. 1
( 9)


This page intentionally left blank

This innovative study of Plato™s ethics focuses on the concept of
virtue. Based on detailed readings of the most prominent Platonic
dialogues on virtue, it argues that there is a central yet previously unno-
ticed conceptual distinction in Plato between the idea of virtue as the
supreme aim of one™s actions and the determination of which action-
tokens or -types are virtuous. Appreciating the “aiming/determining
distinction” provides detailed and mutually consistent readings of the
most well-known Platonic dialogues on virtue as well as original inter-
pretations of central Platonic questions. Unlike most examinations of
Plato™s ethics, this study does not take as its centerpiece the “eudai-
monist framework,” which focuses on the relationship between virtue
and happiness. Instead Aiming at Virtue in Plato argues that the dia-
logues themselves begin with the idea of the supremacy of virtue,
examine how that claim can be defended, and address how to deter-
mine what constitutes the virtuous action.

iakovos vas iliou is Associate Professor of Philosophy at the
Graduate Center and Brooklyn College, City University of New York.

Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo

Cambridge University Press
The Edinburgh Building, Cambridge CB2 8RU, UK
Published in the United States of America by Cambridge University Press, New York

Information on this title: www.cambridge.org/9780521862967
© Iakovos Vasiliou 2008

This publication is in copyright. Subject to statutory exception and to the
provision of relevant collective licensing agreements, no reproduction of any part
may take place without the written permission of Cambridge University Press.
First published in print format 2008

ISBN-13 978-0-511-45574-2 eBook (EBL)

ISBN-13 978-0-521-86296-7 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.
To my parents,
Bill and Irene Vasiliou

Acknowledgements page ix

Introduction 1
1Aiming and determining 1
2Virtue, aims, and eudaimonia 4
3Disputes about virtue and its supremacy 6
4Socrates and Plato on virtuous actions and virtuous characters:
A standard account 10
5 A brief overview of some central principles 15
6 A note on reading Plato (i): The signi¬cance of the dialogue form 18
7 A note on reading Plato (ii): Doctrines and developmentalism 20

1 Socrates and the supremacy of virtue 22
1.1 Introduction 22
1.2 The supremacy of virtue in the Apology 24
1.3 Socrates and moral knowledge 27
1.4 SV and the priority of de¬nition 36
1.5 Socrates™ criticism of his fellow Athenians 39
1.6 Socratic incontinence 42

2 Determining virtue in the here and now: Socrates in the
Apology and Crito 46
2.1 Ill-¬tting remarks in the Apology 48
2.2 The role of Socrates™ divine sign and his decision to avoid public life 51
2.3 Crito™s appeal 56
2.4 Socrates™ response 63
2.5 SV in the Crito 71
2.6 The Laws™ starting assumptions 74
2.7 The arguments of the Laws 77

3 The supremacy of virtue in the Gorgias 91
3.1 The Gorgias and SV 91
3.2 Socrates and rhetoric in the Gorgias 93
3.3 Gorgias, Socrates, and SV 98

viii Contents
3.4 Polus and SV 108
3.5 Callicles and his conception of justice 117
3.6 Callicles™ protreptic 122
3.7 Callicles™ hedonism 128
3.8 Socrates as rhetor 133

4 Trying (and failing) to determine what virtue is 137
4.1 Two commonalities 138
4.2 The dialogues of de¬nition and the “What is F?” question 140
4.3 Aiming and determining in the Euthyphro 145
4.4 Aiming and determining in the Protagoras and Euthydemus 160

5 Socrates and Thrasymachus: Republic 1 166
5.1 Socrates, Cephalus, and Polemarchus 167
5.2 Thrasymachus™ initial account of justice 172
5.3 Thrasymachus™ “de¬nitions” of justice 175
5.4 Cleitophon™s recommendation 178
5.5 Aiming and determining in the “Thrasymachus episode” 182
5.6 Socrates™ defense of SV in Republic 1 187

