner
perceptual incorrigibility (as I claim above) then, if Plato is reducing AI
to absurdity here, he would here be rejecting conceptual incorrigibility
as well, and so turning his back on incorrigibility-based arguments for
forms, leaving the Argument from the Sciences as his primary argument
for the forms (Penner 1987: 40–3).
Consider also the claim admitted by WT: that if Socrates knows both
Theaetetus and Theodorus very well, he would never say to himself
‘Theaetetus is Theodorus’. Why are modern interpreters so likely to grant
that WT is right about this? Surely only because they think that Socrates’
belief that Theaetetus is not Theodorus is about as close as could be to being
analytic in Socrates’ idiolect – true by virtue of the meanings of ‘Theaetetus’
and ‘Theodorus’ in Socrates’ idiolect.27 Put otherwise, it is as if ‘Theaetetus
is Theodorus’ is self-contradictory in Socrates’ idiolect. (Very close to what
Frege at any rate implies Gilgamesh would find self-contradictory: that the
Morning Star is not the Morning Star.) But this depends on our having
knowledge that is as close as it could well be to Socrates knowing incorrigi-
bly the meanings of ‘Theaetetus’ and of ‘Theodorus’ in Socrates’ idiolect.
And why is Frege so certain that Gilgamesh would not think that the
Morning Star is not the Morning Star? Because Gilgamesh is certain that
he is referring to the same thing with both occurrences of the expres-
sion ‘the Morning Star’? Or even that he means the same thing with both
occurrences of the expression? But where does this certainty come from?
From the incorrigible truth that supposedly belongs to our knowledge of
what we mean by a given expression in a particular context? But if so, a
Platonist may ask: does anyone know incorrigibly what they mean on a
particular occasion?
In defence of a negative answer here, I state briefly the Paradox of
Analysis:28
PARAN: In knowing that object which is the meaning of ‘Morning Star’, does
Gilgamesh automatically know also the meaning that the great Alonzo
Church refers to most often in his discussions of Frege?
27
As if in Socrates’ idiolect it should be a rule that with people well known to Socrates, there shall be
only one name to a person, so that a different person in this domain of people well known would
have a different name from any other person in the domain. Wittgenstein does something like this
at Tractatus 5.53.
28
If the Paradox of Identity is the Gilgamesh problem, as it applies to the reference(s) of the expres-
sions such as ‘the Morning Star’ and ‘the Evening Star’, the Paradox of Analysis is the Paradox of
Identity one level up – that is, as applied now to the meanings of expressions that are alleged to solve
the Paradox of Identity. The present paragraph asks: what if meanings themselves are subject to the
very difficulty (one level higher up) that their postulation was to remove. What next? An infinite
hierarchy of such solutions?
Clearly not. If Gilgamesh has no idea which meaning it is that Church
refers to most often, then he does not have incorrigible knowledge of the
object which is what he means on this occasion by ‘the Morning Star’! But
once incorrigible knowledge goes, the entire motivation for representing
the Gilgamesh problem as one that arises at the level of reference, but is
resolved at the level of meaning, has disappeared! For a mistake is, in prin-
ciple, as possible at the level of what object the meaning of ‘The Morning
Star’ is, as it is at the level of reference. The Paradox of Analysis should
then be regarded as just one way to show that we have no incorrigible
knowledge of what we mean, and that the satisfactoriness of the Fregean
response to the Gilgamesh Problem (the Paradox of Identity) is greatly
overestimated.
What I am trying to do here is to establish a prima facie plausibility
at least for the thesis that, although moderns and modern interpreters of
Plato would be quick to reject AKNK and TT, nevertheless those very
moderns may turn out to be committed to relying on entities which are
themselves very much open to just these arguments. And the question I am
raising is this: whether Plato, seeing the kinship of such entities as mean-
ings to Protagorean appearances, would not have supposed that any such
attempts to solve FIB are not going anywhere, if Protagoreanism simply
does not add up as a philosophical thesis.
Rejection of the incorrigibility of Protagorean appearances

As for the incorrigibility of Protagorean appearances, I now offer a sketch
of Plato’s belief that he can refute this doctrine. The refutation is the first
of two main refutations we find in this part of the Theaetetus. At 179c1–d1,
Socrates points out that we can be wrong about how things will appear to
us – about our future inner appearances (how hot I will feel, how sweet or
dry the wine will taste, how in tune the music will sound and how great or
pleasant the food will taste: 178cd). For doctors, vintners, musicians and
cooks will frequently be better judges than we are about even our future
inner experiences. Hence, we do not, as Protagoreans might have hoped,
‘have the kritērion in ourselves’ (178b6–7, c1) for how things will appear. But
as for our present inner experiences, Socrates says, it will be harder to assail
the Protagorean position that here we do ‘have the criterion in our selves’ –
as to how things appear to us and as to our judgments (doxai) about them.
Indeed, he says (178c10–e2 and esp. 179c2–d1), perhaps it is impossible
to assail this position and [as a result] Theaetetus is not beside the mark
in thinking that knowledge is perception (that is, perhaps ‘Knowledge is
206 Terry Penner
Perception’ on the Protagorean reading is true, so that Protagoreanism is
also true). This we know Plato will not grant. Accordingly, at this point,
Socrates says he is going to argue that Theaetetus was beside the mark. So,
contra Protagoreanism, we can be wrong about our present inner appear-
ances. At 179d2–4, he says he will show this by examining Heraclitean flux
(n. 20 above, and pp. 207–8 below), to see if that will guarantee the falsity
of Protagoreanism. If we now look at the conclusion of this argument, we
see Socrates saying that, as a result of the refutation of Heracliteanism, he
has shown that they have refuted Protagoreanism (183b7–c4). But how
on earth can Plato think that refutation of the flux argument will show
the falsity of Protagoreanism? I return to that question shortly. For the
moment it seems to me certain that Plato thinks he has shown Protagoras
wrong even about our appearances (our perceptual impressions).
Before I look at this argument concerning flux, I want to look at
the argument at 184b–186d, which I take to be a supplementary argu-
ment against the Protagorean reading of ‘Knowledge is Perception’.29
In my view, this refutation suggests there cannot be knowledge of the
Protagorean sort without reflection on our (Protagorean) appearances.
Each appearance comes only through one sense modality, whereas, when
we reflect on these appearances, we find that different such appearances –
coming from different sense modalities – may be the same as or dif-
ferent from each other, may each be, each be one, and so forth. To be
appeared-to-blue and to be appeared-to-sweet are two for us – different
appearances – as we shall see all too clearly in TT. But the universals
or relations here (same, different, being, one) cannot come from just one
sense, but must rather apply to all of them. Where could such universals
come from if the only things there are are Protagorean appearances that
we get through only one sense modality? So a knowing Protagoreanism –
one that knows one sensation from another – cannot get the required
universals by any method known to them. Protagoreanism could only
29
Thus I do not take the present argument to be intended as a stand-alone argument against some
general proposition ‘knowledge is perception’. (Here I am grateful to an all-too-brief discussion
with David Sedley at the Durham conference.) That is, I take it that the argument is directed at
an understanding of perception that allows no objects of perception other than such perceptual
appearances as white, black, hard, soft, hot, light, sweet and the like (secondary qualities). (See above,
p. 200, on the suggestion that the aisthēseis in WT are also Protagorean aisthēseis.) No physical
objects, and not even any appearances that are caused by physical objects (which would allow appear-
ances to presuppose the existence of physical objects). Thus in this argument, to perceive is to have
certain perceptual appearances. This is strong evidence, I suggest, that Plato intends 184b–186d to
be taken as a further reductio ad absurdum of the Protagorean understanding of perception.
succeed if there were something like Platonic forms – which is to say that
Protagoreanism is incoherent.30
I return now to the main question here: how does the flux argument
refute Protagoreanism – at least in the essential part that says that we are
always right and can never be wrong about our present inner appearances,
that is, that all our judgements here are incorrigibly true? A brutally brief
answer goes like this. If a subject is to be right from moment to moment
about both:
a. his being appeared to white at t, and
b. his appearance being unchanging in colour between t and (t + δt),
then surely:
c. he will have to be right also about being appeared to white at (t+δt).
But this requires that:
d. there is such a thing as a colour white, unchanging over time
– which is to say, again, that Protagoreanism could only work if there are
such unchanging entities as the colour white – that is, a form. But the exist-
ence of such unchanging objects cannot be granted by Protagoreans, even
as helped out by Heracliteanism.31 In other words, Protagoreanism is inco-
herent. Even Protagoras is committed to the existence of such unchanging
entities; and they make it the case that even our present appearances give
hostages to the future (n. 20 above).
How so? It is a perfectly ordinary occurrence, as colours blend imper-
ceptibly into each other, that it should turn out that actually, after an
interval δt, the subject is being appeared to very light grey at (t+δt). Now,
the Heraclitean could allow that the colour white is constantly changing,
and could easily turn into the colour very light grey during the interval
δt. (‘Look, all that needs to happen is that the colour white [what white-
ness is] is constantly changing, and here changes into what greyness is.)
The problem is that, at this cost, Protagoreans do not have incorrigible
knowledge of how they are being appeared to at t, namely, as character-
ised by the unchanging colour white – where what whiteness is (the colour
white), like what greyness is, does not change over time, but rather ‘abides
30
This conclusion is strikingly similar to the conclusion of the crushing anti-Heraclitean argument
that I discuss in the next two paragraphs.
31
See n. 20 above.
208 Terry Penner
in its own nature’ (182e1–2).32 But if there is no such unchanging white-
ness, yet it is this very unchanging whiteness that Protagoras judges to be
true of his appearance at t, then his position is incoherent. For on his
view, there is no such thing as the unchanging colour white or whiteness!
For Protagoras to be right about his appearances, there need to be such
things as what the colour white is, and what the colour light grey is – things
that are not constantly changing into other colours over time. In fact,
Protagoreanism is committed to the existence of colours not changing
their identity over time – a kind of Platonism! So Protagoreanism’s judge-
ment about a present inner appearance cannot be: ‘There is whiteness
here now’; it can only be – if Heracliteanism applies to universals like
whiteness as well: ‘There’s whiteness-greyness-who-knows-what-the-heck-it-
is here now’ (183a6–b5). This, in spite of the fact that it is the very func-
tion of Heraclitean flux to wipe out incoherences between judgements of
our appearances over time.33
Finally, if this attack on Protagorean incorrigible knowledge of one’s
own inner appearances is successful, and there are no such incorrigibly
known entities to refer to, then this may suggest that Gilgamesh’s belief
shouldn’t be about incorrigibly known meanings; rather, in accordance
with ABT, Gilgamesh’s belief will have to be about the object that is the
Morning Star (that is, the Evening Star) not being identical with the object
that is the Evening Star. But clearly Gilgamesh does hold just that false
belief (that the first object – the Evening Star, that is, the Morning Star –
is not the second). Gilgamesh really does intend that his belief should be
that, if he is in unoverridable error, he does believe that the Morning Star
is not the Morning Star. And Socrates really does intend that, if he is in
unoverridable error, he really does believe that Theodorus is Theaetetus. In
other words, Plato would counsel us to reject the very premises that gener-
ate the Theaetetus problem and the Gilgamesh problem, namely TH4 and
the falsity of FR2.
32
What I have just said about the colour white is confirmed by what Socrates says directly (182e11–12)
about what perception is. Without there being such a thing as what perception is the Protagorean
thesis that ‘Knowledge is Perception’ would fall apart. For that thesis had better not be about
momentary perceptions and momentary bits of knowledge. It is about the (unchanging) real
nature of perception and the (unchanging) real nature of knowledge. Cf. also the contrasts of white
vs. whiteness, and percipient vs. perception at 156e4, 6, setting up the contrasts at 182a7–b7, d1–5,
d8–e12 with 183a2–b5 (including such vs. such-ness, that is, quality).
33
This is the explanation of Protagoreanism being united with Heracliteanism at 152c8–160e4: the
worry that a judgement of an appearance at t might conflict with a judgment at [t + δt] – as when
one’s appearances appear to one unchanging in colour during that interval.
The solutions to FS and NB that Plato needs

for his solution to FIB
But if defenders of Plato are to argue that Socrates can believe that
Theaetetus is Theodorus, they still have to tell us what Plato’s positive
account would be of that false belief – and without referring to anything
but the actual persons or things. This is a non-trivial task. The problem is
that, for Socrates, the false belief under discussion will have to be about the
person who is the one coming, that is, Theaetetus. But, as we have remarked
earlier (pp. 197–8 above), there is no such person. So we might say that
FIB reduces, for Plato, to a problem in the vicinity of the Problem of
Non-Being NB, in that what the belief is about is, so to speak, something
that does not exist. Let us turn to NB.
NB can be seen developing from a theory of existence which says, as do
Parmenides and Plato, that:
NEXNOTH: To not exist is to be nothing at all.
Suppose this account of not-existing is correct. Then when it is true that
something, say, Santa Claus (or Pegasus, or the Golden Mountain), does
not exist, we will have to ask, ‘what is this something that doesn’t exist?’ We
see immediately, from NEXNOTH, that the something in question in any
such case is nothing at all. Something that does not exist? What something?
There is no ‘something’ there to be what does not exist. By NEXNOTH,
any such truth as that Santa Claus doesn’t exist is not a truth about any-
thing at all. This is NB. (A further troublesome consequence: on this view,
there can be no distinction between Santa Claus not existing and Pegasus
not existing.)
Interestingly enough, if we turn to ordinary existing objects that we
talk about, we find that, in the extensional logic employed by most mod-
ern analytical philosophers, the position on what is named, or even on
what is correctly referred to via a definite description, comes to much the
same as on Plato’s view. On such matters, Plato would have thought well
of extensional logic – so far. (There are no meanings, or conventions for
the use of words, for example, in extensional logic.) For the thing named
or correctly referred to is whatever it actually is in the real world, with all
its properties known and unknown, and quite independently of how we
refer to it or think of it. The Cratylus, we have seen in the remarks on the
real nature of cutting, burning and (!) naming (pp. 194–5 above), is quite
ready to override many errors in definite descriptions, errors that make
our descriptions and ways of referring to the object strictly false of the
210 Terry Penner
objects they are intended to single out.34 It is enough if the interlocutors
involved get the right idea about who is being talked about.35 Not all such
errors can be over-ridden, however. Not errors that will lead interlocutors
to identify the wrong thing, or to speak of what does not exist at all. Our
present problem concerns just such an example: the person who is the one
coming, that is, Theaetetus (when in fact it is Theodorus). Still, where the
errors can still be allowed for – perhaps with the real nature of cancer –
and can often, in spite of some errors, still yield the actual, existing object
that is intended, we may say that the definite description is true enough.
(It is true enough when the hearer can pick out what is referred to by the
speaker, even though they both have various things wrong about it.)36
Turning now to how NB gets handled by modern classical logicians,
consider what happens if what was taken to be a name was not a name,
and there is no object named. For the analytical philosopher, this prob-
lem arises right within extensional contexts. But here it is not dealt with
quite candidly. Rather, it is dealt with by fiat: by ruling that the empty
name isn’t a name at all. The empty name, as it were, ‘fouls out’ from
the game of extensional logic – before it even gets on the court. Instead,
some definite description is characteristically sought out as a substitute –
a definite description which, somehow, still captures the error involved in
the use of the empty name. If ‘a’ is the empty name, then ‘a is F’ ‘fouls
out’ of the game of extensional logic and we are given instead the definite
description ‘the thing that is G’,37 getting us the sentence ‘the G thing
is F’ which, in turn, can be transformed into the false existence sentence
‘there exists one and only one thing that is G, and it is F’. This sentence is
34
See my notion of Thrasymachus’ ‘default reference’ (at Penner 2005a: 61–2 n. 28) for what is in fact
the claim that what Thrasymachus intends to refer to when he uses the word dikaiosunē is not: ‘the
weak obeying the strong’, but: ‘the real nature of justice, that is, the weak obeying the strong, that is,
the strong exploiting the weak’. For anyone who thinks (as most commentators do) that there is no
such thing as ‘the real nature of justice, that is, the weak obeying the strong, that is, the strong exploiting
the weak’, the question becomes what Thrasymachus’ default reference is. Is it: ‘the weak obeying the
strong, whatever that may have to do with what justice actually is’? Or is it: ‘the real nature of justice,
even if it is not exactly the weak obeying the strong’? Does Thrasymachus really not want to be taking
part in a discussion of what justice actually is? See also the remarks at nn. 13 and 15 above, and n. 36
below.
35
The same is true of some ‘direct reference’ theorists such as Donnellan.
36
Compare Augustine’s profoundly Platonist remarks on teaching at On the Master 14.45, where he
says the teacher doesn’t teach by conveying to students what is in the teacher’s mind. Rather, as I
read him, he says some words that are sufficient to draw students’ attention to some object or truth
in reality (interiorem scilicet illam veritatem), and students then learn, to the extent they are able
(pro viribus), from attending to that real object or truth. Beyond this, I would add, it matters not
that the teacher has any number of false beliefs about the object or truth to which he or she suc-
cessfully draws the students’ attention.
37
Instead of ‘Santa Claus’ we get ‘the fat guy who gives presents to all rich kids at Christmas’.
false because of the first conjunct. But it is still in the game of extensional
logic. On the other hand, ‘a is F’ is entirely excluded – and without any
explicit facing of NB within extensional logic. (Similar moves apply
to definite descriptions which are intended to be referring expressions,
and do not refer to the actual object intended: a substitute false exist-
ence sentence is supplied.) Thus the modern treatment of, so to speak,
non-existent objects is to exclude empty names by fiat, and substitute
false existence sentences. So far as extensional logic itself is concerned,
non-existence is shoved under the table. (Had it been kept explicit, the
thought might have crossed people’s minds that empty names had to be
dealt with as if they were in false psychological contexts, at which point it
might also have occurred to them that a distinction of principle between
extensional contexts and psychological contexts was distinctly precarious.
For when it is a matter of putting ordinary language into logical form,
what proper name might not turn out, contrary to expectations, to be
empty?)
I add only that the truth or falsity of the resulting existence statements
is determined by what the meanings of the predicates ‘G’ and ‘H’ deter-
mine as falling within their extension. At this point, Plato’s sympathies
with modern logic will also be minimal.
We know from the Sophist that Plato’s own solution to NB is given by
addressing FS. Plato introduces NB very emphatically at Tht. 188c–189b,
Sph. 236d–239c and passim, yet he never offers a solution to it other than
the solution to FS. Presumably he thinks the solution to FS applies, muta-
tis mutandis to NB, in a way I shall try to illustrate. (So the problem of
non-being turns out to be a problem about falsity.)
I now sketch, first, a possible treatment of how Parmenides might have
dealt with NB, and then show how that kind of treatment would work
with Plato’s treatment of FS. Parmenides held that change does not exist.
But how could that be so, if, by NEXNOTH, change is nothing at all?
One move Parmenides might have made here – though it conflicts with
his main thesis that Being is one – is this: ‘change the subject’ from change
(which he thinks does not exist) to the word ‘change’ (which does exist)
and, at the same time, correspondingly ‘change the predicate’ from ‘does
not exist’ to ‘is a mere word’. The idea is that ‘“change” is a mere word’ says
the same thing as ‘change does not exist’. Then, provided that Parmenides
believes that words exist, his problem is solved. Unfortunately for him,
Plato points out that, at this rate, Parmenides would have to believe that
words exist in addition to Being, so that Being could no longer be one
(Sph. 244b–d).
212 Terry Penner
Nevertheless, Plato’s solution to FS is along the same lines as the argu-
ment just given – except that Plato only ‘changes the subject’ to things he
thinks do exist – and exist antecedently to this (and all other) argumenta-
tion. Consider first the true statement that Theaetetus is sitting. Plato will
have thought that this statement just as much takes us to a real state of
affairs in the world as, in extensional logic, the name ‘Theaetetus’ takes us
to a real object in the world. (We might say that what someone is saying
when they say truly that Theaetetus is sitting is that there exists a state of
affairs in which Theaetetus is sitting. Is this not how people think of what
they are saying when they think they are speaking truly?) At any rate, the
point of such statements is surely to take us to real states of affairs in the
world – as the point of names is to take us to real objects in the world. So
I say: Plato seems to suppose that the true statement made using the sen-
tence ‘Theaetetus is sitting’ names a real state of affairs in the world quite
as much as a person who refers to someone using the name ‘Theaetetus’
names a real object in the world. So far, incidentally, there is nothing in
the point of view of extensional logic that would exclude this suggestion.
In effect here, true statements turn out always to be statements about
the existence of states of affairs in the world. We now see, therefore,
that false statements will be statements of non-being. This helps us see
why Plato, in the Sophist gives us his solution to FS as his only answer
to NB. What Plato does is to provide a ‘change of subject’ a little more
elaborate than that from change to the word ‘change’, but still analo-
gous to it.
True, there are all sorts of reasons why an analytical philosopher will
think that true statements do not name states of affairs. Consider only this
difficulty: Plato seems to be headed towards an inability to distinguish
between saying something false and saying nothing at all.38 Granted. Plato
must respond to this charge. And Plato must be able to make a distinc-
tion between Santa Claus not existing, Pegasus not existing and a flying
Theaetetus not existing. Most of the reasons why moderns object to Plato’s
way of thinking here have to do with the fact that moderns take themselves
to have alternatives – alternatives such as reference to the different propos-
itions (that Santa Claus exists, that Pegasus exists and that Theaetetus flies).39
My position, however, is that propositions are quite as foreign to Plato
38
Here as elsewhere, I am much indebted to Ruth Saunders.
39
Some have argued that Sph. 261d–262e shows that sentences differ in logical type from names
because they involve weaving names and verbs together to ‘get somewhere’ (perainein). But this
passage hardly shows any difference between sentences and complex names.
as those meanings that supposedly determine reference in much modern
analytical philosophy. If what Plato says are problems are not problems in
analytical philosophy, may we not wonder whether this is only because of
the postulation of these new entities – entities which I shall claim are new
with Aristotle – namely, meanings and propositions? And if so, whether
these new entities might not be regarded as entities that (in fact) fulfil the
function of solving (or evading) Plato’s two problems, FS and NB, which
Plato needs (pp. 209–11) to back his solution to FIB (pp. 186–209)? This
will enable comparison and contrast of the modern strategy with Plato’s.
I shall accordingly try to show how Plato himself, in the Sophist, handles
false statements (and, by analogy, would handle empty names).
At first sight, it might seem that the correct analysis of the claim
that:
PL1: the statement that Theaetetus is flying is false,
should break the claim up into the two parts:
PL1s: the statement that Theaetetus is flying (subject)
and
PL1p: being false (predicate).
But, as Plato sees all too clearly, this would land him, on his view, with the
claim that:
The situation in which Theaetetus is flying does not exist
– a claim that, by NEXNOTH, we cannot make, at any rate without
further analysis. To speak of what someone is saying who says falsely that
Theaetetus is flying would, on Plato’s view, be to speak of someone saying
(or talking about) nothing at all.
So Plato – apparently without realising or believing that he is, in this,
following what I have said above might be Parmenides’ general strategy –
decides, in effect, that what we need to do here is to break up the claim
differently. We need to ‘change the subject’ – from this supposed thing
(proposition? state of affairs in some other possible world?) that does
not exist, to the following entities, all of which do exist: Theaetetus; the
form (or real nature) of Flying; all the other forms (such as Sitting) that
Theaetetus does partake in; and a certain ‘Greatest Kind’: the form (or real
nature) of Other. That all of these beings exist is expressed by Plato in
214 Terry Penner
terms of their all partaking in another of the Greatest Kinds: the form (or
real nature) of Being.40
With this change of subject, the two-clause claim that:
PL1: the statement that Theaetetus is flying is false
may now be expressed, referring only to things that exist, as follows:
PL2: Theaetetus partakes in Other with respect to Flying
which I take to say that:
PL3: All the forms that Theaetetus partakes in are other than the form of Flying.
Thus Plato has shown us how to ‘change the subject’ from a supposed
thing that does not exist (the state of affairs in which Theaetetus is flying)
to things all of which do exist – things that exist antecedently to the pos-
tulation of any such new entities as meanings or propositions. (People who
do not believe in forms, but believe in properties or universals, or exten-
sions of predicates, can make the same moves, mutatis mutandis.)
So Plato solves the Problem of False Statement in accordance with what
might be a Parmenidean principle, even though Plato also thinks of himself
in the Sophist as going against (committing an act of parricide against) Father
Parmenides in forcing non-being to be (in a way: Sph. 241d–242a, 258c–e).
This last concession raises the question: what is this forcing non-being to be
(in a way)? It certainly isn’t suggesting, in the fashion of Brentano (1874)
and Meinong, that there is a special sense of ‘be’ in which things like Santa
Claus or the Golden Mountain that do not exist have being. This should
be clear from Plato’s opposition to any reference to the ‘opposite of being’,
which we have ‘long ago said good-bye to’ (258e–259a, referring back to
238c). I suggest that what Plato has in mind when he speaks of his act of
parricide in forcing non-being to be (in a way) is precisely his use of the
method of ‘changing the subject’. For by substituting reference to Otherness
for any reference to things that do not exist, he shows us a way for there to
exist sensible versions of, so to speak, claims of non-existence, versions that
only refer to things that antecedently exist. There is no abandoning of the
principle NEXNOTH, and there is also a difference of analysis between its
being false that Theaetetus flies and its being false that the earth is flat.41
40
In Plato, existence is a predicate – though of course only of anything that exists. At the same time,
non-existence is not a predicate of anything (especially not of ‘things that do not exist’). Such is the
moral of the Sophist. See further Penner forthcoming; and O’Brien in this volume.
41
This says, in effect, that the semantic value of true sentences isn’t given by the truth value – as it is
inevitably in the semantics of most analytical philosophers – but rather (in the simplest cases) by
Incidentally, while we are talking about the importance of the form of
Other in the Sophist, I note that Plato evidently thinks that he is dealing
here not only with the Problems of Non-Being and False Statement, but
also with a Problem of Negation (Sph. 251a–259d). Mathematical intui-
tionists have long maintained that there are problems in classical logic
caused by the Law of the Excluded Middle – a key use of negation in
extensional logic. Considerations of space preclude my saying any more
on this exceedingly important topic here. But it should be noted that any
reference to negation is eschewed in the above solution to the Problem of
False Statement. Why? Presumably because Plato sees negation as threat-
ening to force on us reference to states of affairs that do not exist, when
their existence is falsely affirmed. Instead, as already noted, Plato substi-
tutes for the idea of referring to any supposed things that do not exist the
idea of reference to the relation of Otherness – as we see it in Theaetetus
and Theodorus being other, that is, being two. (Two different things.) This
point seems also to be well grasped by the intuitionists.
Choosing between (a) meanings, propositions and the

like and (b) the entities Plato refers to
At first sight, the moderns seem able to handle false statements with utter
ease: it might look as if they think sentences are false just when the propo-
sitions they express are false. I do not think this is how Plato would have
seen it, and I do not suppose that moderns are so naive as to think it is so
simple. We should rather see moderns as proposing that:
PROP1: the sentence that Theaetetus is flying (what someone is uttering when
they say that Theaetetus is flying) is false
just when:
PROP2: the proposition that Theaetetus is flying has the truth value falsehood.
Socrates would have been bound to ask at this point, ‘What in the world
is a proposition anyway let alone the truth value the false?’42
So far, I have treated the proposition in one way in which Frege treats
it – as the meaning or Fregean sense of a sentence which we must go
the real reference of the subject term and the real reference of the predicate term. This avoids the
difficulty that we must suppose all true sentences refer to the same thing.
42
It hardly needs saying that neither names nor sentences are more than noises unless given inter-
pretations, e.g. (among Aristotelians and moderns) interpretations in terms of meanings and
propositions.
216 Terry Penner
to when a sentence is embedded in a psychological context. But there
is another way in which Frege treats it which descends directly from
Aristotle, where Aristotle believes he has discovered the proposition as a
by-product of his creation of the notion of logical form. The proposition
would then engender the ideas of the truth conditions and the logical pow-
ers of a sentence (given by the sentences that follow logically from it and
the sentences from which it logically follows). In these entirely original
supposed discoveries, Aristotle gives birth to perhaps the most consequen-
tial revolution in the history of Western philosophy. We may call it the
logical turn.
In his creation of the supposed science of logic, Aristotle in effect aban-
dons Plato’s ideas about the close connection between truth and existence
(or reality) and the consequent close connection between falsehood and
non-existence (or non-reality). He does so because for purposes of logic –
that is, for purposes of discovering what is true, or what follows, by virtue
of logical form – we need to deal with statements without knowing whether
they are true or false (see above, p. 197). As a result, what Aristotle in effect
does is to declare that:
PROP3: what the proposition expressed by a sentence is shall be the very same
thing whether the state of affairs in the world makes the proposition (and
so the sentence) true, or makes that very same proposition (and sentence)
false.
This is precisely what happens in Aristotle’s syllogistic. The syllogistic
(Prior Analytics 1.1–2, 4–7, esp. 1.4) is designed to tell us what follows from
what, given the form of the sentences involved as premises or conclu-
sions – a single form that (it is supposed) holds both for cases where the
premises and conclusions are true (in proofs) and in cases where some
or all of them may turn out to be false (in ‘dialectical syllogisms’).43 The
idea here is this: if some sentence which we thought was true turns out to
be false, we do not have to go back and do the argument all over again.
(Contrast what happens when what we take to be a name turns out to be
an empty name! Then the whole argument has to be rephrased and gone
through again. Not so with true and false statements in Aristotle’s syllogis-
tic.) This step is not at all unmotivated, given that logic’s utility is thought
to reside in its use in situations where we are ignorant of what is true and
43
Here Aristotle’s position, like that of moderns who take the ‘linguistic turn’, is that what a person
says by means of a given sentence reduces to what that sentence says. (The letters A, B and C in the
syllogism Barbara are dummies for sentence predicates, where the letters are taken to yield the same
thing said whether they are true or false.)
what is false, and want to see under what conditions something would
be true or false, given what other beliefs we may have. But that by itself
hardly renders the notion of the proposition defensible in the end.
So, in Aristotle’s syllogistic, the premises and conclusions are taken to
express the same thing whether they are true or false. This is to say that
they express propositions. If we now put the point in PROP3 using Frege’s
astonishing and persuasive applying of the notion of function to sentences
and their semantics (see Frege 1891, 1892 and 1918), we get that:
PROP4: the actual state of affairs in the world takes the proposition to one of two
abstract objects: the truth values (or semantic values) truth and falsehood.
That is, for a given state of the world, the proposition determines the truth
value.
This tells us that, for Aristotle:
PROP5: what someone is saying by means of a sentence (that is, a statement) always
has two independent parts: (1) the proposition expressed by the sentence
used by the speaker, and (2) its truth value.
Now, PROP5 represents a huge step in the history of metaphysics. No
less a step, I would argue, than the (now rather discredited) introduction
of sense data into the theory of science by Mach and the earlier logical
positivists.44 And no less a step than the postulation of that earlier form
of sense data, namely, Protagorean appearances. (Notice also that once we
have the two independent parts of the statement noted in PROP5, we
have also the notion of the truth conditions of the sentence. Another idea
quite foreign to Plato: that there is a single set of truth conditions whether
the sentence is true or false.)45
So what happens in logical analysis is that – whether what a person
says by means of a given sentence is true or false – we ‘change the subject’
from what the person is saying to the proposition expressed by the sentence.
Plato would have seen this as opting for simplicity over saying what really
is so. For Plato, what really is so is that the true statement straightfor-
wardly refers to or describes the existence of the state of affairs in which,
44
For strict empiricists, an empirical judgement has two parts: the sense datum that you cannot be
wrong about – which is incorrigibly true – and a part where you (‘inductively’) project other sensa-
tions, which projections may be correct or incorrect.
45
Plato, I suggest, would have pointed out that if sentences are your truth bearers, a sentence with
the name Major William Martin, RM, will have one set of truth conditions (referring to that real
person who is Major Martin) if it is true, and another set (making no reference to anyone whatever
who is a major in the Royal Marines) if it is false, because William turns out to be a ‘man who
never was’.
218 Terry Penner
say, Theaetetus is sitting. The false statement takes us to no state of affairs
at all. Instead, all we get is what it is for it to be false that, say, Theaetetus
is flying: Theaetetus partaking in Other with respect to Flying.
Still, it is worth asking: why would Plato reject the alternative account
of moderns?
One way to depict how Plato thinks here is to argue that Plato thinks
that all of these entities – propositions, meanings, Protagorean appear-
ances, logical forms, truth conditions, as well as intentional objects in
Brentano, and so forth – are what I shall call ‘intentional’ entities, and
entities of which it is the case that, if we can do without them we should
do without them. Why? For one thing, if we add to our universe a lot of
unnecessary46 (and new, made-up) entities that in fact have no claim to
existing at all, taking them to exist can result in our finding arguments
that will rule out other entities that really do exist, and are not made up.47
I call something an ‘intentional’ entity if it has all of the following
attributes:
(1) is it an intermediary – intermediary between our thought and the
world (with the objects in it);
(2) it is a neutral entity in the following way: it is the same entity, what-
ever object or state of affairs (if any) it picks out from the world;
(3) it is epistemologically prior to the world, to its truths and to the
objects in it (given that we can know what proposition is involved
in a statement or belief without knowing whether it has the prop-
erty of truth or the property of falsehood; and that we can know
what intentional object or meaning it is that is presupposed in the
use of a referring expression whether or not there is any object in
the real world corresponding to it);48
(4) it is guaranteed to exist in the following way: as long as someone
says something or believes something, it is supposed, there will be
a proposition; and as long as someone tries to refer to something,
there will be an intentional object or a meaning;
(5) it will have very narrow criteria of identity, thus ensuring the
difference between, say, the non-existence of Pegasus and the
46
Unnecessary if Plato can handle NB, FS, and FIB without them, and if we do not believe in logical
form (as I do not) or think we can manage perfectly well without it. I of course expect many to
disagree here.
47
For example, sets or properties may crowd out Platonic forms. See Penner 1987: 251–99 (Clarification
VI).
48
Being incorrigibly true is a strong form of epistemological priority.
non-existence of Cerberus.49 (The point is even clearer for propos-
itions. As for meanings, see above, pp. 204–5, with n. 28 on the
Paradox of Analysis.)
Such is the formidable conjunction of properties I find in all so-called
‘intentional’ objects.
Now, moderns use these ‘intentional’ objects as entities to ‘change the
subject to’ in fulfilling various desiderata, especially in connection with
logic and the logic of psychological contexts. But I do not think that Plato
believed in any such thing as a neutral science of logic, or as the new
‘intentional’ entities necessary to it: logical forms, propositions, meanings
and the like. I picture the older Plato standing by, in his Academy, in won-
derment, as the young Aristotle introduced the idea not only of a neutral
science of persuasion (à la Gorgias) – which Socrates and Plato denied
could be a science50 – but also of a neutral logic. Well, is this introduction
of such new ‘intentional’ entities so necessary to our other purposes that
we must suppose they exist really – even without our making them up?
These entities certainly will be needed if we decide, as in the logical turn
and the linguistic turn, that what we are interested in is what the sentences
of speakers say and entail. But not so clearly if, as in Socratic dialogue, we
are interested in what the speakers are saying by means of their sentences.
Conclusion
Where moderns insist on a difference of logical type between what is said
and what is named, Plato sees (sufficiently) true statements quite as much
taking us to existing states of affairs in the real world as adequate (suf-
ficiently true) names take us to real things in the real world. And where
moderns treat empty names by fiat, and false statements in terms of pos-
tulated ‘intentional’ entities called propositions, Plato treats both empty
names and false statements as instances calling on us to ‘change the sub-
ject’ – but only to other things that exist antecedently to any ‘intentional’
entities. Similarly, where moderns introduce a special logic for psychologi-
cal contexts and tell us that existence is not a predicate in the course of its
introducing a plethora of new ‘intentional’ entities such as propositions
and meanings to be what is referred to – instead of real objects and states
49
Every horse is other than winged creatures, and every dog has exactly one head. I am indebted to
Antonio Chu for the important suggestion that I include (5) – which becomes especially relevant to
my remarks above on the Paradox of Analysis.
50
See what I regard as one of my best papers: the hard-to-find Penner 1988.
220 Terry Penner
of affairs – in psychological contexts, and when doing extensional logic,
Socratic-Platonic Ultra-Realism will say that the problems lie not in psy-
chological contexts, but with that single fault we call, alternatively, ‘falsity’
or non-being51 – whether in extensional contexts or in psychological con-
texts. This Ultra-Realism would have viewed the idea of a neutral logi-
cal form as simply inviting us to turn away from speaking directly of the
objects and states of affairs that exist in the world and are real – and quite
unnecessarily ‘changing the subject’ from them to other made-up entities.
It is true that we are often wrong about what objects and states of affairs
exist and are real. But should this lead us to turn away from the constant
striving for what is real, towards a lot of new made-up entities rather than
simply coping as best we may with falsity and non-being as they arise
from case to case?
I do not deny that there may be other scientific purposes that pay divi-
dends sufficient to offset the costs of postulating such made-up entities as
propositions and meanings. It remains the case, I suggest, that to see what
interlocutors are saying in Socratic and mature Platonic dialogues, we
need to approach those dialogues – including the Theaetetus and Sophist –
with a little more awareness of how much that may be questionable we
interpreters bring to our work from the philosophy of the day. That is, if
we are to learn also from what Socrates and the mature Plato are saying
in those dialogues. We need not feel badly about our being frequently
insufficiently aware of such matters. For just that, together with the prob-
lem of determining what the truth is that Plato intends to pick out (and
intends should override his own ways of putting things), is the interpret-
er’s dilemma in every age.
51
Since I have been arguing that falsity is for Plato a matter of, so to speak, a state of affairs
non-existing, FS and NB come to the same thing.
ch apter t welve
A form that ‘is’ of what ‘is not’

Existential einai in Plato’s Sophist
Denis O’Brien
I
Motivated by an otherwise very understandable desire to study ancient
philosophical texts philosophically, recent commentators have taken to
weeding out from Plato’s dialogues any existential use of the verb einai,
seemingly in deference to the supposedly philosophical principle that
existence cannot be a predicate. The result is disastrous. This is not only
because Plato very clearly does use the verb as a predicate complete in
itself, with a meaning that can properly be described as ‘existential’, nota-
bly in his account of being and non-being in the Sophist, but also because
the principle itself is not what it is all too often thought to be.
In the pages of his Critique of Pure Reason where he takes issue with
what he calls the ontological argument put forward by Descartes in the
Fifth Meditation, Kant does not deny that we can properly say of God that
he ‘is’ or that he ‘isn’t’.1 Kant’s point is that when the verb is used in this
way, and therefore ‘existentially’, it is not being used as what he chooses
to call a ‘real predicate’. In asserting, of whatever may be in question, that
it ‘is’, or that it ‘isn’t’, we assert that the concept is, or is not, instantiated,
that it falls, or does not fall, within the range of our experience. But that is
all we say. The concept does not acquire a fresh property by being instanti-
ated, a property that it would not have if it were not instantiated.
There is of course a good deal more that could be said – a very great
deal more that has been said – about propositions that do, or do not,
entail an assertion of the existence, or the non-existence, of whatever
we are talking about. My (no doubt impossibly abbreviated) thumbnail
My most grateful thanks go to friends who have generously undertaken a critical reading of ear
lier versions of these pages and who have favoured me with their comments: Luc Brisson, Jonathan
Martin, Charles Ramond and Suzanne Stern-Gillet. I regret that the most recent book on the Sophist,
by Paolo Crivelli (2012), reached me too late to be taken account of here.
1
Kant 1787: 620–30.
221
222 Denis O’Brien
sketch of Kant’s position is intended as no more than the briefest pos-
sible protest against the common, but ill-founded, belief that a thesis such
as Kant’s must somehow inhibit any direct assertion of existence, or of
non-existence, and therefore any directly ‘existential’ use of ‘is’ or ‘is not’.
That is not its purpose. Kant is not denying the right of the Psalmist’s Fool
to say in his heart Non est Deus, nor is he accusing the modern Anselm of
a grammatical mistake when he seeks to persuade the Fool of his error. In
either case, whether we say ‘is not’ with the Fool or ‘is’ with Anselm, the
verb is being used as a predicate. A ‘logical’ as distinct from a ‘real’ predi-
cate, if we adopt Kant’s terminology. A ‘predicate’ nonetheless.
This is still true, in modern English, of the composite verb ‘am-was-be’
(originally three different verbs, now considered as parts of one and the same
verb). Despite increasing competition from ‘exist’, a Latinate newcomer to
the language, the two uses of the older verb may still be found, alive and
well, in the opening verse of John Clare’s famous poem, written when he was
confined in Northampton County Asylum: ‘I am: yet who I am, none cares
or knows.’ ‘I am’, an existential use of the verb. ‘Who I am’, a copulative use
of the verb. Clare may have been suffering from bouts of madness at the time
when he wrote the poem, but there is nothing wrong with his use of English.
In the appropriate context, the different parts of ‘am-was-be’ can still be used
as a complete predicate, comparable in meaning to ‘exist’, and are therefore
still viable candidates for translating an existential use of einai in Plato.
To recognise Plato’s use of the verb as a predicate complete in itself, and
therefore as what is commonly called a ‘substantive’ or ‘existential’ use of
the verb, we need only read through the Stranger’s long and painstaking
analysis of the five megista genē that will prove to be being, sameness and
otherness, movement and rest (Sph. 251a5–257a12).2 Three of the five ‘very
great kinds’ are universally participated, insofar as everything that ‘is’ is
both the ‘same’ as itself and different from, or ‘other than’, everything that
is not itself. The two remaining ‘kinds’, movement and rest, are specifi-
cally said to be ‘most contrary’ (250a8); participation in either one of the
pair therefore excludes participation in its opposite.3
In the preliminaries to this analysis, the Stranger first makes the very
simple point that, if there were no participation at all, then neither move-
ment nor rest would participate in being. In that impossible world, so
he asks (252a2–3), will either of them, if it has no share in being, ‘be’?
2
The lineation of references to the text of Plato, throughout this essay, is taken from Burnet’s second
edition (1905: Burnet 1905–1907), not from the more recent edition by Duke et al. (1995).
3
Vlastos’ appeal to ‘Pauline Predication’ does not, as Vlastos claims it does (1981: 270–322 and
404–9), allow movement (the form) to participate in rest. See O’Brien 1995: 103–10.
Plato’s form that ‘is’ of what ‘is not’ 223
A question greeted at once by the pithy reply (252a4): ‘It won’t be.’ The
repeated use of the verb is clearly ‘existential’. Neither movement nor rest
will ‘be’ if neither has any share in ‘being’.
The Stranger returns to that same point, though now cast in positive
form, a moment or two later (254d7–11). Neither movement nor rest can
‘mingle’ with the other, but being ‘mingles with’ both of them: ‘That is
presumably why both of them are.’ The politely dubitative ‘presumably’
(254d10: που) is brushed aside, as it is meant to be. Theaetetus agrees, as
to something that can be taken for granted, that both rest and movement
‘are’. Being ‘mingles with’ both of them; it is assumed that both of them
therefore ‘participate in’ being; that is why both of them ‘are’. The use of
the verb, here again, is clearly ‘existential’.
On the following page, the same lesson is spelled out, no less clearly, a
third time, when by way of summarising the argument that has preceded,
the Stranger runs through the relationship of movement to all the four
remaining genē, including therefore ‘being’ (255e11–256a2). Movement ‘is
altogether other than rest’, and ‘is therefore not rest’. ‘Even so’ (256a1: δέ
γε), despite not being rest, ‘it is, because of its participation in being’.
To which Theaetetus replies, as pithily as he had done before, by simply
repeating the verb: ‘It is.’ The difference between the two uses of the verb,
to the eye and ear of a grammarian, is unmistakable. A copulative use
of the verb, movement ‘is not rest’ (255e14), is followed, clearly and one
might have thought uncontrovertibly, by an existential use of the verb:
because movement participates in being, it ‘is’, it ‘exists’ (256a1 and 2).
Am I therefore pushing at an open door? If I go out of my way to empha-
sise the ‘existential’ use of the verb in this third sentence (256a1) by rehearsing
the Stranger’s use of the same verb in the same way in the preceding pages of
the dialogue (252a2 and 4, 254d10), it is because even so plain and obvious
an instance of ‘existential’ einai has been denied by two of the commenta-
tors commonly thought to have been most successful in throwing fresh light
on Plato’s dialogue by treating the argument ‘philosophically’. Gwil Owen
and Michael Frede both insist that the verb, in this third passage (256a1), is
being used as a copula and has to be heard therefore with a silent comple-
ment supplied from the words that precede or that follow.4 They are plainly
wrong, and I am led to suppose, perhaps too simple-mindedly, that so obvi-
ous an error betrays an ideological preference. A supposedly Kantian version
of the argument favours reading as a copula the verb that, on any straight-
forward construal of the text, is being used as a predicate complete in itself.
4
Owen 1971: 254–5. Frede 1967: 55–9.
224 Denis O’Brien
But do I perhaps hear a frisson of protest? How is it possible to be so
dogmatic? How can we be sure that the verb as used by the Stranger (256a1)
and happily repeated by Theaetetus (256a2) is being used existentially? Even
when occurring without a complement, stated or implied, so I am told by
an anonymous critic, einai may ‘arguably’ be used as a copula. Even if so
anomalous a usage could be substantiated elsewhere, it would do noth-
ing to weaken the ‘existential’ use of the verb in the three passages quoted
above. When the Stranger says that movement ‘is, because it participates
in being’, all he can mean, in the context, is that it ‘is’, that it ‘exists’. He is
not saying, as Owen and Frede suppose, that as a result of its participation
in ‘being’, movement is ‘the same as itself ’ or is ‘other than rest’. If move-
ment, in this context, is ‘the same as itself ’, that is because it participates in
sameness. If it is ‘other than rest’, that is because it participates in otherness
in relation to rest. There is no implication, no possibility even, that in the
three passages I have quoted the verb is being used with any other than an
existential meaning. Movement that participates in ‘being’, in virtue of that
participation, ‘is’ (256a1 and 2). What is true of movement is true also of
rest, specifically included in the first two passages quoted above (252a2–4,
254d7–11), and will be no less true of the remaining forms or genē, all of
which, in the course of the argument, will be said to ‘be’ and to be so many
‘beings’, ‘in virtue of their participation in being’ (256e2–3).
II
But before pursuing even that simple conclusion, I have to pause, to
take account of an objection that will otherwise darken the skies for the
remainder of this brief essay. It is deeply unfashionable to write, as I have
already done twice so far, and as I shall continue to do, not infrequently,
in the pages that follow, of einai having an ‘existential meaning’. Today’s
communis opinio, at least among English-speaking writers (the fashion
does not yet seem to have spread to the continent of Europe), holds that
the undoubted grammatical difference, between an ‘existential’ or ‘sub-
stantive’ use of einai and the use of the same verb as a copula, does not
entail a difference of meaning, any more than does the difference between
the transitive and intransitive use of many a common verb: ‘he’s teaching’,
and ‘he’s teaching French’, an example initially chosen by Lesley Brown,
and repeated, with due qualification, by Myles Burnyeat.5
5
Brown 1986: 52–7 and 1994: 224–8. Burnyeat 2003: 9–10. My quoting Brown and Burnyeat at this
point does not imply that either author adheres to the interpretation of the Sophist put forward
by Owen or Frede. There are considerable differences, which it is no part of my present purpose
to explore. The reader looking for a survey of current literature on the Sophist may like to consult
If I prefer to swim against the tide, it is because I prefer not to separate
use from meaning. I am perfectly happy to acknowledge that, in Greek as
in English, an existential use of the verb leads easily to use of the same verb
as a copula, as it does in the verse already quoted from John Clare (‘I am:
yet who I am, none cares or knows’). In the following pages of this essay, I
shall claim that, in Greek, deliberate play is sometimes made with a single
occurrence of the verb being used in both ways, at the same time. But to
insist, as Burnyeat does, that there is no difference of meaning between the
two uses of the word other than that the first, although ‘complete’, is ‘com-
pletable’, seems to me questionable logic and poor grammar. Questionable
logic: can something be said to be ‘completable’ if it is (already) complete?
‘Completable’ would more naturally be taken to refer to what is not (yet)
complete.6 Poor grammar: can an ‘existential’ use of the verb properly be
defined as a copula that has not yet been given a complement? However
implausible, that would seem to be the implication if, as Burnyeat main-
tains, both uses of the verb are ‘complete’, with only the difference that the
existential use is ‘completable’ by having tacked onto it a complement, so
acquiring the distinctive mark of a copulative use of the same verb.
To avoid such seeming inconsequentiality, we need to look with a more
critical eye at the analogy that Burnyeat has taken over, all too trustingly,
from Lesley Brown: ‘teaching’ and ‘teaching something’, ‘being’ and ‘being
something’. Seductively simple, superficially convincing, the analogy does
indeed suggest that the difference, for either verb, is a difference only of
extension, that the simple verb, the verb used on its own without a com-
plement, is in either case a mere abbreviation of the ‘complete’ use of the
same verb with a complement. So it may well be for the difference between
transitive and intransitive uses of a verb. In the example quoted, we specify,
by our choice of verb, an intermittent activity which is the same whether
or not we specify the object of the action: ‘he’s teaching’, ‘he’s teaching
French’, or again: ‘she’s eating’, ‘she’s eating grapes’. Simple enough. But
does that elementary difference hit off the relationship between different
uses of ‘am-was-be’, as a complete predicate and as a copula?
The answer is not as straightforward as Brown and Burnyeat would
seem to think it is. A transitive verb, even when used intransitively, points
to the nature of its object: we can only teach what is teachable, eat what is
edible. In either case, the ‘meaning’ of the verb, whether used transitively
Michel Narcy’s introduction to the dialogue: Narcy 2012. But he should be warned that none of
the items referred to in the present article has been favoured with a place in Narcy’s ‘Bibliographie
essentielle’.
6
This is the meaning of the single occurrence of the word to be found in the OED, s.v.
226 Denis O’Brien
or intransitively, is the same, with the transitive use of the verb supplying
an object which we may think of as specifying and therefore ‘completing’
the use of the verb without an object. Not so ‘am-was-be’: the use of the
verb as a copula gives us no antecedent knowledge of the complement,
no indication of its nature, other than the specification of person, num-
ber and time conveyed by the inflexion of the verb. Brown’s analogy con-
ceals this important difference. Whereas a transitive verb delimits fairly
strictly the object of the verb (he can ‘teach’ only what is ‘teachable’, she
can ‘eat’ only what is ‘edible’), use of ‘am-was-be’ as a copula gives no
clue, raises no expectation, of what it is that he or she will be said to ‘be’
or not to ‘be’.
We might well have thought that such a difference would exclude use
of the verb as a ‘complete’ predicate. Can a verb that gives no advance
warning of the nature of the complement exclude even the expectation that
there will be a complement? However anomalous it may seem when stated
abstractly, such a use of the verb undoubtedly has its place in the language.
To ask for the verb, when it is used as a complete predicate, to be specified
by the addition of complement (‘… “is”, is what?’), as though the speaker
had cut his sentence short, or had deliberately cast his words as a provoca-
tive ellipse, is to show that we have misunderstood the sentence. We are
asking for information that the speaker may well possess, but that, by his
use of the verb as a complete predicate, he has not chosen to convey.
It does not at all follow that the use of the verb without a complement
is therefore meaningless, still less that it is, grammatically or syntactically,
incomplete. Clare’s ‘who I am’ may perhaps count as a specification of ‘I
am’. But his use of the verb as a copula is not needed to make sense of the
preceding use of the verb. Grammatically and syntactically, Clare’s repeti-
tion of the verb as a copula, ‘who I am’, does not serve to ‘complete’ a use
of the verb that is already a predicate complete in itself. When the verb
and its subject together make up a proposition that we recognise as true
or false, when the verb is being used therefore as a predicate ‘complete in
itself ’, it already has a meaning, independently of whatever specification
may, or may not, attach to the same verb if it happens to occur in the
same context as a copula. A meaning that I refer to as ‘existential’.
How then is the ‘existential meaning’ related to the use of the same
word as a copula? In the concluding pages of what is perhaps his most
famous article, ‘Is existence a predicate?’, G. E. Moore distinguishes the
two uses of the verb, existential and copulative, and does so with his usual
disarming simplicity. When the verb is used as a copula, ‘existence’ is
part of what is asserted. When it is used as a predicate complete in itself,
‘existence’ is the whole of what is asserted.7 Moore’s simple distinction
undercuts Burnyeat’s specious opposition, and will be found to answer
well enough to the needs of the present essay. The ‘copulative’ use of the
verb is not a ‘completion’ of its existential use. The verb has a copula-
tive function and, when used alone as a complete predicate, an existential
meaning. The existential meaning, if Moore is right, is part of what is
asserted by the use of the verb as a copula, but only when the copula has,
so to speak, been activated by the addition of a complement.8
‘The apple is red’, Moore’s example (adapted), asserts as part of its
meaning that there is an apple. With this example, it is easy enough to
see what makes Brown’s analogy so superficially appealing. If he’s ‘teaching
French’, he has to be ‘teaching’, though not conversely (he may be teach-
ing German today). If she’s ‘eating grapes’, she has to be ‘eating’, though
not conversely (she may be eating, but not grapes). If the apple ‘is red’,
it has to ‘be’, though not conversely (it has a colour, but not red). But
we also see in what way the analogy may mislead. We know, or we think
we know, what ‘teaching’ or ‘eating’ means, whether or not the verb is
used with a complement. But do we know what ‘being’ means? When it
is put so bluntly, we rightly hesitate in answering that question, and may
wrongly suppose that the answer can be read off from the analogy: ‘being’
is to ‘being something’, as ‘teaching’ is to ‘teaching something’.
But that perspective risks falsifying the syntax of the verb when used as
a complete predicate. When the Stranger asks if ‘movement is’, Theaetetus
recognises his words as a proposition to which he can, and does, give his
assent. He does so, without adding to the verb a complement, as both
Owen and Frede think he should have done, and therefore must have
done. But no: that is not how language works. When Theaetetus agrees
that ‘movement is’, the young boy’s knowing how to use the word (‘is’ as a
complete predicate), and his knowing how to react to the Stranger’s use of
the word, have to be taken as sufficient proof that the verb, when used as a
7
Moore 1936: 186–8. There are obvious exceptions to Moore’s semantic rule (‘… is impossible’, ‘… is
inconceivable’).
8
The simplicity of Moore’s analysis is a reaction to the technical complexity of Russell’s ‘Theory
of Descriptions’. The initial point made by the two philosophers is perhaps not so very different.
Russell’s famous example ‘The present King of France is bald’, so Moore would presumably want to
say, does in part assert that there is a King of France. If that is not so (if there is no one who answers
to the ‘description’), the proposition as a whole is either false (Russell) or meaningless (Strawson),
and if meaningless, then not properly a proposition at all. For a slightly less breathless comparison
of Russell’s ‘Theory of Descriptions’ with the Stranger’s insistence that a logos, even false, has to be a
logos ‘of something’ (262e5–7), see O’Brien 1995: 76–7.
228 Denis O’Brien
complete predicate, has a meaning, a meaning that can be recognised and
replicated, without recourse to use of the same word as a copula.
In the context of the Sophist, the Stranger’s question has of course
been accompanied by an explanation of why movement ‘is’: movement
‘is’ because it participates in ‘being’. Some of the ramifications of the
Stranger’s explanation, as set out in the course of Plato’s dialogue, will
occupy the remaining pages of this essay. But our enquiry cannot even get
under way unless the Stranger’s question and Theaetetus’ reply are allowed
to have the meaning, the only meaning, that can be given to their words
by a straightforward construal of the syntax of the sentence, in Greek.
The verb is used as a complete predicate, with an existential meaning. The
verb, so used, tells us that something is (‘movement is’), not what it is (‘…
is the same as itself ’), nor even what it is not (‘… is not rest’).
III
Will these ultra-simple propitiatory remarks be able to stem the tide of scholarly
opinion, at least enough to allow talk of an ‘existential use’ and an ‘existential
meaning’? Very likely not. The most powerful voice of the day speaks too firmly
in favour of the current orthodoxy. ‘From the point of view of the modern usage
of “exists”,’ so Kahn writes at one point in his magisterial survey of ‘the verb “be”
in ancient Greek’, ‘the term existential use of εἰμί is a misnomer’.9 In these few
preliminary paragraphs it would be ridiculous to engage with, still less to call
into question, five hundred pages of massed quotations and dense argument.
And yet Kahn’s massive tome is perhaps not the last word on the issue, and
his stricture perhaps not quite all that it may seem to be. ‘A misnomer’: Kahn’s
anathema would have frozen the pen in my hand if I had been aiming to draw
out the meaning of Plato’s εἶναι, used as a predicate complete in itself, from our
modern use of the Latinate ‘exist’. But that is not for one moment my intention.
When I write, as I do in these pages, of an ‘existential meaning’ of the Greek
verb, I do no more than refer to the use of the verb exemplified in Protagoras’
famous declaration: ‘Concerning the gods, I have no means of knowing that
they are, nor do I have any means of knowing that they are not (οὔθ’ ὡς εἰσὶν
οὔθ’ ὡς οὐκ εἰσίν), nor do I know what they are like to look at’ (80 B4 DK).
Kahn’s comment on this text is less trenchant than we might think it
would be. Protagoras’ use of the verb, so Kahn tells us, is a ‘“technical” use of
εἰμί as existential predicate’.10 But how can that be? If the verb has no prop-
erly ‘existential use’ (if to talk of an ‘existential use’ is a ‘misnomer’), how
9
Kahn 2003: 231, with the author’s italics. 10 Kahn 2003: 302.
can it be thought to serve as an ‘existential predicate’? What is the difference
between a ‘technical use’, in scare quotes, and an ‘existential use’, in italics?
Why is the one expression allowed and the other apparently frowned upon?
Why is Protagoras’ ‘technical’ use of the verb as an ‘existential predicate’ only
questionably, if at all, an ‘existential’ use? ‘Technical’ it may or may not have
been, ‘existential’ it certainly is. Protagoras’ bold profession of agnosticism
was meant to be provocative, and would have succeeded only if the ‘existen-
tial use’ and an ‘existential meaning’ of the two verbs, positive and negative
(εἰσίν, οὐκ εἰσίν), forming as they do a complete predicate, had been imme-
diately understandable by a contemporary audience. As no doubt they were.
In the face of so obvious and so deliberate an ‘existential use’ of the verb,
Kahn’s looking askance at that very expression as a ‘misnomer’, apparently
on the grounds that it lends itself to possible anachronism (‘from the point
of view of the modern usage …’), is puzzlingly scrupulous. The scrupulosity
would seem to stem from Kahn’s recurrent suspicion that later philosophical
issues may be found lurking beneath any talk, however traditional, however
conventional, however innocent, of an ‘existential use’ or an ‘existential mean-
ing’ for einai in ancient Greek. Certainly in my case, the suspicion is ground-
less. When I recognise Protagoras’ use of the two verbs as what Kahn calls an
‘existential predicate’, and when I therefore write of both an ‘existential use’
and an ‘existential meaning’ for the two verbs, I am not for one moment,
by drawing on either expression, inviting the reader to enter the world of
Biblical, Islamic or Scholastic thought that Kahn warns against elsewhere.11
Still less am I peddling the ‘Cartesian curse’ which Frege and Wittgenstein
have supposedly freed us from.12 I have no ulterior purpose in mind, religious
or philosophical, when writing, as straightforwardly as I have done and as I
shall continue to do, of both an ‘existential use’ and an ‘existential meaning’
of einai in order to single out, in Plato’s Sophist, occasions where the verb
is specifically used, as it so clearly is in the three passages I have quoted, as
an expression of ‘participation’ in being, with no further complement to the
verb stated or implied, and therefore as a predicate complete in itself.
IV
Granted that fact, the more pressing question is: does it matter? Does a
distinctively ‘existential’ use of einai impinge on the argument? The verb
is commonly used as a copula throughout the dialogue. Does it mat-
ter if it is occasionally the odd man out, by being occasionally given an
11
Kahn 1976. 12 Kahn 2003: 403–4.
230 Denis O’Brien
existential meaning? Yes, it matters very much indeed. It is only if we
have fully acknowledged the Stranger’s existential use of einai that we can
hope to grasp the continuation of his argument, where it will be a ques-
tion, not simply of ‘movement’, but of ‘otherness’, and of the very spe-
cific use of otherness that the Stranger will call upon as part of his reply to
Parmenides.
‘Otherness’, unlike ‘being’, comes into play only when what is ‘other’ is
so in relation to something else (255c8–e2). The principle claimed by the
Stranger when he first introduces otherness as one of the five genē will give
rise, a couple of pages later (257c5–258c5), to a theory of ‘parts’ of otherness.
The second term in the relation of otherness, whatever otherness is ‘other
than’, so to speak claims for itself a specific ‘part’ of the whole form, so that
‘otherness’ exists only as divided into ‘parts’: one ‘part’ of otherness relates
to beauty, another ‘part’ to largeness, another ‘part’ to justice. The ‘parts’ of
otherness, defined in this way, are so many negative forms: ‘non-beautiful’
(‘other than beautiful’), ‘non-large’ (‘other than large’), ‘non-just’ (‘other
than just’), all of them no less ‘real’, so the Stranger is at pains to point out,
than the corresponding positive forms: beauty, largeness, justice.
Theaetetus takes the theory in his stride. He assures the Stranger (258c4–
5), in answer to the latter’s anxious enquiry, that he has no misgivings over
the existence of negative forms, corresponding each one to a ‘part’ of oth-
erness. Not so the modern reader. Nicholas Denyer objects that Plato is
here suffering from a bad attack of what he calls ‘conjunctivitis’, a disease
that Denyer diagnoses as the attempt to define a relationship independ-
ently of the term of the relation.13 But that is exactly the quagmire that the
Stranger has avoided by his specification of the ‘parts’ of otherness. There
is no ‘other’ that is not ‘other than …’, with the specification provided by
whichever of the forms (beauty, largeness, justice) is called upon to act
as the second term of the relationship. The Stranger’s theory might have
been designed (I am inclined to think was designed) to avoid exactly the
objection voiced by his modern critic.
So much has to be made clear if we are to grasp the significance of the
‘existential’ use of einai when, following his account of negative forms,
the Stranger claims to have shown that ‘the nature of otherness is, as well
as being divided up among all the things that are, in their relation to one
other’ (258d7–e1). ‘Is’, in this sentence a participle (258d7: οὖσαν), looks
13
Denyer 1991: 139–45. ‘Conjunctivitis’ is defined as follows (142): ‘Any proposition saying that a
certain subject stands in a certain relation to a certain object is in fact a conjunction of two other
propositions, of which the first says that the subject has that relation, while the second says what
the object is to which the subject is related.’
back to the Stranger’s assertion, only a moment before, that otherness and
all the parts of otherness are to be counted among the many ‘things that
are’ (258a7–9), with a further backward glance to ‘the nature of otherness’
having been included as one of the genē (255d9-e7), all of which ‘are’, inso-
far as, like movement, they participate in being (256e2–3).
On any straightforward reading of the text, the Stranger’s use of the par-
ticiple (258d7: οὖσαν) is therefore, as it would appear to be, an ‘existential’
use of the verb: otherness ‘is’ because, like movement and like all the genē,
it participates in being. Or is this perhaps the elusive use of the verb with-
out a complement, stated or implied, but nonetheless with a copulative
meaning? That would appear to be the happy conviction of a recent com-
mentator, following in the wake of Owen and Frede, fully convinced that
existence cannot be a predicate, and therefore adding to the word a com-
plement nowhere to be found in the Greek text, but even so, to the eye of
the believer, so firmly entrenched in the argument that it can properly be
added to an English version of the text, with the result that the plain parti-
ciple (258d7: οὖσαν) ends up being translated as ‘is something’, for all the
world as though we had in the text, not a plain οὖσαν, but οὖσάν τι.14
The error is an error not only of translation, but of interpretation. Van
Eck’s intrusive ‘something’ is not to be found in the text of the Sophist, for
the very good reason that the Stranger’s ‘otherness’ is not a ‘something’.
Hardly have those words had time to flow from pen to paper before I hear
a sharp expression of dissent. It is a truth universally acknowledged, so I
am warned, that ‘to be is to be something’. The popular slogan, derived
from and companion to the belief that ‘existence cannot be a predicate’,
is indeed commonly thought to be another self-evident and unassailable
truth, and therefore a truth that, whatever the Greek text may or may not
say, cannot be other than what Plato means. But no. As with the existen-
tial use of einai, the words in the text cannot be overridden by appeal to a
principle that is nowhere voiced by the author of the dialogue, and that is
in any case not the unassailable truth that it is commonly thought to be.
Those who protest the loudest that ‘to be is to be something’ are those
who most clearly reveal that they have never taken the trouble to turn the
pages of the Enneads. Plotinus’ supreme principle, the One, is not ‘some-
thing’, but is not therefore non-existent. It ‘is’, and it is not a ‘something’.
But no doubt this is not the time or place to try and shake the deeply
rooted anti-Plotinian prejudice of too many of my English-speaking
colleagues, still less to urge upon them the unfashionable truth that the
14
Van Eck 2002: 73.
232 Denis O’Brien
Enneads provide the philosophical horizon of much that is said, and of
much that is not said, in the Sophist.15 Never mind. With or without the
help of Plotinus, the student of Plato’s Sophist, if he is to succeed in dis-
entangling the argument of the dialogue, has to recognise, however much
it may go against the grain, that otherness ‘is’, but that it is not for that
reason ‘something’.16
Van Eck’s mistranslation is all of a piece with Denyer’s mistaken diagno-
sis of ‘conjunctivitis’. Were otherness ‘something’ independently of its being
divided up into ‘parts’, then there would perhaps be some colour to the
charge that Plato’s Stranger has thought to define a relationship independ-
ently of the term of the relation. But neither argument holds water. In the
theory that the Stranger has put forward, at length and in detail, the term of
the relation does have to be specified for the relationship to have meaning.
Even so, the result of the specification can be expressed only as a negation.
The part of otherness that is opposed to beauty is the ‘non-beautiful’. The
part of otherness that is opposed to justice is the ‘non-just’… To demand
that otherness should nonetheless ‘be something’, is to deprive the Stranger’s
argument of its point and its purpose. There is no otherness that is not a part
of otherness, nor can any part of otherness be specified except as a negation.
I hear again the protest: ‘But to be is to be something.’ Not so. In this
admittedly very special case (but what is philosophy all about if it is not
about special cases?), the question ‘What is it?’ can be answered only by
saying what each part of otherness is not: ‘non-beautiful’, ‘non-large’,
‘non-just’. And so finally: ‘non-being’. For that is the negative form that
the Stranger’s account of the ‘parts’ of otherness has been leading up to.
The ‘part’ of otherness opposed to the form or ‘nature of beauty’ (257d11)
is the negative form ‘non-beautiful’. The ‘part’ of otherness opposed to the
‘nature of being’ (258b1) is the negative form ‘non-being’. A negative form
where the only answer to ‘what it is’, is ‘what is not’.17
The Stranger revels in the paradox. ‘We have brought to light’, so he
tells Theaetetus (258d6–7), ‘the form that there turns out to be, of what is
15
See O’Brien 2012: 30–65.
16
What I have called a ‘slogan’ can be cast in a more professional form: ‘To be is to be the value of a
variable.’ The principle does not therefore have any better right to be read into the text of Plato’s
Sophist. I doubt that Quine’s ontology would leave room for the One of the Enneads, or that it
would at all easily find a place for Plato’s form of otherness. It is true that Plotinus does, in one
place (2.4 [12] 16.3), refer to a part of otherness as ‘something’, but only to reinforce the point that
‘what it is’, is ‘what is not’ (see the continuation of my main text above).
17
For the syntax of this, the first of the Stranger’s two definitions of the form of non-being (258a11–
b3), see O’Brien 1995: 66–71, 2009a: 64–6. The second term of the relationship is to be understood
as ‘the nature of being’, not, as is commonly supposed, a ‘part’ of being. ‘Parts’ of being have no
place in the Stranger’s theory. The forms that, by ‘participating’ in being, are so many ‘beings’
not’ (258d6–7: τὸ εἶδος ὃ τυγχάνει ὂν τοῦ μὴ ὄντος ἀπεφηνάμεθα). The
positive use of the participle (ὄν) is most simply taken, here again, as hav-
ing an existential force. The form in question, though a negative form,
the form of ‘non-being’ or of ‘what is not’, is nonetheless a ‘part’ of other-
ness that participates in being and therefore turns out to ‘be’. With only
a change of case and gender, the participle (258d6: ὄν) looks forward to
the participle that will be repeated, by way of explanation, in the sentence
following (258d7: οὖσαν). In both sentences the purpose of the participle
(258d6: ὄν, 258d7: οὖσαν) is to state not what the form is, but that it is.
Or is that perhaps too simple a construal of the verb when followed
by a substantival expression in the genitive, as it is in the earlier sentence
(258d6: τυγχάνει ὂν τοῦ μὴ ὄντος)? The noun, in such a construction,
may have a predicative role.18 I am not at all averse to the belief that, in our
text, there may well be a deliberate play on the two uses of the verb. The
participle is intended to have the existential force that the matching parti-
ciple will undoubtedly have in the sentence following (258d7: οὖσαν), but
has also a subsidiary role as a copula, giving additional support, as it were,
to the genitive (258d6: τοῦ μὴ ὄντος), otherwise adequately accounted for
as a genitive ‘of definition’, dependent on the noun that precedes (258d6:
εἶδος, ‘form’). We are to understand both that ‘The form of what is not
turns out to be’ (ὄν existential) and ‘The form turns out to be of what is
not’ (ὄν copulative). Hoping to convey something of the deliberate syn-
tactical equivocation, I have made so bold as to add a comma to the trans-
lation. The Stranger revels in having brought to light ‘the form that there
turns out to be, of what is not’ (258d6–7).19
To say so much is not to call into question the exclusively existential
use of the participle in the sentence following (258d7: οὖσαν). On the
contrary, the explanatory particle linking the two sentences (258d7: γάρ)
confirms the existential force of the participle in the sentence preceding
(258d6: ὄν), an existential force that needs to be fully recognised if we are
to appreciate the specific meaning attaching to the participle, preceded by
(256e2–3) are not therefore so many ‘parts’ of being; ‘participation’ is not the same as ‘partition’. For
the Stranger’s second definition (at 258e2–3), see n. 22 below.
18
Thucydides 3.70.5: ἐτύγχανε γὰρ καὶ βουλῆς ὤν. Peithias, the Athenian sympathiser, ‘happened to
be a member of the Council’, or more colloquially, and closer to the Greek: ‘He happened to be on
the Council.’ See Schwyzer and Debrunner 1959: 123, Kahn 2003: 167.
19
A ‘deliberate equivocation’, but not designed to mislead. A ‘double entendre’, but not with a sly
wink. A ‘play on words’, but with only one word to play with. Not quite Kahn’s ‘overdetermined’
or ‘overconstrued’ use of the verb, ‘where a single occurrence of the verb illustrates two or more
distinct sentence types’ (Kahn 2003: 323), since in the Stranger’s definition, so I suggest, the two
meanings are deliberately played off, one against the other. So call it what you will. This, and the
two examples that are to follow, I hope will make my meaning clear.
234 Denis O’Brien
a negative particle, in the second half of the paradox. The form that ‘is’, or
that turns out to ‘be’ (ὄν), is a form of ‘what is not’ (258d6: τοῦ μὴ ὄντος).
What is the meaning of the negation?
V
A spontaneous reaction may be to suppose that the same verb should have
the same meaning when it occurs, as it does here, in the same sentence and
indeed in the same clause (258d6), once with a negative particle and once
without: if an ‘existential force’ is asserted (ὄν), then an ‘existential force’
must be denied (μὴ ὄντος). But that cannot be so. It is too close to contra-
diction for comfort to assert that the form of what does not participate in
being is a form that does participate in being. The Stranger cannot assert, in
one and the same breath, that a form of what does not exist is a form that
does exist. However paradoxical, the Stranger’s paradox is not contradictory.
Contradiction has been avoided by the Stranger’s analysis of negation.
In the preliminaries to his account of otherness and its ‘parts’ (257b1–
c4), the Stranger explains to Theaetetus that the negative particles μή
and οὐ (commonly distinguished as a ‘subjective’ and an ‘objective’ style
of negation, though often with seemingly very little difference between
them) are to be understood as an expression of otherness in relation to
the words that follow them: what is ‘not something’ is ‘other than that’. A
negation of being, ‘non-being’ or ‘what is not’ (257b3: τὸ μὴ ὄν), exactly
the same ‘substantivised’ participle that will be used in the description of
‘the form of non-being’ (258d6: τοῦ μὴ ὄντος), is so, therefore, because it
is ‘other than being’ (257b4: ἕτερον sc. τοῦ ὄντος).
With that explanation of the negative particle, the paradox avoids
contradiction while remaining fully paradoxical. When the Stranger
describes the form that ‘is’ as a form of ‘non-being’ (258d6: τοῦ μὴ ὄντος),
the negation has to be understood with the meaning it had been given in
his preceding account of negation (257b1–c4), and therefore as by impli-
cation a copulative use of the verb, insofar as ‘non-being’, in this context,
is so because it is ‘other than being’. The form of ‘what is not’ turns out
to ‘be’ because it participates in being, as do all the ‘parts’ of otherness,
and therefore ‘is’. At the same time it is opposed to, and is therefore ‘other
than’, the very form in which it participates, and insofar as it is ‘other than
being’, it is therefore ‘non-being’.20
20
In writing of ‘what is not’ (258d6: τοῦ μὴ ὄντος) as ‘by implication a copulative use of the verb’, I
am not adding my widow’s mite to the vast treasury of technical terms serving to classify different
uses of einai in classical Greek. My expression is nothing more than a convenient pointer to the
The meaning claimed for the negation, essential to the paradoxical con-
junction of ‘is’ and ‘is not’, has had the ground thoroughly prepared for
it in the preceding pages of the dialogue. We need to start again from
the successive uses of the verb, copulative and existential, in the Stranger’s
account of movement, when he insists that movement ‘is not’ rest, but
that it ‘is’ nonetheless, in virtue of its participation in being (255e11–256a2).
The Stranger prefaces his assertion that movement ‘is not rest’ (255e14) by
getting Theaetetus to agree that movement ‘is altogether other than rest’
(255e11–13). The addition of the adverb (255e11: παντάπασιν, ‘altogether’)
eases any difficulty that there might otherwise have been in making the
transition from ‘other than …’ to ‘is not …’. Movement and rest have
already been said to be ‘most contrary’ (250a8). Neither therefore partici-
pates in the other (252d2–11). ‘Altogether other than’, indicating therefore,
as it does in this context, total absence of participation, cannot but imply
‘is not’. Movement ‘is not’ rest because it is ‘altogether other than’ rest,
neither form having any participation in its contrary.
But that is only a first step. The sentences following (256a3–b5) repeat
the entailment of ‘other than’ and ‘is not’, but without calling on con-
trariety and without excluding participation. Movement, so the Stranger
tells Theaetetus, is, at one and the same time, both ‘the same’ and ‘not
the same’, ‘the same’ because it participates in sameness, ‘not the same’
because it participates in otherness in relation to sameness. There is no
place here for the adverb. ‘Other than …’ no longer marks lack of par-
ticipation. How could it? ‘Sameness’ is one of the three genē (with ‘being’
and ‘otherness’) that are participated universally. Everything that ‘is’ (with
the possible exception of otherness) is ‘the same’ as itself. Otherness is
therefore still used to specify a negation (‘other than the same’ is ‘not the
same’), but the negation, in relation to sameness, indicates only lack of
identity with the form, not absence of participation in the form.
This is the point of the Stranger’s remark that the two expressions
(‘the same’, ‘not the same’) are not being used ‘in the same way’ (256a12:
ὁμοίως). The negation does not bear directly on the meaning given by the
positive use of the verb. When we say, in this context, that movement is
‘the same’ and is ‘not the same’, we are not first asserting and then deny-
ing that it participates in sameness. To do so would be to contravene what
use of the negation that has been specified by the Stranger in the passage quoted (257b1–c4): what
is ‘not x’ is ‘other than x’, what is ‘non-being’ is ‘other than being’, as distinct from the negation
bearing on an existential use of the verb: ‘is not’, ‘does not exist’. The ‘copulative’ connotation is
inherent in the very notion of otherness: ‘other than …’, ‘is not …’.
236 Denis O’Brien
Aristotle calls the ‘firmest of all principles’, the principle of contradic-
tion.21 If the negation is an expression of otherness, it is so, therefore, only
insofar as ‘otherness’ implies lack of identity. We assert, without contra-
diction, that movement is both ‘the same’ and ‘not the same’, insofar as
movement participates in sameness (and is therefore ‘the same’), but is not
identical to sameness (and is therefore ‘not the same’). The adaptation, the
shift from ‘movement is not rest’ to ‘movement is not the same’, is clear
and clearly deliberate: ‘otherness’, when it bears on a form that is univer-
sally participated, indicates, not lack of participation, but lack of identity.
Whatever participates in ‘other’ in relation to sameness, participates in
sameness, but is not identical to sameness.
The same point is then made of movement in relation first to ‘other-
ness’ (256c5–10) and then to ‘being’ (256c11–d10), again two genē that are
participated universally. Movement qualifies as ‘other’ since it has already
been established that it is ‘other’ than rest (255e11–15) and ‘other’ than
sameness (256a3–6); but at the same time it is ‘not other’, insofar as it is
‘other than otherness’ (256c5–6). Again, a lack of identity with the form,
not a lack of participation in the form. Pari passu, movement is at one and
the same time both ‘being’, since it participates in being (256d8–9), and
‘non-being’, insofar as it is ‘other than being’ (256d5–8). Again, lack of
identity, not absence of participation.
By the time the Stranger comes to describe the form of non-being
(258d5–e3), the ground has therefore been so thoroughly worked over that
the same principle can be virtually taken for granted. In asserting that the
form of non-being ‘is’, we assert that it participates in being. The negation
(the negative particle specifying the form as a form of ‘what is not’) cannot
bear directly on the positive use of the verb; the denial is not a denial of
participation. The negation is still an expression of otherness (‘other than
being’ is ‘non-being’), but again of otherness as marking a lack of identity
with the form, not a total absence of participation in the form.
The difference, the novelty, lies in the point that the paradox now
applies, not to whatever participates in the form, but to the form itself.
We are no longer dealing with movement that participates in being, and
therefore in otherness in relation to being (256c11–d10). We are now deal-
ing directly with the very ‘part’ of otherness that is itself opposed to being,
and that is therefore ‘the form of non-being’ (258d5–e3). The Stranger can
therefore conclude, with a flourish intended to highlight the paradoxical
conjunction of ‘being’ and ‘non-being’: ‘We have dared to say, of the part
21
Aristotle, Metaph. Γ.3, 1005b17–18.
of otherness opposed to the being of each thing, that that very thing is
really what is not’ (258e2–3). The ‘part of otherness opposed to the being
of each thing’ is the form of non-being. The form of non-being ‘is really
what is not’.22
The exaggeration, the ostentatious juxtaposition of ‘is really’ and ‘is not’
(258e3: … ἐστιν ὄντως τὸ μὴ ὄν), is deliberate. Indeed, in the deliberate
doubling of the positive verb by the addition of an adverb that is none
other than a repetition of the verb (ἐστιν ὄντως, untranslatable in English,
where verb and adverb cannot overlap the way they do in Greek), I am
again more than half inclined to see a deliberate play on the existential
and the copulative role of the verb, another example of the Stranger’s
deliberate syntactical equivocation. The form of non-being ‘is really’: verb
and adverb are first heard with a slight pause, to mark the heavy-handed
existential meaning, before being followed by ‘what is not’, with the same
conjunction of verb and adverb (‘is really’) now heard as introducing a
predicate other than the verb, the verb being heard therefore, a second
time round as it were, in its role as a copula. Once again, a comma is the
only help one can give a Greekless reader: ‘… that very thing is really,
what is not’, unless one is to go as far as a repetition: ‘… that very thing is
really, it is really what is not’, perhaps even admitting a change in the plac-
ing of the adverb: ‘… that very thing really is, it is really what is not’.23
The equivocation, if I have diagnosed it aright, is possible only because,
in the preceding pages of the dialogue, the Stranger has clearly established
a difference between the existential use of the verb, ‘is’ as an expression
of participation in being, and ‘is not’ as by implication a copulative use
of the verb, insofar as ‘is not’, in this context, implies ‘other than’, with
‘otherness’ indicating not lack of participation in being, but lack of iden-
tity with being. Fumble that distinction by refusing to recognise an exist-
ential use of the verb, as Owen, Van Eck and Frede have all three done,
22
The two definitions of the form of non-being as a ‘part’ of otherness (258a11–b3 and 258e2–3) are
not the same, insofar as, in the first definition, the second term in the opposition is ‘the nature of
being’ (258b1), whereas here, if we keep to the unanimous reading of the manuscripts, it is ‘the
being of each thing’ (258e2). For the difference (a difference between the form and the form as
instantiated), and for the Neoplatonic origin of the reading adopted in successive Oxford editions
of the text, see O’Brien 1991, 1995: 66–71 and 2009b: 302–7.
23
Is the equivocation deliberate? I am pretty sure that the same doubling up of the verb is used
for conscious effect by Empedocles, in a sequence of verses (31 B17.9–13) where the slight pause,
needed to bring out the syntactical equivocation, is provided by the caesura. The verb in the final
verse of the sequence (v. 13: ἔασιν) is contrasted to change from one to many and from many to
one, specified as a ‘becoming’ two lines before (v. 11: γίγνονται), and is therefore first heard, before
the caesura, with an existential meaning. The same verb is then heard, as it were a second time, but
now as a copula, introducing the expression that completes the verse: ‘fixed unmoving in a cycle.’
238 Denis O’Brien
and the Stranger’s definition of ‘non-being’ will slip through your fingers,
as will the whole strategy of Plato’s dialogue as a reply to Parmenides.
VI
However, before turning to Parmenides, we need to recognise another
‘existential’ use of the verb, in the words following the paradoxical descrip-
tion of the form that ‘is’ of ‘what is not’. ‘Don’t let anyone try to tell us,’ so
the Stranger warns Theaetetus (258e6–7), ‘that proclaiming “non-being” as
the contrary of being we dare to say of that that it is.’ ‘We have long ago
said good-bye,’ he continues (258e7–259a1), ‘to any contrary of being, to
the question of whether it is or it isn’t, whether it has a logos or is altogether
without a logos.’
The Stranger’s ‘farewell’ looks back to the tangled discussion that had
followed his first quotation from Parmenides (237a3–b3), when poor
Theaetetus had been made to wrestle, long and hard (237b4–239c8), with
the seemingly intractable problem of whether or not we can ever speak,
without contradiction, of ‘what is not in any way at all’ (237b7–8), of ‘what
is not in and by itself ’ (238c9), since even in uttering the words ‘what is
not’, we have to have recourse to a singular, a dual or a plural, and therefore
to attach number, which Theaetetus, the budding mathematician, cannot
but agree is included in ‘what is’ (238b1), to what we are vainly trying to
say ‘is not in any way at all’. But why does the Stranger, at this point in the
dialogue (258e6–259a1), hark back to the seeming impasse of the earlier dis-
cussion, and why, in doing so, does he refer to the impossible and incon-
ceivable ‘what is not in any way at all’ (237b7–8) as a ‘contrary’ of being?
The Stranger is here reaping the benefit of his earlier remarks on contra-
riety in relation to movement and rest. From the description of movement
and rest as ‘most contrary’ (250a8), and therefore unable to ‘mingle’ with
each other (252d2–11), we are intended to draw the conclusion that con-
trariety excludes participation. The ‘contrary of being’ would therefore be
whatever does not participate in being. ‘Would be’, because of course there
isn’t anything that doesn’t participate in being. That is why we cannot even
think of it or speak of it. There is nothing there to think of or to speak of.
That is why there can be no question of whether a contrary of being ‘is or
isn’t’, whether it ‘has a logos or is altogether without a logos’, the questions
that had tormented poor Theaetetus in the earlier pages of the dialogue.
The Stranger’s distinction between a ‘form of non-being’ and an impossible
‘contrary of being’ provides a release from the earlier seemingly impossible
aporiai. The buzzing fly has been let out of the bottle. The questions that
seemed to have no answer are no longer questions that have to be asked.
But quite why is that so? We need to return to the distinction between a
copulative and an existential use of the verb. When the Stranger refers to a
‘contrary of being’, looking back therefore to the earlier tangles concerning
‘what is not in any way at all’ (237b7–8) and ‘what is not in and by itself ’
(238c9), he does so because those two expressions encapsulate a negation
of the existential use of the verb, and in so doing point to the restriction
on the use of the negation in the description of the form that ‘is’ of ‘what
is not’. The verb negated in the Stranger’s paradoxical description of the
form of non-being is by implication a copulative use of the verb, ‘is not’
taken as equivalent to ‘is other than’, with otherness restricted to a lack
of identity with being. In arriving at his description of the form that ‘is’
of what ‘is not’, there has been no negation of the verb used existentially.
A negation of the existential use of the verb would drag us back to the
toils and moils of ‘what is not in any way at all’. That whole morass of
self-contradiction (237c8–239a2: how can we ever say that non-being can’t
be spoken of, since we have to speak of it in order to say that it is unspeak-
able?) has been cast aside by the distinction between negation and contra-
riety. In speaking of what is ‘other than being’ as ‘non-being’, a negation
of what is by implication a copulative use of the verb, we are not speaking
of an impossible ‘contrary’ of being, a negation of the existential use of the
verb, ‘what is not in any way at all’ (237b7–8), ‘what is not in and by itself ’
(238c9), a total lack of participation in being, sheer nothingness.
The solution to the earlier aporiai lies therefore in the two different uses
of the negation, and in the two different uses of the verb. A negation of
what is by implication a copulative use of the verb is not a negation of the
same verb used ‘existentially’. In speaking of ‘what is not’ as what is ‘other
than …’, we are not speaking of an impossible ‘contrary’ of being, a nega-
tion of the verb in its existential use. The part of otherness that is opposed
to being and that is in that sense ‘non-being’ (by implication a copulative
use of the verb) nonetheless participates in being, and therefore ‘is’ (an
existential use of the verb). The purpose of the Stranger’s analysis has been
to show that ‘being’ does have a negation, ‘the form of non-being’, but
that it does not have a contrary. The negation of being, as defined by the
Stranger, is not a contrary of being. There is, and there can be, no contrary
to being.24
24
‘Contrariety’ is an essential feature of the Stranger’s analysis, as I hope I have sufficiently made clear
in the preceding pages. On the crucial passage, 257b1–c4 (both ‘equal’ and ‘small’ are ‘non-large’,
but only one of the two is the ‘contrary’ of large), see O’Brien 1995: 57–9 and 91–102, correcting
Owen 1971: 235–6.
240 Denis O’Brien
VII
It is that distinction that provides the kernel to what is commonly supposed
to be the Stranger’s ‘refutation’ of Parmenides. But here the errors of interpre-
tation fall so thick and fast that it is difficult for the still, small voice of sanity
to make itself heard above the general hubbub of confusion and contradic-
tion. Darkness closes in from the moment when the Stranger, following his
first quotation from Parmenides’ poem, asks Theaetetus if we ‘dare speak’
of ‘what is not in any way at all’ (237b7–8). The young boy, so innocent, so
guileless, replies that yes, of course we ‘dare’: ‘Why on earth not?’ (237b9)
How could he answer otherwise? How could the boy deny that we ‘dare’
speak of ‘what is not in any way at all’, when that is what the Stranger him-
self has just done in framing the question? It is only after an intense session
of question-and-answer, lasting for all of two pages or more (237b2–239c8),
that the boy will be brought to see that no, that one cannot think, that one
cannot properly even speak, of ‘what is not in any way at all’.
This stretch of the dialogue (237b2–239c8) is treated by one of Plato’s
more recent commentators, Noburu Notomi, as all part of Plato’s ‘refuta-
tion’ of Parmenides.25 But how can that be so? Parmenides had himself
argued that it is impossible to ‘think’ or ‘speak’ of ‘what is not’. The poem
is replete with warnings that ‘non-being’ cannot be ‘known’ (28 B2.6–8
DK), is neither ‘speakable’ nor ‘thinkable’ (B8.8–9), is ‘unthinkable’ and
‘nameless’ (B8.17–18). The immediate impact of the arguments put for-
ward by Plato’s Stranger, in the pages following his first quotation from the
poem (237a3–b2), is no different: if ‘what is not’ is construed as ‘what is
not in any way at all’, then it cannot be spoken or thought of. This whole
stretch of dialogue (237b2–239c8) is a virtual re-play of the arguments put
forward by Parmenides in his poem. Far from ‘refuting’ Parmenides, the
Stranger, in these pages, argues vigorously in his support.26
There is of course a powerful arrière pensée to the Stranger’s endorsement
of Parmenides’ arguments. The explanation of his (provisional) agreement
with Parmenides is to be found in the qualification ‘not in any way at all’
25
Notomi 2007b: 170–81. The references to myself in Notomi’s article are wildly inaccurate. See
O’Brien 2011: 213 n. 11.
26
For the structure of the poem, insofar as it can be recovered from the generous supply of frag-
ments, including the whole of the prologue (thirty-two verses), quoted by Sextus, and all of the
so-called Way of Truth (forty-nine verses), quoted by Sextus and Simplicius, see O’Brien 1987.
For Parmenides’ insistence that we cannot think or speak of ‘what is not’ (the ‘non-being’ of the
second Way), see O’Brien 2000: 19–43, correcting the misconception in G. E. M. Anscombe 1969
(Parmenides’ ‘non-being’ wrongly construed as ‘impossible pictures’ and ‘impossible states of
affairs’).
(237b7: μηδαμῶς). So long as ‘non-being’ is taken to be tantamount to
‘what is not in any way at all’, and therefore indistinguishable from what the
Stranger will later refer to as a ‘contrary’ of being (258e6–259a1), then it is
indeed as unthinkable and unspeakable as Parmenides had said it was. It is
only when ‘non-being’ is no longer reduced to ‘what is not in any way at all’,
and is therefore no longer conflated with an impossible ‘contrary’ of being,
that the Stranger will part company with Parmenides. When he claims, as
he will do after repeating his quotation from the poem (258c6–259b7), that
we can indeed both think and speak of ‘what is not’, it is because the world
the Stranger has introduced us to is no longer the world of Parmenides,
and because the ‘non-being’ in question is therefore no longer an impos-
sible ‘contrary’ of being, but the ‘non-being’ that is a ‘part’ of otherness, the
‘non-being’ that has been defined as a form that ‘is’ of ‘what is not’.
It is that difference that explains the Stranger’s repeated quotation of
the same two verses from Parmenides’ poem (B7.1–2). The two verses, in
their original context in the poem, are a condemnation of the ‘opinions
of mortals’, their contradictory belief, so Parmenides claims, that ‘things
that are not, are’. The words are therefore first quoted as an illustration
of the seemingly impossible conjunction of ‘being’ and ‘non-being’ that
would supposedly be needed if there were ever to be such a thing as a
logos that is false (236e1–237b3). When the same verses are quoted a second
time, it is to mark the radical change that has been brought about by the
Stranger’s brilliant discovery of ‘a form that is, of what is not’ (258d1–e5),
a form that, after much adaptation, will be claimed as an explanation of
how a logos can, after all, state falsely, of things that ‘are not’, that they ‘are’
(263d1–4).27
The meaning given to the verses (B7.1–2) is therefore not at all the same
in the two places. The first time round, the verses usher in the question,
whether we can think or speak of ‘what is not in any way at all’ (237b7–8),
a feat that the Stranger joins forces with Parmenides in declaring to be
impossible (237b4–239c8). When the same verses are repeated, many pages
later (258d1–e5), the Stranger is able to quote, but now with approval,
the claim that ‘things that are not, are’ because the words are no longer
contradictory, provided that the ‘non-beings’ in question are taken as a
description of all the many things that ‘are not’ because they participate in
27
‘After much adaptation’: the proposition ‘things that are not, are’, adopted by the Stranger as a uni-
versal truth (258d1–259b7: whatever participates in being, and therefore ‘is’, has also to participate
in otherness in relation to being, and therefore ‘is not’), will require a whole new stretch of argu-
ment (259b8–263d5) before it can be put to use as an explanation of the false logos (263a5–b13: how
we can say, falsely, of Theaetetus, who is sitting down, that he is flying). See O’Brien 1995: 72–88.
242 Denis O’Brien
the form of non-being. If ‘things that are not, are’ (B7.1), it is because the
form of non-being is itself a form that ‘is’ of ‘what is not’ (258d5–7).
VIII
The structure of the dialogue hinges on that distinction. When the Stranger
calls a provisional halt (at 239c4–8) to his first long session of questioning,
following his initial quotation from the poem, he suggests that the reso-
lution of the seemingly insurmountable difficulties confronting anyone
who attempts to overturn Parmenides’ condemnation of ‘what is not’ will
have to take the form of a compromise (241d5–7). Non-being will have
to agree that it ‘is, to a certain extent’. ‘Being’, contrariwise, will have to
agree that ‘it is not, in a way’. The compromise, as regards non-being, is
what the Stranger will claim to have achieved when he arrives at his para-
doxical definition of a ‘form of what is not’ that nonetheless ‘turns out to
be’ (258d6), a form that ‘is really, what is not’ (258e2–3). The paradoxi-
cal conjunction of ‘being’ and ‘non-being’, in these two sentences, is the
Stranger’s answer to what had seemed, at the time, the almost impossible
requirement that non-being should ‘be, to a certain extent’. Non-being
‘is, to a certain extent’, provided that it is non-being as defined by the
Stranger, a ‘part’ of otherness opposed to, but nonetheless participating in
being, therefore both a form that ‘is’ and a form of ‘what is not’.28
Once we have grasped that distinction, we see how meaningless it is
to describe the Stranger’s critique of Parmenides as an outright ‘refuta-
tion’ of his illustrious predecessor, a simplification endlessly and mind-
lessly repeated by students of Plato, old and young. The Stranger has
introduced a distinction where Parmenides had none. We can speak of
‘what is not’, but only as the Stranger has defined it, a negation of being,
but not a contrary of being. In saying so much, the Stranger is not for one
moment asserting what he had earlier denied. He is not for one moment
saying that we can think, or properly even speak, of a contrary of being, of
what has no participation in being, of ‘what is not in any way at all’. Still
less therefore is he claiming to ‘refute’ Parmenides by deliberately ‘leaving
open’ the possibility that ‘what is not in any way at all’ might nonetheless,
somehow, just possibly ‘be’.29
28
I quote from the more familiar passage where the Stranger first states his project for an escape from
the preceding aporiai (241d5–7). When he re-states his project several pages later (254c8–d2), the
wording is even closer (almost, but not quite, identical) to the words that will follow his second
quotation of Parmenides’ two verses (258e2–3).
29
Notomi 2007b: 184. ‘Plato deliberately leaves open the question about the being of what in no way is’
(the author’s own italics). For Plato’s supposed ‘refutation’ of Parmenides, see O’Brien 2000 and 2011.
The Stranger’s ‘farewell’ to a ‘contrary’ of being (258e6–259a1) has exactly
the opposite purpose. A ‘contrary’ of being, a total absence of any partici-
pation in being, cannot, by any stretch of the imagination, ever be said to
‘be’. If the Stranger claims to have shown that what is not ‘is, to a certain
extent’ (cf. 241d5–7), it is because he has shown that ‘the form of non-being’
is a form that ‘turns out to be’ (258d6), that the form of non-being ‘really
is’, and yet ‘is really what is not’ (258e2–3). That conjunction of ‘being’ and
‘non-being’ is possible only because the ‘form of non-being’ is not a ‘con-
trary’ of being. The ‘farewell’ to a contrary of being (258e6–259a1) marks
the climax of the Stranger’s argument precisely because it shows up the dif-
ference between what does and what does not participate in being, between
the negation bearing directly on an existential use of the verb and therefore
excluding any participation in being, and a negation expressing the absence
of identity that is a necessary condition of participation in being.
To suppose that the ‘contrary’ of being might itself be allowed to ‘be’
shows a complete misunderstanding of the argument. The thrust of the
Stranger’s argument, the subtlety of his reply to Parmenides, turns on the
distinction between a ‘contrary’ of being, ‘what is not in any way at all’,
and the form of non-being that, although a form of ‘what is not’, is none-
theless a form that ‘turns out to be’. To have failed to grasp that distinction
is to have failed to grasp the whole point and purpose of the dialogue.
IX
Have others succeeded where Notomi has so signally failed? Notomi has
failed to distinguish a ‘form of non-being’ from a ‘contrary of being’. Owen
stumbles over the relation of ‘non-being’ to ‘being’. When the Stranger
declares (257a4–6) that being ‘is not’ all the many things that partake in
being, Owen concludes that ‘being itself ’ is therefore ‘non-being’.30 That
is not what the Stranger says, and it cannot be what he implies. In the
Sophist, to ‘be’, no less than ‘not to be’, requires participation in a form,
the form or ‘nature of being’ in the one case, the ‘form of non-being’ in
the other. With the obvious exception of ‘being itself ’, everything that
‘is’, by participating in ‘being’, participates also in ‘non-being’, insofar as
it participates in otherness in relation to being, and is therefore not iden-
tical to being. On both counts, the form of being has to be the obvious
30
Owen 1971: 232–3. ‘Being itself ’: although ‘not being’ all the many things that participate in being,
being ‘is one thing, itself ’ (257a5). The reference is clearly to being as one of the genē, and therefore
to the form or ‘nature’ of being (258b1). Here and in the paragraphs that follow, I therefore abbrevi-
ate as ‘being itself ’.
244 Denis O’Brien
exception. The form of being does not participate in itself in order to ‘be’.
Still less can it participate in otherness in relation to itself in order ‘not to
be’. ‘Being itself ’ cannot be, as Owen thinks it is, a ‘non-being’.
Confusion inevitably arises if ‘self-predication’ is thought to
entail ‘self-participation’. There is an exceptionally clear statement of
self-predication when the Stranger summarises his theory of the parts
of otherness (258b8–c5). Theaetetus agrees without demur that ‘the large
is large’ and that ‘the beautiful is beautiful’ (258b10–c1: the verb is the
so-called ‘philosophical’ use of the imperfect), with the clear implication,
carried over from the preceding words, that this is so because, in either
case, the form is ‘in possession of its own nature’ (258b10). That simple
explanation excludes self-participation. A form cannot be said to ‘partici-
pate’ in the ‘nature’ that it already possesses. The form of beauty is indeed,
itself, ‘beautiful’, but not because it participates in itself.
So much is made additionally clear when the Stranger adopts the con-
verse argument as proof of the need for a form of ‘otherness’. No one
form, so the Stranger tells Theaetetus, is different from any other ‘in virtue
of its own nature’, but only ‘as a result of its participation in the form of
otherness’ (255e4–6). The two passages are clearly complementary. A form
is what it is ‘in virtue of its own nature’ (258b10–c1); but ‘what it is’ is
not sufficient explanation of ‘what it is not’ (255e4–6). Every form there-
fore participates in otherness in relation to every other form, with the one
obvious exception of the form that is ‘itself ’.
This is the point of the Stranger’s assertion that being ‘is not as often
and in as many ways as other things are’ (257a4–6). ‘Being’ participates
in otherness in relation to all the forms that participate in being. But that
assertion does not for one moment imply that being participates in other-
ness in relation to itself, nor therefore is there any implication, as Owen
thinks there has to be, that ‘being itself ’ is ‘non-being’. A ‘non-being’, in
this context (256d5–257a7), can be constituted only by whatever partici-
pates in being, and therefore in otherness in relation to being. Two condi-
tions that being itself, in the nature of things, cannot satisfy.
Owen has been led to his impossibly contradictory conclusion, not only
by his over-all misunderstanding of the Stranger’s theory of ‘being’ and
‘otherness’, but more immediately by his failure to follow the argument in
the preceding lines of the dialogue, where all the things that are said to ‘be’
are described as so many ‘beings’, and all the things that are said ‘not to be’
are counted as so many ‘non-beings’ (256d5–e7). Owen concludes that, since
‘being itself’ will be said ‘not to be’ all the things that participate in being
(257a4–6), therefore ‘being itself ’ must be a ‘non-being’. It requires only a
moment’s thought to see that that cannot be so, and that it is not merely
the result of chance that the Stranger does not make the transition from ‘is
not’ to ‘non-being’ when he talks specifically of ‘being’ in the latter sentence
(257a4–6). ‘Being’ cannot be ‘other than itself ’. ‘Being’ cannot participate in
otherness in relation to being. It cannot be itself and not itself.
Owen has misconstrued the passage because he has failed to recognise
that ‘is not’, as an expression of otherness, may be construed as ‘non-being’
only when ‘being’ is the term of the relationship. ‘Other than beauty’
yields the negative form ‘non-beautiful’. ‘Other than justice’ yields the
negative form ‘non-just’. But only ‘other than being’ yields the negative
form ‘non-being’. The genē that, in the preceding lines of the dialogue,
have been designated as both ‘beings’ and ‘non-beings’ (256d5–e7) are so
because they participate both in ‘being’ and in otherness in relation to
being. ‘The nature of otherness,’ so the Stranger tells Theaetetus (256d12–
e1) ‘works on each one of them … bringing it about that each one is other
than being,’ and in so doing ‘makes each one of them non-being.’ The
one exception to that rule is, and has to be, being itself. ‘Being’ cannot be
‘worked upon’ by otherness in order to be ‘made other than being’, and
therefore ‘other than’ itself. Being ‘is not’ all the many things that partici-
pate in being (257a4–6). But it is not therefore ‘other than’ itself, and is
not therefore itself ‘non-being’.31
X
To free ourselves from such error and confusion, we need to start again,
preferably from the existential use of einai.32 In the Sophist as in the Phaedo,
31
For the syntax of the sentence as a whole (256d12–e1), and especially for the explanatory role of
the particle in the opening expression (256d12: κατὰ πάντα γάρ), see O’Brien 1995: 91–102. Van
Eck 2002, attempts to resuscitate Owen’s interpretation, 1971: 233 n. 20 (= 1986: 111 n. 19), but is
hampered by his difficulty in understanding both Plato’s Greek and my French. See O’Brien 2011:
218 n. 16.
32
The paragraphs that follow do no more than outline the barest philosophical bones holding
together the central part of the dialogue (237a2–259b7), stopping short of the account of a false
logos (259e4–263d5), where the analysis of ‘being’, ‘non-being’ and ‘contrariety’ will be significantly
different. I deliberately adopt a more down-to-earth example of positive and negative predication
(‘beautiful’ and ‘not beautiful’), applying therefore to the first of the ‘negative’ forms (257d7–11)
the principle that the Stranger states for movement in relation to ‘sameness’ (256a10-b5: ‘the same’
and ‘not the same’) and to ‘otherness’ (256c5–10: ‘other’ and ‘not other’). My hope is that this will
make the theory clearer; but have I perhaps made it appear only more paradoxical even than it is?
Please note: when I write, in these final pages, of the form as ‘instantiated’, an expression com-
monly reserved for concepts or universals, I do no more than refer to the distinction drawn in the
Phaedo (102a10–103c6) between the form ‘in nature’ (‘largeness itself ’) and the form ‘in us’ (large-
ness ‘instantiated’ in Simmias or in Phaedo).
246 Denis O’Brien
whatever is beautiful participates in the form or ‘nature’ of the beauti-
ful. In the Sophist, the same is true of being: whatever ‘is’ participates in
the ‘nature’ of being. Today’s philosopher may complain that this is to
treat ‘being’ no less than ‘beauty’ as an attribute, and therefore as what
Kant would call a ‘real’ predicate. So be it. Criticising Plato in the light of
later philosophy may, or possibly may not, prove a useful exercise (largely
depending on how it is done); in either case, we have to start by working
out what Plato has to say, and in the Sophist the Stranger is undoubtedly
made to say that movement ‘is’ because it participates in being.
The difficulty Plato intends to solve, since this is not, or at least not
primarily, an aporetic dialogue, arises with the use of the negation. If the
negation bears directly on an existential use of the verb, then the meaning
can be only that there is no participation in being, and Plato shares with
Parmenides the robust conviction that this is no meaning at all. What
doesn’t exist, doesn’t exist: there is nothing there to think of or to speak
of. To explain the negation falling on a copulative use of the verb, the
Stranger introduces his new theory of ‘parts’ of otherness. Whatever ‘is
not beautiful’ participates in a part of otherness opposed to beauty, a ‘part’
that ‘is’, since otherness and all its parts participate in being, but that can-
not be specified except as a negation: ‘non-beauty’ or ‘the non-beautiful’.
So far so good. But the theory has a less obvious use as a corollary to
the theory of forms existing independently of whatever objects participate
in the form. Whatever is beautiful, revealing therefore to the eye of the
Platonist its participation in a form of beauty, cannot be identical to the
form in which it participates, since in that case there would be no form
existing independently of its instantiation. The part of otherness opposed
to beauty is therefore called upon to establish a distinction between the
form and its many particulars, insofar as whatever participates in beauty
has also to participate in otherness in relation to beauty, and is therefore,
in that special sense, ‘not beautiful’.
This extension of the theory may well appear counter-intuitive. The
beautiful young girl sitting opposite me in the metro very obviously par-
ticipates in the form of beauty. It requires a firm effort on the part of
the convinced Platonist to recognise that, in some abstruse sense, she is
‘not beautiful’, insofar as she ‘is not’ the form in which she so very obvi-
ously participates. But so it has to be. Participation in a form has to be
accompanied by participation in otherness in relation to the form, and
therefore has to be accompanied by a use of the negation that is obviously
not a denial of participation (all the more obviously so, now that she has
noticed my admiring glance and sent me a watery smile).
So much for ‘beauty’. ‘Being’ is trickier. For the Stranger’s theory to
work, negation, in order to produce the negative form, has to bear on
‘being’, but not on ‘being something’ and not directly on the existential
use of the verb. Not directly on the existential use of the verb, since in that
case we are taken straight back to ‘unknowable’ and ‘unspeakable’ noth-
ingness. But not on the copulative use of the verb taken together with its
complement (‘is something’), for the complement has its own negation:
whatever ‘is not x’ partakes in otherness in relation to x.
The negation has to bear on the verb independently of its use as a
copula, but not as a straightforward denial of its existential meaning. Yes,
that is tricky. But it is something that Plato’s reader has to think his way
around and into, if he is to succeed in picking his way through the argu-
ment of the dialogue. Movement ‘is’ because it participates in being, ‘is
not’ because it is not identical to being, and therefore has to participate in
otherness in relation to being, the negation indicating lack of identity, as
it does for participation in any other form, including beauty. But with the
added paradox that, since ‘being’, unlike ‘beauty’, is universally partici-
pated, the very part of otherness that is opposed to being has to participate
in being, so providing the Stranger with the open sesame that he needs for
his reply to Parmenides, a ‘form of non-being’ that ‘turns out to be’.
It is true that, in this brief summary, I have strayed from the detail of
the text in taking ‘beautiful’ and ‘not beautiful’ as my initial example of
negative and positive predication. The Stranger’s own examples are taken
from forms that are universally participated (sameness, otherness), with
the result that whatever is ‘not the same’ or ‘not other’ has also to be ‘the
same’ and to be ‘other’. Unless it is also participated universally (which
seems unlikely in Plato’s world), that will not be true of beauty. Whatever
participates in beauty has to participate in otherness in relation to beauty
and is in that sense ‘not beautiful’. But for a form that is not participated
universally, the converse is not true: the negation may point, not only to
lack of identity, but also to absence of participation, as it does when move-
ment is said to be ‘altogether other’ than rest. So it is that we arrive at the
form of the paradox set out above: whatever is ‘beautiful’ has to be ‘not
beautiful’; but what is ‘not beautiful’ does not have to be ‘beautiful’. The
other people I can see in the carriage are plain, if not ugly: ‘not beautiful’
without being ‘beautiful’. Only the young girl, paradoxically, is both.
The paradox, in that form, does not apply to ‘being’ and to ‘non-being’,
since both ‘being’ and ‘non-being’ are forms universally participated.
Everyone in the carriage both ‘is’ and, in the Stranger’s special sense,
‘is not’. They ‘are’ because they participate in being, and they ‘are not’
248 Denis O’Brien
because they participate in otherness in relation to being. In either case,
they participate in a form, but the two forms, ‘being’ and ‘non-being’, are
not, and cannot be, symmetrical. ‘Being’ cannot participate in ‘non-being’
since it cannot ‘be’ both itself and ‘not itself ’. But the form of ‘non-being’
cannot but ‘be’, since as a ‘part’ of otherness, even if it is the very part of
otherness opposed to being, it cannot but participate in being. If it didn’t,
it would ‘be’ – sit venia verbo – sheer nothingness, not a form at all.
Veer to one side or another of that narrow line and you end up in one or
other of the errors portrayed in the concluding pages of this essay. Identify
the form of non-being with a straightforward negation of the existential
meaning of the verb, and the Stranger will end up asserting, of ‘what is
not in any way at all’, that it ‘is’ (Notomi’s error). Identify the form of
non-being with a negation of the copulative use of the verb joined to any
and every complement, so that ‘non-being’ is so because it is ‘other than’
and therefore ‘is not’ any one of all the vast variety of different forms that
participate in being, and you will end up asserting, of ‘being itself ’, that
it is ‘non-being’ (Owen’s error). Start from Plato’s own assumption that
an existential use of einai has to be subjected to the same analysis as ‘is
the same’ or ‘is beautiful’, with one specific part of otherness, and only
one, opposed to ‘being’, whether to the form or to the instantiation of the
form, while at the same time taking into account the different extension
of forms that are, and forms that are not, participated universally, and you
will, if you pay close attention to both syntax and argument, avoid both
errors. You may even come within shouting distance of the essentials of
Plato’s reply to Parmenides.
ch apter th i rteen
Truth and story in the Timaeus-Critias