6 The bene¬ts of injustice 192
6.1 De¬ning justice and the project of the Republic 192
6.2 The classi¬cation of goods 194
6.3 Understanding Glaucon™s example 201
6.4 The origin of justice according to the many 203
6.5 The bene¬ts of injustice 206

7 Early education and non-philosophers in the Republic 212
7.1 Overview 212
7.2 The signi¬cance of early education 215
7.3 A tension in the account of early education 219
7.4 Philosophers and non-philosophers in the Republic 232

8 Aiming at virtue and determining what it is 247
8.1 Just actions and the just soul in Republic 4 247
8.2 Just persons 254
8.3 The virtue of non-philosophers 259
8.4 The promise of an answer to determining questions 267
8.5 The role and signi¬cance of Books 8 and 9 272

9 Epilogue 282

Bibliography 286
Index locorum 295
General index 305

Many friends, colleagues, and family members have helped me at various
stages of this project by reading and commenting on sections, discussing
relevant issues, and/or providing intellectual and emotional support. My
longest-standing intellectual debt is to my father, Bill Vasiliou. He is respon-
sible for my pursuing philosophy in general and Greek philosophy in par-
ticular. I have discussed the ideas in this book with him over several years,
and his encouragement has been unfailing. He represents for me a con-
temporary example of what it means for philosophy to inform a person™s
I also single out Brad Inwood and Jennifer Whiting. Brad was the ¬rst
person to read a complete draft of this book and his insightful comments
helped me to focus its overall aim more precisely. Jennifer read and com-
mented extensively on an earlier version of the entire manuscript, with an
attention to detail and a philosophical acumen that I have bene¬ted from
ever since she supervised my dissertation on Aristotle. She has saved me
from many errors, large and small, and, whatever its ¬nal merit, has made
this a much better book than it would have been.
I am also especially grateful to: Jonathan Adler, Marc Angers, Robert
Arrington, Rachel Barney, Alex Boro, David Bothner, Friedemann
Buddensiek, John Cooper, Matthew Evans, Elizabeth Gaffney, Christopher
Grau, Edward Harris, Thomas Hurka, Jucifer, Bill Keough, Jonathan
Mandle, Wolfgang Mann, Michael Menser, Diane Moroff, Alexander
Nehamas, Josh Ober, John Partridge, George Rainbolt, Anthony Schneider,
Andrea Schulz, Dominic Scott, Sergio Tenenbaum, Joe Tonetti, Sarah
Tonetti, Irene Vasiliou, Dina Vasiliou, George Vasiliou, Katja Vogt, Willy
Wiener, Eugenia Worman, and Nathaniel Worman.
In addition I would like to thank audiences at the American Philo-
sophical Association, Paci¬c Division, the State University of New York
at Albany, the New York Colloquium in Ancient Philosophy, Franklin
and Marshall College, the University of Toronto, Brooklyn College, and
x Acknowledgements
The Graduate Center, City University of New York. I owe particular thanks
to the members of my graduate seminar on Plato™s Republic in the Spring
2006 term at The Graduate Center, City University of New York.
I would like to acknowledge the Professional Staff Congress, City Uni-
versity of New York for research grants, which supported a course release
and research supplies, and especially the Ethyl Wolfe Faculty Fellowship
in the Humanities at Brooklyn College, which provided a necessary year™s
release from teaching in order to write the initial draft of this book.
I am grateful to Michael Sharp, Elizabeth Noden, Jodie Barnes, Iveta
Adams, and the anonymous referees from Cambridge University Press. I am
also indebted to Jake Berthot for kindly allowing me to use a reproduction
of his painting, Room, on the cover.
Finally I would like to thank Nancy Worman, my friend, intellectual
partner, and wife, whose discerning eye and courageous character have
greatly improved this book in particular, and my work and life in general.