Sarah Broadie
Within the dialogue Timaeus, Plato sets out his theory of how the world
of nature, including the human race, came into being in accordance with
a divine rational plan. Strangely enough, this rational cosmology comes
packaged with a fantastic legend of which Plato is also the author: this is
the myth of hyper-archaic Athens and Atlantis. The cosmology and the
Athens-Atlantis myth are combined in a highly integrated way, for the text
presents them as two stages of one vast story. The centrepiece of the sin-
gle story is (or is to be) a narrative about the great confrontation between
hyper-archaic Athens and Atlantis, and the cosmology relates to this by way
of preface. First the character called ‘Timaeus’, an astronomer from Italian
Locris, narrates in great scientific detail the creation of the universe down
to and including the creation of man. Then the character called ‘Critias’,
an Athenian, takes over and (in the companion dialogue, Critias) begins
to narrate the Athens-Atlantis tale in great historical detail (Ti. 27a2–b9;
Criti. 106b6–7). It is a tale of epic events which supposedly took place at
the dawn of human time, soon after the ordered universe and its contents
had as it were settled down from having been divinely created.1
This narrative continuity between the monologue of Timaeus and that
of Critias sets up a puzzle about their status as fiction or non-fiction.
Given the continuity, one would expect that on some level or from some
perspective this status would be the same for both. Now, we are bound
to assume that Plato meant the cosmology as a serious set of proposals
about the natural world and its origin, and not as an essay in fiction.2
It is a great pleasure to have this opportunity to help celebrate Christopher Rowe’s achievements,
especially in the field of Plato studies, not only as a scholar but also as a leader in promoting ancient
philosophy to students at every level and round the world.
1
For important discussions of different aspects of Plato’s Atlantis story, see Vidal-Naquet 1986; Gill
1977, 1979b, 1980, 1993; Pradeau 1997; Morgan 1998; Rowe 1999; Johansen 2004; Nesselrath 2006.
2
I put on one side the question whether Plato meant the quasi-temporal aspects of the cosmopoiesis
as a sort of fiction. The picture of the resulting universe is plainly not fictional; nor is the general
249
250 Sarah Broadie
Certainly Plato’s contemporaries and followers took it as offering a picture
of the physical world in which we live: they criticised and in some cases
extended it on the basis of what they took the physical and metaphysical
realities to be. On the other hand, the Athens-Atlantis story is fiction,
and clearly intended as such. It is or should be plain to us today that it is
fictional, and there is compelling reason to think that Plato’s original audi-
ence too would have recognised it as such without ado.3 It follows that
the perspective from which both monologues get the same answer to the
question ‘Non-fictional or fictional?’ is not that of Plato and his real-world
contemporaries and followers; nor is it that of real-world modern schol-
ars and historians. The only perspective from which a uniform answer is
possible is that of the dialogue-world itself or its characters. Of these we
have already mentioned Timaeus and Critias; the others are Socrates and
Hermocrates.
The uniform status they assign is positive. All four solemnly treat or
accept both the monologues, that is, both parts of the one great story,
as non-fictional or veracious.4 And in the dialogue-world these charac-
ters certainly do not stand for foolishness and gullibility. On the contrary,
they are presented as highly cultivated and intellectually weighty.5 So what
are we to make of this confection of Plato’s, this weaving together of the
obviously fictitious story of Athens and Atlantis with a cosmology that is
premise that this world is as perfect as any physical system can be, and is the product of divine
intelligence.
3
Consider the fact that in the Athens-Atlantis story the heroic feat of Athens is remarkably similar
to that whereby historical Athens fended off the Persians at Marathon in 490 BC. This strongly sug-
gests that, in the Timaeus-Critias dialogue-world, what we think of as the early fifth-century battle
of Marathon has never taken place; this is because in any given Athenian history, real or fictional,
there would be, so to speak, room for only one event having that extraordinary profile. Since, in
the history that features Atlantis, the event in question is supposed to have occurred nine thousand
years before Solon (who lived c.640–c.560 BC), Plato’s contemporary real-world readers would have
known straight off that the story featuring Atlantis is a fiction. By the same token, they would have
recognised straight off that the dialogue-world in which this story is embedded is fictional too.
These and other questions relevant to this paper are discussed in detail in Broadie 2012, esp. ch. 5.
4
I here use ‘veracious’ as the contrary of ‘fictional’, although strictly the contrary is ‘would-be vera-
cious’. In my view, none of the characters utters a word challenging the veracity, in the above sense,
of either monologue. (This position has to be defended in connection with Ti. 26e4–5, on which see
more in the main text below.) At 20d8, Critias speaks of the Atlantis story, including the associated
account of its transmission, as ‘true’ (alēthēs); cf. 21d8 and 26d1. At 26e4 Socrates calls it an alēthinos
logos, contrasting this with a plastheis muthos (‘made up story’). At 20c6–d1, Hermocrates speaks as if
he accepts the Atlantis story precisely as Critias offers it. It seems clear that Timaeus too is on board
with the others on this. At Criti. 106b8–108a4, Critias complains that it is much easier to criticise
the historian than the cosmologist because the former’s subject-matter is so much more familiar, but
this in no way suggests a difference in respect of would-be veracity.
5
The Socrates-character – whose own gravitas seems impeccable – vouches for the intellectual stature
of Timaeus, Critias and Hermocrates at Ti. 19e8–20b1.
Truth and story in the Timaeus-Critias 251
obviously not intended as mere fiction, the resultant complex being, as
a whole, received as serious and non-fictional by characters depicted as
themselves serious, sane, intelligent and well educated?
Here is one possibility: Plato is sending his real-world audience the
message that they should trust their own judgements on what in his
dialogues ought to be accepted as mere fiction and what ought to be
accepted as intended truth. That is, readers are to trust their own judge-
ments as distinct from following the lead of any personage in a Platonic
dialogue-world. In that of the Timaeus-Critias the characters appear as
trustworthy, even authoritative. Perhaps this is more clearly true of two
of them, Timaeus and Socrates, than it is of the others, Critias and the
mostly silent Hermocrates. But any such difference hardly matters. For
these characters present a united front in accepting with equal serious-
ness both the blatantly fictional story voiced by Critias and the scientifi-
cally intended cosmology voiced by Timaeus.6 The characters seem to be
collectively oblivious to any difference in veraciousness. They are shown,
therefore, as not providing reliable guidance on the question ‘fiction or
non-fiction?’ to anyone in the world of Plato and Plato’s readers. This con-
clusion takes on additional significance given that one of them is Socrates,
represented here as a master of political philosophy. Perhaps, then, Plato
is telling us here that in any given dialogue we should judge for ourselves
and not automatically tend to favour a position or stance simply because,
in that dialogue, Socrates is shown endorsing it.
Well, this may be one thing that Plato wants to do, but the
Timaeus-Critias raises other questions too about truth and story-telling.
To pursue some of these a few more stage-setting details are necessary.
More must be said about the characters, more about the Atlantis tale, and
more about how this tale comes to be told in the dialogue-world. Only
then shall we be in a position to form any interesting hypothesis on why
Plato chose to create a dialogue-world in which the Atlantis tale is told.
So first the tale: fiction in the real world, true history in the dialogue.
Countless people over the centuries have been fascinated, credulously or
not, by the mythical entity Atlantis.7 In fact, of course, the story has two
mythical protagonists. The other one, at least as riveting for Plato’s con-
temporary readership, is hyper-archaic Athens, an Athens which at the
beginning of human history had been personally founded by Athena,
6
Although, as many commentators have pointed out, some elements of the cosmology seem plainly
more ‘mythical’ (whatever exactly we mean by that) than others, much of its material falls straight-
forwardly under the headings of astronomy, chemistry, anatomy, physiology and so on.
7
For an illuminating short history of the reception of the story, see Vidal-Naquet 2007.
252 Sarah Broadie
the goddess of wisdom (Ti. 24c4–d3, Criti. 109c4–d2). In the Timaeus a
preview of the story (Plato never got round to finishing the full version)
sketches how, nine thousand years before Solon, the sea-power of Atlantis,
an enormous island west of Gibraltar, had taken over North Africa and a
large part of Europe and then launched a campaign to subjugate all the
peoples of the Mediterranean; how Athens, deserted by her allies, stood
alone against the forces of Atlantis, defeated them in battle, and thereby
saved the Greek world from enslavement; how, soon afterwards, earth-
quakes and cataclysms overwhelmed the ancient Athenians and totally
submerged Atlantis, so that all record and remembrance of what had hap-
pened was lost from the Greek world; how a record nonetheless did sur-
vive in Egyptian temple archives for the subsequent millennia; how the
Egyptian priests revealed the story to Solon on his foreign travels; and
how from Solon it passed to the present Critias’ great-grandfather and so
through the generations to this Critias himself (Ti. 20d7–25e2).
A little more now about the names of the characters and the dramatic
setting. Virtually the only thing we are shown about the setting is that
the discourse takes place in Athens during a festival of Athena.8 As for the
names of the four characters: two, at least, carry lamentable associations
which Plato must surely have meant to evoke. The use of a ‘Critias’ and
a ‘Hermocrates’ must have been calculated to remind early audiences of
two of the most humiliating episodes of Athenian history in the late fifth
century. ‘Hermocrates’ was the name of the Syracusan general who led the
victorious campaign against the disastrously ill-judged Athenian exped-
ition to Sicily, launched in 415 bc. ‘Critias’ was a name in Plato’s mother’s
family, passed on from grandfather to grandson over many generations.
The name would have brought to mind, in particular, the Critias who
was an older cousin of Plato’s – a literary figure, a philosopher and at
one time an associate of Socrates (he appears in the Charmides and the
Protagoras). This Critias led the short-lived but extremely brutal oligarchic
regime known as the Thirty Tyrants, which took over at Athens in 404
bc when Plato would have been in his early twenties. This Critias died
in the fighting that toppled the Thirty, and for long afterwards under the
restored democracy his name was amongst the most loathed and despised
in Athens. The opprobrium lingered in strength for decades, well down
into the middle of the fourth century when Plato in his seventies would
have been writing the Timaeus-Critias. It is enough for the present purpose
8
On the festival, see Ti. 21a1–3 and 26e3. That Athens is the location is obvious from several touches,
but see especially 27b3.
to suppose that the names ‘Hermocrates’ and ‘Critias’ are meant to bring
to mind those miserable episodes and the individuals just mentioned,
whether or not it also makes sense to suppose they meant to refer to those
individuals.9
By contrast, the sketch of the hyper-archaic Athens facing and repuls-
ing the encroaching imperialism of Atlantis is of course meant to remind
Plato’s audience of what was traditionally regarded as the finest hour of
real-life Athenian history: the moment in 490 BC when Athens almost
completely unaided stood up to and defeated the Persian invaders at
Marathon. It was from this moment that many historians have traced the
start of the rise of Athens to its brilliant fifth-century pre-eminence – a
pre-eminence easy with hindsight to associate with the over-ambitious
ventures that contributed to the later humiliations already mentioned.
Given all this, as many scholars have noted, it makes sense to look upon
Plato’s Atlantis story as more than anything a story about Athens – directly
about a mythical proto-Athens, obliquely about historical Athens of the
comparatively recent past – combining in flashes of association her proud-
est with her most shameful historical moments.10 Accordingly, it makes
sense to focus particularly, as I shall now do, on the two Athenian char-
acters of the dialogue-world: the one called ‘Socrates’ and the one called
‘Critias’.
Since, as I have suggested, the name of the one called Critias must at
least be meant to evoke the historical Critias who was a philosophical
friend of the historical Socrates, it is relevant to mention how this rela-
tionship ended. Xenophon tells us that this Critias fell out with Socrates
and from then on nursed deep enmity against him (Mem. 1.2.29–31). We
gather from Plato’s Apology that later on, when Critias’ junta had come
to power, they ordered Socrates to take part in an illegal arrest which was
probably to lead to the summary execution of the citizen concerned: they
wanted this person out of the way because he was rich and they aimed
to take possession of his wealth.11 No doubt Critias would have been
9
There has been a long-running debate on the identity of the Critias-character: see e.g. Taylor 1928:
23–5; Cornford, 1935: 1; Lampert and Planeaux 1998; Nails 2002: 106–13 (s.vv. ‘Critias III’ and
‘Critias IV’); Nesselrath, 2006: 43–50. For a discussion from quite a different standpoint of the
significance of Plato’s reference to Critias, see Tulli in this volume (‘The family of Plato and the
enquiry of Socrates’).
10
According to the hypothesis of Giuseppe Bartoli, revived in the twentieth century by Vidal-Naquet
1986, historical Athens is represented by both sides in the confrontation with Atlantis: hyper-archaic
Athens stands for Marathonian Athens versus Persia whereas Atlantis stands for the subsequent
imperialist Athens versus the states she made or aimed to make her tributaries.
11
Pl., Ap. 32c4–d8; for references in other ancient authors see Burnet 1924: 135–7.
254 Sarah Broadie
delighted to involve Socrates in this crime, but Socrates refused to have
anything to do with it – a stance that could have cost him his life at the
time. It is ironical that, despite this disassociation of Socrates from Critias,
the odium surrounding the name of Critias rubbed off also (in many peo-
ple’s minds) on Socrates too. This must have contributed something to the
political dynamics that got Socrates indicted and condemned for impiety
in 399 bc. In fact, more than fifty years after his execution for this sup-
posed crime public speeches were still being made in Athens pointing the
finger at Socrates as the sophist who had nurtured traitors and tyrants
such as Critias.12
If Plato’s choice of names for the characters of the Timaeus-Critias is
meant to bring back such grim associations, there is a distinct irony in
the way in which Plato has brought these characters together in his
dialogue-world. The situation is this. The proceedings begin with Timaeus,
Critias, Hermocrates and Socrates assembling one day on from ‘yesterday’.
Yesterday, it emerges, Socrates had entertained the other three with a pol-
itical discourse which is now summarised; the summary makes clear that
Socrates had been presenting an ideal polity whose institutions correspond
in quite close detail to institutions of the ideal state expounded in Plato’s
Republic (Timaeus 17c1–19b2).13 ‘Today’ it falls to the other three to recipro-
cate: collectively they are to host a ‘return-feast’ of discourse for the enter-
tainment of Socrates.14 So, for example, Socrates says to them (20b7–c3):15
After considering the matter together among yourselves, you agreed to
recompense me today with hospitable offerings of discourse (ta tōn logōn
xenia); so here I am adorned for the occasion and as ready as anyone could
be for what I am to receive.
And on being told their precise plan for the return-feast, he says (27b7–8):
‘It looks as if it will be unstinting and brilliant, the feast of discourse (tēn
tōn logōn hestiasin) which I am about to get in return.’ The main dish of
the return feast is to be Critias’ story of Athens and Atlantis. Critias gives
Socrates a preview of it just to make sure that the full version, intended
as an immensely detailed narrative tapestry, will be acceptable to him.
However, the run-up to the full version will consist in a cosmology, the
12
See Aeschines, Against Timarchus 173. This speech was given in 345 BC.
13
However, internal evidence shows that this discourse is neither the speech by Socrates that
constitutes the Republic, nor the discourse reported therein which he held the day before with
Thrasymachus, Glaucon, Adeimantus, etc. See Cornford 1935: 4–5.
14
This theme is established at the very beginning of the Timaeus, when Socrates greets the others as
‘yesterday’s guests, today’s hosts’ (17a2–3).
15
Translations are my own, based on the text in Burnet 1900–1907: vol. 4.
responsibility of Timaeus. Thus the cosmology is presented, anyway
initially, as a preliminary to the major component of the return-feast.
By the same token, Timaeus figures as an ancillary contributor to the
return-feast, the main contributor being Critias with his long version of
the Athens-Atlantis story.16 The situation, in sum, is this: Socrates with
his political discourse of ‘yesterday’ is the raison d’être of all that is offered
to Socrates ‘today’; and the figure centrally responsible for today’s collec-
tive homage is the Critias-character. These are the principals in this play:
Socrates and the Critias-character.
The exchanges that establish this scenario are conducted in a tone of
highest ceremonial courtesy, with the themes of gratitude, recognisance,
reciprocity and recompense repeatedly to the fore. Among the various
things all this is meant to convey is surely a bitter reminder of the manner
in which, in real history, Plato’s older cousin Critias interpreted his debt
to Socrates; and of how, in real history, what ‘returned’ to Socrates from
Critias was the poisonous backwash of the latter’s deserved disgrace.
After more than two thousand words of historical description, including
a good measure of pseudo-history, the reader may be wondering whether
there will be some philosophy in this discussion. There may be material
for moralising about particularities of late fifth- and early fourth-century
Athenian history; but where is the philosophy? Well, the main point,
which I now begin to approach, is that by means of this strange and fan-
tastic scenario Plato brings his spotlight to bear on the question of why
after all it matters whether an account is true or on the mark, rather
than false or off the mark. This is not precisely a question about the con-
trast ‘veracious/fictional’ which has concerned us up to now. It is rather
the question of why it matters whether a would-be veracious – that is,
non-fictional – account is or is not true. No one will dispute that this is a
philosophical question.17
16
Notoriously, the long version breaks off in mid-sentence before the action between Atlantis and
Athens really gets going. If Plato had told the entire story of this in the same detail as he tells the
preliminaries in Criti. 108e1–121c5 (cf. Ti. 26c5–7), the long version might have been two or three
times the size of the cosmology. Again according to initial appearances, Critias’ long version was
to be followed by a contribution from Hermocrates (Criti. 108a5–c6), the nature of which is never
made clear. Thus if we take the plan at face value, the size of the cosmology would have been
dwarfed by that of the completed whole. With most scholars, I assume that Plato did originally
plan a complete three-part whole but was forced, or decided, to abandon the project. Some, how-
ever, think that he intended all along to break off the Critias ‘early’. On this hypothesis the cosmol-
ogy, as it turns out, would correspond to a much larger proportion of the designed-as-a-fragment
return-feast, and Timaeus would be the main rather than an ancillary contributor. On the ‘unfin-
ishedness’ question, see Nesselrath 2006: 34–41.
17
It is explored comprehensively and deeply by Bernard Williams 2002.
256 Sarah Broadie
It is, however, a question of such vast generality that a shrewd philoso-
pher might suspect that a correspondingly general answer is doomed to
be either skewed or vacuous. Would it not be more fruitful to lay aside
such an extraordinarily abstract question as the value of truth as such,
and instead turn to examining one or another specific human activity
where the question of truth might be expected to arise? We could call such
activities truth-relevant practices. In my view, Plato in the Timaeus-Critias
is concerned with two truth-relevant practices: that of recounting some
glorious episode in our country’s past, and that of laying out some kind of
ethical or political ideal. In concrete reality these two practices are some-
times conflated, and the potentiality for conflating them is something that
Plato helps us to recognise by the way in which he stages Critias’ telling of
the Athens-Atlantis story.18 What Plato is aiming at (there are, of course,
other major things he is aiming at as well, but I shall not be concerned
with them on this occasion) is, first, to bring out the difference between
the practices just mentioned and, secondly, to confront in this context the
question of why truth should matter. The confrontation is effected in and
through the Critias-character, who is constructed as a contradiction that is
unaware of itself. On the one hand – and this he is certainly aware of – he
behaves as if it really matters to him that the story he tells is true; and on
the other hand he unknowingly behaves and has been behaving (in the
world of the dialogue) as if its truth makes no difference.
So I shall argue; but first we need some further literary details. These
have to do with a connection between the characters of Socrates and
Critias which has not yet come properly into view. These characters are
not only linked by being the two Athenians in the dialogue-world, and
by bringing to mind a comradeship that turned to alienation between the
pair of historical persons of whom their names remind us. The characters
are also connected by the fact that they represent a shared focus on an
identical object. What is this object? Well, it is the city of perfect institu-
tions and pre-eminently good citizens which Socrates expounded in his
ethical-political discourse ‘yesterday’; and at the same time it is also the
wonderful Athens which in Critias’ story defended Greece against the
18
Given the Marathonian profile of the Athens-Atlantis confrontation, Plato must amongst other
things be sending a message to any contemporary Athenians inclined still to bask in (or batten on)
the victory of 490 BC; cf. the surely satirical replay of this at Menexenus 240c–e. The Timaeus-Critias
version carries the added sting that even if the real Athens of Marathon had enjoyed – which
of course she did not – a philosophically ideal constitution, continuing nostalgic obsession with
that achievement could be no substitute for real political analysis of the needs and possibilities of
Athens in the mid-fourth century.
aggression of Atlantis nine thousand years ago. One and the same entity
is accessed by the two Athenian characters under different guises or from
different perspectives, and the difference between the perspectives is what
is philosophically momentous. Getting his readers to identify and reflect
on this difference was, I believe, one of Plato’s great purposes (even if not
his only one) in composing the whole scenario.
The nub of the scenario is a remarkable coincidence. Socrates has let
the others know that the recompense for yesterday’s discourse that would
really satisfy him would be a story showing the citizens of his ideal polis
rising to meet a great military crisis with exploits of warlike courage and
wise diplomacy. Socrates says that he lacks the power to provide himself
with such a particularised work of the imagination. His own competence
is for a priori philosophising about the basic structure of the civic ideal;
but a quite different sort of ability is needed to tip this structure over the
brink of abstract necessity into the contingency of some particular series of
supposed events so as to show in concrete detail how ideal citizens would
bear up ethically in face of the demands and opportunities generated by
a great emergency. Such a task, says Socrates, is also beyond the capacities
of poets and sophists. By contrast, he says, his three companions on the
present occasion are between them well qualified to provide him with the
sort of spectacle in words for which his soul craves. And indeed they have
undertaken to meet his wish, he says, so that he now sits back in happy
expectation (Ti. 19b3–20c3). Now, it just so happened yesterday, when
Socrates was expounding his abstract ideal, that Critias (we are told) was
reminded of an already existing story about real events – a story that he
heard long ago as a child from his grandfather Critias, and one that won-
derfully fits the bill. This is the story of ancient Athens and Atlantis. Why
was Critias reminded of this family-transmitted story? Because it depicts
ancient Athens as having had certain specific institutions, and those insti-
tutions pretty well match the ones proposed in the Socratic ideal devel-
oped yesterday (25e2–5).
On this basis, Critias says words to this effect: ‘So, Socrates, we’ll postu-
late an identity between (on the one hand) the citizens of your own purely
hypothetical ideal19 and (on the other hand) the real, historical, Athenians,
our own forefathers, who long ago fought Atlantis [see 26c7–d5]. Here
is the plan for recompensing you: Timaeus will provide a cosmogony
whose last phase will explain basic human nature; you, Socrates, have
already yesterday given us the theory of the best city and the education
19
This is meant to paraphrase ὡς ἐν μύθῳ διῄεισθα σύ (26c8) and οὓς διενοοῦ (26d2).
258 Sarah Broadie
of its citizens; and it is for me, Critias, the next speaker along, to receive
from Timaeus human beings, then focus on those whom I receive from
you, Socrates, namely a group educated in the ideal city, and to confer on
these educational products of yours the status of citizens of this Athens of
ours (bringing them before our assembled company for that purpose as if
before a law court), by identifying them with those Athenians of long ago
[see 27a2–b6].’
This identity between the Socratic ideal citizens and the citizens of
Critias’ ancient Athens is not presented as an independently given fact.
Instead, the identity is conferred by Critias’ postulate, stipulation or proc-
lamation,20 this being ratified by the others’ quiet acceptance. But Critias
stipulates in the light of independently given resemblances between the
two sides of the identity. On the one hand, as we have seen, it is a given
that the Socratic polis has certain institutions and that Critias’ ancient
Athens has similar ones. And on the other hand it is a given that each
is superlatively excellent. We know this in the case of the Socratic polis
because it has been constructed as an ideal (see 17c2–3) by a master of such
constructions. In the case of ancient Athens, the fact of this city’s extraor-
dinary excellence is simply part of the record inherited by Critias, and this
whole record passes for authoritative with everyone in the dialogue-world.
Just as they all accept as given (by contrast with stipulated) truth the
Critian story’s description (this is how we might classify it) of the ancient
Athenian institutions, so they accept as given its evaluation (as we might
distinctly classify it) of that city and its conduct. And this evaluation is a
paean of praise unqualified. According to it, the ancient Athenians were
the finest and best human beings in the world (23b7–8), and the city was
‘best in war and in every respect superlatively well governed beyond all
others (eunomōtatē diapherontōs). Its deeds are said to have been finest,
and finest its civic institutions, of all the cities beneath the heavens of
which we have heard tell’ (23c5–d1).21 Athena the founder chose for it a
place whose climate would bring forth men of superlative intelligence
(phronimōtatous andras): ‘She, the goddess, being a lover of war and
a lover of wisdom (philopolemos te kai philosophos), chose the place that
would bring forth men most similar to her’ (24c4–d3). Being better gov-
erned than even their observance of the goddess’ laws would suggest, the
ancient Athenians ‘excelled over all mankind in every virtue, as one would
20
See θήσομεν, 26d1, and ποιῆσαι πολίτας, 27b2–3.
21
Critias is reporting the Egyptian priest’s address to Solon. The Egyptian priests have kept records of
great deeds carried out in all lands, not just in the Greek and Egyptian territories (23a1–5).
expect from those brought to birth and educated by gods’ (24d3–6).22 The
records tell of many great and wonderful Athenian deeds, but there is
one – the victory over Atlantis – that ‘surpasses all in magnitude and in
virtue’ (24d6–e1). This was the city’s greatest deed and would have deserv-
edly been its most renowned one had the news of it not been lost to the
Greek world (21d4–7).23
As I have said, the characters are united in accepting the story as true
(see n. 4). What is meant by ‘true’ here? I think that we should under-
stand it in the ordinary straightforward factual way as meaning that things
actually were as the story says they were. In other words, from the charac-
ters’ perspective, the things recounted in the story did not happen merely
mythically ‘once upon a time’: they happened at places belonging in the
same spatial system, and at times belonging on the same time-line, as the
place and the time of the conversation between the characters themselves.
Ancient Athens, from the characters’ perspective, was where the Athens is
in which the four friends are meeting, and the ancient war with Atlantis
stands in a definite temporal relation to events belonging to the real his-
tory of this present-day Athens, for instance, the career of Solon.24
One could ask whether we are meant to take it that in the fictional
world of the Timaeus-Critias the Athens-Atlantis story is true, as distinct
from being merely accepted as true by the characters. But this seems to be
a distinction without a difference, given the absence from that world of
any more authoritative perspective. More authoritative would have been
a surviving first-hand record from an observer who for a few decades had
watched life round the Mediterranean nine thousand years before Solon:
or – better still – a surviving first-hand record from someone who actu-
ally took part in the Athens-Atlantis episode. If such a perspective were
somehow inserted into the Timaeus-Critias dialogue-world, then what
it conveyed, as stemming from a witness’s own experience in the distant
22
At 24d3 the priest in Critias’ account flatteringly switches to the second person plural, thus identi-
fying Solon and his fellow countrymen with the original Athenians.
23
See also Criti. 112e4–6, which describes the ancient Athenians as possessing ‘beautiful qualities
of body and every sort of excellence of soul’, for which they ‘were famous through the whole of
Europe and Asia and had the greatest renown of anyone in that period’; cf. Ti. 25b5–6.
24
In the dialogue-world the events involving proto-Athens and Atlantis belong to what Williams 2002
calls the objective as opposed to the mythical past. Treating a past-tensed statement as factually or
historically true is bound up with assuming that what it reports (a) occurred at a definite (whether
or not known) time-interval from the present, and (b) was itself once present (Williams 2002:
53–7 and 148–71). It emerges in the next paragraph that Plato has constructed the Timaeus-Critias
dialogue-world in a way that invites trying to drive a wedge between (a) and (b). On the stress
on the factual, quasi-historical dimension of the story, see also Tulli in this volume (‘The story of
Atlantis and the facts of the past’).
260 Sarah Broadie
past, might falsify within that world the priestly story that Critias and
the others would accept thousands of years afterwards. But it seems to
be fundamental to the scenario as Plato has actually constructed it that
none, or anyway none of the human, cognitive standpoints therein – and
here we include those of the priests and Solon and Critias’ ancestors – is
the standpoint of a first-hand witness of the happening – or, alternatively,
non-happening – of the events portrayed in the Athens-Atlantis story.
Essentially, the only access anyone has to those events is through a record
or report about the past handed on by someone else. And this, I believe,
allows room for the supposition (we are dealing, after all, with a fictional
world) that, even though Critias and his friends locate those events at a
definite temporal interval – nine thousand years plus a few – from their
own present, the events have never themselves been present! Yet how could
they ever have happened without at least then being present? How could
they happen without being present to the heroes and villains at the centre
of them? How, unless they were once present, could they give rise to this
chain of true, but never first-hand, testimony about themselves? These are
just some of the questions that we may expect to see stirred up when a
philosopher starts to circle round the notion of truth.
At any rate, I shall take it that in the dialogue-world the Athens-Atlantis
story is factually true and not just accepted as such by the characters. But
are the characters justified in believing it? Well, the account of the story’s
transmission is fantastic by real standards. On the one hand, natural disas-
ters ensured that very soon after the great events happened remembrance
of them was erased from the Greek world. On the other hand, divine inter-
vention plus the unique natural conditions of Egypt ensured that there
the record of them was preserved intact. ‘Divine intervention’, because a
thousand years after the great events the goddess Athena founded a city,
Saïs, at the head of the Nile delta, creating for it institutions like those
she designed for proto-Athens. Whereas Saïs, it seems, was never involved
in its own version of the Athens-Atlantis confrontation – such a thing
would be unthinkable in Egypt – it possessed a sort of substitute: a writ-
ten account of that original confrontation was among the records guarded
by the priests of Saïs. Reading between the lines, we can only suppose that
Athena was somehow the source of that account and the agent of its pres-
ence in Saïs. For eight thousand years, Greek total ignorance and Egyptian
divinely provided knowledge of the Athens-Atlantis war existed as it were
in parallel universes. That ended only when the Athenian Solon, ‘wisest
of the Seven Sages’ (20d8), visited Saïs and triggered the priestly revela-
tion to him of the depth of his and his fellow countrymen’s ignorance
concerning the magnificent character and superlative achievement of their
very own ancestors. Solon then brought the story back to its home, to
Athens, where he told it to a kinsman of his, the great-grandfather of the
Critias now telling the tale to Socrates (20e7–25d6).
We might want to explain as follows the fact that our four rather august
characters accept this ludicrous transmission-story without the slightest
sign of doubt: ‘Well, in the dialogue-world epistemic standards are so low,
or at any rate so different from real-world standards, that serious, edu-
cated, men – such as Timaeus in particular, who is said by Socrates to have
held the highest political positions in Locris and to be a consummate phi-
losopher (20a1–5) – make what, in that world, counts as a sound and reli-
able judgement when they accept without demur the transmission-story.’
However, if we attribute to Timaeus such a stunted epistemic standard in
connection with the Athens-Atlantis tale and its transmission, then we
have to be prepared to show how we can consistently expect him to aspire
to a scientific standard deserving our respect (even if it is not the most
demanding possible epistemic standard) when he comes to present his
cosmology.25
A way out of this particular difficulty may open up if we consider that
two people may be equally willing to accept that some account is factually
true, and yet have very different attitudes on whether, or how much, it
matters that it is true. If the question of truth is unimportant to me, it
may not be irrational of me to accept the account as true even though
I am aware that the grounds for such acceptance are epistemically very
poor. If the importance to me of truth varies with the nature of the mater-
ial, then the epistemic standards I apply may reasonably be much lower
for some areas than for others.26
25
In the methodological preamble to the cosmology, Timaeus says that his listeners should expect
from him logoi that are eikotes (‘likely’) or at any rate as eikotes as anyone else’s on the various cos-
mological topics: they should not expect logoi that are irrefutable and invincible (the attribute of
successful demonstrations in mathematics). See Burnyeat 2005 on the meaning of to eikos in the
Timaeus. Burnyeat argues convincingly that the term, which has often been interpreted in a nega-
tive sense in this context, conveys an epistemic standard to be positively aimed at; thus for a logos
to be eikōs is an achievement. Burnyeat at no point suggests that what should count as eikos for
Timaeus when doing cosmology in the dialogue-world could seriously part company with what
should count as eikos for Plato in the real world when doing cosmology through Timaeus. For
further discussion see Broadie 2012, esp. ch. 2.
26
These sketchy remarks gesture towards the vigorous contemporary debate on epistemic contextual-
ism, on which see Rysiew 2007. The point about e.g. Timaeus is that he can be ‘intensely relaxed’
over giving in to someone’s prompt to accept as true the Atlantis story if this for him is ‘low stakes’
material, while holding himself to a high standard of justification when it comes to endorsing
cosmological proposals. The situation is slightly complicated by the fact that, in his methodological
preamble to his monologue, Timaeus contrasts ‘truth’ (alētheia) with pistis (however we understand
262 Sarah Broadie
For a brief moment Timaeus and his epistemic standards have been
to the fore, but it is time to return to our main characters, Critias and
Socrates. They and the others, in one accord without hesitation, give the
Athens-Atlantis story (and the story of its transmission) full marks for
being factually true; but, as we have noted, it need not matter to them
all to the same degree whether the marks are deserved. I am going to sug-
gest that Critias is uncritical because the truth of the tale matters to him
so much: that is to say, it matters to him so much that the tale be true.27
His identity is all bound up with the connection it affords him to the
glory of hyper-archaic Athens, and also with the connection to his more
recent ancestors effected by the inter-generational ritual of receiving and
re-telling the story – it is so much easier to be intensely serious about
receiving a story and passing it on if there are no flickers of doubt about
its truth. By contrast, Socrates is uncritical because the story’s truth is not
something that matters to him. We have noted how Plato engineers a sort
of convergence between Socrates and Critias; we shall soon see that their
shared focus places them infinitely far apart. That is what begins to emerge
from the penultimate exchange between those two before Timaeus begins
the cosmological speech.
Critias has just finished giving Socrates the preview of the Athens-Atlantis
story, and now comes out with his proposal that they convert the citizens
of the ‘purely hypothetical’ Socratic ideal into the citizens of the ancient
Athens that faced Atlantis (26c5–d5). The citizens of the Socratic ideal are
to be as it were replaced by the citizens of ancient Athens. The context
calling for this replacement is precisely Socrates’ announced desire to be
shown citizens of his own ideal immersed in heroic activity. For Socrates
to accept the replacement is ipso facto for him to receive the imagina-
tive spectacle that he hankers after: this is because the citizens of Critias’
ancient Athens come complete with a ready-made story in which they
do various particular heroic things under particular circumstances. (Those
pistis here, it is epistemically inferior to alētheia), and he lines up ‘truth’ exclusively with eternal,
intelligible being as opposed to becoming. Thus in terms of that preamble, the Atlantis story has no
hope of being ‘true’ even if things took place just as it relates. (Another difference is that, whereas
the truth predicated of the Atlantis story is – or so I claim – a relation between a logos and a fact or
set of facts, the truth Timaeus speaks of in his methodology seems to be a state of mind or a cogni-
tive condition: see Burnyeat 2005.) I suggest that just as the epistemic standard of acceptance varies
for Timaeus depending on the genre of the account he is considering, so does his use or concur-
rence with others’ use of ‘true’ and ‘truth’. On the different genres represented by Socrates, Timaeus
and Critias, and the way the differences set each up as a distinct kind of authority uncriticisable by
the others, see Broadie 2012: ch. 2, section 3.
27
The point, in effect, is that Critias wants the story to be true – not that he open-mindedly regards
the question whether it is true or not as an important one.
circumstances and the human responses to them are part and parcel of
what you access when you access Critias’ ancient Athenians, whereas
Socrates’ abstract citizens offer no purchase for any such additions.) Critias
then asks Socrates whether he welcomes this story (26d5–e1). Obviously
Socrates is going to welcome it, if only because it is just such a narrative as
he desires. In fact, however, Socrates utters more than a mere ‘Yes, please’:
he exclaims that the story cannot be bettered, and this for two reasons,
the second of which he seems to hold more lightly than the first. The first
reason is that ‘its close link to the goddess makes it splendidly appropriate
for her festival which is now taking place’ (26e3–4). Here is his second rea-
son: ‘that it is not a fabricated story but a true account is an enormously
important consideration, I suppose’ (τό τε μὴ πλασθέντα μῦθον ἀλλ’
ἀληθινὸν λόγον εἶναι πάμμεγά που, 26e4–5).
The remark carries a hint of sarcasm, emanating from πάμμεγά που
(‘is enormously important, I suppose’). The words convey a slight curl of
the lip, a sketched caricature of someone being overwhelmingly impressed
by something. Thomas Johansen has suggested that Plato hereby makes
Socrates express polite scepticism about the account’s factual truth.28 But if
Plato had intended this, he would surely have made Socrates say with the
same hint of sarcasm: ‘that it is not a fabricated story but a true account
is eminently plausible (vel sim.), I suppose.’ Alternatively, he could have
made Socrates express scepticism about its truth by casting doubt on
Critias’ authority, for instance, by exhibiting overtly insincere support for
Critias’ claim that it is factually true: ‘that you, Critias, put this story for-
ward as true is an enormously important consideration, I suppose.’ But the
actual sentence does not refer to Critias (or to anyone) as the source of the
claim. The straightforward interpretation, instead, is this: Socrates’ overtly
insincere sketch of being tremendously impressed is directed towards the
fact that the account is true (not just a fabricated story).29 In short, while
showing that he does accept it as true – straightforwardly true – he says
in effect: ‘and that’s tremendously important – I don’t think.’ His message
is: ‘Yes, the story is true, not fabricated – but that’s not something to get
excited about.’
So Socrates is shown conveying that it is unimportant that the story is a
true one. Fair enough: he had only wanted an account of noble actions by
28
Johansen 2004: 45–6.
29
It has been suggested that Socrates’ phrase μὴ πλασθέντα μῦθον ἀλλ’ ἀληθινὸν λόγον draws a
contrast with his own discourse of yesterday. If that is right we should translate πλασθέντα μῦθον
as ‘tale that is a construct’, i.e. a piece of philosophical theorising rather than a report of historical
fact. I think, however, that the contrast is with a fictional historical story.
264 Sarah Broadie
the ideal polis that would feed his imagination; for that, it did not have to
be or even claim to be about real events. Whether the story he gets today
is a slice of actual history matters to him no more today than it would
have mattered yesterday – when he was constructing his abstract ethical
and political ideal – whether a city of these specifications ever had existed
in real time and space. Constructing it was the work of fact-independent
reason. The reasoning was fact-independent because it started from value:
it started by asking about the nature of the best city-state and, from that,
given fundamental assumptions about human nature, it worked towards
description: description in empirical terms of what such a city would be
like, what its institutions would be. And since reasoning was the only
mode of access to this empirical description, the discourse must have pro-
ceeded by argued explanations. So every conclusion about the empirical
features of the city would have come on the back of an explanation of why
each feature was right or best. Socrates could only find out what is politi-
cally good by finding out why it is. And since this is Socrates, we also know
the why of his entire investigation into what is politically good and why:
its purpose will have been to uncover a standard either showing what to
aim for in real human affairs, or (if that was in practice impossible in the
Athens of his time) at least providing a framework for a critique of exist-
ing institutions. But the advancement of this great purpose, along with
the a priori reasoning by which Socrates has sought to implement it, owes
nothing to the fact, even the known fact, that a particular city meeting the
standard has once existed on earth. As Kant said of the moral law and ‘the
Holy One of the Gospel’, we cannot derive the former from contempla-
tion of the latter, since the former is the standard for recognising even the
supremely exemplary status of the latter. And therefore, whether or not
Kant meant to draw this implication, even though we accept without dis-
pute the historical existence of the latter, that existence and the knowledge
of it make no positive difference to philosophy’s search for the standard.
Let us now turn back to the Critias-character. Plato has contrived to
make it perfectly obvious, without putting the words into any charac-
ter’s mouth, that the ultra-ancient history which this Critias brings to
the feast has never seen the public light of day in Athens since it arrived
there in Solon’s possession several generations ago. The saga has been told
and retold through the generations of Critias’ family, but it has remained
within the family walls. How do we know this? Because Critias’ telling of
the story is to be the leading element of a return gift for Socrates. You can-
not give something as a gift to someone if the person already has it at her or
his disposal. If the story had been commonly known in the Athens where
Socrates and Critias meet in the dialogue-world, Socrates would already
have been in full possession of it, as would many others. He would never
have craved for a story of that general kind, because he would already have
had one available to him. Now, all this means that, since Solon brought
the Athens-Atlantis story back with him from Egypt, no one has done
what we might call anything serious with the story. Certainly, no one has
used it with a view to the common good or common enlightenment. The
story – remember, in the dialogue-world it both is true and passes mus-
ter among serious people as true – has not in all the generations since
Solon been put to practical or philosophical use by anyone in the fam-
ily that received it; that is, put to use as template for a political ideal or
standard of critique.30 If it had, the story or news of it would have got
out and about in Athens. For we are hardly to suppose that kinsmen of
the character Critias (kinsmen who included Solon) would have been a
bunch of nobodies to whom no one would pay any attention: in real his-
tory, Critias and Plato were scions of a great aristocratic family. So, in the
dialogue-world, Critias and his recent forebears, members of a grand fam-
ily, have received and passed on the story as a mere private heirloom. The
story since returning to Athens has been no more productive of ethical
effects there or in Greece at large than the ancient events depicted were
able to be once proto-Athens and Atlantis, and all Greek memory and
record of them, had been totally destroyed by floods and earthquakes.
So: just as Critias as a young child received the story – a true story
passed on and received as true – from his grandfather, so now all these
years later he has no greater ambition for it than to recount it to a few
select friends, to recount the content as it was told to him, and to recount
it just for the sake of recounting so that the story should continue to exist
as a story. The story’s truth is not a jumping-off point for any new kind of
endeavour: it is as if this truth is a perfect end in itself – so precious that it
would be sacrilege to expect it to lead beyond itself to anything more than
new rounds of telling.
It is all very well for Socrates to welcome the story passively, as enter-
tainment, rather than find in it some kind of ethical or political message
concerning the nature of the best city-state. Socrates has no need to extract
his ethical and political ideal from a saga – even a saga reliably testified to
be reliable – about a once-upon-a-time wonderful city that perished long
30
The sole unfulfilled possibility envisaged concerning this story was envisaged by Critias’ grandfa-
ther when he told a friend that if Solon (who composed poetry) had worked the Athens-Atlantis
story into a poem – instead of attending to the political upheavals in Athens – his reputation as a
poet would have eclipsed those of Hesiod and Homer (21c2–d3). See also Tulli in this volume.
266 Sarah Broadie
ago. Socrates has already articulated that ideal himself in the abstract by
hard reasoning. And surely the Socrates of Plato’s dialogues holds that in
principle everyone from within themselves can come within reach of phil-
osophical truth: you do not need to be a member or friend of some given
family, although perhaps you need to have come under the influence of
someone like Socrates.
Not only has Critias, like his predecessors, done nothing with or
about the story that would have caused ripples in the wider world. In
the Timaeus preview he also comes across as wholly uninterested in the
story’s potential for philosophical illumination or illustration.31 He tells it
as a string of declarations side by side: Athens existed nine thousand years
ago; it had such and such institutions; the Athenian citizens were the best
human beings who have ever walked the earth, and their exploit vis à vis
Atlantis was the greatest achievement ever. It never occurs to him to try to
explain why and how this greatness was grounded in those institutions. We
get the impression that if the authority of the Egyptian priests had trans-
mitted to Solon a story in which the ancient Athenian greatness remained
a constant but the ancient Athenian institutions were importantly differ-
ent, that story would have been passed down in Critias’ family with just
the same trusting respect as the actual one. In that case, of course, Critias
would not have been reminded of his family story by the Socratic ideal
that was presented ‘yesterday’: he would have had no reason to be amazed
at the coincidence of content between his story and Socrates’ discourse
(25e2–5). He would not have had this link with Socrates. But see how lit-
tle use this Critias makes of his link with Socrates. We know from the text
that he never says to Socrates after hearing yesterday’s discourse, ‘That is
wonderful, Socrates, not least because from your reasons and explanations
I now at last understand why the ancient Athenians were such superla-
tive human beings.’ And this is evidence – I think the strongest possible
evidence – that the Critias-character simply lacks interest in getting to
understand the nature and basis of the human virtues.
It is part of the story that the goddess Athena founded hyper-archaic
Athens, designing it to be as conducive as possible to wisdom and military
prowess in the citizens. So anyone at all interested in understanding what
Athena’s mind was aiming for, and the precise reasoning by which she
reached her conclusions about the best political constitution, could have
31
And the situation is not very different in what we have of the long version. There are, however, sig-
nificant differences between the Critias-character as represented by the preview and as represented
by the long version, but space forbids discussion of them here.
started to learn about this through attending to the reasoning of Socrates.
For no one has found fault with Socrates’ discourse of yesterday. Either
Critias realises that Socrates’ argument (even if unbeknownst to Socrates)
must have tracked the goddess’s reasoning, but this realisation leaves
Critias cold; alternatively, Critias has failed to put two and two together
so as to make this connection between Athena and Socrates. But that sort
of obtuseness is surely the mark of a mind uninterested in why any wise
being, whether Athena or Socrates, would deem a certain kind of city the
best.
It fits in with this that Plato has Critias say the following to Socrates –
Socrates, whose discourse yesterday is the raison d’être of today’s whole
show (26b2–7):
Just see how true the saying is that children’s lessons have a wonderfully
memorable quality. Take me, for instance: what I heard yesterday I don’t
know that I could completely lay hold of again in memory; but in matters
where a long, long time has passed since my listening, I should be utterly
amazed if anything has escaped me.
Critias then proceeds to recall how his grandfather told him the
Athens-Atlantis story all those years ago, declaring that all the details have
remained indelibly in his mind. As for Socrates’ discourse, Critias has just
let everyone know that as far as he is concerned it went in one ear and out
the other – halting en route only just long enough to jog remembrance of
his childhood story.
Just before that piece of self-revelation, Critias said this to Socrates
(25e2–5):
While you were speaking yesterday about the political constitution and the
citizens whom you were describing, I began to marvel in recollection of the
very things which I now relate; I began to realise how miraculously some
chance had brought it about (ὡς δαιμονίως ἔκ τινος τύχης) that in most
points you were not off target (οὐκ ἄπο σκοποῦ) <but> in coincidence
with what Solon said.
In other words: Critias already knew the nature of the superlatively good
city from the evidence of the story he had been carrying round inside him
since his childhood. As he listened to Socrates yesterday he was amazed
by the luck whereby Socrates’ ideal turned out to coincide with ancient
Athens. By ‘a miraculous chance’ Socrates was ‘in most points not off tar-
get’. Here Critias shows that he does not really believe in reason at all
(which fully explains why he is not, and was not, interested in the ethical
and political reasonings of yesterday). For Socrates, as he sees it, hit on
268 Sarah Broadie
the right answers – that is, the answers that are the same as the ones that
you can read off from Critias’ story – by luck. And so he happily con-
gratulates Socrates on just happening to land in the right places – ‘in most
points you were not off target’. This is very strange homage to Socrates the
philosopher. Oblivious to the a priori power of Socratic reason (which is
nothing other than universal human reason alive to its own capacities for
analysis and construction in the field of ethics and politics), this homage
offers to Socrates – as a pearl of great price, as the paradigm or ‘target’ by
which Socrates’ political theory is to be assessed – a story rendered true
(but not in the least more valuable therefore to Socrates) by an almost
unimaginably remote event that might as well not have occurred for all
the ethical and philosophical difference it has since made. The dissocia-
tion between the characters Socrates and Critias could scarcely be greater.
By this sardonic construction, partly perhaps a response to the painful
record of interactions between the historical Socrates and the historical
Critias who was his one-time companion, Plato invites reflection on why
it matters whether something is true.
ch apter f ou rteen
The Atlantis poem in the Timaeus-Critias

Mauro Tulli
Introduction
After his archonship in 594–592 BC, Solon went to visit Amasis in Egypt
and Croesus in Lydia. Herodotus (1.30.1–5) states that the purpose of the
journey is to acquire learning. But learning about what? Plato, in the
Timaeus (23d–24e), links the story about Atlantis with a noble tradition
that Solon discovered within the walls of Saïs, near the Nile delta. As con-
crete support for this tradition, Critias, Socrates’ interlocutor at the start of
the Timaeus, points to a body of ‘sacred writings’ preserved by the priests
of Neith as evidence of a record covering more than nine thousand years.1
We can speak of more than nine thousand years, because the story about
Atlantis tells of events that the sacred writings date and record precisely.
One of the priests, a very old man, recalls these events in order to demon-
strate that the memory of Solon is extremely limited and that Athens has,
in fact, a very rich past and an ancient social structure shared with Egypt
of which present-day Athens is ignorant. Solon listens, and recognises the
great significance of the story told about primeval Athens and Atlantis.
After his return to Athens, he is ready both to give an oral retelling of the
story, and to prepare a poem on the topic.
Critias recalls one such oral retelling of the story on an occasion in the
house of Dropides (Ti. 20d–21d). Solon’s narration of the Atlantis story
was addressed to Critias the son of Dropides. As it happens, Socrates’
interlocutor, the grandson of Critias, is also the son of Dropides, both
1
For Luce 1969: 13–44, the ‘sacred writings’ preserved by the priests of Neith did indeed come from
Egypt, and the Atlantis story reflects ancient knowledge of Minoan Crete. For a critique of this
view, see Gill 1976, 1980: viii–xii. For West 1992: 164–5, the Atlantis story derives from Eudoxus
of Cnidus, who recovers the memory of the eruption of Thera preserved in Egypt. For Giovannini
1985, the story is based on a combination of an ideal image of Sparta and memory of the earthquake
at Helice in the Peloponnese. Nesselrath 2002: 17–24 suggests that behind the story about Atlantis,
we can glimpse the results of research in the Academy. For a critical review of attempts through the
ages to find the real Atlantis and to establish its concrete basis, see Vidal-Naquet 2007.
269
270 Mauro Tulli
named Critias in accordance with a custom that is both ancient and mod-
ern.2 ‘Solon once said this’ (Σόλων ποτ’ ἔφη, 20e1): perhaps only one
occasion is involved and we should suppose that the telling of the story
really only occurred once.3 Certainly, Solon did not actually publish the
story he heard about Atlantis. This is clear because both in the Timaeus
and in the Critias Socrates had been unaware of it, and Hermocrates says
he only heard it for the first time the day before, along with Timaeus.
However, Plato refers to the preparation of a poem at various points in
this context.
In the Timaeus (20d–21d), on the third day of the Apaturia, the initiation
festival at Athens, the older Critias, grandfather of Socrates’ interlocutor,
says more about Solon and the story. After Solon’s return to Athens, his
general tendency to regard literary production as a secondary occupation
(parergon), along with the political crisis and other problems, forced him
to interrupt a poem about Atlantis. Hence the conversion of the story
into a poem was not finished, although it was certainly started: ‘If he had
completed the logos that he brought here from Egypt …’ (Ti. 21c5–6). The
fact that Plato uses the term logos (‘story’/‘account’/‘discourse’) does not
count against the hypothesis that Solon planned to compose the story in
the form of a poem.4 The topic of the discussion here is excellence in lit-
erary production. The older Critias stresses to the young Amynander that,
if Solon had completed the poem, neither Hesiod nor Homer would have
been more famous than Solon. In other words, the poem about Atlantis
would have been superior to the Theogony or Works and Days, and superior
to the Iliad or the Odyssey – a strong claim!
In the Critias (112e–113b), Plato refers again to the idea that Solon was
planning to write a poem about Atlantis (‘intending to use the story for
his own poem’), and also suggests that the poem was to be based on a writ-
ten record (grammata). After the description of Attica and Athens, Critias
discusses the nature of the set of names he will use in his account of the
people and places of Atlantis. The names emerged, we are told, as the
result of an initial translation from the language of Atlantis into the lan-
guage of Egypt by the priests of Neith, and also of a second translation by
2
We can identify Dropides among the fragments of Solon as the addressee of an encomium (fr. 22a
West). Cf. Bultrighini 1999: 273–97.
3
A hypothesis whose significance will become clear as we proceed. The story and the writings on
which it is said to be based are fictional, as becomes clear in a series of ways in the course of both the
Timaeus and the Critias.
4
The older Critias begins his report of the story by referring to the children performing various songs
by Solon (21a–d): songs, in the form of a poem. See further Renehan 1976: 88–92.
The Atlantis poem in the Timaeus-Critias 271
Solon from the language of Egypt into Greek, on the basis of the meaning
(dianoia) of the names. A second translation by Solon.5 But how can this
process be documented? Critias himself does not see any problems in the
alleged process of transmission, because (he says) he possesses the relevant
written records (grammata) of Solon, which were previously the property
of his grandfather, Critias, and were studied carefully by him when he was
a child.
The core of the story of Solon is already present in the Critias
(109d–110c), in the description of Attica and Athens. In the course of this
account, Critias uses ancient names such as Cecrops or Erechtheus, and
explains how these names survived the passage of time. Why does Plato
place this stress on ancient names for both protagonists? The reference
to a set of names confirms the claim that Solon was preparing a poem.
Aristotle states in Poetics 9 (1451a36–b23) that a set of names for the key
personae is indispensable to compose an epic or dramatic poem.6 For the
author of an epic or dramatic poem, a set of names offers the best instru-
ment to express the general sense (to katholou) of the poem, whether they
are invented or based on history or tradition.
Thus, in both the Timaeus and the Critias, Plato refers to a poem by
Solon about Atlantis which may not have been finished, but had certainly
been started. He cites a key feature, the set of names, and he alludes to
a written record (grammata), which is preserved and can be consulted –
preserved by the older Critias in the sixth century and by Critias, the
interlocutor of Socrates, in the fifth century. But after Plato, the text con-
cerning Atlantis completely disappears.7 Should we consider the poem to
be an early victim of the lamentable selection that time has made of the
heritage of the classical age?
In reality, we have to deny the existence of a poem by Solon about
Atlantis. Plato, the only source, provides not the slightest external
5
At Timaeus 21e–22b, the possible translation of ‘Neith’ into Greek is ‘Athena’; in the Critias
(113e–114d), the possible translation of ‘Gadirus’ into Greek is ‘Eumelus’. Gambarara 1984: 73–5,
98–9 finds here an echo of Herodotus 4.8.1–3. For a translation from the language of Egypt into
Greek, see also Phdr. 274c–275b: the city of King Thamus is called ‘Thebes’ in Greek and its king,
‘Ammon’, is called ‘Thamus’ in Greek. See Heitsch 1993: 188–203. Themistocles, after Solon, learned
the language of the barbarians ‘as far as he could’, according to Thucydides (1.138.1–2). See further
Canfora 1989: 209–20.
6
Aristotle draws a contrast with philosophy in this respect. Philosophy is capable of expressing a
concept that a poem or a play may develop, but without the use of specific names. See further Horn
1988, Halliwell 1992.
7
With reference to Solon as the source of the Atlantis story, Martina 1968: 22–31 cites with Plato only
Strabo 2.3.6 and Plutarch, Solon 26.1–4 and 31.3–32.2, both of whom explicitly depend on Plato. See
also Gentili and Prato 1988: 61–126.
272 Mauro Tulli
evidence. Plato also provides many passages that cannot be reconciled
with the existence of a poem by Solon about Atlantis.
Critias indicates in the Timaeus (25d–26e) the efforts he made to recall
the story about Atlantis, which he heard at a festival long ago at the house
of his grandfather Critias. Although he mentions several details, none
of them leads us to suppose that there were writings (grammata) left by
Solon. The factor he stresses is the role of memory, and the effort he made
all through the night to recall the story about Atlantis. Its recovery, he
says, is not easy, but it is possible, because the things that a person learns
when he is a child remain in his memory, and because of the pleasure that
Critias took in listening to the story about Atlantis and the readiness of
his grandfather to teach him.8 Why place this stress on memory, if Critias
actually possessed the written record (grammata) of Solon? The contrast
between the two accounts is undeniable.
In the Critias (112e–113b), in the very same sentence that mentions the
grammata of Solon, Critias refers to the decisive role of memory. As a
child, Critias studied the writings (grammata) of Solon, and that is why
he remembers the set of names he uses.9 If Critias still has the writings
of Solon in his possession, why does Plato make him claim to make this
effort? In the Phaedrus (228de), by contrast, Socrates suddenly reveals the
text of Lysias, which Phaedrus was deliberately hiding in order to pass
the speech off as his own. This element is not used in the same way in
the Critias, and was quite absent from the Timaeus. Why does Critias not
produce now the relevant writings by Solon and just read them aloud?
Scholars tend to presuppose complete consistency within every text of
classical Greek literature and find a problem in even the slightest vari-
ation, even if it is excusable and intelligible. Consequently, they forget the
peculiar condition of the Critias, which seems to be an unfinished text and
is, indeed, perhaps the most ancient unfinished text transmitted by the
medieval tradition. In the Critias, for reasons explained later, Plato men-
tions the grammata of Solon, even though, as noted earlier, this cannot be
reconciled with the initial outline of the origin of the story.10 However,
already in the Timaeus, in his report of the older Critias (Ti. 20d–21d),
Plato refers to an unfinished, but undoubtedly started, poem. In the
Timaeus, a complete text, the poem about Atlantis is distinctly visible at
8
See Pradeau 1997: vii–xx. 9 See Nesselrath 2006: 243–8.
10
Rosenmeyer 1956 suggests that the story of Atlantis was created by Plato in two stages: the Timaeus,
with its lengthy exordium, was created after the Critias, and the two works were combined with
some discrepancies.
first, though in a shadowy form, but it soon disappears, and is replaced by
the theme of memory. In the Critias, it takes on a more concrete form, as
a set of writings (grammata), but here too it rapidly disappears, replaced
by memory.
Is this a problem without any solution? In reality, even a minimal
acquaintance with Plato is enough to remove any sense of awkwardness.
The picture of the afterlife offered in the Republic cannot be reconciled
with the picture offered in the Phaedo. We find the same phenomenon
even within the same dialogue. Herodicus of Babylon (fr. 2 Düring)
argued long ago that it was impossible that the encounter of Socrates with
Gorgias, Polus and Callicles in the Gorgias could take place both imme-
diately after the battle of Arginusae in 406 BC (472d–474c), and immedi-
ately after the death of Pericles in 429 BC (502d–503d).11 In the Menexenus,
during the funeral speech given by Socrates, Plato mentions events that
take place after the death of Socrates. He does not respect chronologi-
cal sequence and in this sense he makes a mistake.12 At least, he makes a
mistake if he is judged by the standards of strict accuracy and consistency.
But a literary production, in particular a literary production in ancient
Greece, does not need to obey the standards of absolute consistency. For
example, it is not difficult to show that the picture of the afterlife is differ-
ent in the Phaedo and in the Republic because the perspective changes, and
the philosophical content of the dialogue changes.13
But if the existence of a poem about Atlantis is not plausible, why does
Plato mention it in the Timaeus and the Critias alongside the repeated
emphasis on the crucial role of memory? I believe that he does so to bring
out the relationship between the Atlantis story and the facts of the past,
between his family and the enquiry of Socrates, and between the story of
Atlantis and literary tradition.
The story of Atlantis and the facts of the past

In the Timaeus and the Critias, the story of Atlantis responds to the
requirement that Socrates sets out at Ti. 19b–20c: that of providing a pic-
ture of the ideal polis in action, the polis that Socrates had depicted the
day before, and thus of going beyond simply analysing its nature to dis-
playing its character in action.14 ‘The day before’ (χθές, 17a2, b2, c1): is this
11
See Düring 1941: 46–54. Reconstruction of the relevant text of Herodicus of Babyon is difficult. See
Marchiori 2001.
12
See Tulli 2003. 13 See Annas 1982.
14
On the implications of this passage for Plato’s intentions in the story, see Gill 1977, 1980: xiv–xvi.
274 Mauro Tulli
the day of the Republic, spent by Socrates with Adeimantus and Glaucon?
Here in the Timaeus, Plato mentions the day spent by Socrates together
with Timaeus, Critias and Hermocrates, but not with Adeimantus and
Glaucon. But the rapid outline of the ideal polis that he provides immedi-
ately afterwards, apart from some significant omissions, derives from the
account of the ideal polis of the Republic.15 In the context of the Timaeus
and the Critias, the purpose of this recapitulation of the Republic is cer-
tainly not a straightforward one, but has a broader significance. Plato indi-
cates that he wants to situate the ideal polis of the Republic in the world of
‘becoming’ (genomena), and to present concrete details in a way that can
be reconciled with theoretical abstraction.
In the Timaeus (20d–21d), Critias claims to offer decisive support for
Socrates’ account in the form of the Atlantis story. He suggests that the
Atlantis story deserves to be taken seriously because it was told by Solon,
the wisest of the Seven Sages. Immediately afterwards, Critias mentions
the similar judgement about Solon offered by the young Amynander.16 But
in order to situate the ideal polis of the Republic in the world of becom-
ing, and to present concrete details in a form that can be reconciled with
theoretical abstraction, it is necessary to start from the basis of the kind of
empirical truth that can be verified. The myth, the story about Atlantis,
is offered as providing this kind of support, which is indispensable for the
reconstruction of the past. It is no accident, then, that Socrates speaks
positively about the efforts being made by his interlocutors to go beyond
the realm of the myth and to locate the story of Atlantis in the world of
concrete facts (26e–27b).
Plato thus acknowledges, in this and other ways, the importance that
empirical and verifiable truth has in the reconstruction of the past. This
may be a surprising feature, because Plato, the author of the ideal polis,
is sometimes generally considered to have no interest in the reconstruc-
tion of the past.17 But such a view of Plato runs counter to many features
of his works. In the first instance, Plato’s decision to present his reflec-
tions in the form of conversations between precisely located individuals
15
See Callahan 1977. On points of difference between the outline of the ideal city in the Timaeus
and the Republic, see Clay 2000: 168–72. A very striking difference is the omission of the
philosopher-rulers: see Rowe 1997. But the outline can also be seen as having a proleptic function:
the Atlantis story prefigures the ‘second-best’ polis of the Laws (739b–e). See Naddaf 1994.
16
Thus Socrates, wisest of human beings, according to the god of Delphi in the Apology (20c–23c), is
the ultimate addressee of the Atlantis story told by Solon, the wisest of the Seven Sages. See Laplace
1984.
17
Romilly 1951: 297–305, following Wilamowitz 1969: 126–7, goes so far as to deny that Plato had any
knowledge of the text of Thucydides. See also Giordano 2000.
such as Socrates and Protagoras, Gorgias and Alcibiades reveals an inter-
est in the reconstruction of the past. Of course, the past here is that of
philosophy, which is not always compatible with historical evidence and
very often fictional. But the past of philosophy is given a precise setting,
with many details about the houses of Callias and Agathon, the palaestra,
the law court and the gymnasium. Also, an interest in the reconstruc-
tion of the past is central for the funeral speech in the Menexenus which,
by frank admission, derives from that of Pericles in Thucydides, Histories
2.18 In addition, some scholars recognise the influence of Herodotus and
Thucydides on Plato, especially on Book 3 of the Laws, with its perceptive
analysis of constitutions.19
Plato’s interest in the reconstruction of the past is also clear in the
Atlantis story in a number of ways. These include his recognition of the
problem posed for enquiry by discontinuities in civilisation and the need
to take account of facts about land formation and ancient legends to give
a plausible picture of the past. It is against this broader background that
we can make sense of Critias’ insistence in the Timaeus (22b–24c) and
the Critias (109b–111d) on the concrete foundation of his reconstruction.20
This dimension of Plato’s concerns helps to explain why Critias mentions
an unfinished, but undoubtedly started, poem by Solon about Atlantis.
This is in spite of the importance that he attributes to memory, and even
if this introduces contradictions in his story. This is surely why, when
Plato mentions a key nucleus of the story, the set of names, he cites the
record created by Solon (grammata), presented as a body of evidence that
is preserved and can be consulted. On the basis of this alleged body of evi-
dence, myth is given a solid foundation, and Socrates’ vision of the ideal
polis in the Republic is linked with the world of concrete facts in the past.
Certainly, the world of concrete facts in the past guarantees the possibility
of the ideal polis in the future. It is the perspective Aristotle indicates in
chapter 9 of his Poetics (1451a36–b23).
The family of Plato and the enquiry of Socrates

But who actually possesses the poem, according to Plato? Scholars, natu-
rally, stress the significance of Socrates’ interlocutors in works such as the
Charmides or Symposium, the Parmenides or Philebus. In general, within the
18
See Pradeau 2002: 14–35. 19 See Weil 1959: 42–54.
20
See Gill 1980: xx–xxi. It is not easy to see, in Socrates’ reaction to Critias’ insistence, the ‘slight curl
of the lip’ Broadie discovers in this volume (p. 263 above).
276 Mauro Tulli
dialogue, the process of Socratic enquiry and its success or failure express
the education, attitudes and capacities in philosophy that the interlocu-
tors show.21 The treatment of the topic of rhetoric in the Gorgias reflects
the differing qualities of Gorgias, Polus and Callicles.22 In the Phaedo, the
discussion turns to the immortality of the soul because Socrates is engaged
in dialogue with Simmias and Cebes.23 Against this background, it is note-
worthy that in Republic Book 2, after the inconclusive debate about the
nature of justice with Thrasymachus, Socrates conducts the enquiry about
the ideal polis with Adeimantus and Glaucon, the brothers of Plato. Thus,
when Socrates is engaged in the boldest of his speculative endeavours,
Plato finds interlocutors for Socrates in his own family, who are presented
as capable of understanding the concept of the ideal polis.24
In the Timaeus and in the Critias, in order to satisfy Socrates’ wish to
have a picture of the ideal polis in action, Plato considers it appropri-
ate to deploy other figures – Timaeus, Critias, Hermocrates – rather than
his brothers, Adeimantus and Glaucon. This is perhaps because, in the
Timaeus and in the Critias, he moves from a purely abstract ideal, the best
polis, to differing types of description of empirical reality, the concrete his-
tory of the Atlantis story on the one hand and the analysis of the natural
universe on the other. Socrates’ opening request stresses the importance of
having the right kind of spokesman for the task in hand. If Plato’s broth-
ers are no longer to play this role, then what criterion should be employed
to choose the interlocutors for this purpose? Socrates, commenting at
Ti. 19b–20c on his three interlocutors, highlights aspects of their ability
as well as their education and upbringing (paideia kai trophē, 19c6) that
make them suited for this role. The sentence is both striking and unquali-
fied, and seems to extend to Critias.25 In any case, as in the Republic, by
21
See Gill 1996a: 288–9. 22 See Kahn 1996: 125–47.
23
On the importance of Socrates’ interlocutors in Plato, see Blondell 2002: 1–52.
24
Should we say Plato’s own family, or rather Plato himself, behind the personae of his family? The
latter hypothesis, though plausible, is not straightforward and has no supporting evidence. See
Stella 1998.
25
On the language of Socrates’ critical evaluation of the interlocutors, see Regali 2012: 43–71.
Specifically, Timaeus is praised because of his noble lineage, outstanding experience in the polis
and an unsurpassed knowledge of philosophy. Hermocrates is praised because he has a competence
based on nature and education. But what about Critias? There is only a very brief comment on
him: ‘All of us here, in Athens, know that he is no amateur (idiōtēs) in any of these things.’ Why
only this brief comment? There are different possible explanations, and this question is linked with
the continuing debate about his exact identity. Perhaps Plato thinks it is not necessary to explain
the significance of his family background, just as this is signalled with a bare gesture as regards
Adeimantus and Glaucon in the Republic (367e–368c).
including Critias in this group, Plato affirms the significance of his own
family. But here scholars raise a serious problem. Can Plato really allow
Socrates to depict in this way Critias son of Callaeschrus, the notoriously
brutal member of the ‘Thirty Tyrants’?26 Or should we suppose that the
Critias presented here is the one who falls chronologically between the
older Critias son of Dropides, and the tyrant, the son of Callaeschrus (the
grandson of the former and the grandfather of the latter)? Alternatively,
again, should we suppose that Plato leaves it ambiguous whether he is
referring to the grandfather of Perictione, his own mother, or the tyrant,
the cousin of his own mother?27
On any of these interpretations, Plato links his own family with
Socrates’ enquiry. Specifically Critias (whatever his precise family relation-
ship to Plato) is presented as being an appropriate person to offer a picture
of the ideal polis in action, even if it is by means of a story whose truth we
cannot really accept.28 But Plato’s attribution of significance to his family
extends still further. His family is presented as supporting the account of
the ideal polis, as in the Republic, but in this case with concrete evidence,
embodied in the writings of Solon, which validate an allegedly concrete
account.29 The reference to the writings, discussed earlier, is thus not a
negligible detail. From the story contained in the writings of Solon, the
hope arises of an ideal polis that could be transferred from the facts of the
past to the facts of the future and could thus transform the character of
Athens, the city of Solon, Critias and Plato himself.
26
See further Broadie in this volume, according to whom the personal and intellectual defects of
Critias the tyrant form an important part of the background for making sense of the Atlantis
story.
27
In the Timaeus (21a–d), the situation is this: when he was very young, Critias heard the story from
Critias son of Dropides, who was very old; Critias son of Dropides heard the story from Solon,
perhaps before 558 BC. Critias the tyrant, the son of Callaeschrus, was born in 460 BC, about one
hundred years later. The genealogical tree is not easy to understand. Perhaps Plato proceeds here on
the basis of a forty-year generational pattern, so that Critias son of Dropides would be ninety years
old when Critias was ten. See further Gill 1980: 39, Strauss 1993: 87–97, Brisson 1994: 32–49.
28
It is surely no accident that Critias son of Dropides recalls Solon on the third day of the Apaturia,
which is the Athenian feast of initiation into the family, in a broad sense – that is, into the phratria.
See Nilsson 1951: 150–70. The aetiological myth linked with this feast hinges on the astuteness of
Melanthus who, with the protection of a god, kills Xanthus. This may be meant as a hint by Plato
about the character of the story. Gill 1980: 40 sees an indication that this is a deceitful narrative,
marked by astuteness. In any case, by mentioning the third day of the Apaturia, Plato sets the
Critias within the framework of the initiation feast of his family.
29
In the Timaeus (25d–26e), with his story about primeval Athens and Atlantis, Critias intends to
transfer the ‘guardians’ of Socrates’ ideal polis to the concrete facts of the verifiable past: ‘We shall
say that those citizens that you had in mind were our true ancestors’ (ἐκείνους τοὺς ἀληθινοὺς
εἶναι προγόνους ἡμῶν, 26d2–3). Thus, Critias places the ‘true’, ‘actual’ guardians firmly among the
concrete facts of the verifiable past. See Tulli 1994.
278 Mauro Tulli
The Atlantis story and literary tradition