Throughout translations are my own, but I have been signi¬cantly helped
and in¬‚uenced by the translations in John Cooper™s Plato: Complete Works,
David Gallop™s Defence of Socrates, Euthyphro, Crito, and Paul Shorey™s
There is some overlap between chapter three, section four of this book
and section 2 of my “Disputing Socratic Principles: Character and Argu-
ment in the ˜Polus Episode™ of the Gorgias” in Archiv f¨ r Geschichte der
Philosophie 84.3 (2002), 245“72.

1 a iming a nd de t e rm i ni ng
In the Cleitophon, a short and strange dialogue attributed to Plato, the char-
acter “Socrates” speaks only twice. He accuses the eponymous interlocutor
on the one hand of telling people that it is a waste of time to associate
with him, while on the other of lauding contact with Thrasymachus, the
infamous character from Book 1 of Plato™s Republic. Cleitophon replies
that Socrates has not heard the whole story: he is in certain respects deeply
impressed by Socrates, but in other ways sharply critical. Always open to
correction and betterment, Socrates is happy to hear Cleitophon™s com-
plaints and the last four Stephanus pages of the work consist solely of a
speech by Cleitophon in which he sharply distinguishes between two tasks:
(1) persuading a person that virtue1 is more important than anything else;
and (2) saying precisely what virtue is. According to Cleitophon, Socrates
does an excellent job, better than any other person, at persuading and
exhorting people to pursue virtue and the care of their souls (407a7,
410b4“6), but he is utterly unhelpful when it comes to saying what virtue
actually is. Because Socrates is so useless with this substantive question,
Cleitophon is forced to conclude that either Socrates™ ability to champion
virtue does not in any way imply that he knows what virtue is, or else
Socrates is simply unwilling to tell him. It is Socrates™ failure on this sec-
ond issue that leads Cleitophon to turn to Thrasymachus (410c“d). The
dialogue ends with no response from Socrates.
Cleitophon™s speech suggests a distinction between two sorts of ethical
principles: what I call “aiming principles” and “determining principles.”
An aiming principle tells the agent what overall aim she ought to have in
acting, for example, to do the virtuous action; because this particular aiming

I translate ˆretž as “virtue” or “excellence,” varying only for stylistic reasons.