At the same time, Plato discovers in his family a literary heritage. Critias
possesses a literary heritage that derives from Critias son of Dropides. The
writings of Solon ‘in my house’, though fictional, thus figure at the end of
Plato’s genealogical tree. In this sense, and by an exchange of roles, Plato
implicitly attributes the writings of Solon to himself.
The identification of the Atlantis story with the alleged poem of Solon
gives added significance to the passages noted earlier about the quality of
the potential achievement of Solon. The sentence uttered in the Timaeus
(21a–d) by the older Critias to the young Amynander can be seen as imply-
ing a specific and large ambition. There is no author more renowned than
Solon – read: ‘Plato’. The poem about Atlantis would have been superior
to the Theogony or Iliad. Of course, the excellence of Solon is here hypo-
thetical, an unfulfilled condition. But the content of the poem of Solon is
the same as that of the Atlantis story of Plato.
There is a further implication in the parallelism that Plato establishes
between himself and Solon in this respect. We might well be reminded
of the myth of Thamus and Theuth in the Phaedrus (276b–277a). In that
myth a significant motif is the relationship between what is ‘play’ (paidia)
and what is ‘serious occupation’ (spoudē). Literary production is there pre-
sented as ‘play’ and as a relaxation from the ‘serious’ matters of investiga-
tion, for instance, about the nature of justice.30 The attitude that the older
Critias attributes to Solon is similar: literary production is treated as a sec-
ondary occupation (parergon) compared to more serious matters – mean-
ing here not the theory of justice but the attempt to put this into practice,
as Solon did at Athens. The parallelism is undeniable. Can it be accidental
that the story about Atlantis was not completed either by Solon or Plato?
Thus, the story that Plato announces in this way, the one started in the
Critias, is superior to the Theogony or the Iliad. In some way, it reveals
the form of a poem. Perhaps this is a surprising result. After the rejection
of mimesis in the Republic, after the ban that is placed on Homer in the
Republic, does Plato admit here that, for the purposes of philosophy, it
is worthwhile composing a poem, a form of literary production which
falls centrally within the sphere of mimesis? The ambition is all the more
surprising in the first section of the Timaeus, a text that Plato links closely
30
Cf. Rowe 1988: 208–12. Occasional or marginal activity does not necessarily mean inferior activity.
Erler 1994 points to a similar problem in the Politicus (286d–287b). See also Jouët-Pastré 2006:
139–73.
to the Republic. In particular, the request of Socrates for a picture of the
ideal polis in action (19b–20c) echoes strongly the critique of mimesis in
the Republic.31
However, scholars also point out that the rejection of mimesis in the
Republic is not absolute. In Book 10, admittedly, Plato states that the pres-
tige of Homer is not based on knowledge. Rather, Homer is three degrees
removed from the truth, and offers pale shadows (596a–599b) or ‘appear-
ances’ (phantasmata), not based on knowledge of reality (601b–607a).32
But, at the same time, Plato indicates that presenting images based on
truth by means of mimesis is possible. This is, precisely, the task of the
good artist (agathos zōgraphos), who recognises the pattern of the god
(471c–473b and 484a–485a). The author who possesses this kind of know-
ledge offers a result that has value as a genuine reflection of the justice, the
beautiful and the temperate in itself (500b–502a).33
The good kind of mimesis is based on the knowledge that derives from
the kind of education (paideia) and upbringing (trophē) that Socrates pos-
sesses. It is not difficult to find an underlying consistency between the
concept of mimesis in the Republic and the story of Atlantis in the Timaeus
and Critias. This consistency prepares the way for the definition of dia-
logue in the Laws (817a–d) as the ‘truest tragedy’ or the ‘finest drama’
for the polis.34 The devaluation of the literary production of Hesiod and
Homer implied in the Timaeus (19b–20c) means the rejection of the
kind of literary production that is not based on the kind of education
and upbringing that Socrates possesses. In the Atlantis story, by contrast,
Plato derives its content directly from the programme that is designed to
produce the best education and most complete upbringing. This is the
programme of the Republic depicted by Socrates, in response to which
Critias offers his account of the Atlantis story. It is surely no accident that
in the Timaeus Socrates is keen to see Timaeus, Hermocrates and Critias
carry out the kind of project he asks for, since he sees in them the com-
bination of natural ability and education that provides the foundation for
such mimesis.35
31
Welliver 1977: 8–21 finds here an ironical comment on the literary ability of Critias. But on this
hypothesis, see Gill 1979a.
32
See Nehamas 1982.
33
See further Ferrari 1989: 120–41, Halliwell 2002: 37–71, Naddaff 2002: 37–66.
34
Gaiser 1984: 103–23 offers strong arguments here to interpret the definition of the ‘state’ or of
the ‘constitution’ as the definition of the dialogue that builds the plot of the ‘state’ or of the
‘constitution’.
35
See Dalfen 1974: 231–8.
280 Mauro Tulli
Also, it is surely no accident that Plato recalls in the Critias (106b–108a)
the good kind of mimesis envisaged in the Republic. The Atlantis story
is the outcome of this kind of mimesis, an outcome that is not certain
to appeal to its listeners, as Critias points out, because people are highly
critical when human beings are the object of mimesis. As regards the por-
trayal of the cosmos, the ignorance of the listeners places a limit on their
expectations. But, if human beings are the object of mimesis, they can
easily detect any deficiencies. Hence, there arises the need to paint human
beings accurately according to the requirements of likelihood (to eikos)
and what is appropriate (to prepon).36
This is another respect in which Plato aims in this story at a kind of
mimesis which is both based on theoretical understanding or knowledge
and which also translates a general concept into concrete and specific
facts (the world of ‘becoming’). In this respect, Plato in a sense combines
the two approaches that Aristotle distinguishes in chapter 9 of his Poetics
(1451a36–b23). Here, as noted earlier, Aristotle draws a contrast between
the kind of discourse that remains at the level of the particular and that
which aims to illustrate the general or universal.37 Thucydides’ writings are
taken as the exemplar of the first kind of mimesis, whereas literary pro-
duction and philosophy are presented as exemplifying the second kind.
If we attend only to one element of the Atlantis story, the fact that it
expresses in narrative form the core programme of the Republic, we can
see the Atlantis story as anticipating one side of Aristotle’s dichotomy,
namely the portrayal of the universal.
But this leaves out of account the interest of Plato in the reconstruc-
tion of the past, and his conviction that the writings of Solon constituted
the foundation for establishing factual or verifiable truth.38 If, in order to
situate the ideal polis of the Republic in the world of becoming, Plato has
to choose an idiom that is that of Thucydides, what follows for the status
of the Atlantis story? What follows is that the story takes on a distinc-
tive character, between particular and general content, in a way that cuts
across the dichotomy of Aristotle.
36
The idea that likelihood (to eikos) and what is appropriate (to prepon) need to be achieved in
accounts of the natural cosmos is a related theme in the Timaeus. (27d–29d), as noted by Burnyeat
2005.
37
See Erler 1998. On the links between Plato and Aristotle on this point, see Halliwell 1986: 109–37,
and Arrighetti 2006: 183–270.
38
Gill 1993 suggests that the Atlantis story should be seen as a kind of ‘foundation myth’ for primae-
val Athens: the ideal polis in action.
It is perhaps for this reason that we find the story sometimes pre-
sented in a form that evokes literary codes falling between (or outside)
reconstruction of the past and philosophy, namely the hymn or the
encomium and the drama. It may not be accidental that the hymn to
the gods and the encomium of good people are the only types of literary
production explicitly exempted from exclusion from the ideal polis after
the critique of mimesis in Republic 10 (601b–607a), a concession that
is fully in line with the positive role envisaged for literary production
and the other arts in Republic 3 (401b–403c). These are literary codes
which combine in their own way particular and general content. For
instance, in the Timaeus (20d–21d) we find indications that a prerequis-
ite for an image of the ideal polis in action is that it takes the form of
an encomium. Critias offers the Atlantis story both as a hymn to the
goddess and an encomium to the city of Athens. In the Critias (108bc),
Hermocrates also urges Critias to compose an encomium and a hymn.
But, in the same context, Socrates substitutes the language of drama for
that of encomium. In this respect, the good kind of mimesis that he rec-
ommends is transferred to the context of the theatre, with Timaeus and
Critias presented as competing before an audience, a motif picked up by
Hermocrates and Critias.39
Does Plato in these respects anticipate the mixture of literary codes
that is normally seen as a distinctive characteristic of the Hellenistic
age? As emphasised earlier in the Timaeus (21a–d), the comment of the
older Critias to the young Amynander invites comparison between the
Atlantis story and the works of Homer and Hesiod. In the Critias too
(108b–d), the Atlantis story has strong literary connotations as a poem: if
Hermocrates urges Critias to compose an encomium and a hymn, Critias
responds by also invoking Memory, the mother of the Muses. Alongside
the concrete allusions built into the narrative of the Atlantis story in the
Critias, scholars have detected also numerous allusions in the story to
Homer especially.40
39
See further Gill 1977 on these allusions to Greek literary codes and their relationship to the good
kind of mimesis in the Republic.
40
For Friedländer 1954: 300–5, the paradigm for the picture of Atlantis is the oriental polis of
Herodotus: for instance Ecbatana (Hdt. 1.98.1–6) or Babylon (1.178.1–187.5). But, behind this,
Szlezák 1993 also sees the picture of Troy. The idea that the narrative is the result of the memory
inspired by the Muses depends on the Iliad (2.484–93). The Atlantis story builds on the length of
the war of Troy, nine years in the Iliad (2.134–8), extending the scale of the time since the war with
Atlantis to nine thousand years. See further Gaiser 1968: 260–70.
282 Mauro Tulli
Against this background, we can make sense of the symbolic function
of the writings that Critias both indicates and conceals in the Timaeus and
the Critias. Onto these writings Plato projects the relationship between
past literary production and the new kind of mimesis that he sees as valid.
The unfinished, but certainly started, poem of Solon is compatible with
the unfinished but certainly started Critias of Plato, which is superior to
the literary production of Hesiod and Homer.
ch apter f i f teen
Friendship and justice in the Laws

Malcolm Schofield
Introduction
In the opening chapter of the books on friendship in the Nicomachean
Ethics, Aristotle says (EN 8.1, 1155a22–6; cf. EE 7.1, 1234b22–3):
It seems that cities too are held together by friendship, and that lawmakers
are concerned more with that than with the virtue of justice. For concord
(homonoia) seems to be something similar to friendship – but concord is
what they are most of all aiming at, and stasis, which involves enmity, is
what most of all they are trying to get rid of.1
Aristotle talks of lawmakers. But many of his readers will think imme-
diately of Socrates playing legislator in the Republic with Glaucon and
Adeimantus, and of the passage in Book 4 where self-discipline (sōphrosunē)
is identified as the homonoia that he has just described: ‘a natural harmony
of worse and better as to which of them should rule, whether in city or in
each individual’ (R. 4, 432a). It is hard to get away from the thought that
this remark – and with it the whole political project of the Republic – lies
at the back of Aristotle’s claim.
One might have supposed that when he talks here of lawmakers,
Aristotle must primarily have in mind those legislators who frame con-
stitutions. The obvious candidates? Presumably, Solon and Lycurgus (cf.
Pol. 2.12, 1273b27–34). However, Aristotle’s references in the Politics to
their achievements and intentions as legislators do not refer particularly or
explicitly to the promotion of homonoia or the avoidance of stasis.2 More
I thank Christopher Gill for editorial proddings, and audiences in Lille and Durham for their com-
ments on earlier versions of this essay, which is offered to a friend who is also something of a fighter
for (among other things) justice, as well as a Plato scholar of distinction.
1
All translations from the Greek are my own except where otherwise indicated.
2
For example, in his account of the ‘middle’ politeia (social and political system) in Pol. 4.11 Aristotle
claims that it is because the middle class is dominant in large cities that they are relatively free of
stasis; that it is thanks, again, to the weight and numbers of the middle class that democracies are
283
284 Malcolm Schofield
generally, the survey in Book 2 of the Politics of previous attempts at devis-
ing constitutions has little to say explicitly about the place that friendship
or concord occupies among the aims of legislators, whether those aims are
more theoretical or more practical. The one obvious exception is, in fact,
Aristotle’s treatment of Plato’s Republic, and of the theory of community
worked out in Book 5 of the dialogue. Commenting on the reason why
his Socrates introduces the notorious system whereby, among the guards,
children and women are held in common, he says (Pol. 2.4, 1262b7–10):
We think friendship is the greatest of goods for cities (for then people will
be least likely to engage in stasis); and Socrates sings the praises particularly
of the idea that the city should be a unity – something that is thought to be
achieved by friendship (philia), as he too claims it will be.
Probably, Aristotle has his eye principally on the passage at R. 5, 462a–465d,
where Socrates makes the good of political unity turn on the emotional
solidarity between citizens who regard one another as members of a single
family, and feel one another’s joys and pains as their own. Actually, there
is no attempt at that point in the dialogue to conceptualise this solidarity
explicitly in terms of philia (friendship). However, when Socrates spoke
earlier of the need for unity in the city (4, 423b–424b), the communistic
scheme of life referred to in passing was summed up in the proverbial
phrase koina ta philōn: ‘friends share all they have’.
The political philosophy of friendship

When in Politics 2.6 Aristotle considers Plato’s Laws, he does not return to
the theme of friendship. Yet in none of Plato’s contributions to political
philosophy is the need for a lawgiver to devise the constitution with regard
to friendship stressed more emphatically, or by more explicit insistence on
the value of philia, than in the Laws. Indeed, this is the topic – or rather
the conclusion – of the very first sustained argument of the entire dialogue.
This argument consists in an attempt by the Athenian Stranger to persuade
his Cretan and Spartan interlocutors that it is a mistake to conceive of the
supreme objective governing a legislator’s constitutional provisions as vic-
tory in war over other cities – on the Hobbesian assumption that the war
of all against all is inevitable, because this is simply human nature.
more stable and survive longer than oligarchies; and that the best legislators – Solon and Lycurgus
receive prominent mention – are drawn from the ‘middle citizens’ (Pol. 4.11, 1296a18–21). The impli-
cation at least is that their excellence as lawgivers is indicated by their middle-class aversion to those
extremes that foment instability and stasis. But there is no talk of philia or homonoia.
Friendship and justice in the Laws 285
The argument against what is represented as the Cretan and Spartan
position is not altogether straightforward. What the Athenian Stranger
seeks to get agreement on is this (Lg. 626e–628c):
(1) There is victory over oneself as well as victory over an external
enemy.
(2) The notion of victory over oneself can be applied not just to the indi-
vidual but to the household, the village and the city.
(3) No person or group counts as superior to himself or itself unless a
better element controls a worse – thus ‘victory’ here carries an evalu-
ative significance.
(4) Within a group (for example, the family) justice is secured if the
worse are made willingly to accept the rule of the good (rather than
if the bad are destroyed by the use of force).
(5) Justice and legislation are, however, at their best when they go fur-
ther than that – when following reconciliation rather than victory,
a system of laws is introduced which secures peace and friendship
between all parties.
Given (5) in particular, the idea of victory needs radical revaluation. To
revert to the case of a city, even victory over itself (that is, of the better ele-
ments over the worse) is not to be regarded as among the things that are
best, but merely as something necessary in the circumstances – necessary,
presumably, if the city is to attain any degree of health at all (the medical
analogy here is Plato’s own). The best a legislator could achieve would be
a city that needs no purgation in the first place – where the citizens vol-
unteer their cooperation in friendship; where it is not a matter of worse
elements doing what the better elements instruct them to do, even if they
do it willingly (628c–e).3
‘Of the forms of justice,’ says Aristotle in the opening chapter of the
Nicomachean Ethics’ treatise on friendship, ‘the most important is thought
to be that bound up with friendship’ (EN 8.1, 1155a28).4 As often, Plato in
the Laws and Aristotle in the Ethics and Politics share a similar outlook.5
3
I suppose that best might be seen as an ideal so ambitious as to be utopian: only realisable in a
community such as that constituted by the guards of Book 5 of the Republic, operating as a single
family sharing its joys and sorrows. On the other hand, prima facie it would seem likely that the
Stranger’s idea of legislating for friendship is one that will be geared to what is possible in Magnesia,
not restricted to the sort of city of ‘the gods or sons of gods’ that Book 5 of the Republic is famously
now said to represent (see 739a–e).
4
See my discussion of this and related material in the Eudemian Ethics in Schofield 1998: 37–40.
5
Whether this is because both reflect discussion in the Academy in the 350s, or because the Laws
shaped much of Aristotle’s thinking in his early maturity, or because Plato in the Laws was learning
Further illumination on what the Athenian Stranger means by taking the
justice that will secure friendship as the legislator’s goal for the city comes
in a key passage in Book 6, in the relatively short section of the Laws
on what we might call the constitution proper (751a–768e). The passage
begins at 756e with the remark that the politeia must always occupy a
middle position between monarchy and democracy. The Stranger makes
friendship his basis for arguing the point (particularly as it applies to the
selection of office-holders): taking it for granted that friendship is a main
objective for the legislative enterprise. This is in line with what in Book 5
he had articulated as ‘the hypothesis of the laws: that our people should be
as happy as possible, and as far as possible friends with each other’ (743c;
my italics, of course). In that context he had seen the injustice and conse-
quent litigation arising from extremes of wealth and poverty as the main
threat to such harmony, in the context of a vigorous moral onslaught on
money and a money economy (5, 741e–744a).
In our Book 6 passage the Stranger enunciates first some terms on
which friendship cannot be secured: (i) slaves and masters cannot be
friends (so pure monarchy – which is essentially despotic – will not be
able to deliver it); (ii) as for citizens, given that they include the good and
the bad (or the better and the worse), you will not get friendship between
them if they are accorded equal honours, that is, given equal access to public
office (so pure democracy will not work either). The reason that neither
despotism nor egalitarianism will make for friendship is that both provoke
stasis-generating resentments. The solution is not to abandon the idea of
equality, but to employ predominantly a different and better model of
equality: proportionate or (as 744c puts it) ‘symmetrical’ equality, ‘the
truest and best equality’. This equality is achieved by according to equals
what is equal – that is, to those with great virtue the opportunity of great
offices, to those at the opposite end of the scale appropriately reduced
opportunities (757bc).6
‘Natural’ equality thus identified (in provocatively anti-democratic
terms, of course) is precisely what political justice consists in (757c). It is
what legislation must make its objective in any city whatever: not tyranny
or popular control, but always justice, understood as ‘nature’s version
of equality’, as applied to those who are unequal (757cd). But however
natural, proportionate equality will not on its own guarantee friendship.
from the young Aristotle, or because of a combination of two or more of these options, is unfortu-
nately something we are never likely to know.
6
On the different models of equality in play here, the discussion in Harvey 1965 is still worth
consulting.
Every city will sometimes need to employ ‘the equality of the lot’, or what
Aristotle would call arithmetic equality, to assuage the discontents of the
many. When that happens, one will just have to pray that providence or
luck will engineer the most just outcome: that is, one in which it simply
turns out by chance that the lot allocates the most important offices to the
best people, and vice versa (757d–758a). When that outcome is achieved,
then the old saying: ‘equality creates friendship’ (757a) will have had its
truth confirmed. And the Athenian Stranger envisages its realisation as
a possibility not only in a utopia, but in principle in any city whatsoever
where the lawgiver understands the saying properly, and keeps it firmly in
his sights.7
This discussion of proportionate equality as the basis of political justice
is where Plato works out in theoretical detail his proposal in Book 1 that
the legislator will seek to promote friendship through his constitutional
provisions. The key is equality: the kind of equality that (in the language
of Book 3) gives due recognition both to the freedom that all citizens pos-
sess as citizens (which is why no form of despotism can be countenanced),
and to the virtue and wisdom of those who possess superior virtue and
wisdom (see especially 693a–e; cf. 701d). The dialogue’s theory of politi-
cal justice, as set out in Book 6 (and summarised in the preceding para-
graphs), bears a striking resemblance to Aristotle’s thinking in Book 5 of
the Nicomachean Ethics and in the Politics.8 Aristotle, too, sees this kind
of justice as what is needed if conditions ripe for stasis are not to develop,
and he too regards disparities in the distribution of wealth as a major cata-
lyst for the perception of injustice that triggers stasis (see Politics 5.1–2). He
also thinks that superior virtue and wisdom are the most important bases
of entitlement to political participation (see Politics 3.12–13).
What is no less striking is that Aristotle does not associate political jus-
tice, as Plato does, with the production of friendship. Aristotle certainly
has an idea of civic or political friendship; and he does connect that idea
with justice (see especially EE 7.9–10). However, he regards civic friend-
ship as the social glue of mutual advantage between individuals who are
personally acquainted, seeing it as exhibited above all in exchange and
7
So the expectation enunciated in n. 3 above – that the Stranger adumbrates at 628c–e an ideal realis-
able in Magnesia – is confirmed.
8
The Laws’ theory is palpably unlike the account of political justice in the Republic, which makes
‘doing one’s own job’ by the different classes within the city the key idea (R. 4, 433a–434c), and has
nothing to say about equality of any sort, or about selection of persons for public office. Whether
or not Plato in the Laws has changed his mind about political justice, he has certainly harnessed the
concept to a different theoretical agenda. See further Vlastos 1977: 17–34.
commerce, and not as the outcome of the political settlement worked out
by the legislator in his constitutional provisions. For Aristotle, friendship
in any proper sense requires a mutual knowledge and concern that can-
not be achieved by legislative means. What citizens share in common as
citizens, and what lawgivers do strive to bring about, is, at best, something
else: homonoia, concord (see again EN 8.1, 1155a22–6; cf. 9.6).9
Friendship, wisdom and freedom