2 Aiming at Virtue in Plato
principle is so important to Socrates2 and Plato, I shall give it a name: the
“supremacy of virtue” (henceforth, SV). SV says that doing the virtuous
action trumps any other aim one may have in acting. An aiming principle
functions in two ways: as an “explicit aim” and as a “limiting condition” for
action.3 When SV functions as an explicit aim, an agent who adheres to SV
will explicitly aim to do the virtuous action above all. In other situations,
however, SV may operate as a “limiting condition.” When acting for some
end other than virtue (for example, pleasure or ¬nancial gain), SV requires
that the agent nevertheless not act in a way that is contrary to virtue. The
role of SV as a limiting condition is expressed in Socrates™ well-known
statement that “it is never right to do wrong.”4 We can now see how one
can follow SV in all actions, without that implying the implausible view
that in every action one ought explicitly to aim at acting virtuously. Many
actions may be morally neutral, but what is crucial about the agent who
adheres to SV is that she will never knowingly act in a way contrary to
Consider, by contrast, an agent who holds a different aiming principle
than SV; let™s call it the “supremacy of survival (SS).” According to SS a
person should aim at surviving above all. SS too may function both as an
explicit aim and as a limiting condition. Sometimes the adherent of SS
will explicitly deliberate about which action will ensure his staying alive. In
addition, SS may also function as a limiting condition insofar as the agent
committed to SS will not (intentionally) perform any action that leads to
his death when, for example, he is aiming at pleasure; “It is never right to
act in a way that leads to one™s death” would be the expression of SS as a
limiting condition. If an action does not lead to death, then the adherent
of SS is allowed to choose to do it or not on whatever grounds he likes (as
far as SS is concerned). As we shall see later in the book, SS is a view that
Socrates frequently disparages. In general, then, an aiming principle sets
the supreme aim of an agent™s action: for SV, the supreme aim is virtue;
for SS, the supreme aim is survival.5 Henceforth when I say that an agent
who is committed to SV “aims to act virtuously above all,” I mean that
as shorthand for “makes acting virtuously the supreme aim of her actions
2 Unless explicitly noted otherwise, “Socrates” refers to the character in Plato™s dialogues, not to the
historical ¬gure. I discuss my approach to the dialogues below.
3 I borrow the term “limiting condition” from Herman (1981/1993), 14“17.
“To do wrong” translates ˆdike±n, which is also sometimes translated “to do injustice.” It is important
to remember that the word carries with it the broader connotation of wrongdoing in general. This
will be particularly important in the discussions of the Crito, Gorgias, and Republic. The just action,
the right action, and the virtuous action are the same.
5 For a hedonist, the supreme aim is pleasure.
Introduction 3
so that virtue will sometimes be the explicit aim of her actions and will
always at least function as a limiting condition on actions which aim at
other ends.”
SV does not by itself rule in or out any non-evaluatively described action-
type, and it says nothing about how to determine what the virtuous action
actually is, which is precisely Cleitophon™s complaint. I thus distinguish
between establishing the supreme aim of an agent™s action (which is the
function of an aiming principle) and the distinct issue of how to determine
what action is going to constitute the virtuous action in some circumstance.6
Thus merely adhering to SV leaves open what sorts of considerations may
be relevant in any particular deliberation about what to do, as well as what
particular action such deliberation might yield; commitment to SV ensures
only that, barring error about which action is virtuous, the agent™s action will
be the virtuous one or at least not contrary to virtue. The pleasure or pain the
action causes oneself or others, the ¬nancial cost, the risk one runs of life or
death, may all be relevant considerations in determining what the virtuous
action actually is here and now. SV simply but importantly maintains that
a person™s aim must always be to act virtuously above all (understood as
explained above), and not to save her life, or to cause pleasure, or to generate
¬nancial gain, or even to follow the law.
By contrast, a determining principle (e.g., a proposed moral rule such
as “Never kill anyone”) would be one that actually determines which action
or action-type is forbidden or required; once you adopt “Never kill any-
one” as a principle, then, at least as far as that principle is concerned, if
an action involves killing someone, it is forbidden. The role of a deter-
mining principle may be played by a principle (a moral rule), but it may
also be ful¬lled by other means, reasonable or unreasonable, such as intu-
ition, tarot cards, following a virtuous person, and so on. I shall refer
to questions about which token actions or action-types are virtuous as
determining questions, because they involve determining what the virtuous
action actually is, whereas I shall call questions about what supreme aim
one should have in acting aiming questions. According to Cleitophon, then,
while Socrates has a clear answer to an aiming question (and he is apparently
very persuasive about this), he utterly fails to offer answers to determining

6 There are parallel, but less interesting, determining questions at issue with SS as well: “Here and
now what token action will save my life?”, or, when SS functions as a limiting condition, “Will this
token action lead to my death?” These are less interesting than determining questions about virtue
because they are not ordinarily dif¬cult to answer nor the subject of dispute.
4 Aiming at Virtue in Plato
It is well known that the ethics of Plato and Aristotle do not offer us
determining principles.7 We look in vain in their writings for particular
moral rules, containing only non-evaluative terms, that determine which
actions are virtuous and which vicious. Indeed, for many contemporary
scholars, it is an advantage of the ancients that they do not fall into what
is seen as the trap of trying to supply determining principles, but instead
focus on people™s characters. By developing a virtuous character, the agent
will act virtuously because of the distinctive outlook on the world that she
has acquired. I have one comment about this here. We might agree that
there are no determining principles in Plato and Aristotle, but nevertheless
believe that they argue about moral principles of a different type, namely,
aiming principles. Socrates™ claim that one should look to virtue above all
in action and that it is never right to do wrong is such a principle (SV).8
SV, however, both because of its generality and because it contains an eval-
uative term, does not by itself resolve the problem of determining what the
virtuous or right action is, either in general or in some concrete circum-
stance. Cleitophon is understandably frustrated. He has been successfully
persuaded to commit himself to virtue, but then SV leaves him without
any way of determining what virtue is. But if we distinguish between aim-
ing and determining principles we can at least qualify the claim that Plato
rejects moral rules or principles in general: while he may deny that one can
supply determining principles, this does not imply that he rejects all uni-
versal moral principles, for he is concerned with and argues for an aiming
principle, SV.9
This book argues that in the ethics of Socrates and Plato virtue is crucially
conceived of as an aim, and that this is contrasted with determining ques-
tions about virtue, which seek to know what virtuous action is in general
or in speci¬c instances. I examine how the aiming/determining distinction
structures Plato™s conception of virtue in what are typically referred to as the
“early” and “middle” dialogues. I concentrate in detail on how arguments
in Plato about SV differ signi¬cantly from those about what virtue is, and
show that the dialogues themselves distinguish between them.

2 virtu e, a im s, a nd e u da i m on i a
Almost all contemporary discussions of ancient ethics importantly and use-
fully take eudaimonia or “happiness,” as it is traditionally translated, as the
supreme aim of action, and then explain how different ancient theories
7 An exception is the hedonic calculus proposed at the end of the Protagoras. See 4.4 for discussion.
(References such as 4.4 refer to chapter four, section four.)
8 9 I argue that this holds for Aristotle as well, see Vasiliou (2007).
See chapter one.
Introduction 5
¬ll in its content. The nature of eudaimonia is the central topic of most
philosophical discussions about ancient ethics. It is a common principle
of such studies that, beginning with Plato, what all ancient moral theo-
ries have in common despite their particular differences is a “eudaimonist
framework.”10 Its fundamental question is: “What sort of life ought one to
lead?” (i.e. “What is eudaimonia?”). The main components of philosoph-
ical conceptions of eudaimonia are virtue (both moral and intellectual),
pleasure, and the “external goods,” the last being Aristotle™s expression for
goods of the body, such as health and beauty, and material goods broadly
speaking, such as wealth, good luck, noble birth, and good reputation.
Different philosophers and philosophical schools then argue about which
combination of these goods constitutes happiness. The locus classicus for
the eudaimonist framework is, of course, Aristotle™s Nicomachean Ethics.
He begins by offering a highly abstract and formal account of happiness,
and then seeks to specify its content. The situation with Plato is some-
what less clear, although he is still understood as belonging within this
A reader might reasonably think that there is a kind of aiming/
determining distinction that operates at the level of eudaimonia. We take as
our starting point that all people aim to do well and live well, that is, to live
happily, and then we seek to determine what happiness is. One might say,
then, that the aiming principle is “the supremacy of happiness” and that
determining principles tell us what happiness consists in. I refrain, however,
from using “aiming” and “determining” this way in the context of eudai-
monia and its determination. Of course I do not deny that our ordinary
language (and ancient Greek as well) speaks of aiming at happiness and of
determining what it is. But it is signi¬cant that the posited “aiming prin-
ciple,” “the supremacy of happiness,” is practically speaking a tautology,
as Plato and Aristotle themselves admit. Aristotle says that there is general