There is one other context in the Laws where the Athenian Stranger devel-
ops the theme of friendship as the goal of the legislator’s construction of a
constitution. This is the extended passage at the end of Book 3 (the book
was called ‘The Lessons of History’ by Ernest Barker, followed by Trevor
Saunders in the Penguin translation), running from 693a almost to the last
page (701e).10 The general theme is the same as in the Book 6 passage: the
need to find a mean between despotism and democracy. This time, this
idea is presented as what is required if friendship, wisdom and freedom (a
value stressed here more than in Book 6) – the triad of values the lawgiver
should be taking as his ideals – are to be achieved.
Friendship has not hitherto been a theme in Book 3. To understand
why it is first introduced into the discussion just where and when it is, we
need to devote some time to reviewing the way the book gets launched
and develops subsequently. At the very outset, the Athenian Stranger
proposes an investigation into the origin of politeia – abruptly, and with-
out explanation why or how this might help in the ongoing enquiry into
politeia (676a). The proposal launches the highly selective and speculative
history of Greece that follows. The point of the exercise gets a bit clearer at
682e –683c when, with the introduction of Sparta into the historical nar-
rative, the Athenian flags a connection with Book 1 and a resumption of
the discussion of the idea of a good politeia that his interlocutors claim is
exemplified in the Spartan and Cretan constitutions (626a–c).11
This discussion is harnessed to a complex account of the alliance that the
three Dorian states of Sparta, Argos and Messene made with each other,
9
See my discussion in Schofield 1998: 40–3 (contra e.g. Cooper 1990: 235).
10
See Barker 1918: 307; Saunders 1970: 118, 143.
11
Editors pick up particularly the reference to the beginning of Book 1’s discussion of drinking par-
ties (636e). But the Athenian mentions this as the point at which the dialogue left its initial focus
on Cretan and Spartan institutions (629a). He makes it clear that they are now making a fresh
start upon that enquiry, conceived as the attempt to get clear about what makes for a good politeia
(683b).
its disintegration, and the differences in their governmental arrangements
which led to Sparta’s success as a champion of Greek freedom and the dis-
graceful conduct of the other two (683c–693a). This is how the Athenian
delivers on the agenda he announces at 683b: to understand ‘what is good
in the way these settlements were established, and what is not; which laws
keep things stable, where they are stable, which cause destruction, where
there is destruction; and what changes would need to be made in order to
make a city flourish’. In other words, it has eventually transpired that his-
tory is invoked as a way of investigating what makes a politeia successful.
What is said to emerge from the discussion over the next few pages is
the fact of a single desire common to all humanity: ‘that what happens
should happen in obedience to one’s own will’ (687c). But the interlocu-
tors agree that we ought to be praying for wisdom in what we make the
object of our willing (687e). The Stranger recalls the argument of Book 1
against the view that war should be the goal of a legislator’s lawgiving, and
the position he himself had embraced there: that laws should be framed
by reference not just to courage, but to the whole of virtue and above all
to the leading virtue: wisdom (phronēsis) (687e–688b). From this point
on, the focus is on what factors undermine or enhance the chances of
attaining wisdom in the exercise of power, whether in a state or on the
part of a ruler. What caused the ruin of kingdoms such as Argos and
Messene was not cowardice or incompetence in warfare, but deficiencies
in the rest of virtue, and above all ignorance of the most important things
in human life.
The principal danger is in allowing rule to be absolute: a certain recipe
for self-destruction. The whole sequence of thought culminates in a mem-
orable formulation of the thesis that power corrupts, and absolute power
corrupts absolutely – ultimately through the workings of that ‘greatest
of diseases, folly’ (691cd; cf. 688a–689e). The Athenian argues that of the
three Dorian states that exercised most power in the Peloponnese in his-
torical times, Argos and Messene had inadequate experience of legislation
(their laws and politeia were quickly corrupted: 685a) and so they suc-
cumbed to tyranny, and only Sparta had devised a constitution capable
of controlling regal power. Consequently, only Sparta among the Dorian
states was able to act as it should have done in helping in the repulse of the
Persian invader in the early fifth century. Of those states, only Sparta was
prepared to rescue Greece from the threat of enslavement. Only Sparta
was not herself a despotic regime (691d–693a).
I now return to my point of departure in this section of the chapter.
As highlighted earlier, the Athenian proposes recognition of a triad of
values – freedom, wisdom, friendship – as the proper basis for legislation,
rather than using legislation to establish ‘great and undiluted’ forms of
rule (693b). He claims that, although that might look something quite
different from making wisdom the objective, in fact it adds up to the
same thing (693bc). There is already material in the discussion of the igno-
rance and folly of despotism that may give a clue as to how that claim is
intended. What he stresses there is that wisdom – whether in the indi-
vidual soul or within society at large – requires the establishment of a
properly founded concord between its elements, psychic or social as the
case may be (689a–e). Initially (689b), he speaks primarily of the folly
in evidence when pleasures and pains or the mass of the people disobey
reason (in the city in its guise as law). But subsequently he focuses on ‘the
dissonance which is stupidity at its greatest’ of the ruler who through his
arrogance pays no attention to ‘due measure’ (691a–d). A wise ruler, then,
will be successfully exercising wisdom only if he presides over a popu-
lace that respects his measured judgement and lives in harmony with him
accordingly. The lawgiver’s promotion of such wisdom will, in that sense,
be in effect the same thing as promoting friendship or harmony between
ruler and ruled.12
Nonetheless, while these elements in the preceding discussion do some-
thing to prepare the ground for the introduction of the Athenian’s triad,
the inclusion of freedom in the list remains unexpected, despite the refer-
ence to Greek resistance to the Persian invasion. When running through
the different forms that claims to rule might take, he had identified as
apparently strongest the idea that ‘the person without knowledge should
follow, whereas the person who has understanding should lead and rule’
(690bc). And that expresses succinctly the general view taken throughout
the Laws. Willing obedience, not freedom, is what the Athenian Stranger
often prizes most, as for example in his prescription for rulers themselves
in Book 4: ‘Where the law is master over rulers, and the rulers are slaves to
the law, there I see salvation’ (715d).
I suspect that, with the introduction of the triad (freedom, wisdom and
friendship) at this juncture in Book 3, a highly ingenious literary and his-
toriographical manoeuvre is afoot. Plato is temporarily about to allow into
his text an alternative political vision, in which the Persia of Cyrus the
Great will figure as heroic exemplar of the triad. But before we look at the
12
And adequate legislation for this is what is needed. It is not enough for both parties – ruler and
ruled – to swear oaths (692b), the one to exercise authority properly, the other to respect that
authority provided the ruler kept his word (as in the original foundation of the Doric states
(684a)).
account of Cyrus’ Persia, we need to ask ourselves where the triad comes
from if the immediately preceding text of Book 3 does not supply enough
of the answer. This will require an excursus into late fifth-century Greek
propaganda.
Diodorus Siculus reproduces four oracles in verse which the Spartan
lawgiver Lycurgus is said to have received from the Delphic oracle. The
second of these runs as follows in the Loeb translation (DS 7.12.2–4; cf.
Strabo 10.4.10):
Two paths there be which farthest parted are,
One leading on to freedom’s honoured halls,
The other to the house of slavery which
All mortals shun. The former path is trod
By those of manly soul and concord sweet [eratēs homonoias];
And on this way I charge you lead the folk;
The latter is the path of loathsome strife
And weak delusion: This the way which thou
Must guard against most carefully.
In these lines homonoia and courage are represented as the conditions of
freedom and thus as a triad of values that the lawgiver needs to make
prime objectives in framing his constitution.
As with most of the other oracles that Diodorus reproduces, scholars
agree that this cannot be a composition of the archaic or early classical
period. There has long been general acceptance of the hypothesis that this
one – with its interest in homonoia and its explicit use of the verb homo-
noein – may have been fabricated at precisely the period (the end of the
Peloponnesian War) when the Athenians started talking about homonoia
as a remedy for stasis and for their own current divisions. Eduard Meyer
proposed that behind Lycurgus’ freedom oracle, we may perhaps see the
hand of the Spartan king Pausanias. Pausanias is said by Strabo to have
deployed a whole series of oracles as propaganda against the other Spartan
king Lysander, who had had him exiled (Str. 7.5.5). No less significantly,
Xenophon reports that after the defeat of the Thirty Tyrants (403 BC), and
following Lysander’s associated eclipse, Pausanias did his best to get the
Athenians to embrace reconciliation and concord (HG 2.4.35).13
If Meyer’s conjecture is correct, and Lycurgus’ oracle on freedom was
indeed in circulation by the end of the fifth century, then it would be
no surprise if Plato were echoing, but at the same time rewriting, its
triad of values in our passage of the Laws. As with the oracle, freedom
13
See Meyer 1892: 211–44; further discussion in de Romilly 1972: 205–8.
is here a fundamental preoccupation: the emphasis on freedom is what
is most striking in both texts. But in the Athenian’s version, friendship
and wisdom take the place of concord and, more importantly, courage or
manliness. In other words, we get precisely the substitution of what the
Athenian Stranger calls the leading virtue – wisdom – for the courage that
Spartans and Cretans mistakenly suppose to be what virtue consists in.
Naturally, the Stranger does not find this non-Spartan triad in the
Spartan constitution, even though he has made much of its measured
character, referring to the admirable system of checks on regal power that
the Spartans have devised (Lg. 691d–692b). Instead, he locates freedom,
wisdom and friendship in the first instance in the Persian monarchy. This
is paradoxical in itself, but is doubly paradoxical because the triad is artic-
ulated here in a form that emphasises its democratic elements.14 This is so,
despite the fact that Persia in the time of Cyrus the Great is introduced
as one of two illustrations of the concept of a mixed or measured con-
stitution, balancing kingship with popular participation (693e, 701e; the
Athens that fought the Persians is the other). Cyrus is represented as giv-
ing his subjects freedom to speak their minds, thereby putting common
soldiers more on an equal footing with their commanders, so that they
were on friendly terms with each other. At the same time, freedom is said
to have provided Cyrus’ subjects with the opportunity to make a con-
tribution to common deliberation through ‘their power of thinking’. In
short, by treating his subjects not as slaves but as free persons Cyrus pro-
moted equality and friendship, and also the potential for ‘sharing in nous
(reason or understanding)’ (694ab).
Cyrus’ Persia sounds for all the world like the Athens celebrated by
Pericles in the funeral speech. It is as though the Athenian is saying: the
Periclean ideal was realised not in his Athens (the Athens over which
he presided at the start of the Peloponnesian War), but in the Persia of
long ago, paradigm of monarchic rule. This is, of course, tantamount to
saying: it has never been realised at all. Plato must know perfectly well
that this romantic (and implicitly and simultaneously anti-Athenian
and anti-Spartan) view of Cyrus’ Persia is fiction. His likeliest model is
Xenophon’s Cyropaedia. It is indeed generally accepted that he is using
the Cyropaedia at this juncture in Book 3. For example, a few lines further
on from the explanation of the way the Persian system instantiates his
14
The salience of democratic elements in the Persian monarchy under Cyrus as here represented is
well observed by Schöpsdau 1994: 453, 459–62, who at 462 notes the possibility of an anti-Isocratean
agenda on Plato’s part.
non-Spartan triad of values, the Athenian Stranger comments on Cyrus’
complete failure to ‘lay hold on a correct education (paideia)’ (Lg. 694c).
This remark is best read as dismissing Xenophon’s account in Book 1 of the
Cyropaedia of Cyrus’ upbringing (too like the Spartan educational regime
for Plato’s taste) as misconceived in its enthusiasm, while at the same time
reinforcing the message – central in the Laws as in the Republic – that edu-
cation is fundamental to the health of any social and political system.15
But the Cyropaedia is not only or mostly about education. Books 7 and
8 of the work have things to say about Cyrus’ enlightened approach to
government that are much more in tune with the picture the Athenian
paints at 694ab – the picture first of relations between soldiers and
their commanders, and then of friendship as a keynote of Cyrus’ Persia.
Xenophon begins Book 8 by having Cyrus emphasise that soldiers are not
slaves, but free men, doing willingly ‘what seems to be of most import-
ance’ (Cyr. 8.1.4). In Book 7, he has had Cyrus invite those around him to
‘give advice on what someone spots as most advantageous’ (Cyr. 7.5.47).
Cyrus in much of his speech and behaviour in Books 7 and 8 is at pains
to address his associates as friends and to treat them accordingly in his
deliberations, when he judges them ‘the most deserving sharers in exer-
tions and in good things’ (Cyr. 7.5.71). It is hard to resist the conclusion
that Plato took this material in the Cyropaedia as the basis of his case for
the Athenian Stranger’s treatment of Persia under Cyrus as governed with
a view to freedom, wisdom and friendship.16
Athenian realities
Plato’s choice of the Persian monarchy to illustrate what is essentially a
kind of democratic ideal comes as something of a shock, even if a mixed
constitution is envisaged, not one where ‘extreme’ freedom prevails (it
remains a form of monarchy). His treatment of Cyrus’ enlightened rule
is in marked contrast with the subsequent account of Athens ‘under its
ancient constitution’ at the time of the Persian Wars (Lg. 698a–699d).
Ancient Athens with its own version of a mixed constitution is presented
as the other historical example of a system that achieved a blend of free-
dom and monarchy. But whereas Cyrus, according to the Stranger, was a
15
Cf. e.g. Schofield 2006: 35–43.
16
For a full discussion of Plato’s use of the Cyropaedia in this section of Book 3, see Schöpsdau 1994:
457–68.
ruler who deliberately fostered freedom, wisdom and friendship, in Athens
things were quite different.
What the Stranger stresses is fear: two sorts of fear. First, there is aidōs,
a sense of shame,17 causing the Athenians to live as ‘willing slaves’ of their
laws (698b, 700a), a description which makes them sound more like
Spartans than Athenians. We recall Demaratus’ words to Xerxes on the
eve of his expedition against Greece in 480 BC (Hdt. 7.164.4):
When the Spartans fight individually, they are second to none, but when
they fight in a body they are best of all. The reason is that though they are
free, they are not completely so, because they have a master over them – the
law – which they fear more than your subjects fear you.
But as well as being motivated by a sense of shame, the fear of dis-
grace in the eyes of one’s friends (at 647b, the greatest contribution of
aidōs is said to be in regard to ‘victory and salvation in war’), the Athenian
Stranger argues that the Athenians at the time of their victory over Xerxes’
fleet at Salamis were motivated also by fear of the enemy to unite in
self-defence. This is something which, along with aidōs, ‘instilled friend-
ship for each other’ (698b–699c). There is no mention of freedom in the
narrative that he develops. The accent is on fear – shame at the thought
of disgrace imbued in them by their subjection to laws already in place,
and terror at the prospect of the enemy – the catalyst for the solidarity of
friendship: ‘because of these factors we found that between us there had
grown intense friendship’ (698c).
In this assessment of the Athenian situation, Plato exploits – but also
subverts – tradition. The idea that Athens flourished in the Persian Wars
because she was still then living under a moderate Solonian democracy
(not the extreme democracy of later times) was, evidently, a favourite
theme among Athenian aristocrats of the fourth century. And the notion
that the Athenians of the age of the Persian Wars responded to the exter-
nal threat by forging homonoia among themselves seems to have gained
currency by the early fourth century – perhaps as an ingredient in the ide-
ology of the amnesty after the defeat and expulsion of the Thirty Tyrants
in 403 BC. We find Andocides referring to the notion three years later (On
the Mysteries 107). According to the so-called ‘decree of Themistocles’, an
inscription generally regarded as a forgery produced for political purposes
17
Correctly so translated by Rowe 2007b: 86 n. 2, 87. The concept has been carefully introduced and
explained in Book 1 (646e–650b), and the Stranger explicitly relates his deployment of it here to
the earlier discussion (699d). Other translators render it differently, e.g. as ‘awe’, ‘reverence’, ‘con-
science’, ‘modesty’, ‘la retenue’.
long after Themistocles’ time, Athenian leaders who had been ostracised
in the years after Marathon were recalled to Salamis ‘so that all Athenians
may be united (homonoountes) in resisting the barbaroi’.18
But Plato’s representation of the Athenians has nothing heroic about
it, in marked contrast with other treatments of Marathon and Salamis,
notably his own in the pastiche funeral oration of the Menexenus.19 There,
those who fought at Salamis are said to have schooled the other Greeks –
by their virtue, it is implied – not to fear the barbarians and their numbers
of ships and men (Mx. 240e–241c). What the Athenian Stranger makes a
point of, as Christopher Rowe has emphasised, is the cowardice of at least
some of the Athenians involved in defending the city against the Persians
at the time of Salamis. (Not all, by any means: in general, fear of the
enemy enhanced the sense of shame that made people submit willingly
to the laws, 698b.) The coward is free – the one ironic occurrence of the
word in his narrative – of the fear of disgrace that is shame. But sheer ter-
ror induced him to cooperate with others and rally to the cause (699c).20
Here the Spartan insistence on the need for courage (reflected in the triad
of values enshrined in Lycurgus’ freedom oracle) is dramatically undercut.
The Athenian Stranger’s point is not that the Athenians did not over-
come the enemy at Salamis, or that they were lacking in wisdom. His
narrative of the fear that the Persian advance induced in them is studded
with expressions indicating what they ‘thought’ or ‘learnt’ or ‘expected’, or
‘heard’ or could ‘see’; it ends with a description of them ‘together conceiv-
ing of one hope of salvation’ and ‘discovering that escape lay with them-
selves alone and with the gods’ (699ab). Of course and above all, they
achieved the principal goal of a city: friendship with each other (698c,
699c). The pre-eminent importance of friendship (in line with the dia-
logue’s ‘hypothesis’)21 was already implicitly reinforced on the last page
of the Stranger’s Persian narrative, when he articulates his diagnosis of
the reason for the Persians’ corruption in subsequent generations. He
18
ML 23.44–5. See Romilly 1972: 203–5.
19
See Rowe 2007b: 91–103, Schofield and Griffith 2010: xviii–xxiii for a defence of the interpretation
of the Menexenus as pastiche, none of which is to be taken as Plato’s own view of Athenian history.
20
See Rowe 2007b: 85–91, to whose construal of the Greek of this key sentence I am in general
indebted. But I disagree with his reading of the deos (fear) that is said to have gripped the cow-
ard and got him to join in the defence of his country. Like Schöpsdau 1994 ad loc. (but unlike
Saunders 1970), Rowe takes it to be (exceptionally for a coward) aidōs: which seems to me diffi-
cult, given that we have just been expressly told that a coward is in that respect fearless, and given
that the main emphasis of the Stranger’s narrative is on the terror the Persian expeditionary force
inspired in the Athenians.
21
See Lg. 5, 743c, quoted at p. 286 above.
explained there that ‘by taking away the freedom of the common people
too completely, and by according undue weight to the authoritarian ele-
ment, they destroyed the basis for friendship and community in the state’
(697cd).
Conclusion
Fear, wisdom, friendship: not quite the triad that the Stranger enunciated
in specifying the aims that the legislator should have in mind in attempt-
ing to devise a balanced constitution (693b–d), and to which he will recur
at the end of his historical excursus (701d). But those aims constitute an
ideal. In Cyrus’ Persia, Plato imagines a scenario in which a wise monarch
gives his people a democratic freedom and equality of speech and thought
that generates friendship and a willingness to make common cause in
the face of danger (694ab, a vision perhaps already anticipated at least
in its general spirit in the opening argument of Book 1, 627e–628a). As
the Republic might have put it, this is voluntary friendship, formed as the
outcome of actively embraced engagement in political deliberation (cf. R.
3, 399bc). Just as Atlantis may be seen as one – ugly – version of imperi-
alistic Athens,22 so Cyrus’ Persia comes close to embodying the idealised
Athens of Pericles’ funeral speech. Plato seems to be giving it house room
as an aspiration. In other words, he thinks it would be best if members of
a political community could freely and of their own choice contribute to
society under the guidance of a wise king. That is what it would be like for
reason to be in control from the start.
But the account that the Stranger gives of the Athenians at the time of
Salamis suggests not only a typically Platonic critique of his compatriots’
inveterate romanticisation of their history, but perhaps something of more
general import. In real historical circumstances, the wisdom and solidarity
that a community is most likely to be able to achieve consist in percep-
tions, expectations and practical improvisation, and through their exercise
the involuntary friendship (cf. R. 3, 399ab) that develops in the course of
response to an emergency, forced by circumstances. Here fear – a reflex of
the lower part of the soul – in its two guises (terror at the prospect of dan-
ger and shame at misbehaving) is what may allow reason to take charge.
As the Stranger stresses, terror enhanced a subjection to laws and rulers
that had already – cowards excepted – been voluntarily accepted, some-
22
Here with most scholars (but not Rowe 2007b: 102 n. 64), I follow Vidal-Naquet 1986.
thing Plato evidently sees as a necessary condition for reason to be able to
operate on any more ambitious scale.23
If we return briefly to the theory of political justice worked out in Laws
6 (pp. 286–7 above), we can now see it as carving a via media between
Cyrus’ Persia and Athens at the time of Salamis. The Laws puts its faith in
law, not in wise kings (this is not the Republic). So the egalitarian pater-
nalism of a Cyrus must remain a utopian dream. But at the same time
the dialogue hopes that we might be able to do better than the Athenians
did, as they coped with the imminent arrival of the invader. Not that the
significance of aidōs or of the willing enslavement to the laws exemplified
by them is to be downplayed. Aidōs is introduced as a key focus for moral
education in the opening phases of the dialogue (646e–650b); while vol-
untary submission to law strikes a keynote that will resound throughout
the rest of the Laws and is, of course, precisely what the whole appar-
atus of ‘preludes’ to laws is designed to elicit.24 But at Athens friendship
was achieved only by happenstance. Ordinary human nature is capable
of more. Wise legislation, operating with the principle of proportionate
equality so as to achieve a carefully considered blend of ‘monarchy’ and
‘democracy’, would (within the limits of what is humanly feasible) prod-
uce political justice. And a just society that is equal in the right sense
will deliver more reliably the rational harmony between its different elem-
ents that constitutes a form of friendship which Plato particularly values
(cf. Grg. 508a; R. 4, 443cd).
23
This is a point at which Platonic and Epicurean thought prove not to be that far apart: what the
Stranger says about the effects of fear at Athens bears some similarity to Hermarchus’ theory of
the initial formation of communities and the subsequent introduction of law. See Porphyry, On
Abstinence 1.7–12.
24
On preludes, see Bobonich 1991.
Bibliography
Adam, J. (ed.) (1963) The Republic, with critical notes, comm. and appendices,
2nd edn. Cambridge.
Ademollo, F. (2011) The Cratylus of Plato: A Commentary. Cambridge.
Alline, H. (1915) Histoire du texte de Platon. Paris.
Annas, J. (1981) An Introduction to Plato’s Republic. Oxford.
(1982) ‘Plato’s myths of judgment’, Phronesis 27: 119–43.
(1995) The Morality of Happiness. Oxford.
(1999) Platonic Ethics: Old and New. Ithaca.
Annas, J. and Rowe C. (eds.) (2002) New Perspectives on Plato, Modern and
Ancient. Cambridge, Mass.
Anscombe, G. E. M. (1969) ‘Parmenides, mystery and contradiction’, Proceedings
of the Aristotelian Society 69: 125–32. [Reprinted in The Collected Philosophical
Papers of G. E. M. Anscombe, vol. I: From Parmenides to Wittgenstein
(Minneapolis, 1981), 3–8.]
Armstrong, J. M. (2004) ‘After the ascent: Plato on becoming like god’, OSAPh
26: 171–83.
Arrighetti, G. (2006) Poesia, poetiche e storia nella riflessione dei Greci. Pisa.
Bailey, D. J. (2012) ‘Megaric Metaphysics’, AncPhil. 32: 303–21.
Balansard, L. (2012) Enquête sur la doxographie platonicienne dans la première par-
tie du Théétète. Sankt Augustin.
Barker, E. (1918) Greek Political Theory: Plato and his Predecessors. London.
Beaney, M. (ed.) (1997) The Frege Reader. Oxford.
Bekker, I. (ed.) (1814) Anecdota Graeca, vol. I. Berlin.
(1826) Platonis et quae vel Platonis esse feruntur vel platonica solent comitari
scripta graece omnia, vol. IV. London.
Bett, R. (1999) Pyrrho: His Antecedents and his Legacy. Oxford.
Bloch, E. (2002) ‘Hemlock poisoning and the death of Socrates: did Plato tell the
truth?’, in Brickhouse and Smith (2002): 255–78.
Blondell, R. (2002) The Play of Character in Plato’s Dialogues. Cambridge.
Bloom, A. (1991) The Republic of Plato, tr. with notes and an interpretative essay,
2nd edn. New York.
Bluck, R. S. (1958) ‘The Phaedrus and reincarnation’, AJPh 79: 156–64.
Bobonich, C. (1991) ‘Persuasion, compulsion and freedom in Plato’s Laws’, CQ
41: 365–88.
298
Bibliography 299
Boys-Stones, G. (1999) ‘Phaedo of Elis and Plato on the Soul’, Phronesis 49: 1–23.
Brancacci, A. (2010) ‘Aristotele e la teoria del sogno del Teeteto’, in Platone: La teoria
del sogno nel Teeteto, ed. G. Mazzara and V. Napoli, Sankt Augustin: 43–60.
Brennan, T. (2003) ‘Stoic moral psychology’, in The Cambridge Companion to the
Stoics, ed. B. Inwood, Cambridge: 257–94.
Brentano, F. (1874) ‘The distinction between mental and physical phenomena’, in
Chisholm (1960): 39–61.
Brickhouse, T. C. and Smith, N. D. (eds.) (2002) The Trial and Execution of
Socrates: Sources and Controversies. Oxford.
(2010) Socratic Moral Psychology. Cambridge.
Brisson, L. (1974) Le même et l’autre dans la structure ontologique du Timée de
Platon. Paris.
(1991) ‘Le corps selon Platon’, in Joubaud (1991): 7–17.
(1994) Platon: Les mots et les mythes, 2nd edn. Paris.
(2003) ‘Le corps des dieux’, in Les dieux de Platon, ed. J. Laurent. Caen: 1–23.
Broackes, J. (2009) ‘Αὐτὸς καθ’ αὑτόν in the Clouds: was Socrates himself a
defender of separable soul and separate Forms?’, CQ 59: 46–59.
Broadie, S. (2001) ‘Soul and body in Plato and Descartes’, PAS 101: 295–308.
(2012) Nature and Divinity in Plato’s Timaeus. Cambridge.
Broadie, S. and Rowe, C. (2002) Aristotle, Nicomachean Ethics, tr. with introduc-
tions and comm. Oxford.
Brown, L. (1986) ‘Being in the Sophist: a syntactical enquiry’, OSAPh 4: 49–70.
(1994) ‘The verb “be” in Greek philosophy: some remarks’, in Companions to
Ancient Thought, vol. 3: Language, ed. S. Everson. Cambridge: 212–36.
Bultrighini, U. (1999) Maledetta democrazia: Studi su Crizia. Alexandria.
Burkert, W. (1972) Lore and Science in Ancient Pythagoreanism, trans. E. L. Minar
Jr. Cambridge, Mass. (Originally published as: Weisheit und Wissenschaft:
Studien zu Pythagoras, Philolaos und Platon. Nuremberg, 1962.)
Burnet, J. (ed.) (1900–1907) Platonis opera (5 vols.). Oxford.
(1905) Platonis opera, vol. 1, 2nd edn. Oxford.
(1909) Platonis opera, vol. 3, corrected edn. Oxford.
(1911) Plato, Phaedo. Oxford.
(1924) Plato’s Euthyphro, Apology of Socrates, and Crito, with notes. Oxford.
(1930) Early Greek Philosophy, 4th edn. London.
Burnyeat, M. F. (1987) ‘Platonism and mathematics: a prelude to discussion’, in
Mathematik und Metaphysik bei Aristoteles, ed. A. Gräser, Bern: 213–40.
(ed.) (1990) The Theaetetus of Plato, with a tr. by M. J. Levett revised by M. F.
Burnyeat. Indianapolis.
(1997) ‘First words’, PCPhS 43: 1–19.
(1999a) ‘Utopia and fantasy: the practicability of Plato’s ideally just city’, in
Fine (1999): 2.175–87.
(1999b) ‘Culture and Society in Plato’s Republic’, The Tanner Lectures on Human
Values 20: 217–324.
(2000) ‘Plato on why mathematics is good for the soul’, in Mathematics and
Necessity: Essays in the History of Philosophy, ed. T. Smiley, Oxford: 1–81.
300 Bibliography
(2002) ‘Plato on how to speak of what is not’, in Le Style de la pensée: Recueil
de textes en hommage à Jacques Brunschwig, ed. M. Canto-Sperber and P.
Pellegrin, Paris: 40–66.
(2003) ‘Apology 30B2–4: Socrates, money and the grammar of γίγνεσθαι’, JHS
123: 1–25.
(2004) ‘Aristotelian revisions: the case of De sensu’, Apeiron 37: 177–80.
(2005) ‘Eikos mythos’, Rhizai 2: 143–65.
Bury, R. G. (ed.) (1932) The Symposium of Plato, with comm. Cambridge.
Callahan, J. F. (1977) ‘Dialectic, myth and history in the philosophy of Plato’, in
Interpretations of Plato, ed. H. F. North, Leiden: 64–85.
Cambiano, G. (2007) ‘Come confutare un libro? Dal Fedro al Teeteto di Platone’,
AntPhilos 1: 99–122.
Campbell, L. (ed.) (1883) The Theaetetus of Plato, with notes, 2nd edn. Oxford.
Canfora, L. (1989) Una società premoderna. Bari.
Carone, G. R. (2005a) Plato’s Cosmology and its Ethical Dimensions. Cambridge.
(2005b) ‘Mind and body in late Plato’, AGPh 87: 227–69.
Chantraine, P. (1968) Dictionnaire étymologique de la langue grecque: Histoire des
mots (4 vols.). Paris.
Chappell, T. (2004) Reading Plato’s Theaetetus. Sankt Augustin.
Chiesara, M. (ed.) (2001) Aristocles of Messene: Testimonia and fragments, with tr.
and comm. Oxford.
Chisholm, R. (ed.) (1960) Realism and the Background of Phenomenology. Glencoe,
Illinois.
Clay, D. (2000) Platonic Questions: Dialogues with the Silent Author. University
Park, Pa.
Cooper, J. M. (1984) ‘Plato’s theory of human motivation’, HPhQ 1: 3–21.
(1990) ‘Political animals and civic friendship’, in Aristoteles’ ‘Politik’ (papers
from the ninth Symposium Aristotelicum), ed. G. Patzig, Göttingen: 220–41.
(ed.) (1997) Plato: Complete Works, with D. Hutchinson (associate ed.).
Indianapolis and Cambridge.
(1999) ‘The psychology of justice in Plato’, in Cooper, J. M., Reason and
Emotion: Essays on Ancient Moral Psychology and Ethical Theory, Princeton:
138–49.
Cornford, F. M. (1935) Plato’s Theory of Knowledge: The Theaetetus and the Sophist
of Plato, tr. with comm. London.
(1937) Plato’s Cosmology: The Timaeus of Plato, tr. with comm. London.
Cousin, V. (ed.) (1864) Procli commentarium in Platonis Parmenidem. Paris.
Crivelli, P. (2012) Plato’s Account of Falsehood: A Study of the Sophist. Cambridge.
Cross, R. C. and Woozley, A. D. (1964) Plato’s Republic: A Philosophical
Commentary. London.
Cushman, R.E. (1958) Therapeia: Plato’s Conception of Philosophy. Chapel Hill.
Dalfen, J. (1974) Polis und Poiesis. Munich.
De Vries, G.J. (1969) A Commentary on the Phaedrus of Plato. Amsterdam.
Deigh, J. (2010) ‘Concepts of emotion in modern philosophy and psychology’, in
Goldie (2010): 17–40.
Bibliography 301
Del Forno, D. (2007) ‘La produzione delle anime immortali nel Timeo’, in La
sapienza di Timeo: Riflessioni in margine al Timeo di Platone, ed. L. M.
Napolitano, Milan: 281–312.
Denyer, N. (1991) Language, Thought and Falsehood in Ancient Greek Philosophy.
London and New York.
Di Marco, M. (ed.) (1989) Timone di Fliunte: Silli. Rome.
Diès, A. (1932) Platon: Oeuvres complètes, vol. VI: La République I–III. Paris.
Dodds, E. R. (1951) The Greeks and the Irrational. Berkeley, Los Angeles and
London.
Döring, K. (ed.) (1972) Die Megariker: Kommentierte Sammlung der Testimonien.
Amsterdam.
(ed.) (1988) Der Sokratesschüler Aristipp und die Kyrenaiker. Mainz.
Dorion, L.-A. (1995) Aristote: Les Réfutations sophistiques, introduction, tr. and
comm. Paris.
(2000) ‘Euthydème et Dionysodore sont-ils des Mégariques?’, in Robinson and
Brisson (2000), 35–50.
(2011) ‘The rise and fall of the Socratic problem’, in The Cambridge Companion
to Socrates, ed. D. R. Morison, Cambridge: 1–23
Dover, K. J. (ed.) (1980) Plato: Symposium, with introduction and notes.
Cambridge.
Duke, E. A. et al. (eds.) (1995) Platonis opera, vol. I. Oxford.
Dümmler, F. (1901) Kleine Schriften, vol. I. Leipzig.
Düring, I. (1941) Herodicus the Cratetean: A Study in Anti-Platonic tradition.
Stockholm.
Eggers Lan, C. E. (1992) ‘Zeus e anima del mondo nel Fedro’, in Understanding the
Phaedrus (papers from the second Symposium Platonicum), ed. L. Rossetti,
Sankt Agustin: 40–6.
(1995) ‘Body and soul in Plato’s anthropology’, Kernos 8: 107–12.
El Murr, D. (forthcoming) ‘Αὐτὸ καθ᾽ αὑτό: la genèse et le sens d’un philos-
ophème platonicien’, in Autos: Idipsum: Figures de l’identité d’Homère à
Augustin, ed. D. Doucet and I. Koch, Aix en Provence.
Emlyn-Jones, C. (ed.) (2005) Plato: Republic 1–2.368c4, with introduction, tr. and
comm. Oxford.
Erler, M. (1998) ‘Idealità e storia: la cornice del Timeo e del Crizia e la Poetica di
Aristotele’, Elenchos 19: 5–28.
(1994) ‘Episode und Exkurs in Drama und Dialog’, in Orchestra: Drama,
Mythos, Bühne (Festschrift H. Flashar), ed. A. Bierl and P. von Möllendorff,
Stuttgart and Leipzig: 318–30.
Ferrari, G. R. F. (1989) ‘Plato and poetry’, in The Cambridge History of Literary
Criticism, vol. I: Classical Criticism, ed. G. A. Kennedy, Cambridge:
92–148.
(2007) ‘The three-part soul’, in The Cambridge Companion to Plato’s Republic,
ed. G. R. F. Ferrari, Cambridge: 165–201.
Fierro, M. A. (2001) ‘Symp. 212a2–7: desire for the truth and the desire for death
and a god-like immortality’, Méthexis 14: 23–43.
302 Bibliography
(2003) Plato’s Theory of Desire in the Symposium and the Republic. PhD thesis,
Durham.
(2006) ‘Platón y los privilegios de los amantes’, Nova Tellus 24: 167–95.
(2008) ‘La concepción platónica del éros en la República’, Diánoia 53, 60:
21–52.
(2010) ‘La concepción del éros universal en Fedro’, in Identidad y diferencia, vol.
II: El pasado y el presente: Sección: Antigüedad y Medioevo, ed. J. Labastida and
V. Aréchiga, Mexico: 11–25.
Fine, G. (ed.) (1999) Plato (2 vols.). Oxford.
Fournier, H. (1946) Les verbes ‘dire’ en grec ancien. Paris.
Fowler, H. N. (tr.) (1925) Plato, vol. 9. (Loeb Classical Library). Cambridge,
Mass. and London.
Frede, M. (1967) Prädikation und Existenzaussage: Platons Gebrauch von “… ist…”
und “… ist nicht…” im Sophistes. Göttingen.
Frege, G. (1891) Über Funktion und Begriff. Jena. [Tr. P. Geach and M. Beaney as
‘Function and concept’, in Beaney (1997): 130–48.]
(1892) ‘Über Sinn und Bedeutung’, Zeitschrift für Philosophie und philosophische
Kritik 100: 25–50. [Tr. M. Black as ‘On Sinn and Bedeutung’, in Beaney
(1997): 151–71.]
(1918) ‘Der Gedanke: Eine logische Untersuchung’, Beiträge zur Philosophie des
deutschen Idealismus 1: 58–77. [Tr. P. Geach and R. H. Stoothoff as ‘Thought’,
in Beaney (1997): 325–45.]
Friedländer, P. (1954) Platon, vol. I: Seinswahrheit und Lebenswirklichkeit, 2nd
edn. Berlin.
Fritz, K. von (1931) ‘Megariker’, Paulys Real-Encyclopädie der Altertumswissenschaft,
ed. G. Wissowa et al. Supplementband V. Stuttgart: cols. 707–24.
Fronterotta, F. (2007) ‘Carone on the mind-body problem in late Plato’, AGPh
89: 231–6.
Furth, M. (1968) ‘Elements of Eleatic ontology’, JHPh 6: 111–32.
Gaiser, K. (1968) Platons ungeschriebene Lehre, 2nd edn. Stuttgart.
(1984) Platone come scrittore filosofico. Naples.
Gallop, D. (1975) Plato: Phaedo, tr. with notes. Oxford.
Gambarara, D. (1984) Alle fonti della filosofia del linguaggio. Rome.
Gentili, B. and Prato, C. (1988) Poetarum elegiacorum testimonia et fragmenta, vol.
I, 2nd edn. Leipzig.
Giannantoni, G. (ed.) (1958) I Cirenaici. Florence.
(1990) Socratis et Socraticorum reliquiae (4 vols.). Naples.
(1996) ‘Il concetto di percezione nella filosofia cirenaica’, in Lezioni Socratiche,
ed. G. Giannantoni and M. Narcy, Naples: 181–203.
Gill, C. (1976) ‘The origin of the Atlantis myth’, Trivium 11: 1–11.
(1977) ‘The genre of the Atlantis Story’, CPh 72: 287–304.
(1979a) review of Welliver (1977), CR 29: 163–4.
(1979b) ‘Plato’s Atlantis story and the birth of fiction’, Ph&Lit 3: 64–78.
(1980) Plato: The Atlantis Story. Bristol.
Bibliography 303
(1993) ‘Plato on falsehood – not fiction’, in Lies and Fiction in the Ancient
World, ed. C. Gill and T. P. Wiseman, Austin and Exeter: 38–87.
(1996a) ‘Afterword: dialectic and the dialogue form in late Plato’, in Gill and
McCabe (1996): 183–311.
(1996b) Personality in Greek Epic, Tragedy, and Philosophy: The Self in Dialogue.
Oxford.
(1998) ‘Ethical reflection and the shaping of character’, Proceedings of the Boston
Area Colloquium in Ancient Philosophy 12: 193–225.
(2000a) ‘Protreptic and dialectic in Plato’s Euthydemus’, in Robinson and
Brisson (2000): 133–43.
(2000b) ‘The body’s fault? Plato’s Timaeus on psychic illness’, in Reason and
Necessity: Essays on Plato’s Timaeus, ed. M. R. Wright, London: 59–84.
(2002a) ‘Dialectic and the dialogue form’, in Annas and Rowe (2002): 145–71.
(2002b) ‘The Death of Socrates’, in Brickhouse and Smith (2002): 251–5.
[Reprinted from CQ 23 (1973): 25–8.]
(2004) ‘Plato’s Republic: an ideal culture of knowledge’, in Ideal and Culture
of Knowledge in Plato, ed. A. Becker, W. Detel and P. Scholtz, Stuttgart:
37–55.
(ed.) (2005) Virtue, Norms and Objectivity: Issues in Ancient and Modern Ethics.
Oxford.
(2006) ‘The Platonic dialogue’, in Gill and Pellegrin (2006): 136–50.
(2007) ‘The Good and mathematics’, in Pursuing The Good: Ethics and
Metaphysics in Plato’s Republic, ed. D. Cairns, F.-G. Hermann and T. Penner,
Edinburgh: 251–74.
(2013) ‘What is the point of the tripartite psyche in Plato’s Republic?’, in
Notomi and Brisson (2013: 161–7).
Gill, C. and McCabe, M. M. (eds.) (1996) Form and Argument in Late Plato.
Oxford.
Gill, C. and Renaud, F. (eds.) (2010) Hermeneutic Philosophy and Plato: Gadamer’s
Response to the Philebus. Sankt Augustin.
Gill, M. L. and Pellegrin, P. (eds.) (2006) A Companion to Ancient Philosophy.
Oxford.
Giordano, L. (2000) ‘Alcibiade maestro: la recezione di Tucidide nell’Alcibiade I’,
SIFC 18: 55–70.
Giovannini, A. (1985) ‘Peut-on démythifier l’Atlantide?’, MH 42: 151–6.
Goldie, P. (ed.) (2010) The Oxford Handbook of the Philosophy of Emotion.
Oxford.
Goldstein, J. A. (1968) The Letters of Demosthenes. New York and London.
Greene, W. C. (ed.) (1938) Scholia Platonica. Haverford.
Griswold, C. (1986) Self-Knowledge in Plato’s Phaedrus. New Haven.
(ed.) (1988) Platonic Writings, Platonic Readings. New Haven.
Grote, G. (1865) Plato and the Other Companions of Sokrates, vol. II. London.
Guthrie, W. K. C. (1962) A History of Greek Philosophy, vol. I: The Earlier
Presocratics and the Pythagoreans. Cambridge.
304 Bibliography
Hackforth, R. (1952) Plato’s Phaedrus, tr. with introduction and comm.
Cambridge.
(1955) Plato’s Phaedo, tr. with introduction and comm. Cambridge.
Halliwell, S. (1986) Aristotle’s Poetics. London.
(1992) ‘Pleasure, understanding, and emotion in Aristotle’s Poetics’, in Essays on
Aristotle’s Poetics, ed. A. Oksenberg Rorty, Princeton: 241–60.
(2002) The Aesthetics of Mimesis. Princeton.
Harris, J. (2000) ‘Intimations of immortality’, Science 288, no. 5463: 59.
Harvey, F. D. (1965) ‘Two kinds of equality’, C&M 26: 101–46.
Hawtrey R. S. W. (1981) Commentary on Plato’s Euthydemus. Philadelphia.
Heidegger, M. (2001) Gesamtausgabe II. Abteilung: Vorlesungen 1919–1944, vols.
36–37: Sein und Wahreit. Frankfurt.
Heitsch, E. (1993) Platon: Phaidros, tr. and comm. Göttingen.
Helm, B. W. (2010) ‘Emotions and motivation: reconsidering Neo-Jamesian
accounts’, in Goldie (2010): 303–23.
Hicks R. D. (ed.) (1925) Diogenes Laertius: Lives of Eminent Philosophers, with tr.
(2 vols.) (Loeb Classical Library). Cambridge, Mass.
Hirmer, J. (1897) ‘Entstehung und Komposition der platonischen Politeia’,
Jahrbücher für Classische Philologie, supplementary vol. 23: 579–678.
Horn, H.-J. (1988) ‘Zum neunten Kapitel der aristotelischen Poetik’, RhM 131:
113–36.
Huffman, C. A. (ed.) (1993) Philolaus of Croton: Pythagorean and Presocratic, with
tr., comm. and interpretative essays. Cambridge.
Hunter, R. (2004) Plato’s Symposium. Oxford and New York.
Irwin, T. (1995) Plato’s Ethics, Oxford.
Janaway, C. (1995) Images of Excellence: Plato’s Critique of the Arts. Oxford.
Johansen, T. K. (2000) ‘Body, soul and tripartition in Plato’s Timaeus’, OSAPh
19: 87–111.
(2004) Plato’s Natural Philosophy: A Study of the Timaeus-Critias. Cambridge.
Joseph, H. (1948) Knowledge and the Good in Plato’s Republic. Oxford.
Joubaud, C. (1991) Le corps humain dans la philosophie platonicienne: Étude à
partir du Timée. Paris.
Jouët-Pastré, E. (2006) Le jeu et le sérieux dans les Lois de Platon. Sankt Augustin.
Jowett, B. (1892) The Dialogues of Plato, tr. (5 vols.), 3rd edn. London.
Kahn, C. H. (1976) ‘Why existence does not emerge as a distinct concept in
Greek philosophy’, AGPh 58: 323–34.
(1996) Plato and the Socratic Dialogue: The Philosophical Use of a Literary Form.
Cambridge.
(2002) ‘On Platonic chronology’, in Annas and Rowe (2002): 93–127.
(2003) The Verb ‘Be’ in Ancient Greek, 2nd edn. Indianapolis. [A reprint of the
first edition of 1973, ‘with a new introductory essay’.]
Kant, I. (1787) Critik der Reinen Vernunft, 2nd edn. Riga. [Reprinted in Kant’s
Gesammelte Schriften, vol. III. Berlin, 1911.]
Kaplan, D. (1968) ‘Quantifying in’, Synthèse 19: 178–214.
Karasmanis, V. (1988) ‘Plato’s Republic: the Line and the Cave’, Apeiron 21:
147–71.
Bibliography 305
Kerferd, G. B. (1981) The Sophistic Movement. Cambridge.
Klagge, J. C. and Smith, N. D. (eds.) (1992) Methods of Interpreting Plato and his
Dialogues (OSAPh supplementary volume). Oxford.
Krinks, P. (2011) Developing a Defence of Erōs: The Unity of Plato’s Symposium. PhD
thesis, London.
Krohn, A. (1876) Der platonische Staat. Halle.
Lampert, L. and Planeaux, C. (1998), ‘Who’s who in Plato’s Timaeus-Critias and
why’, RMeta 52: 87–125.
Lane, M. (2007) ‘Virtue as the love of knowledge in Plato’s Symposium and
Republic’, in Scott (2007): 44–67.
Laplace, M. (1984) ‘Le Critias de Platon, ou l’ellipse d’une épopée’, Hermes 112:
377–82.
Lear, J. (1980) Aristotle and Logical Theory. Cambridge.
Lee, D. and Johansen, T. K. (2008), Plato: Timaeus-Critias, tr. with annotations
(Penguin Classics). London and New York.
Leroux, G. (2002) Platon: La République, tr., introduction and notes. Paris.
Long, A. A. (1996) ‘Socrates and Hellenistic philosophy’, in Long, A. A., Stoic
Studies, Cambridge: 1–34. [Reprinted from CQ 38 (1988): 150–71.]
Lorenz, H. (2006) The Brute Within: Appetitive Desire in Plato and Aristotle.
Oxford.
Luce, J. V. (1969) The End of Atlantis. London.
McCabe, M. M. (1993) ‘Persistent fallacies’, PAS 93: 73–93.
(1998) ‘Silencing the sophists: the drama of Plato’s Euthydemus’, Proceedings of
the Boston Area Colloquium in Ancient Philosophy 14: 139–68.
(2002a) ‘Developing the good by itself: critical strategies in Plato’s Euthydemus’.
Internet Journal of the International Plato Society 2. http://gramata.univ-paris1.
fr/Plato/article20.html.
(2002b) ‘Indifference readings: Plato and the Stoa on Socratic ethics’, in Classics
in Progress: Essays on Ancient Greece and Rome, ed. T. P. Wiseman, Oxford:
363–98.
(2005) ‘Out of the labyrinth: Plato’s attack on consequentialism’, in Gill (2005):
189–214.
(2006) ‘Does your Plato bite?’, in Agonistes (Festschrift Denis O’Brien), ed. J.
Dillon and M. Dixsaut, Aldershot: 107–19.
(2009) ‘Escaping one’s own notice knowing: Meno’s paradox again’, PAS 109:
233–56.
McDowell, J. (1973) Plato: Theaetetus, tr. with notes, Oxford.
McGibbon, D. (1968) ‘The fall of the soul in Plato’s Phaedrus’, CQ 14: 56–63.
Malcolm, J. (1962) ‘The Line and the Cave’, Phronesis 7: 38–45.
Marchiori, A. (2001) ‘Libro V: traduzione e commento’, in Ateneo: I Deipnosofisti,
vol. I, ed. L. Canfora, Rome: 524–40.
Martina, A. (ed.) (1968) Solon: Testimonia veterum. Rome.
Meiggs, R., and Lewis, D. M. (eds.) (1988) A Selection of Greek Historical
Inscriptions to the End of the Fifth Century B.C. Revised edn. Oxford.
Meyer, E. (1892) Forschungen zur Alten Geschichte, vol. I. Halle.
Moline, J. (1978) ‘Plato and the complexity of the psyche’, AGPh 60: 1–26.
306 Bibliography
Moore, G. E. (1936) ‘Is existence a predicate?’, Aristotelian Society, Supplementary
Volume 15: 175–188. [Reprinted in G. E. Moore, Philosophical Papers (London
and New York, 1959): 115–26.]
Moreau, J. (1939) L’âme du monde de Platon aux stoiciens. Paris.
Morgan, K. (1998): ‘Designer history: Plato’s Atlantis story and fourth-century
ideology’, JHS 118: 101–18.
Muller, R. (ed.) (1985) Les Mégariques: Fragments et témoignages, Paris.
Murphy N. R. (1951) The Interpretation of Plato’s Republic. London.
Naddaf, G. (1994) ‘The Atlantis myth: an introduction to Plato’s later philosophy
of history’, Phoenix 48: 189–209.
Naddaff, R. (2002) Exiling the Poets: The Production of Censorship in Plato’s
Republic. Chicago.
Nails, D. (2002) The People of Plato: Prosopography of Plato and the other Socratics.
Indianapolis.
Narcy, M. (1984) Le Philosophe et son double: un commentaire de l’Euthydème de
Platon. Paris.
(1994) Platon: Théétète, tr. with introduction and notes. Paris.
(2007) ‘Che cosa è un dialogo socratico?’, in Il Socrate dei dialoghi, ed. G.
Mazzara, M. Narcy and L. Rossetti, Bari: 21–32.
(2012) ‘Platon – Sophiste’, in Dictionnaire des philosophes antiques, vol. 5a, ed. R.
Goulet, Paris: 700–6.
Nehamas, A. (1982) ‘Plato on imitation and poetry in Republic 10’, in Plato
on Beauty, Wisdom, and the Arts, ed. J. Moravcsik and P. Temko, Totowa:
47–78.
Nesselrath, H. G. (2002) Platon und die Erfindung von Atlantis. Leipzig and
Munich.
(2006) Platon: Kritias, tr. and comm. Göttingen.
Nightingale, A. W. (2004) Spectacles of Truth in Classical Greek Philosophy: Theoria
in its Cultural Context. Cambridge.
Nilsson, M. P. (1951) Cults, Myths, Oracles, and Politics in Ancient Greece. Lund.
Notomi, N. (2005) [The Birth of the Philosopher: People around Socrates] (in
Japanese). Tokyo.
(2006) ‘Plato’s metaphysics and dialectic’, in Gill and Pellegrin (2006):
192–211.
(2007a) ‘La metafisica come risveglio dell’anima: una lettura etica del Fedone
di Platone’, in Interiorità e anima: La psyche in Platone, ed. M. Migliori, L.
Napolitano Valditara and A. Fermani, Milan: 201–10.
(2007b) ‘Plato against Parmenides: Sophist 236D–242B’, in Reading Ancient
Texts, vol. I: Presocratics and Plato, ed. S. Stern-Gillet and K. Corrigan,
Leiden and Boston: 167–87.
(2010) ‘Socrates versus sophists: Plato’s invention?’, in Socratica 2008: Studies in
Ancient Socratic Literature, ed. L. Rossetti and A. Stavru, Bari: 71–88.
Notomi, N., Bonazzi, M., Dorion, L.-A., Van Ackeren, M. and Hatano, T. (2009)
‘“Socratic” dialogues’, Plato: the Internet Journal of the International Plato
Society 9. http://gramata.univ-paris1.fr/Plato/ article88.html.
Bibliography 307
Notomi, N. and Brisson, L. (eds.) (2013) Dialogues on Plato’s Politeia (Republic)
(selected papers from the ninth Symposium Platonicum). Sankt Augustin.
Nussbaum, M. (1986) The Fragility of Goodness: Luck and Ethics in Greek Tragedy
and Philosophy. Cambridge.
O’Brien, D. (1987) Le Poème de Parménide: Texte, traduction, essai critique (‘en
collaboration avec J. Frère pour la traduction française’) = P. Aubenque (ed.)
Études sur Parménide (Paris, 1987), vol. 1.
(1991) ‘Platon et Plotin sur la doctrine des parties de l’autre’, Revue Philosophique
de la France et de l’Étranger 181 (année 116): 501–12.
(1995) Le Non-Être: Deux études sur le ‘Sophiste’ de Platon. Sankt-Augustin.
(2000) ‘Parmenides and Plato on what is not’, in The Winged Chariot: Collected
Essays on Plato and Platonism in Honour of L. M. de Rijk, ed. M. Kardaun
and J. Spruyt, Leiden, Boston, Cologne: 19–104.
(2009a) ‘Plato the Pythagorean: a critical study of Kenneth Sayre, Plato’s
Late Ontology: A Riddle Resolved’, The International Journal of the Platonic
Tradition 3: 58–77.
(2009b) ‘Plotin: la question du mal’, Cahiers d’Études Lévinassiennes 8:
295–316.
(2011) ‘The Stranger’s “farewell” (258e6–259a1)’, in Plato’s Sophist (proceed-
ings of the seventh Symposium Platonicum Pragense), ed. A. HavlíČek and F.
Karfík, Prague: 199–220.
(2012) ‘Plotinus on the making of matter, part III: the essential background’,
The International Journal of the Platonic Tradition 6: 27–80.
Ostenfeld, E. N. (1982) Forms, Matter and Mind: Three Strands in Plato’s
Metaphysics. The Hague, Boston and London.
(1987) Ancient Greek Psychology and the Modern Body-Mind Debate. Aarhus.
Owen, G. E. L. (1971) ‘Plato on not-being’, in Plato: A Collection of Critical
Essays, vol. I: Metaphysics and Epistemology, ed. G. Vlastos, Garden City, NY:
223–67. [Reprinted in G. E. L. Owen, Logic, Science and Dialectic: Collected
Papers in Greek Philosophy, ed. M. Nussbaum (London, 1986): 104–37.]
Peirce, C. S. (1878) ‘How to make our ideas clear’, Popular Science Monthly 12:
286–302.
Penner, T. (1971) ‘Thought and desire in Plato’, in A Collection of Critical Essays,
vol. II, ed. G. Vlastos, New York: 96–118.
(1987) The Ascent from Nominalism: Some Existence Arguments in Plato’s Middle
Dialogues. Dordrecht.
(1988) ‘Socrates on the impossibility of belief-relative sciences’, Proceedings of
the Boston Area Colloquium in Ancient Philosophy 3: 263–325.
(2005a) ‘Platonic Justice and what we mean by “justice”’, in Plato: the Internet
Journal of the International Plato Society 5. http://gramata.univ-paris1.fr/
Plato/article60.html.
(2005b) ‘Socratic ethics: ultra-realism, determinism, and ethical truth’, in Gill
(2005): 157–87.
(2006a) ‘Plato’s ethics: early and middle dialogues’, in Gill and Pellegrin
(2006): 151–69.
308 Bibliography
(2006b) ‘The forms in the Republic’, in A Companion to Plato’s Republic, ed. G.
Santas, Oxford: 234–62.
(2007a) ‘What is the Form of the Good the Form of? A question about the
plot of the Republic’, in Pursuing the Good: Ethics and Metaphysics in Plato’s
Republic, ed. D. Cairns, F.-G. Herrmann and T. Penner, Edinburgh: 15–41.
(2007b) ‘The death of the so-called “Socratic elenchus”’, in Gorgias-Meno
(papers from the seventh Symposium Platonicum), ed. M. Erler and L.
Brisson, Sankt Augustin: 3–19.
(2011) ‘Socratic ethics and the Socratic psychology of action: a philosophical
framework’, in The Cambridge Companion to Socrates, ed. D. R. Morrison,
Cambridge: 260–92.
(forthcoming), ‘Plato takes aim at intentionality’.
Penner, T. and Rowe, C. (2005) Plato’s Lysis. Cambridge.
Poratti, A. (2010) Platón: Fedro, tr. with introduction and notes. Madrid.
Pradeau, J.-F. (1997) Le monde de la politique. Sankt Augustin.
(2002) Plato and the City: A New Introduction to Plato’s Political Thought, trans.
J. Lloyd. Exeter.
Price, A. W. (1995) Mental Conflict. London and New York.
Primavesi, O. (2007) ‘Ein Blick in den Stollen von Skepsis: vier Kapitel zur frühen
Überlieferung des Corpus Aristotelicum’, Philologus 151: 51–77.
Reeves, H., Rosnay, J. de, Coppens, Y. and Simonnet, D. (2008) La historia más
bella del mundo: Los secretos de nuestros orígenes, tr. O. L. Molina. Barcelona.
Regali, M. (2012) Il opera e il demiurgo. Teoria e prassi della produzione letteraria
nel Timeo e nel Crizia di Platone. Sankt Augustin.
Renaud, F. (1999) Die Resokratisierung Platons: Die platonische Hermeneutik
Hans-Georg Gadamer. Sankt Augustin.
Renehan, R. (1976) Studies in Greek Texts. Göttingen.
Riginos, A. S. (1976) Platonica: The Anecdotes Concerning the Life and Writings of
Plato. Leiden.
Robin, L. (ed.) (1954) Platon: Oeuvres complètes, vol. IV part 3: Phèdre, with tr. Paris.
Robinson, T. M. and Brisson, L. (eds.) (2000) Plato: Euthydemus, Lysis, Charmides
(papers from the fifth Symposium Platonicum). Sankt Augustin.
Romilly, J. de (1951) Thucydide et l’impérialisme athénien. 2nd edn. Paris.
(1972) ‘Vocabulaire et propagande ou les premiers emplois du mot homonoia’,
in Mélanges de linguistique et de philologie grecques offerts à Pierre Chantraine,
ed. F. Bader, Paris: 199–209.
Rosenmeyer, T. G. (1956) ‘Plato’s Atlantis myth: Timaeus or Critias?’, Phoenix 10:
163–72.
Rosenthal, E. (ed.) (1956) Averroes’ Commentary on Plato’s Republic, with tr. and
notes. Cambridge.
Rossetti, L. (1973) ‘Socratica in Fedone di Elide’, StudUrb(B) 47: 364–81.
(1980) ‘Ricerche sui dialoghi socratici di Fedone e di Euclide’, Hermes 108:
183–200.
Rowe, C. (ed.) (1988) Plato: Phaedrus, with tr. and commentary, 2nd edn.
Warminster.
Bibliography 309
(1991) ‘L’argument par “affinité” dans le Phédon’, RPhilos 181: 463–77.
(ed.) (1993) Plato: Phaedo, with introduction and notes. Cambridge.
(ed.) (1995) Plato: Statesman, with tr. and comm. Warminster.
(1996) ‘The Politicus: structure and form’, in Gill and McCabe (1996): 153–78.
(1997) ‘Why is the ideal Athens of the Timaeus-Critias not ruled by philoso-
phers?’, Méthexis 10: 51–7.
(ed.) (1998) Plato: Symposium, translated with introduction and comm.
Warminster.
(1999) ‘Myth, history, and dialectic in Plato’s Republic and Timeus-Critias’, in
From Myth to Reason? Studies in the Development of Greek Thought, ed. R.
Buxton, Oxford: 263–76.
(2001) ‘The concept of philosophy (philosophia) in Plato’s Phaedo’, in Plato’s
Phaedo (proceedings of the second Symposium Platonicum Pragenese), ed. A.
Havlíček and F. Karfík, Prague: 34–47.
(2002) ‘Comments on Penner’, in Annas and Rowe (2002): 213–25.
(2003a) ‘Socrates and Plato on virtue and the good: an analytical approach’, in
New Images of Plato, ed. G. Reale and S. Scolnicov, Sankt Augustin: 253–64.
(2003b) ‘“All our desires are for the good”: reflections on some key Platonic
dialogues’, in Plato Ethicus: Philosophy is Life, ed. M. Migliori and L. M.
Napolitano Valditara, Sankt Augustin: 265–72.
(2005) ‘Les parties de l’âme et le désir du bien dans la République’, in Études sur
la République, ed. M Dixsaut and F. Teisserenc, Paris: 209–223.
(2006a) ‘Interpreting Plato’, in A Companion to Plato, ed. H. Benson, Oxford:
13–24.
(2006b) ‘The Symposium as a Socratic dialogue’, in Plato’s Symposium: Issues
in Interpretation and Reception, ed. J. H. Lesher, D. Nails and F. Sheffield,
Cambridge, Mass.: 9–22.
(2007a) Plato and the Art of Philosophical Writing. Cambridge.
(2007b) ‘Plato and the Persian wars’, in Cultural Responses to the Persian Wars,
ed. E. Bridges, E. Hall and P. J. Rhodes, Oxford: 85–104.
(2010) ‘L’argument relatif à la pleonexia dans le premier livre de la République:
deux conceptions de ce qu’est “avoir avantage”’, in Aglaïa, autour de Platon:
Mélanges offerts à Monique Dixsaut, ed. A. Brancacci, D. El Murr and D. P.
Taormina, Paris: 291–305.
(2012a) Plato: The Republic, tr. with annotations (Penguin Classics). London.
(2012b) ‘Socrates on reason, appetite and passion: a response to Thomas C.
Brickhouse and Nicholas D. Smith, Socratic Moral Psychology’, The Journal
of Ethics 16: 305–24.
(2013) ‘On justice and the other virtues: whose justice, whose virtues?’, in
Notomi and Brisson (2013): 49–59.
Ryle, G. (1966) Plato’s Progress. Cambridge.
Rysiew, P. (2007) ‘Epistemic contextualism’, in Stanford Encyclopedia of Philosophy.
http://plato.stanford.edu/entries/contextualism-epistemology/.
Saunders, T. J. (1970) Plato: The Laws, tr. with annotations (Penguin Classics).
Harmondsworth.
310 Bibliography
Schofield, M. (1998) ‘Political friendship and the ideology of reciprocity’, in
Kosmos: Essays in Order, Conflict and Community in Classical Athens, ed. P.
Cartledge, P. Millett and S. von Reden, Cambridge: 37–51.
(2006) Plato: Political Philosophy. Oxford.
(2007) ‘Metaspeleology’, in Scott (2007): 216–32.
Schofield, M. and Griffith, T. (2010) Plato: Gorgias, Menexenus, Protagoras, tr.
with annotations. Cambridge.
Schöpsdau, K. (1994) Platon: Nomoi (Gesetze). Buch I–III, tr. and comm.
Göttingen.
Schwyzer, H., and Debrunner, A. (1959) Griechische Grammatik, vol. II: Syntax
und syntaktische Stilistik, 2nd edn. Munich.
Scott, D. (1987) ‘Platonic anamnesis revisited’, CQ 37: 346–66.
(2007) Maieusis: Essays on Ancient Philosophy in Honour of Myles Burnyeat.
Oxford.
Sedley, D. (1989) ‘Teleology and myth in the Phaedo’, Proceedings of the Boston
(1995) ‘The dramatis personae of Plato’s Phaedo’, in Philosophical Dialogues:
Plato, Hume, Wittgenstein, ed. T. J. Smiley, Oxford: 3–26.
(1996) ‘Three Platonist interpretations of the Theaetetus’, in Gill and McCabe
(1996): 79–104.
(1999) ‘The ideal of godlikeness’, in Fine (1999): 2.309–28.
(2004) The Midwife of Platonism: Text and Subtext in Plato’s Theaetetus.
Oxford.
(2005) ‘Plato’s tsunami’, Hyperboreus 11: 205–14.
(2007a) ‘Philosophy, the forms, and the art of ruling’, in The Cambridge
Companion to Plato’s Republic, ed. G. R. F. Ferrari, Cambridge: 256–83.
(2007b) Creationism and its Critics in Antiquity. Berkeley.
(2009) ‘A Thrasyllan interpretation of the Theaetetus’, in The Oxyrhynchus
Papyri 73: 65–71.
Selinger, E. (2009) ‘Cyborgs’, in A Companion to the Philosophy of Technology,
ed. J. K. B. Olsen, S. A. Pendersen and V. F. Hendricks, Chichester and
Malden, Mass.: 154–6.
Slings, S. R. (ed.) (2003) Platonis Respublica. Oxford.
Stallbaum, G. (ed.) (1836) Platonis opera omnia, vol. VI. Gotha.
Steel, C. (ed.) (2007–2009) Procli in Platonis Parmenidem commentaria (3 vols.).
Oxford.
Steinthal, H. (1993) ‘μόγις und ἐξαίφνης: Platon über die Grenzen des Erkennens’,
in Antike Texte in Forschung und Schule (Festschrift Willibald Heilmann), ed.
C. Neumeister, Frankfurt: 99–105.
Stella, M. (1998) ‘Socrate, Adimanto, Glaucone: Racconto di ricerca e rappresen-
tazione comica’, in Platone: La Repubblica, ed. M. Vegetti, Naples: 233–79.
Strang, C. (1986) ‘Plato’s analogy of the cave’, OSAPh 4: 19–34.
Strauss, B. S. (1993) Fathers and Sons in Athens: Ideology and Society in the Era of
the Peloponnesian War. Princeton.
Bibliography 311
Striker, G. (1996) ‘Plato’s Socrates and the Stoics’, in Striker, G., Essays on
Hellenistic Epistemology and Ethics, Cambridge: 316–24.
Szlezák, T. A. (1993) ‘Atlantis und Troia, Platon und Homer: Bemerkungen zum
Wahrheitsanspruch des Atlantis-Mythos’, ST 3: 233–7.
Taylor, A. E. (1928) A Commentary on Plato’s Timaeus. Oxford.
Thompson, E. S. (1901) The Meno of Plato, with introduction and notes.
London.
Tsouna, V. (1998) The Epistemology of the Cyrenaic School. Cambridge.
Tulli, M. (1994) ‘Il Crizia e la famiglia di Platone’, SCO 44: 95–107.
(2003) ‘L’Atene di Aspasia: tradizione del racconto e ricerca dell’ ideale nel
Menesseno di Platone’, in Evento, racconto, scrittura nell’antichità classica, ed.
A. Casanova and P. Desideri, Florence: 91–106.
Van Eck, J. (2002) ‘Not-being and difference: on Plato’s Sophist 256D5–258E3’,
OSAPh 23 : 63–84.
Vegetti, M. (ed.) (1998) Platone: La Repubblica, vol. I: Libro I. Naples.
Vidal-Naquet, P. (1986) ‘Athens and Atlantis: structure and signification of a
Platonic myth’, in Vidal-Naquet, P., The Black Hunter: Forms of Thought and
Forms of Society in the Greek World, tr. A. Szegedy-Mascak, Baltimore and
London: 263–84. (Based on ‘Athènes et l’Atlantide: structure et signification
d’un mythe platonicien’, REG 77 (1964): 420–44.)
(2007) The Atlantis Story: A Short History of Plato’s Myth, tr. J. Lloyd. Exeter.
Vlastos, G. (1977) ‘The theory of social justice in the polis in Plato’s Republic’,
in Interpretations of Plato: A Swarthmore Symposium, ed. H. North, Leiden:
1–40.
(1981) Platonic Studies, 2nd edn. Princeton.
(1991) Socrates: Ironist and Moral Philosopher. Cambridge.
Weil, R. (1959) L’“Archéologie” de Platon. Paris.
Welliver, W. (1977) Character, Plot and Thought in Plato’s Timaeus-Critias. Leiden.
West, M. L. (ed.) (1992) Iambi et elegi Graeci ante Alexandrum cantati (2 vols.),
2nd edn. Oxford.
White, N. P. (1979) A Companion to Plato’s Republic. Oxford.
(2002) Individual and Conflict in Greek Ethics. Oxford.
Wilamowitz, U. von (1969) Platon, vol. II. 4th edn, Dublin and Zürich.
Williams, B. A. O. (2002) Truth and Truthfulness. Princeton.
Zilioli, U. (2007) The Challenge of Relativism: Plato’s Subtlest Enemy. Aldershot.
(2012) The Cyrenaics. Durham.
Christopher Rowe
P u b l i c at i o n s 1969–2012
1969
‘The Relation between the Eudemian and Nicomachean Ethics of Aristotle’, PhD
diss. Cambridge. [Written under the supervision of H. J. Easterling.]
1971
The Eudemian and Nicomachean Ethics: A Study in the Development of Aristotle’s
Thought. Cambridge. [Revised version of the PhD thesis q.v. above, 1969.]
‘The meaning of φρόνησις in the Eudemian Ethics’, in Untersuchungen zur
Eudemischen Ethik (papers from the fifth Symposium Aristotelicum), ed. P.
Moraux and D. Harlfinger, Berlin: 73–92.
1972
‘Conceptions of colour and colour symbolism in the ancient world’, Eranos-Jb 41:
327–64.
review of G. Vlastos (ed.), Plato: A Collection of Critical Essays, vol. I: Metaphysics
and Epistemology, in JHS 92: 218–19.
1974
‘God, man and nature: ancient Greek views on the foundation of moral values’,
Eranos-Jb 43: 255–91.
review of S. Lilja, The Treatment of Odours in the Poetry of Antiquity, in JHS 94:
206–7.
review of R. Sorabji, Aristotle on Memory, in JHS 94: 194–5.
1975
‘A reply to John Cooper on the Magna Moralia’, AJPh 96: 160–72.
review of T. Ebert, Meinung und Wissen in der Philosophie Platons, in JHS 95:
200–1.
subject review (Ancient Philosophy), in G&R 22: 97–9.
1976
An Introduction to Greek Ethics. London. [Reprinted 1993.]
‘One and many in Greek religion’, Eranos-Jb 45: 37–67.
312
Christopher Rowe publications 313
1977
‘Aims and methods in Aristotle’s Politics’, CQ 27: 159–72.
‘Conceptions of colour and colour symbolism in the ancient world’, in Color
Symbolism: Six Excerpts from the Eranos Yearbook 1972, ed. A. Portman et al,
Zurich: 23–54. [Reprint of ‘Conceptions of colour’, q.v. 1972.]
review of W. W. Fortenbaugh, Aristotle on Emotion, in JHS 97: 183.
review of C. C. W. Taylor, Plato: Protagoras, in PhilosQ 27: 353–4.
1978
Essential Hesiod (Theogony 1–232, 453–733, Works and Days 1–307), with introduc-
tion and notes. Bristol. [Reprinted 1988.]
1979
‘The proof from relatives in the Peri Ideon: further reconsideration’, Phronesis 24:
270–81.
‘Justice and temperance in Republic IV’, in Arktouros: Hellenic Studies Presented to
Bernard M. W. Knox on the Occasion of his 65th Birthday, ed. G. W. Bowersock,
W. Burkert and M. C. J. Putnam, Berlin: 336–44.
review of W. Leszl, Il ‘De Ideis’ di Aristotele e la teoria platonica delle idee, in CR
29: 77–9.
review of R. J. Sullivan, Morality and the Good Life, in JHS 99: 179.
1981
‘The character of Aristotle’s writings on politics’, in Proceedings of the World
Congress on Aristotle (held at Thessaloniki, 7–14 August 1978), Athens: 93–7.
review of E. M. Wood and N. Wood, Classical Ideology and Ancient Political
Theory, in JHS 101: 179–80.
1982
with M. Welbourne and C. J. F. Williams, ‘Knowledge, perception and memory:
Theaetetus 166b’, CQ 32: 304–6.
review of T. Irwin, Plato: Gorgias, in JHS 102: 249–50.
review of A. Kenny, Aristotle’s Theory of the Will, in JHS 102: 250–3.
review of P. J. Rhodes, A Commentary on the Aristotelian Athenaion Politeia, in
JACT Bulletin Review 59: iii.
1983
‘“Archaic thought” in Hesiod’, JHS 103: 124–35.
‘De Aristotelis in tribus libris Ethicorum dicendi ratione: particles, connectives and
style in three books from the Aristotelian ethical treatises’ (in four parts), LCM
8: 4–11, 37–40, 54–7, 70–4.
‘Plato on the Sophists as teachers of virtue’, HPTh 4: 409–27.
‘The nature of Homeric morality’, in Approaches to Homer, ed. C. A. Rubino and
C. W. Shelmerdine, Austin: 248–75.
314 Christopher Rowe publications
review of J. L. Ackrill, Aristotle the Philosopher and W. K. C. Guthrie, A History of
Greek Philosophy, vol. VI, in JHS 103: 175–6.
review of H.-D. Voigtländer, Der Philosoph und die Vielen, in CR 33: 140.
review of Paul Woodruff, Plato, Hippias Major, in JACT Bulletin Review 62: ix.
review of M. Woods, Aristotle’s Eudemian Ethics, Books I, II, and VIII, in CR 33: 60–1.
1984
Plato. Brighton and New York. [Second edn., London 2003.]
‘Plato: the search for an ideal form of state’, in Political Thought from Plato to
NATO, ed. B. Redhead, London: 18–29.
review of R. Bodéüs, Le philosophe et la cité, in CR 34: 209–10.
review of E. Hussey, Aristotle’s Physics Books III and IV, in JACT Bulletin Review
64: vi.
review of J. M. Rist, Human Value, in CR 34: 59–61.
review of R. F. Stalley, An Introduction to Plato’s Laws, in PhB 25: 195–7.
1985
review of B. Dumoulin, Recherches sur le premier Aristote, in JHS 105: 188–9.
review of R. Sorabji, Time, Creation and the Continuum, in JACT Review 2nd
series 2: 27–8.
‘The sparser Stagirite’, The Times Literary Supplement 8 February: 150. [Review of
Barnes (ed.), Aristotle: Complete Works.]
1986
(ed.), Plato: Phaedrus, with tr. and comm. Oxford. [Second edition, 1988.]
‘The argument and structure of Plato’s Phaedrus’, PCPhS 32: 106–25.
review of W. W. Fortenbaugh, Quellen zur Ethik Theophrasts, in CR 36: 321–2.
review of G. Klosko, The Development of Plato’s Political Theory, in JACT Review
2nd series 5: 31.
‘After Aristotle’, The Times Literary Supplement 4 July: 730. [Review of M.
Schofield and G. Striker (eds.), The Norms of Nature.]
1987
‘Many-coloured Homer?’, in Omnibus: 30–2.
‘Platonic irony’, Nova Tellus 5: 83–101.
‘Who was Socrates?’, Cogito (The Journal of the Cogito Society): 9–11.
review of C. L. Griswold, Self-Knowledge in Plato’s Phaedrus, in The Washington
Book Review 21–22: 32.
review of M. C. Nussbaum, The Fragility of Goodness, in PhB 28: 73–6.
‘Living dialogue’, in THES 13 March: 18. [Review of C. L. Griswold, Self-
Knowledge in Plato’s Phaedrus.]
‘Incitements to philosophy’, The Times Literary Supplement 3 April: 368. [Review
of D. Bostock (ed.), Plato’s Phaedo.]
1988
‘The meaning of φρόνησις in the Eudemian Ethics’, in Schriften zur aristotelischen
Ethik, ed. C. Mueller-Goldingen, Hildesheim: 253–72. [Reprint of ‘The mean-
ing of φρόνησις’, q.v. above, 1971.]
‘A reply to John Cooper on the Magna Moralia’, in Schriften zur aristotelischen
Ethik, ed. C. Mueller-Goldingen, Hildesheim: 371–83. [Reprint of ‘A reply to
John Cooper’, q.v. above, 1975.]
‘Public and private speaking in Plato’s later dialogues’, in Platón: Los dialogos tar-
díos (papers from the third Symposium Platonicum), ed. C. Eggers Lan, Mexico
City: 125–37.
‘Plato: Die Suche nach einer idealen Staatsform’, in Von Plato bis Popper, ed. B.
Redhead and J. Starbatty, Bonn: 23–37. [Tr. of ‘Plato: the search for an ideal
state’ q.v. above, 1984.]
with A. Gómez-Lobo and C. Eggers Lan, ‘Discusión: la autenticidad de la
Carta VII’, in Platón: Los dialogos tardíos (papers from the third Symposium
Platonicum), ed. C. Eggers Lan, Mexico City: 161–8.
review of P. Accattino, L’anatomia della città nella Politica di Aristotele and of J.
Touloumakos, Die theoretische Begrundung der Demokratie in der klassischen Zeit
Griechenlands: Die demokratische Argumentation in der ‘Politik’ des Aristoteles, in
CR 38: 282–4.
review of G. R. F. Ferrari, Listening to the Cicadas, in CR: 223–5.
review of R. Véron, Platon: Une introduction à la vie de l’esprit, in CR 38: 425.
1989
‘φρόνησις’, Historisches Wörterbuch der Philosophie, vol. VII, ed. J. Ritter. Basel.
with E. I. McQueen, ‘Phaedo, Socrates and the chronology of the Spartan war
with Elis’, Méthexis 2: 1–18.
‘Plato’s use of irony: a case study’, in Sprachaspekte als Experiment: Beiträge zur
Literaturkritik in Antike und Neuzeit, ed. T. Viljamaa, S. Jäkel and K. Nyholm,
Turku: 83–97.
‘Reality and utopia’, Elenchos 10: 317–36.
‘The unity of the Phaedrus: a reply to Heath’, OSAPh 7: 175–88.
review of T. L. Pangle, The Roots of Political Philosophy, in CR 39: 194–5.
‘Sitting down with a know-nothing’, The Times Literary Supplement 31 March:
341. [Review of D. Bostock, Plato’s Theaetetus.]
1990
‘Philosophy, love and madness’, in The Person and the Human Mind: Issues in
Ancient and Modern Philosophy, ed. C. Gill, Oxford: 227–46.
‘The good for man in Aristotle’s ethics and politics’, in Studi sull’etica di Aristotele,
ed. A. Alberti, Rome: 193–225.
‘How scholarship works’, Pegasus (The Journal of the Department of Classics and
Ancient History in the University of Exeter) 33: 28–30. [Abridged version of Rowe
and McQueen q.v. above, 1989.]
‘Forms of ecstasy’, in Polis (Newsletter of the Society for the Study of Greek Political
Thought) 10: 105–12. [Review article on M. L. Morgan, Platonic Piety.]
review of P. Coby, Socrates and the Sophistic Enlightenment, in JHS 110: 225–6.
review of T. J. Saunders (ed.), Early Socratic Dialogues and of L. Brisson, Platon:
Lettres, in JHS 110: 224–5.
review of R. Sorabji, Matter, Space and Motion, in JACT Review 2nd series 7: 25.
‘Rational ends’, The Times Literary Supplement 11 May: 507. [Review of R. Kraut,
Aristotle on the Human Good.]
‘Variations in Plato’s style’, review of G. R. Ledger, Re-counting Plato, in The Times
Higher Education Supplement 14 September: 25.
1991
‘Aristotele sulla felicità: lo sviluppo di un ragionamento’, in Da democrito a Collingwood:
studi di storia della filosofia, ed. A. Ingegno (= Pubblicazioni del Dipartimento di
Filosofia e Scienze Sociali dell’Università di Siena 2), Florence: 25–42.
‘Ethics in ancient Greece’, in A Companion to Ethics, ed. P. Singer, Oxford:
121–32.
‘Aims and methods in Aristotle’s Politics’, in D. Keyt and F. D. Miller (eds.), A
Companion to Aristotle’s Politics, Oxford: 57–74. [Revised version of ‘Aims and
methods’ q.v. above, 1977.]
1992
‘L’argument par “affinité” dans le “Phédon”’, RPhilos 181: 463–77.
‘La data relativa del Fedro’, in Understanding the Phaedrus (papers from the sec-
ond Symposium Platonicum), ed. L. Rossetti, Sankt Augustin: 31–9.
‘Parasite or fantasist? The role of the literary commentator’, Cogito (Journal of the
Cogito Society): 9–18. [Inaugural lecture delivered in the University of Bristol
on 3 December 1990.]
‘Philosophy and literature: the arguments of Plato’s Phaedo’, Proceedings of the
Boston Area Colloquium for Ancient Philosophy 7: 159–81.
‘On reading Plato’, Méthexis 5: 53–68.
‘Reflections of the sun: explanation in the Phaedo’, in The Language of the
Cave (= special issue of Apeiron), ed. A. Barker and M. Warner, Edmonton:
89–101. [A shorter, adapted version of ‘Explanation in Phaedo’ q.v. below,
1993.]
1993
(ed.) Plato: Phaedo. Cambridge.
and P. Nicholson (eds.), Plato’s Statesman: Selected Papers from the Third Symposium
Platonicum = Polis (Newsletter of the Society for the Study of Greek Political
Thought), 12/1–2. [A collection of papers not included in Rowe (ed.), Reading
the Statesman, q.v. below, 1995.]
‘Explanation in Phaedo 99c6–102a8’, OSAPh 11: 49–69.
review of R. A. McNeal, Law and Rhetoric in the Crito and of R. Bosley and M.
Tweedale (eds.), Aristotle and his Medieval Interpreters, in CR 43: 440–2.
‘Booknotes: Plato’, Phronesis 38: 214–22.
1994
and T. Penner, ‘The desire for good: is the Meno inconsistent with the Gorgias?’,
Phronesis 39: 1–25.
‘Bad and appalling constitutions in Plato’s Statesman’, Polis (Newsletter of the
Society for the Study of Greek Political Thought) 13: 119–32. [Review article on J.
Annas and R. Waterfield (eds.), Plato: Statesman.]
review of P. L. Donini, Ethos: Aristotele e il determinismo, of N. Sherman, The Fabric
of Character, of R. Wardy, The Chain of Change, of E. Belfiore, Tragic Pleasures,
and of A. O. Rorty (ed.), Essays on Aristotle’s Poetics, in JHS 114: 170–4.
review of J. V. Luce, An Introduction to Greek Philosophy, in Hermathena 156:
73–4.
review of C. D. C. Reeve, Socrates in the Apology and of T. C. Brickhouse and N.
D. Smith, Socrates on Trial, in JHS 114: 191–2.
1995
(ed.) Reading the Statesman (papers from the third Symposium Platonicum). Sankt
Augustin. [Includes ‘Introduction’ at 11–28.]
(ed.) Plato: Statesman, with tr. and comm. Warminster. [Corrected edn., Oxford
2005.]
‘Contre Platon: philosophie et littérature dans le Phédon’, in Contre Platon 2:
Renverser le Platonisme, ed. M. Dixsaut, Paris: 271–91. [Tr. of ‘Philosophy and
literature’ q.v. above, 1992.]
1996
‘A reply to van Eck’, OSAPh 14: 227–40.
‘Stile e forma nel Filebo’, in Il Filebo di Platone e la sua fortuna (Atti del Convegno
di Napoli, 4–6 novembre 1993), ed. P. Cosenza, Naples: 19–28.
‘The Politicus: structure and form’, in Form and Argument in Late Plato, ed. C.
Gill and M. M. McCabe, Oxford: 153–78.
entries on ‘Cebes of Thebes’, ‘Echecrates’, ‘Elis, school of ’, ‘dialogue (Greek)’,
‘logos’, ‘Phaedon of Elis’, ‘Simmias of Thebes’, ‘soul’ and ‘transmigration’, in
The Oxford Classical Dictionary, ed. S. Hornblower and A. Spawforth, 3rd edn.
Oxford: s.vv.
review of Anthony Kenny (ed.), The Oxford History of Western Philosophy, in PhR
105: 525–7.
review of R. B. Rutherford, The Art of Plato, in JACT Review 2nd series 20:
19–20.
review of D. Scott, Recollection and Experience, in THES 11 October: 24.
1997
‘Platonic aesthetics and psychology’, in Routledge History of Philosophy, vol. 1:
From the Beginning to Plato, ed. C. C. W. Taylor, London: 425–55.
‘The good, the reasonable and the laughable in Plato’s Republic’, in Laughter Down
the Centuries, ed. S. Jäkel, A. Timonen and V.-M. Rissanen, vol. III, Turku:
45–54.
‘Why is the ideal Athens of the Timaeus-Critias not ruled by philosophers?’,
Méthexis 10: 51–7.
(trans.), Plato’s Statesman, in J. Cooper (ed.), Plato: Complete Works, Indianapolis:
294–358. [Tr. revised from that in Plato, Statesman q.v. above, 1995.]
review of J. Annas and R. Waterfield (eds.), Plato: Statesman, in CR 47: 277–9.
review of E. A. Duke, W. F. Hicken, W. S. M. Nicoll, D. B. Robinson and J. C.
G. Strachan (eds.), Platonis opera, vol. I, in CR 47: 272–4.
review of P. Murray, Plato on Poetry, in JACT Review 21: 31.
‘Observing the Greeks observing themselves’, in Times Literary Supplement 24
October: 13. [Review of J. Brunschwig and G. E. R. Lloyd (eds.), Le savoir
grec.]
‘Gorgias on my mind’, Times Higher Education Supplement 10 October: 29.
[Review of R. Wardy, The Birth of Rhetoric.]
1998
Il Simposio di Platone: Cinque lezioni con un contributo sul Fedone e una breve
discussione con Maurizio Migliori e Arianna Fermani (27–29 March 1996,
University of Macerata). Sankt Augustin.
(ed.) Plato: Symposium, with tr. and comm. Warminster.
‘Democracy and Sokratic-Platonic philosophy’, in Democracy, Empire, and the
Arts in Fifth-Century Athens, ed. D. Boedeker and K. A. Raaflaub, Cambridge:
241–53.
‘Matar a Sócrates: los pensiamentos tardíos de Platón acerca de la democracia’, in
Theoria (Rivista del Colegio de Filosofía, UNAM, Mexico City) 6: 53–74. [Spanish
version of ‘Killing Socrates’ q.v. below, 2001.]
‘On Plato, Homer and archaeology’, Arion 5: 134–44.
‘Socrates and Diotima: Eros, immortality and creativity’, Proceedings of the Boston
‘The uses and disadvantages of Socrates’, Histos 2: 216–29. http://research.ncl.
ac.uk/histos/ [Inaugural lecture delivered at Durham University on 3 February
1999, and published in March of that year (the Histos issue date is nominal).]
review of P.-M. Morel, Platon et l’objet de la science, in CR 48: 211–12.
1999
(ed.) Plato: Statesman, with tr. and comm. [Revised version of Plato: Statesman
q.v. above, 1995, including the tr. used in Plato: Statesman q.v. above, 1997.
Corrected edn.: Oxford 2005.]
Plato: Statesman, tr. with introduction. Indianapolis. [The tr. is reprinted from
Plato: Statesman q.v. above, 1997.]
‘La forme dramatique et la structure du Philèbe’, in La fêlure du plaisir: Études sur
le Philèbe de Platon, ed. M. Dixsaut, Paris: 9–25.
‘Myth, history and dialectic in Plato’s Republic and Timaeus-Critias’, in From Myth
to Reason? Studies in the Development of Greek Thought, ed. R. G. A. Buxton,
‘The speech of Eryximachus in Plato’s Symposium’, in Traditions of Platonism:
Essays in Honour of John Dillon, ed. J. J. Cleary, Aldershot: 53–64.
review of M. Hoffmann, Die Entstehung von Ordnung: Zur Bestimmung von Sein,
Erkennen und Handeln in der späteren Philosophie Platons, in CR 49: 426–7.
‘Booknotes: Socrates and Plato’, Phronesis 44: 72–82 and 242–52.
2000
and M. Schofield (eds.), The Cambridge History of Greek and Roman Political
Thought. Cambridge.
‘Aristotle for and against democracy’, in Political Equality and Justice in Aristotle
and the Problems of Contemporary Society, ed. D. N. Koutras, Athens: 408–16.
‘Socrates and his twin in Plato’s Symposium’, Omnibus 41: 30–2.
‘Socrates, Plato, and why did I buy those chocolates?’, Durham First 12: 6–7.
‘The Lysis and the Symposium: aporia and euporia?’, in Plato: Euthydemus, Lysis,
Charmides (papers from the fifth Symposium Platonicum), ed. T. M. Robinson
and L. Brisson, Sankt Augustin: 204–16.
‘The nature and function of Platonic utopianism’, in Utopias, ed. F. Le Saux and
N. Thomas, Durham: 37–53.
‘Uccidere Socrate: le idee di Platone sulla democrazia nei dialoghi “tardi”’, Dianoia
(Annali di Storia della Filosofia, Dipartimento di Filosofia, Universita di Bologna)
5: 15–37. [Italian version of ‘Killing Socrates’ q.v. below, 2001.]
‘Preface’, in Bibliography on Plato’s Laws by T. J. Saunders and L. Brisson, revised
edn. with additional bibliography on the Epinomis, Sankt Augustin: 9–10.
review of A. Capizzi, Paradigma, mito, scienza, in JHS 120: 176.
review of G. Fendt and D. Rozema, Platonic Error: Plato, a Kind of Poet, in JACT
Review, 2nd series 27: 34.
review of D. Frede: Platon: Philebos, in CR 50: 582–3.
review of M. S. Lane, Method and Politics in Plato’s Statesman and of N. Notomi,
The Unity of Plato’s Sophist, in CR 50: 490–3.
‘Booknotes: Plato and Socrates’, Phronesis 45: 159–73.
2001
‘Killing Socrates: Plato’s later thoughts on democracy’, JHS 121: 63–76.
‘Modelli di ermeneutica platonica nei secoli XIX e XX’, in Plato: The Internet
Journal of the International Plato Society 1. http://gramata.univ-paris1.fr/Plato/
article16.html [For paper version see below, 2008.]
‘Socrates’, in Fifty Major Thinkers on Education from Confucius to Dewey, ed. J. A.
Palmer, London: 5–10.
‘The concept of philosophy (philosophia) in Plato’s Phaedo’, in Plato’s Phaedo:
Proceedings of the Second International Symposium Platonicum Pragense, ed. A.
HavlíČek and F. Karfík, Prague: 34–47.
‘Not coming after’, review of K. Algra, J. Barnes, J. Mansfeld and M.
Schofield (eds.), The Cambridge History of Hellenistic Philosophy, in TLS 23
March: 22.
2002
and S. Broadie, Aristotle: Nicomachean Ethics: Translation, Introduction, and
Commentary. Oxford.
and J. Annas (eds.), New Perspectives on Plato, Modern and Ancient. Washington.
‘Reply to Penner’, in Rowe and Annas, New Perspectives on Plato (as above):
213–25.
‘Handling a philosophical text’, in The Classical Commentary: History, Practices,
Theory, ed. R. K. Gibson and C. S. Kraus, Leiden: 295–318.
‘Zwei oder drei Phasen? Der Mythos im Politikos’, in Platon als Mythologe: Neue
Interpretationen zu den Mythen in Platons Dialogen, ed. M. Janka and C.
Schäfer, Darmstadt: 160–75.
‘Socrate, les lois et les “Lois”’, RFHIP 16: 259–73.
‘Just how Socratic are Plato’s “Socratic” dialogues? A response to Charles
Kahn, Plato and the Socratic Dialogue’, in Plato: The Internet Journal of the
International Plato Society 2. http://gramata.univ-paris1.fr/Plato/article30.
html.
review of A. Capra, Ἀγὼν λόγων: Il “Protagora” di Platone tra eristica e commedia,
in AJPh 123: 521–4.
review of Z. Planinc (ed.), Politics, Philosophy, Writing, in CR 52: 370–1.
2003
‘“All our desires are for the good”: reflections on some key Platonic dialogues’,
in Plato Ethicus: Philosophy is Life, ed. M. Migliori and L. M. Napolitano
Valditara, Sankt Augustin: 265–72.
‘Plato’, in The Cambridge Companion to Ancient Philosophy, ed. D. Sedley,
‘Plato on knowing and merely believing’, in Ideal and Culture of Knowledge in
Plato, ed. W. Detel, A. Becker and P. Scholz, Stuttgart: 57–68.
‘Plato, Socrates and developmentalism’, in Desire, Identity and Existence: Studies
in Honour of T. M. Penner, ed. N. Reshotko, Kelowna: 17–32.
‘Reply to Roger Crisp’, in Plato and Aristotle’s Ethics, ed. R. Heinaman, Aldershot:
79–86.
‘Reply to Richard Kraut’, in Plato and Aristotle’s Ethics, ed. R. Heinaman,
Aldershot: 168–76.
‘Socrates, the laws and the Laws’, in Plato’s Laws: From Theory to Practice (papers
from the sixth Symposium Platonicum), ed. S. Scolnicov and L. Brisson, Sankt
Augustin: 87–97. [English version of ‘Socrate, les lois et les “Lois”’ q.v. above,
2002.]
‘Socrates and Plato on virtue and the good: an analytical approach’, in New Images
of Plato, ed. G. Reale and S. Scolnicov, Sankt Augustin: 253–64.
‘The Politeiai of Zeno and Plato’, in The Philosophy of Zeno, ed. T. Scaltsas and A.
S. Mason, Larnaca: 291–308.
‘The status of the “Myth” in Plato’s Timaeus’, in Plato Physicus: Cosmologia e
antropologia nel Timeo, ed. C. Natali and S. Maso, Amsterdam: 21–31.
‘The treatment of anger in Aristotle’s Ethics and Rhetoric’, in On Aristotle’s Poetics
and the Art of Rhetoric, ed. D. N. Koutras, Athens: 366–76.
‘Τα µέρη της ψυχής και η επιθυµία του αγαθού στην Πολιτεία’, Philosophia:
33: 129–42.
review of Paul W. Ludwig, Eros and Polis, in JACT Review 2nd series 34: 32.
2004
‘Éthique et métaphysique platoniciennes ou pourquoi il faut abandonner la classifi-
cation en dialogues de jeunesse, de maturité et de vieillesse’, in PhilosAnt 4: 131–
50. [French version of ‘Plato, Socrates and developmentalism’, q.v. above, 2003.]
‘Giving Socrates his due: the interpretation of Plato’s Republic’, Lampas (Tijdschrift
voor Nederlandse Classici) 37: 10–23.
‘Hedonism in the Protagoras again: Protagoras, 351b ff.’, in Plato’s Protagoras (pro-
ceedings of the third Symposium Platonicum Pragense), ed. A. Havlícek and F.
Karfík, Prague: 133–47.
‘Socrates, Plato, and Aristotle on desire, “friendship” and the good’, in Socrates:
2400 Years Since his Death (399 B.C.–2001 A.D.), ed. V. Karasmanis, Athens:
121–6.
‘Socrate e Simonide’, in Il Protagora di Platone: struttura e problematiche, ed. G.
Casertano, Naples: 460–73.
‘The best life according to Aristotle (and Plato): a reconsideration’, in The Ways
of Life in Classical Political Philosophy (papers from the third Meeting of the
Collegium Politicum, Madrid), ed. F. L. Lisi, Sankt Augustin: 121–33.
‘The case of the missing philosophers in Plato’s Timaeus-Critias’, in WJA 28(b):
57–70.
review of Christopher Bobonich, Plato’s Utopia Recast in Notre Dame Philosophical
Review: 12 August. http://ndpr.nd.edu.
review of Lloyd Gerson, Knowing Persons, in Ancient Philosophy 24: 464–9.
review of W. Hamilton and C. Emlyn-Jones (eds.), Plato: Gorgias, in The Journal
of Classics Teaching 3: 73.
review of R. A. Naddaff, Exiling the Poets, in The Journal of Classics Teaching 1: 58.
review of M. R. Wright (ed.), Reason and Necessity, in CR 54: 316–17.
2005
and Terry Penner, Plato’s Lysis. Cambridge.
Plato: Phaedrus, tr. (Penguin Books). Harmondsworth. [Substantially revised ver-
sion, with new notes, of the tr. that appeared in Plato: Phaedrus, q.v. above,
1986.]
‘Le traitement des constitutions non idéales dans le Politique’, in Eph 3: 385–400.
‘Hommes et monstres: Platon et Socrate parlent de la nature humaine’, in
Agonistes. Essays in Honour of Denis O’Brien, eds. J. Dillon and M. Dixsaut.
Aldershot: 139–56.
‘Les parties de l’âme et le désir du bien dans la République’, in Études sur la
République de Platon, vol. 2, ed. M. Dixsaut, Paris: 209–24. [French version of
‘Τα µέρη της ψυχής’, q.v. above, 2003.]
‘Needs and ethics in ancient philosophy’, in The Philosophy of Need (= Royal
Institute of Philosophy Supplement 57), ed. S. Reader, Cambridge: 99–112.
‘O estilo de Aristóteles na Ethica Nicomachea’, in A Ethica Nicomachea de
Aristóteles = Analytica 8: 13–29.
‘What difference do forms make to Platonic epistemology?’, in Virtue, Norms, and
Objectivity: Issues in Ancient and Modern Ethics, ed. C. Gill, Oxford: 215–32.
‘What might we learn from the Clitophon about the nature of the Academy?’, in
Pseudoplatonica, ed. K. Döring, M. Erler and S. Schorn, Stuttgart: 213–24.
‘The good and the just in Plato’s Gorgias’, in Platon über das Gute und die
Gerechtigkeit, ed. D. Barbarić, Würzburg: 73–92.
‘Mettre Socrate à mort: la position de Platon sur la démocratie dans les derniers
dialogues’, in M. Fattal (ed.), La philosophie de Platon, vol. 2, Paris: 299–325.
[French version of ‘Killing Socrates’ q.v. above, 2001.]
review of G. Fine, Plato on Knowledge and Forms, in PhilosQ 55: 504–6.
2006
‘Interpreting Plato’, in A Companion to Plato, ed. H. H. Benson, Oxford: 13–24.
‘Socrates in Plato’s dialogues’, in A Companion to Socrates, ed. S. Ahbel-Rappe
and R. Kamtekar, Oxford: 159–70.
‘The literary and philosophical style of the Republic’, in A Guide to Plato’s Republic,
ed. G. Santas, Oxford: 7–24.
‘The Symposium as a Socratic dialogue’, in Plato’s Symposium: Issues in Interpretation
and Reception, ed. J. Lesher, D. Nails, and C. C. Frisbee. Washington: 9–22.
‘The treatment of non-ideal constitutions in the Politicus: further consider-
ations’, in Laws and Rights in the Ancient Greek Tradition (Philosophical Inquiry
Supplementary Volume), ed. G. Anagnostopoulos, Thessaloniki: 106–21.
2007
Plato and the Art of Philosophical Writing. Cambridge.
‘A problem in the Gorgias: how is punishment supposed to help with intellec-
tual error?’, in Akrasia in Greek Philosophy: From Socrates to Plotinus, ed. C.
Bobonich and P. Destrée, Leiden: 19–40.
‘La concezione dell’anima in Repubblica IV: che cosa manca esattamente alla “via
più breve”?’, in Interiorità e anima: la psychè in Platone, ed. M. Migliori, L. M.
Napolitano Valditara and A. Fermani, Milan: 245–53.
‘Plato and the Persian wars’, in Cultural Responses to the Persian Wars: Antiquity to
the Third Millennium, ed. E. Bridges, E. Hall and P. J. Rhodes, Oxford: 85–104.
‘Plato the Socratic’, Méthexis 20: 145–57.
‘The form of the good and the good in Plato’s Republic’, in Pursuing the Good:
Ethics and Metaphysics in Plato’s Republic, ed. D. Cairns, F.-G. Herrmann and
T. Penner, Edinburgh: 124–52.
‘The moral psychology of the Gorgias’, in Gorgias-Menon (papers from the seventh
Symposium Platonicum), ed. M. Erler and L. Brisson, Sankt Augustin: 90–101.
‘The place of the Republic in Plato’s political thought’, in The Cambridge
Companion to Plato’s Republic, ed. G. R. F. Ferrari, Cambridge: 27–54.
review of A. Kenny, A New History of Western Philosophy, vol. 1: Ancient Philosophy,
in CR 57: 310–11.
review of R. Knab, Platons Siebter Brief, in ExClass 11: 291–8.
2008
‘The good and the just in Plato’s Gorgias’, Philosophical Inquiry 30: 55–75. [Reprint
of ‘The good and the just’, q.v. above, 2005.]
‘Modelli di ermeneutica platonica nei secoli XIX e XX’, in Platone e Aristotele:
Dialettica e logica, ed. M. Migliori and A. Fermani, Brescia: 343–56. [Reprint of
‘Modelli di ermeneutica platonica’, q.v. above, 2001.]
review of R. Bodéüs, Le véritable politique et ses vertus selon Aristote, in PhilosAnt
8: 274–8.
review of T. A. Szlezák, Das Bild des Dialektikers in Platons späten Dialogen, in
Gnomon 80: 198–204.
2009
‘Greek political theory’, in The Oxford Handbook of Hellenic Studies, ed. G. Boys-
Stones, B. Graziosi and P. Vasunia, Oxford: 401–9.
‘The charioteer and his horses: an example of Platonic myth-making’, in Plato’s
Myths, ed. C. Partenie, Cambridge: 134–47.
and E. OKell, ‘The “research-teaching nexus” and the learning-teaching rela-
tionship: who’s in charge?’, in Arts and Humanities in Higher Education 8:
180–90.
entries on ‘Democritus’, ‘Heraclitus’, ‘Marcus Aurelius Antoninus’, ‘Plato’,
‘Plotinus’, ‘Pythagoras’, ‘Socrates’ and ‘Stoicism’, in The Oxford Companion to
English Literature, ed. D. Birch, 7th edn. Oxford: s.vv.
‘Reading Socrates in Plato’s dialogues’, Hermathena 186: 25–41.
review of S. C. Rickless, Plato’s Forms in Transition, in Mnemosyne 62: 298–301.
2010
The Last Days of Socrates: Euthyphro, Apology, Crito, Phaedo, tr. with notes.
(Penguin Classics.) London.
‘Boys, kingship, and board-games: a note on Plato, Politicus 292e–293a’, in Il
quinto secolo: Studi di filosofia in onore di Livio Rossetti, ed. S. Giombini and F.
Marcacci, Perugia: 529–33.
‘L’argument relatif à la pleonexia dans le premier livre de la République: deux con-
ceptions de ce qu’est “avoir avantage”’, in Aglaïa, autour de Platon: Mélanges
offerts à Monique Dixsaut, ed. A. Brancacci, D. El Murr and D. P. Taormina,
Paris: 291–305.
‘How to read Plato: some rival approaches’, The Journal of Greco-Roman Studies
(The Korean Society of Greco-Roman Studies, Seoul) 42: 161–79.
‘On grey-haired babies: Plato, Hesiod, and visions of the past (and future)’,
in Plato and Hesiod, ed. G. Boys-Stones and J. H. Haubold, Oxford:
298–316.
‘Plato’, in The Oxford Encyclopaedia of Ancient Greece and Rome, vol. 5, ed. M.
Gagarin, New York: 297–307.
‘Self-examination’, in The Cambridge Companion to Socrates, ed. D. Morrison,
‘The relationship of the Laws to other dialogues: a proposal’, in Plato’s Laws: A
Critical Guide, ed. C. Bobonich, Cambridge: 29–50.
‘Thomas Taylor the Platonist’, in Argumenta in dialogos Platonis Teil I:
Platoninterpretation und ihre Hermeneutik von der Antike bis zum Beginn des 19.
Jahrhunderts, ed. A. Neschke-Hentschke, Schwabe: 315–25.
‘The good and the just in Plato’s Gorgias’, in Socrates, Plato, Aristotle: In Honour of
G. Santas, ed. G. Anagnostopoulos, Athens: 169–89. [Reprint of ‘The good and
the just’, q.v. above, 2005.]
2011
‘Socratic citizenship’, in The Notion of Citizenship in Greek Philosophy, ed. E.
Moutsopoulos and M. Protopapas-Marneli, Athens: 58–69.
‘The good and the just in Plato’s Gorgias’, in Socratic, Platonic and Aristotelian
Studies: Essays in Honor of Gerasimos Santas, ed. G. Anagnastopoulos, New
York: 145–63. [Reprint of ‘The good and the just’, q.v. above, 2005.]
‘The treatment of the soul in Republic IV: exactly what does the “shorter” way
(435C–D) miss?’, in Inner Life and Soul: Psychê in Plato, ed. M. Migliori, L. M.
Napolitano and A. Fermani, Sankt Augustin: 225–31. [An English version of
‘La concezione dell’anima in Repubblica IV’, q.v. above, 2007.]
Classics and the Life of Plocrates, Classical Association Presidential Address 2011.
Sherborne.
review of D. N. McNeill, An Image of the Soul in Speech, in PhilosQ 61: 633–4.
2012
Plato: Republic, tr. with notes. (Penguin Classics.) London.
‘Socrates in Plato’s Laws’, in Presocratics and Plato: A Festschrift in Honor of Charles
H. Kahn, ed. V. Karasmanis and A. Hermann, Las Vegas: 367–87.
‘Aristotle’s Eudemian Ethics on loving people and things’, in The Eudemian Ethics
on the Voluntary, Friendship, and Luck (The Sixth Keeling Colloquium in
Ancient Philosophy), ed. F. Leigh, Leiden: 29–41.
‘Socrates on reason, appetite and passion: a response to Thomas C. Brickhouse
and Nicholas D. Smith, Socratic Moral Psychology’, The Journal of Ethics 16:
305–24.
‘The status of the myth of the Gorgias, or: Taking Plato seriously’, in Plato and
Myth: Studies on the Use and Status of Platonic Myths, ed. C. Collobert, P.
Destrée and F. Gonzalez, Leiden: 187–98.
‘O tema da raiva na Retórica e na Ética de Aristóteles’, in Archai: revista de estudos
sobre as origenes do pensamento ocidental 9 (2012), 11–16. [Revised, Portuguese
version of ‘The treatment of anger in Aristotle’s Ethics and Rhetoric’, q.v. above,
2003.]
Index of ancient passages
Aeschines Nicomachean Ethics, 86

Against Timarchus 1.6, 1096b32–4, 122n. 3
173, 254n. 12 3.8, 1116a15–21, 83n. 25
Alcinous 5, 287
Didaskalikos 5.3, 1131a31–b6, 123n. 7
4, 169 8.1, 1155a22–6, 283, 288
Andocides 8.1, 1155a28, 285
On the Mysteries 9.6, 288
107, 294 On Sense Perception
anonymous 438a5–7, 94n. 12
Commentary on the Theaetetus Poetics
coll. iii.50 – iv.6, 170 9, 1451a36–b23, 271, 275, 280
Apuleius Politics
On Plato 2.4, 1262b7–10, 284
1.1 [182], 1n. 2 2.6, 284
Archytas 2.12, 1273b27–34, 283
fragments (47 DK) 3.12–13, 287
B3, 91n. 5 4.11, 1296a18–21, 284n. 2
Aristippus 5.1–2, 287
fragments (ed. Giannantoni, SSR IVA) Prior Analytics
174, 178n. 21, 181n. 24 1.1–2, 216
181, 178n. 21 1.4–7, 216
Aristocles Rhetoric
fragments (ed. Chiesara) 3.1, 1404a5–7 18n. 18
T5, 178n. 21, 183 Sophistical Refutations, 131
Aristophanes 171b23, 164
Clouds 179a33–b4, 172n. 14
504, 60 Aristoxenus
Aristotle fragments (ed. Wehrli)
Eudemian Ethics 18, 56n. 19
3.1, 1229a13, 83n. 25 19, 56
7.1, 1234b22–3, 283 Arius Didymus
7.9–10, 287 On the Philosophical Schools
Magna Moralia (ed. Mullach)
1.20, 1191a5–10, 83n. 25 59.1, 25–7 90n. 2
Metaphysics Athenaeus
Α.6, 987b6–9, 67n. 41 Deipnosophistae
Γ.3, 1005b17–18, 236n. 21 11, 505d, 1n. 2
Γ.3, 1005b19–23, 148 11, 506a, 1n. 2
Μ.3, 1078b30–2, 67n. 41 11, 507cd, 1n. 1
Ν.2, 1089a11, 18n. 18 12, 544ab, 178n. 21, 181n. 24
325
326 Index of ancient passages
Augustine 9, 155n. 13, 171n. 11
On the Master 29, 171n. 11
14.45, 210n. 36 fragments (ed. Giannantoni, SSR IIA)
Aulus Gellius 3, 155n. 13, 171n. 11
Attic Nights 22, 151n. 5
14.3.3, 70n. 3 27, 151n. 5
34, 171n. 11
Cicero 107, 171n. 11
Academica Eusebius
2.20, 183 Preparation for the Gospel
2.76, 182 14.18.32, 178n. 21
Letters to Quintus 14.19.3–4, 183
3.5.1–2, 70n. 2 14.19.5, 183
On Ends
1.26, 178n. 21 Galen, ps.-
On Laws, 70n. 2 History of Philosophy
Tusculan Disputations 7, 151n. 5
5.8–9, 56 Gorgias
Helen
Damascius 13, 13n. 7
On First Principles (ed. Westerink)
3.169.5–22, 169n. 8 Heraclitus
Diodorus Siculus fragments (22 DK)
Library of History B22, 57n. 21
7.12.2–4, 291 B35, 57n. 21
Diogenes Laertius B40, 57n. 21
Lives of Eminent Philosophers B129, 57n. 21
1.12, 57 Herodicus of Babylon
2.30, 155n. 13, 171n. 11 fragments (ed. Düring)
2.86, 178 2, 273
2.105, 58 Herodotus
2.106, 151n. 5, 170 Histories
2.107, 171n. 11 1.30.1–5, 269
2.108, 171, 171n. 11 1.98.1–6, 281n. 40
2.124–5, 58n. 25 1.178.1–187.5, 281n. 40
3.35, 1n. 2 4.8.1-3, 271n. 5
3.49–51, 169n. 6 7.164.4, 294
8.1, 56n. 20 Homer
8.8, 57 Iliad
8.46, 56 2.134–8, 281n. 40
2.484–93, 281n. 40
Empedocles Odyssey
fragments (31 DK) 4.271–89, 180n. 23
B17.9–13, 237n. 23
B17.27–9, 91n. 5 Iamblichus
B128–30, 91n. 5 On the Pythagorean Life, 57
Eubulides 35.251, 56n. 19
fragments (ed. Döring) 36.267, 56n. 19
63, 155n. 13
fragments (ed. Giannantoni, SSR IIB) Lucian
1, 171n. 11 Auction of Lives
12, 155n. 13 22, 172n. 14
Euclides
fragments (ed. Döring) Parmenides
8, 171n. 11 fragments (28 DK)
B2.6–8, 240 109c4–d2, 252
B7.1–2, 241 109d–110c, 271
B7.1, 242 112e–113b, 270, 272
B8.8–9, 240 112e4–6, 259n. 23
B8.17–18, 240 113e–114d, 271n. 5
B8.49, 91n. 5 Crito
Pausanias 45b, 58n. 24
Description of Greece 51a, 63
2.13.1–2, 56 Euthydemus
2.13.2, 56n. 20 271a3, 155
2.4.35, 291 271b9–c1, 154
Pindar 272b10, 152
Olympian Odes 272c, 136
1.87, 45n. 61 274e9–275a2, 155
Plato 275d–277c, 141n. 27
Alcibiades I 275e5, 132n. 6
127d–133c, 63n. 37 276–8, 139
128a–d, 63n. 37 276a–278e, 148
130e, 69 278b3–7, 160
Apology 281, 148
17c2, 22 281de, 135, 141, 141n. 25
20c–23c, 274n. 16 282a–d, 141n. 27
29ab, 60 285e, 132n. 6
29de, 65 285e–287a, 132n. 7
29d7–e2, 62 287ab, 132n. 7
30ab, 62 287b3, 132n. 6
30a, 62, 63 288–291, 148
30b, 63 288d–293a, 141
30b2-4, 62n. 36 288d, 141
32c4–d8, 253n. 11 289a4–6, 141
36cd, 63 289b, 142
36c, 63 289b1–3, 144n. 31
39d, 63 289b4–6, 142
41e–42a, 65 289c, 141
Charmides 290–1, 148
172c–174d, 124 290d, 144
Cratylus 290d5–7, 144
384c10–385e3, 194 290e–293a, 145
385e4–386a4, 194, 194n. 8 290e, 135, 141
386a, 162n. 22 291b, 144
386d8–387c4, 194 291d7, 77
387a, 194 292, 131
400c, 40n. 46 292a4–e5, 78
429e, 197 292b–d, 144
430a–433b, 197 292b1–2, 78
433e4–5, 194n. 8 292d, 144
Critias 292e, 145, 153
106b–108a, 280 292e6–293a6, 78
106b6–7, 249 293a–303a, 133
106b8–108a4, 250n. 4 293a, 135, 141
108a5–c6, 255n. 16 293b–296d, 133
108b-d, 281 293b–d, 136
108bc, 281 293b, 133
cd108e1–121c5, 255n. 16 293b8–c1, 136, 136n. 15
109b–111d, 275 293b9–c1, 136n. 16
Plato (cont.) 3, 676a, 288
293c8–d1, 136 3, 682e–683c, 288
293d, 133 3, 683b, 288n. 11, 289
293d4–6, 137n. 18 3, 683c–693a, 289
293d4–5, 137 3, 684a, 290n. 12
293d6, 137 3, 685a, 289
293d8, 137n. 20 3, 687c, 289
293e–294a, 137, 137n. 19 3, 687e–688b, 289
294b–d, 137n. 19 3, 687e, 289
295a6–9, 137 3, 688a–689e, 289
295a8, 138n. 21 3, 689a–e, 290
295b2, 138n. 21 3, 689b, 290
295c4–7, 164 3, 690bc, 290
296a–d, 138 3, 691a–d, 290
296d, 134, 135 3, 691cd, 289
296d3–4, 138 3, 691d–693a, 289
297b–299a, 133 3, 691d–692b, 292
298a–c, 133 3, 692b, 290n. 12, 292
298bc, 134 3, 693a–701e, 288
298c, 134 3, 693a–e, 287
304b5, 153 3, 693b–d, 296
304b7–c1, 153 3, 693bc, 290
304e6–7, 154 3, 693b, 290
307b6–8, 156 3, 693e, 292
Euthyphro 3, 694ab, 292, 293, 296
7c–e, 97n. 17 3, 694c, 293
Gorgias 3, 697cd, 296
448d7–10, 12n. 5 3, 698a–699d, 293
449bc, 12n. 3 3, 698b–699c, 294
449b9–10, 13n. 8 3, 698b, 294, 295
449c7, 11n. 2 3, 698c, 294, 295
467b–468c, 86n. 29, 125n. 9 3, 699ab, 295
467c5–468c8, 80 3, 699c, 295
472d –474c, 273 3, 699d, 294n. 17
492d–494a, 86n. 29 3, 700a, 294
493a, 40n. 46 3, 701d, 287, 296
502d–503d, 273 3, 701e, 292
507c–508a, 86n. 29 4, 715d, 290
508a, 91n. 5, 104, 297 4, 721d7–722a2, 16n. 12
Hippias Major 4, 722a8–9, 13n. 9
287d–289e, 179n. 22 4, 723d5–e3, 19
376a3–4, 12 4, 735a–738e, 19
Hippias Minor 5, 739a–e, 285n. 3
363cd, 12n. 3 5, 739a–e, 274n. 15
376b4–6, 198 5, 741e–744a, 286
Laws, 70n. 2 5, 743c, 286, 295n. 21
1, 626a–c, 288 5, 744c, 286
1, 626e–628c, 285 6, 751a–768e, 286
1, 627e–628a, 296 6, 757a, 287
1, 628c–e, 285, 287n. 7 6, 757bc, 286
1, 629a, 288n. 11 6, 757cd, 286
1, 636e, 288n. 11 6, 757c, 286
1, 646e–650b, 294n. 17, 297 6, 757d–758a, 287
1, 647b, 294 6, 768c8–d1, 20
3, 275 7, 817a–d, 279
10, 887b2–3, 15 66e2–5, 30n. 11
Lysis 67a–c, 65
204a6–7, 158n. 17 67a, 61n. 33
223b5–8, 73n. 11 67c–e, 59
Menexenus 67c, 61n. 33
240c–e, 256n. 18 67de, 59
240e–241c, 295 67d4, 61n. 33
Meno 67d9, 61n. 33
75cd, 152n. 6 68bc, 64
75d3–7, 160n. 19 68b4, 30n. 11
76e3, 11n. 2 68b8–69d3, 82, 84
77b–78b, 125n. 9 68cd, 64, 65
77b6–78b6, 80 69de, 59
80e-81a, 152n. 6 72e–77b, 67
86a7, 19n. 20 74a–c, 203n. 26
87–8, 141n. 26 76c, 61n. 33
89a3, 87n. 31 76d–77a, 67
Parmenides 76e5–9, 67
136e1–3, 24 78b–80b, 67
Phaedo 79c, 29, 68
57a–59c, 55n. 16 79d1–7, 68
57ab, 55 80e–81a, 59
57a, 54 81b–e, 64
57c5–6, 54n. 15 81c–82a, 30n. 13
58d, 58 82a11–b3, 83
59a, 60 82b–d, 64
59bc, 170 82bc, 65
59b, 55 82c, 64
61c, 59 82c2–8, 82n. 24
61de, 56, 57 82d–84b, 65
61d, 59 82d, 65
62c, 59 82e–83c, 30
63b–69e, 59 82e, 40n. 46
63b, 59 83e5–7, 82
63d, 59 84b, 65
63e–64b, 59 85e–86d, 58
64a–c, 59 88b, 30, 59
64ab, 60 88c–89a, 55n. 16
64a, 60 89a–90d, 55
64b–65a, 65 89b, 55
64c–65a, 59 91c–95a, 58
64c, 61 95bc, 59n. 31
64c6, 61 97c–99c, 129
64d, 61 100b, 66
64e–65a, 61 102a, 55n. 16
65a–67b, 59 102a10–103c6, 245n. 32
65a, 40, 65 105c–e, 60
65b, 29 106d, 60
65d–66a, 66 107c8–d2, 64
65de, 66 114c, 28n. 5, 65
66a, 29 114d8–115a3, 82
66b–67b, 64 115b, 65
66b–d, 40 115b5-c1, 64
66b5–d3, 30 115c, 69
66d, 64 118a15–17, 69
Plato (cont.) 255c, 43
Phaedrus 257b–279b, 38
228de, 272 264c2–5, 22
230e–234c, 38 265a, 16
237b–241d, 38 267b3–5, 13n. 8
241e5–8, 16n. 13 270b, 38
242c6, 16 273e, 26
242d2, 16 274c–275b, 271n. 5
243e–257b, 38 276b–277a, 278
244b3–5, 13 278d4–6, 11
245c–246a, 39, 44 Philebus
245e, 39, 40 17a, 152n. 6
246ab, 40 23b7–8, 15n. 11
246a, 45 41b8, 15n. 11
246a7, 45n. 59 Protagoras
246bc, 47 313c5, 158
246b, 41 317b, 157
246b2, 45n. 59 317b4–5, 158
246cd, 39 320c2–328d2, 12
246c, 33n. 20, 40, 43, 44, 44n. 58, 45, 46 328d5–6, 12n. 6
246de, 42 328d9–e1, 12n. 6
246d, 46n. 63 329b, 12n. 3
246e–247a, 47, 47n. 66 334a1–c6, 12
246e–247b, 46 334c, 12
246e, 45 334d6–7, 13n. 8
246e5, 45 334e2–3, 13
247a, 48 336b3, 12n. 5
247a4, 47n. 69 349b–360e, 112
247b, 41n. 48, 44, 46, 47, 48n. 74 352d–357e, 112
247c, 44n. 57, 45, 47, 47n. 70, 48 353a–360e, 115
247d, 47n. 70 353c–354e, 112n. 10
247e–248a, 45 356b–357a, 113
247e, 43 358b–360d, 113n. 11
248a–249b, 40n. 45 359c–360d, 113
248a, 42n. 50, 48 Republic
248b, 42n. 50 1, 334b–335e, 112n. 9
248c–e, 40, 42 1, 344a1, 91
248cd, 46 1, 349a–350c, 112n. 9
248c, 43 1, 352d–353e, 112n. 9
248c7, 41n. 49 1, 354c1–3, 73n. 11
249a, 43 2, 367e–368c, 276n. 25
249b, 40n. 44, 49 2, 276
249cd, 49 2, 368c–369a, 90
249c, 43n. 55, 49 2, 377b–378e, 33
250c, 39, 40, 46n. 63 3, 394d8–9, 22n. 27
250c7–8, 25 3, 399ab, 296
250d, 42 3, 399bc, 296
250e, 42 3, 401b–403c, 281
251a, 42, 43 3, 401e–402a, 119n. 23
251bc, 43, 45n. 60 3, 412de, 120n. 25
251c, 43 3, 413c–414b, 120n. 25
252c, 47n. 66 4, 423b–424b, 284
253b, 47n. 66 4, 430c3, 84n. 25
254a–e, 44 4, 430c8, 84n. 25
254e, 44 4, 430e4–8, 79
4, 432a, 283 6, 505a4–b1, 200n. 16
4, 433a–434c, 287n. 8 6, 505a5–6, 126n. 11
4, 435cd, 76 6, 505a7, 126n. 11
4, 436cd, 137n. 16 6, 505bc, 124
4, 438a, 116n. 20 6, 505b5–c11, 77n. 19
4, 439e5–440a5, 79 6, 505b5, 77
4, 442b10-c2, 84n. 25 6, 505b6, 77
4, 442c10, 91 6, 505b8-c5, 77
4, 442d, 90 6, 505c6–d1, 77
4, 443cd, 297 6, 505d–506a, 125, 126n. 11, 127
4, 443d–444a, 114n. 14 6, 505de, 125
5, 450b1, 19n. 20 6, 505e1–5, 200n. 16
5, 450b6–7, 14 6, 505e1–2, 79
5, 450e1–451a1, 11 6, 505e3–4, 126n. 11
5, 454a, 152n. 6 6, 506a6, 126
5, 454a7, 151n. 4 6, 506bc, 123n. 6
5, 454b5–6, 151n. 4 6, 506c, 78
5, 462a–465d, 284 6, 506d, 123
5–7, 471c-541a, 127 6, 506e3–4, 98n. 21
5, 471c–473b, 279 6, 507a3–4, 98n. 21
5, 471c, 75n. 16 6, 507b, 123
5, 472a, 127 6, 507c–508b, 100
5, 476cd, 68 6, 508a, 108
5, 476d1, 95n. 14 6, 508e3–4, 98
6, 129 6, 509a5, 123
6, 484a–504a, 113 6, 509b, 98, 124, 127
6, 484a–485a, 279 6, 510a, 100
6, 484a2–b1, 20 7, 514a–521b, 92, 104
6, 485a–504d, 119 7, 514a–517a, 68
6, 485de, 111, 113, 115, 120 7, 514b, 100
6, 485d6–e5, 79 7, 514b5, 100n. 26
6, 485d6, 79 7, 514b9, 99
6, 485e1, 113 7, 514c1–515a1, 99
6, 490a8–b7, 80n. 23 7, 515a5, 94
6, 490b3–4, 76 7, 515c2, 100
6, 498d8–499a2, 119n. 23 7, 515c6–d7, 107
6, 499a, 152n. 6 7, 515c6, 105
6, 499b6–500a6, 77 7, 515d–516a7, 105
6, 500b–502a, 279 7, 515d3, 95
6, 500b-d, 66n. 39 7, 515e1, 100
6, 500c–501b, 119n. 23 7, 515e4, 105
6, 500d5–9, 83 7, 516a, 100
6, 500d6, 91n. 3 7, 516a4, 105
6, 502c–506b, 75 7, 516b, 100
6, 502e–503a7, 120n. 25 7, 516b8–d2, 95n. 14
6, 503e1–2, 120n. 25 7, 516b9-c2, 100
6, 503e3, 76 7, 516c, 104
6, 504a, 76 7, 516d, 105n. 35
6, 504b1–5, 19 7, 516e8, 95n. 14
6, 504cd, 76 7, 517a8–b1, 94
6, 504c2–3, 17 7, 517bc, 122n. 2
6, 504c9–d8, 76n. 18 7, 517b3, 100
6, 504d–505a, 76 7, 517c1–5, 101
6, 504e–505b, 124, 125 7, 517c3, 95, 102, 127
6, 505ab, 126n. 11 7, 517d9, 96
Plato (cont.) 10, 597c, 100
7, 518b–519b, 111, 113n. 12, 119 10, 601b–607a, 279, 281
7, 518b7–d2, 80 10, 602c–608b, 36n. 34
7, 518b7–c2, 83 10, 611a10–612a7, 88
7, 518c–519b, 115, 119 10, 611b–612c, 115n. 19
7, 518c, 95, 115 10, 611b–612a, 111, 113, 114n. 14, 115, 117, 118
7, 518c4–d1, 76, 95n. 13 10, 611b–d, 40
7, 518c5, 113n. 13 10, 611b5–7 88
7, 518d9–519b5, 81, 84n. 25 10, 611b10, 118
7, 518d9–e2, 113n. 12, 119, 120n. 26 10, 611e1, 115
7, 518d9-e1, 87n. 31 10, 612a3–4 88n. 33
7, 518d10–11, 83 10, 612a4, 118
7, 518d10, 81 10, 613ab, 66n. 39
7, 518d11–e1, 87n. 31 10, 618b–620a, 114n. 17
7, 518e1, 81, 84 10, 619b-d, 84n. 25
7, 519ab, 113n. 13, 120n. 26 Sophist
7, 520bc, 105 216cd, 157
7, 520b, 87n. 30, 106 217a4, 157
7, 520c1–7, 97 217c3–6, 21
7, 521c–541b, 92, 105 217e1–3, 21
7, 522–5, 143 223d6–11, 158
7, 524c, 107 236d–239c, 211
7, 524d10–525a5, 107 236e1–237b3, 241
7, 527d8, 76 236e–239c, 186
7, 529b, 107 237a2–259b7, 245n. 32
7, 529d7–530b4, 107 237a3–b3, 238
7, 530bc, 106 237a3–b2, 240
7, 530d, 108 237b2–239c8, 240
7, 532bc, 92 237b4–239c8, 238, 241
7, 532b6–d1, 106 237b7–8, 238, 239, 240, 241
7, 532cd, 103n. 30 237b7, 241
7, 532c6, 76 237b9, 240
7, 533d1–2, 76 237c8–239a2, 239
7, 533d2, 105 238b1, 238
7, 535a–536a, 120n. 25 238c, 214
7, 536d, 105 238c9, 238, 239
7, 537, 144 239c4–8, 242
7, 537c9–d2, 120n. 25 241d–242a, 214
8, 543c5–6, 20 241d5–7, 242, 242n. 28, 243
8, 554c–e, 114n. 15 244b–d, 211
8, 561e, 114n. 15 250a8, 222, 235, 238
9, 575a, 114n. 15 251a–259d, 215
9, 580bc, 114n. 15 251a5–257a12, 222
9, 580d–581c6, 82 252a2–4, 224
9, 581b, 120n. 27 252a2–3, 222
9, 582b–d, 120n. 27 252a2, 223
9, 585c1–2, 120n. 27 252a4, 223
9, 585d–586b, 120n. 27 252d2–11, 235, 238
9, 586e, 114n. 15 253c7–8, 26
9, 586e4–587a1, 120 253d1–e9, 11
9, 590cd, 117 254c8-d2, 242n. 28
9, 590c7-d6, 84n. 25 254d7–11, 223, 224
9, 591b–d, 114n. 14 254d10, 223
9, 592b, 104, 108 255c8–e2, 230
10, 596a–599b, 279 255d9–e7, 231
255e4–6, 244 258e7–259a1, 238
255e11–256a2, 223, 235 259b8–263d5, 241n. 27
255e11–15, 236 259e4–263d5, 245n. 32
255e11–13, 235 261c–263d, 186
255e11, 235 261d–262e, 212n. 39
255e14, 223, 235 262e5–7, 227n. 8
256a1–2, 223, 224 263a5-b13, 241n. 27
256a1, 223, 224 263d1–4, 241
256a2, 224 268b1–5, 11
256a3–b5, 235 268b7, 11
256a3–6, 236 268d2, 100n. 26
256a10-b5, 245n. 32 Statesman
256a12, 235 265a–266e, 17
256c5–10, 236, 245n. 32 267a2, 17, 18n. 19
256c5–6, 236 275b1–6, 16
256c11–d10, 236 277b1–3, 16
256d5–257a7, 244 277b6, 16
256d5–e7, 244, 245 283b1–3, 16
256d5–8, 236 283e–284b, 13
256d8–9, 236 285d5–6, 23
256d12–e1, 245, 245n. 31 286a7, 23
256d12, 245n. 31 286b7–11, 24
256e2–3, 224, 231, 233n. 17 286c1, 18n. 17
257a4–6, 243, 244, 245 286d–287b, 278n. 30
257a5, 243n. 30 286d1–2, 22
257b1–c4, 234, 235n. 20, 239n. 24 286d7–8, 18
257b3, 234 286e5, 18, 24
257b4, 234 302b8, 18n. 18
257c5–258c5, 230 Symposium
257d7–11, 245n. 32 180c–185c, 31
257d11, 232 185e–188e, 32
258a7–9, 231 189c–193e, 32
258a11-b3, 232n. 17, 237n. 22 192de, 32
258b1, 232, 237n. 22, 243n. 30 201d–212a, 32
258b8–c5, 244 204d–206b, 32
258b10–c1, 244 206a–209e, 31
258b10, 244 206c, 31
258c–e, 214 210ab, 31
258c4–5, 230 210a–212a, 31
258c6–259b7, 241 210bc, 31
258d1–259b7, 241n. 27 210d, 31, 43
258d1–e5, 241 211a–212a, 93n. 10
258d5–e3, 236 211c, 31
258d5–7, 242 211d, 31
258d6–7, 233 215a–222b, 32
258d6, 233, 234, 234n. 20, 242, 243 Theaetetus
258d7–e1, 230 151d–186a, 187
258d7, 230, 231, 233 151d3, 20n. 21
258e–259a, 214 151e –152c, 173
258e2–3, 233n. 17, 237, 237n. 22, 242, 242n. 151e3–152a4, 172
28, 243 152a2–157c3, 169
258e2, 237n. 22 152a6–7, 162
258e3, 237 152a6, 162, 162n. 22
258e6–259a1, 238, 241, 243 152a9, 162n. 22
258e6–7, 238 152c5–6, 173
Plato (cont.) 166e1, 162
152c8–160e4, 208n. 33 167a4, 161
152c10, 173 167e1, 164
152d1–e1, 172 167e4–5, 160
152d2–e1, 173 167e5–6, 160
152d6, 174n. 16 167e6–168a2, 165
152e2–153d5, 172 167e7–168a1, 160
153d8–154b8, 173 168b4, 162, 163n. 25
153e7–154a2, 174n. 17 168b5–6, 162
154a2, 174 168b8–c2, 163
154e8, 20n. 21 168b8–c1, 151n. 4
156a–157e, 185 169a9, 170
156a–157c, 167, 183n. 26 169b2–3, 170n. 9
156a–160e, 167 169d3, 20n. 21
156a2–157c3, 172, 173, 182, 183 169d6–8, 161
156a3, 167, 170, 173, 179 169d9, 162
156a5, 173 169e, 151
156a6–7, 173 169e7, 151n. 3, 162
156a7–b2, 173, 183n. 26 170a3–4, 162
156b–157c, 182, 184 170a3, 151n. 3, 162
156b3–4, 173 170a5, 151n. 3
156b4–6, 178 171b, 151
156b7–c3, 183n. 26 171b7–8, 151n. 3
156c6–157a5, 173, 174 171d3–5, 163
156d4–5, 183n. 26 171d6–7, 161
156e4, 208n. 32 171e–172b, 161n. 20
156e6, 208n. 32 172b8–c1, 25
157ab, 201n. 20 172b9–c1, 173
157a8–157b1, 174, 177 172c3–177c5, 173
157b3–4, 174 172c3–4, 25
157b4–7, 174 172d8–9, 13
157b9–c2, 174, 177 173b5, 20n. 21
157c1, 183n. 26 173b6–7, 11
159a–c, 201n. 20 173b8–c4, 20n. 23
161c1–162a3, 172 175e1, 26
161c2, 162 176ab, 66n. 39
161c3–5, 163 177c1–2, 25
161c3, 162 177c5, 20n. 21
162a5–6, 161 178b6–7, 205
162d, 150 178c1, 205
162d6–e2, 150n. 1 178c10–e2, 205
164c1, 20n. 21 178cd, 205
164c10, 171 178de, 161n. 20
164d1, 171 179c1–d1, 205
165a2, 162 179c2–4, 181
165b2–4, 172 179c2–d1, 205
165b9–c2, 172 179c3, 181
165d6, 160 179d–183b, 179
166a1–168c5, 172 179d2–4, 206
166a7–b1, 161n. 21 182a7–b7, 208n. 32
166b2–4, 181 182b3–4, 174n. 16
166b4–5, 172n. 15 182d1–5, 208n. 32
166c, 150 182d8–e12, 208n. 32
166c7–9, 150n. 2 182e1–2, 208
166d1, 150n. 2 182e11–12, 208n. 32
183a2–b5, 208n. 32 20b7–c3, 254
183a6–b5, 208 20c6-d1, 250n. 4
183b7–c4, 206 20d–21d, 269, 270, 272, 274, 281
184b–186e, 167, 179, 182, 184, 185 20d7–25e2, 252
184a3–b2, 23 20d8, 250n. 4, 260
184b–186d, 206, 206n. 29 20e1, 270
184b–e, 184 20e7–25d6, 261
184d1–5, 180, 185 21a–d, 270n. 4, 277n. 27, 278, 281
185a1–6, 181 21a1–3, 252n. 8
185c9–10, 179 21c2–d3, 265n. 30
185e1–2, 180 21c5–6, 270
186b1, 182 21d4–7, 259
186b11–c2, 181 21d8, 250n. 4
186c1–10, 180 21e–22b, 271n. 5
186d2–5, 181 22b–24c, 275
187a9–b1, 19, 20n. 21 23a1–5, 258n. 21
187c1–3, 17n. 16 23b7–8, 258
187e1–2, 20n. 21 23c5–d1, 258
187e–188c, 199 23d–24e, 269
188a–c, 172 24c4–d3, 252, 258
188a7–8, 199 24d3–6, 259
188a10–b1, 198, 199 24d3, 259n. 22
188b1, 172 24d6–e1, 259
188b3–c9, 199 25b5–6, 259n. 23
188b3–5, 198 25d–26e, 272, 277n. 29
188c–189b, 186, 211 25e2–5, 257, 266, 267
188c2–3, 172, 199 26b2–7, 267
189b–190e, 203 26c5-d5, 262
189e, 188 26c5–7, 255n. 16
190d5–6, 198 26c7-d5, 257
190e–195b, 190 26c8, 257n. 19
192a1–b2, 198 26d1, 250n. 4, 258n. 20
193c–194b, 190 26d2–3, 277n. 29
194c–195b, 191 26d2, 257n. 19
195e–196c, 191 26d5-e1, 263
196b8–10, 198 26e–27b, 274
196c1–2, 198 26e3–4, 263
196c4–d2, 198 26e3, 252n. 8
200a11–12, 20n. 21 26e4–5, 250n. 4, 263
200c3–4, 20n. 21 26e4, 250n. 4
201c–210d, 187 27ab, 90
210d4, 20n. 21 27a2-b9, 249
Timaeus 27a2-b6, 258
17a2–3, 254n. 14 27b2–3, 258n. 20
17a2, 273 27b3, 252n. 8
17b2, 273 27b7–8, 254
17c1–19b2, 254 27d–29d, 280n. 36
17c1, 273 28a–29a, 91n. 3
17c2–3, 258 28b, 33
18a1, 91 28c, 33n. 20
19b–20c, 273, 276, 279 29cd, 33n. 20
19b3–20c3, 257 30d, 33n. 20
19c6, 276 31b, 33
19e8–20b1, 250n. 5 31c, 34
20a1–5, 261 32c, 34, 91n. 5
Plato (cont.) 68e–69a, 35n. 27
33b–34a, 33n. 19 69c–72d, 118n. 22
33bc, 34 69c–73a, 36n. 33
34ab, 34 69c, 37
36c, 35 69d–71e, 90
36d, 35 82a6, 91
37b, 35 82e3, 91
37c, 35 83a5, 91
37d, 35 86d–87b, 91
38c–39e, 33n. 19 87a7–b8, 91
38c, 35 87c–90d, 38n. 40
38d, 35 88e3, 91
38e, 35 88e5, 91
39a, 35 88e6, 91
39bc, 102 89e3–90a2, 86
39b3, 102n. 28 90a–d, 118n. 22
40a, 35 90a2–d7, 86
40bc, 34n. 23 90d, 104
40b, 35 91d–92c, 104
40cd, 104 Plotinus
40c, 35 Enneads
40d–41a, 33n. 20 1.2.3.8, 83n. 25
40d, 35, 104 2.4.16.3, 232n. 16
41a-b, 44n. 58 Plutarch
41a, 34 Against Colotes
41b, 34 1120CD, 184
41d–42d, 37, 91 1120E, 176n. 18
41de, 37 Life of Solon
41e, 49n. 76 26.1–4, 271n. 7
42ab, 36 31.3–32.2, 271n. 7
42a8, 109 Porphyry
42b, 37 On Abstinence
43a, 36n. 31 1.7–12, 297n. 23
44ab, 91 Proclus
44b7, 109 On the Parmenides (ed. Steel)
44de, 36n. 33 654.15–26, 169n. 7
44e, 37 657.5–10, 169n. 7
46e7–47c4, 87 658.9–15, 154n. 9
47ab, 103, 108 658.22–3, 154n. 9
47bc, 91 Protagoras
47b7–c1, 19 fragments (80 DK)
47c, 104 B4, 228
47c2, 105n. 34
48a, 35n. 27, 36 Sextus Empiricus
48b–d, 109 Against the Professors
48cd, 103n. 31 7.13, 155n. 13
49bc, 104 7.190–200, 175
49c, 36 7.191–2, 175
51c–52a, 17 7.194–5, 175
51c7–d1, 17 7.194, 177
51c7, 18n. 18 7.195–6, 177
51e, 109 Outlines of Pyrrhonism
53d, 103n. 31 1.31–4, 201n. 17
54c, 104 Solon
59cd, 109 fragments (ed. West)
59d, 109 22a, 270n. 2
Stobaeus 1.138.1-2, 271n. 5
Eclogues (ed. Wachsmuth and Hense) 2, 275
2, 49.18–22, 90n. 2 2.37.1, 108n. 40
Strabo 3.70.5, 233n. 18
Geography
2.3.6, 271n. 7 Xenophon
7.5.5, 291 Cyropaedia,
10.4.10, 291 70n. 3
Suda 1, 293
ε.3539 (Εὐκλείδης), 151n. 5 7.5.47, 293
7.5.71, 293
Theophrastus 8.1.4, 293
On Sense Perception Memorabilia
49–50, 94n. 12 1.2.29–31, 253
Thucydides 1.2.48, 58n. 24
Histories 3.11.17, 58n. 24
Index of topics
aidōs (‘shame’), 293–5, 297 vs. becoming, 123

Alcinous, 169 belief, 7
Alline, H., 70, 73 see also falsity
allodoxia (‘interchange of opinions’) Berkeley, G., 201n. 21
see Argument Bloom, A., 17n. 15, 23n. 30
analytical philosophy, 2, 186 body, 4, 27–50
anamnesis (‘recollection’), 25, 30, 67, 138 as obstacle, 4, 29–31
Andocides, 294 as vehicle, 30, 32–8, 44–7
Annas, J., 124 care for, 61–5
Antatticist, 70 heavenly, 35, 44, 48n. 72, 50, 104
appearances, 6 mortal-immortal, 33–6, 39–40, 49–50
scope of, 99–102 Brentano, F., 214
vs. reality, 122–9, 176, 279 Brown, L., 224–6
Argos, 288–9 Burge, T., 195n. 11
Argument Burnyeat, M. F., 92, 105, 224–6, 261n. 25
from Being and Not-Being, 187
from Interchange, 187, 203 Cambiano, G., 163n. 24
from Knowing and Not-Knowing, 187, Campbell, L., 169
199–200 Cave image, 5, 68, 90–109, 127
Aristippus, 7, 167, 178, 179, 181 Chappell, T., 170
Aristocles, 183–4 Cicero, 56, 70, 70n. 2, 182
Aristophanes, 60 circularity, 18–21, 78, 144–6
Aristotle, 3, 4, 8, 51, 67, 67n. 41, 86, 122, 131, 148, Clare, J., 222, 225, 226
149, 164, 177, 193, 195n. 10, 197, 213, 215–17, contextualism, epistemic, 261–2, 261n. 26
236, 271, 275, 280, 283–4, 283n. 2, 285, contrariety, 238–9
285n. 5, 287–8 conventionalism, linguistic, 194–5
Athena, 258, 260, 266 Cornford, F. M., 91
Athens, 8, 251–4, 253n. 10, 292, 293–7 cosmology, 5, 90–2, 104–5, 106–9, 249, 255
Atlantis story, 8–9 courage, 291–2
status of, 249–68, 269–82 Critias, 8–9, 252–4, 276
transmission of, 259–61, 264–5 Cross, R. C., 25n. 31
Augustine, 210n. 36 Cyrenaics, 7, 167–85
Aulus Gellius, 70 Cyrus, 292–3, 296–7
Averroes, 75n. 16
Damascius, 169
beauty, 42 death, 59–61
form of, 39, 43, 49 Delphic oracle, 291
being, 8 democracy, 286–7, 297
and non-being, 8, 211–12, 234–7, 240–5 Denyer, N., 230
as predicate, 7 see also einai Descartes, R., 22n. 28, 221
338
Index of topics 339
desire, 79, 113 friendship (philia), 9, 93, 160, 283–97
for the good, 79, 122–9
despotism, 286 Gadamer, H. G., 53n. 10
dialectic, 6, 23–4 Galen, 70
expertise in, 143–4 Giannantoni, G., 151n. 5
dialogue, Platonic Gilgamesh problem, 188
dialogue form, 3, 14 Gill, C., 93
see also digression good
frame, 53, 130–2, 148 human vs. absolute, 122
unity in, 3, 6, 21–2 see also form
Diès, A., 22n. 26 Gorgias, 1n. 2, 13n. 7, 157
digression, 3, 4, 10–26, 74–5, 85–6
and progression, 18–26 harma (‘carriage’), 45–6, 49n. 76
Diodorus Siculus, 291 Heidegger, M., 122
Dodds, E. R., 27 Heraclitus, 57, 172
Donnellan, K., 195n. 11, 210n. 35 and ‘flux’, 207–8
Dorion, L.-A., 155–7, 164 Hermarchus, 297n. 23
doxomimetic (‘imitative of opinion’), 11 Herodicus of Babylon, 273
due measure see measure Herodotus, 275
Hippias, 157
education, 118–20 Hirmer, J., 71
eikasia (‘conjecture’), 96–8, 96n. 16 homonoia (‘concord’), 283–4, 291–2, 294–5
eikos (‘likelihood’), 261n. 25, 280 Hume, D., 201n. 21
einai (‘to be’), 221–48
Empedocles, 91n. 5, 237n. 23 identity, 134
empiricism, 217n. 44 criteria of, 218
equality, 286–7, 297 substitution of, 188
eristic, 6–7, 150–66, 153n. 8, 171 imitation see mimesis
erōs, 28, 31–2, 38–9, 42–4 incorrigibility
essence see form conceptual, 203
Eubulides, 171, 171n. 11, 171n. 12 perceptual, 201–2, 205–8
Euclides, 7, 150–2, 151n. 5, 155–7, 170–2 indifferents, promoted, 131
existence, 67, 197, 209, 216 intellectualism, 6, 76, 80–1
and predication, 186n. 1, 214n. 40, 221–48 see also Socrates
see also NEXNOTH intermediaries, 218
eyes, 95, 108 intuitionism, mathematical, 215
irony, 153
fallacy, theory-laden, 132–3 Isocrates, 18n. 18
falsity
and identity beliefs (FIB), 186–220 Johansen, T., 126, 263
and statement (FS), 209–19 justice, 93, 117–18, 127, 285–8, 287n. 8, 297
fear, 293–6
form Kahn, C. H., 228–9
and essence, 123 Kant, I., 201n. 19, 201n. 21, 221–2, 246, 264
negative, 229–34, 246–8 Kaplan, D., 192n. 6, 195n. 11
of the Good, 5, 6, 75–8, 80, 93, 95, 98–9, knowledge
100–2, 108–9, 122–9 alienable, 142
Platonic, 4, 17, 40, 52, 65–9, 213–14 causation, 139
separation of, 66 conditions for, 145–6
Frede, M., 223 externalist account of, 140
freedom, 26, 288–93 higher-order, second-order, 143–4, 145–6
Frege, G., 7, 188, 200, 204, 215, 217, 229 internalist account of, 145–7
and Fregean senses, 191, 192 qualifications of, 6, 135–9
and Frege-style theories, 189–90 reflective, 145–6, 145n. 32
340 Index of topics
knowledge (cont.) of analysis, 204, 204n. 28
sophistic, 135 of identity, 204n. 28, 205
value of, 139, 141–3 Parmenides, 8, 211, 230, 240–2, 246
Kripke, S., 195n. 11 past
objective vs. mythical, 259n. 24
law, 285–6, 290, 297 reconstruction of, 8–9, 273–5, 280
of excluded middle, 149, 215 Peirce, C. S., 201
legislation, 283–4 Penner, T., 51
Leroux, G., 20n. 22 perception, 7, 94, 173–85
Line image, 5, 94 Persia, 292–3, 296
linguistic turn, 193–8 Phaedo, 1n. 2, 58
logic philia see friendship
and extensional contexts, 188 Philolaus, 57
and logical form, 215–17 Phlius, 4, 55–7
and logical positivists, 217 Plato
and psychological contexts, 188, 219 and conception of philosophy, 10–11, 53
neutral, 7, 219 and development of thought, 1, 3, 110–21
love see erōs and family of, 275–7
luck, 267 and literary ambition of, 278–82
Lycurgus, 283, 291–2 and psychology, 110–21
Plotinus, 231, 232n. 16
Mach, E., 217 Plutarch, 184
Marathon, battle of, 250n. 3, 253 polis (‘city’)
mathematics, 92–3, 104–9, 143–4 ideal, 21, 256–9, 275, 276–7
McDowell, J., 26n. 32 politeia (‘constitution’), 104, 283n. 2, 288
meanings, 202–3, 212–13 principle
as reference-conditions, 189, 193–8 Non-Contradiction [PNC], 6, 133–4, 134n. 11,
measure 136–7, 136n. 16, 140, 146, 147–9, 149n. 34
and art of measurement, 113 Proclus, 154, 169
and due-measure, 10, 13–14, 22 Prodicus, 13, 157
and man-measure principle, 161, 201 propositions
Meinong, A., 214 and truth value, 193, 215–19
memory, 7, 181, 272 Protagoras, 6–7, 12–14, 155, 157–9, 160–5,
traces, 190–1 168–70, 172–5, 228–9
Messene, 288–9 and appearances, 200–8
Meyer, E., 291 Pythagoras, 4, 56–7
mimesis (‘imitation’), 278–80 Pythagoreans, 55–7
monarchy, 286–7, 297
Moore, G. E., 226–8 Quintilian, 18n. 18
negation, 215, 232, 234–8, 246–7 reason, 296–7

neutral entity, 218 recollection see anamnēsis
NEXNOTH (To not exist is to be nothing at reference, 189
all), 209–10 rhetoric, 10, 22
non-being see being Rowe, C., 23n. 29, 52, 85, 88, 110–21, 124, 128–9,
Notomi, N., 243 295, 295n. 19
numbers, 191–2 Russell, B., 227n. 8
Ryle, G., 22, 22n. 25
ochema (‘vehicle’), 46–7, 49n. 76
omniscience, 136, 138, 148 Salamis, 294–6
ousia (‘essence, being’), 14, 123 Schofield, M., 92, 93, 105
Owen, G. E. L., 223, 243–5 Sedley, D., 110–21, 127–9
self-knowledge, 140, 144–7
paidia (‘play’), 278 self-participation, 244
paradox, 132–3, 171–2 self-predication, 244
Index of topics 341
sēma (‘tomb’, ‘sign’), 39–42, 40n. 46 sun
semantic value, 193n. 7 image of, 5, 94, 100, 123
sense data, 217 light of, 5, 102–4
sentences, 7, 195n. 10, 216–18 syllogism, 216–17
Sextus Empiricus, 175–9
shame see aidōs tautology, 145, 146
skill Thrasyllus, 70
making, 141–3 Thucydides, 275, 280
using, 141–3 truth
Socrates conditions of, 216–19,
and the elenchus, 7 261n. 26
and intellectualism, 4, 5, 41n. 49, 74–89, value of, 255–68
110–21, 128
and interlocutors, 55, 253–5 ultra-realism, 195, 220
vs. Plato, 3, 4, 5, 51–4, 68, 84–5, 115–16, 120,
128–9 Van Eck, J., 231–2
and psychology, 3, 110–21 virtue
Solon, 8, 250n. 3, 260, 264, 269–72, 274, 278, bodily, 81–3
283, 294 demotic, 83–4, 84n. 25
sōma (‘body’), 40n. 46 intellectual, 4, 81–4
see also body Vlastos, G., 222n. 3
sophists, 157–9
soul, 3 wax tablet
care for the, 4, 61–5, 82 argument, 187, 190–1
and immortality, 52, 58, 61, 88 model, 7, 202n. 23
and tripartition, 3, 4, 5, 74, 84–5, 86–9, 110–21 Williams, B. A. O., 255n. 17
and world-soul, 35, 36, 47 wisdom (phronesis), 6, 65, 77–8, 84, 87n. 31,
Sparta, 288–9, 291–2 288–93, 297
speech Wittgenstein, L., 204n. 27, 229
appropriate length of, 10–14 Woozley, A. D., 25n. 31
circular vs. linear, 18–21 writing, 150–1
see also digression
Sterne, L., 18, 21n. 24 Xenophon, 292–3

George Boys-Stones, Dimitri El Murr, Christopher Gill - The Platonic Art of Philosophy-Cambridge University Press (2013)

Uploaded by

Document Information

Original Description:

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

George Boys-Stones, Dimitri El Murr, Christopher Gill - The Platonic Art of Philosophy-Cambridge University Press (2013)

Uploaded by

Copyright:

Available Formats

more information - www.cambridge.

This is a collection of essays written by leading experts in honour of

g e o rg e b oy s - s to n e s is Professor of Ancient Philosophy at

List of contributors page vii

Ge orge Boy s- Stones is Professor of Ancient Philosophy at Durham

According to an anecdote preserved by Athenaeus, Socrates once reported

Macrology and digression

Macrology and brachylogy

Macrology and digression

Two conceptions of the body in Plato’s Phaedrus

The negative attitude to the mortal body in

The sōma as an ‘obstacle’ (empodion) in the Phaedo

The negative attitude to physical love in the Symposium

The positive view of the body in the Timaeus: the body

The imperishable body of the gods as a vehicle for rational activity

The teleological conception of the mortal body of human beings

The possible acquisition of a god-like body by philosophers

Understanding the Phaedrus’ puzzling attitude to the body

The positive dimension: the mortal body as a medium

The philosopher’s acquisition of a god-like kind of body in the Phaedrus

The possible acquisition of a god-like body by the philosopher In spite of

Socrates in the Phaedo

Re-examining Socrates in Plato

Socrates the philosopher

A narrative about Socrates

The ideal ‘philosopher’

Philosophy of soul and form

Practice for death

Care for the soul

Metaphysics as the awakening of the soul

Socratic intellectualism in the Republic’s

This chapter was originally entitled ‘Socratic intellectualism in Republic

Timaeus in the cave

Two projects in the Cave?

The scope of the appearances inside the cave

The sun as a source of insight

Cosmology in the philosopher’s curriculum

Socrates, crucially, says that the admirer of the heavens is a geometer;

Reflective commentary (1)

In this discussion, and in the companion piece by Dimitri El Murr, the

Reflective commentary (2)

Waving or drowning? Socrates and the

The structure of the Euthydemus

Vertical and lateral order

The third wave

The sophists to the rescue

Knowledge and its conditions

Making and using

Regresses, tautologies and circles – the toils of self-knowledge

Socrates versus the sophists

Coda: Socrates … and Aristotle

Why was the Theaetetus written by Euclides?

Socrates’ desire for eristic: the Euthydemus

Socrates on what a sophist is

The disappearance of the distinction between

The Wooden Horse

The Theaetetus as a peirastic dialogue

The choice of characters: Euclides

The subtler thinkers

Sextus Empiricus on the Cyrenaics

No criterion is common to human beings, common names are assigned to

According to Cyrenaic epistemology, we are incorrigibly aware of our own

The Wooden Horse