10 See, e.g., Annas (1993). This claim does not include the Cyrenaics, who Annas argues constitute the
exception that proves the rule.
11 Vlastos (1991), 203, refers to “the Eudaemonist Axiom” and claims that “once staked out by Socrates,
[it] becomes foundational for virtually all subsequent moralists of classical antiquity.” Some of those
who bring to bear the eudaimonist framework most strongly recognize that Plato does not raise the
same explicit questions about happiness as Aristotle, although they still interpret the dialogues, both
“early” and “middle,” as eudaimonist. For example, Irwin (1995), 248, writes: “At the beginning of
the [Nicomachean] Ethics, Aristotle sees that it is important to form some conception of happiness
before trying to decide whether different claims about how to acquire happiness are justi¬ed. We
noticed that the Socratic dialogues do not take up Aristotle™s question. In the Republic Plato does not
take it up either, but we must try to identify assumptions about happiness that convince him that the
just person is happier than the unjust.” Brickhouse and Smith (1994), 103, claim that the “Principle
of Eudaimonism,” the view that “a thing is good only insofar as it is conducive to happiness,” is “at
the heart of Socratic ethics.” They cite no texts to justify this.
6 Aiming at Virtue in Plato
verbal agreement on this question (NE 1.4, 1095a17“20).12 Everyone wishes
to do well; no one would say that he wishes to do badly, no matter how
warped or ¬‚awed his conception of doing well may be. Contrast this with
the genuine aiming principle, SV. To say that one should act virtuously
above all is a substantive and controversial claim, and I shall restrict the
concept of an aiming principle to such claims. Thus the question “What
is eudaimonia?” is an “aiming question” insofar as it asks what a person™s
supreme aim should be. But for it to count as an aiming principle, one
would have to state in a contentful way what one™s supreme aim is, for
example, virtue or pleasure. Saying that one™s supreme end is living well
or doing well is not yet to make a substantive claim about what one is
aiming at. For this reason, I do not regard “the supremacy of happiness” as
an aiming principle. Given that SV is the aiming principle, the question
“What is virtue?,” either in general or in some concrete circumstance, then
counts as a determining question.
I restrict the terms in this way because the focus of this study is virtue,
not eudaimonia. Socrates, Plato, and Aristotle believe that eudaimonia is
essentially bound up with virtue, even if each differs about the relationship
of virtue to eudaimonia.13 They all maintain SV, and so they believe that
one ought to aim at virtue above all; none of these philosophers holds that
one should ever act contrary to virtue. In contrast to most contemporary
work, this study begins from the idea that virtue is supreme (as I shall
argue Socrates himself does), considers how this claim is defended, and
then asks how we determine what it is. I hope to show that the focus on
virtue as an aim yields new interpretations of central Platonic dialogues and
leads us to lesser-known passages within these texts that have not attracted
notice in part because of the almost universal focus on eudaimonia and the
eudaimonist framework.

3 d ispu tes a bou t v irt ue a n d i ts su prem acy
According to the above distinction I separate two types of deliberation. In
one sort, let™s call it “aiming deliberation,” we assume that a person has
determined, somehow, that one action is the virtuous one and a different
action is, for example, ¬nancially pro¬table, but contrary to virtue. A person
might then wonder which she ought to do, the virtuous action or the
12 Plato does as well: for example, Euthyd. 278e.
13 I say “essentially bound up” as a way of remaining neutral about the precise nature of the relationship
between them “ whether virtue is necessary and/or suf¬cient for happiness, whether it is virtuous
activity rather than virtue, and so on.
Introduction 7
pro¬table one. The familiar question of the moral skeptic arises at this
point: why should I do the virtuous action, she asks, rather than the non-
virtuous but ¬nancially pro¬table one? Socrates believes that virtue as an
aim ought always to trump whatever other aim we may have in acting;
one should never act contrary to virtue (SV). The principle SV, however,
in no way answers the skeptic. By providing the supreme aim of choice
it simply tells the skeptic and everyone else which of the two actions to
choose. Now Socrates™ claim that virtue is more important than anything
else has been interpreted, loosely, as saying that he thinks that one should
not care at all about money, physical health, death, and so on, which is
taken to mean that he would never take such things into consideration in a
deliberation whose aim is to do the virtuous thing. But we shall see that this
is incorrect. What we might call, by contrast, “determining deliberation”
is quite different from aiming deliberation. Once we have accepted that
virtue is the supreme aim (and thus excluded at least one kind of moral
skeptic from the conversation), virtue is not then also a consideration in
the deliberation about what constitutes acting virtuously in this or that
circumstance. External goods, however, will be.14
While no one disputes that she wants to be happy nor does any one need
to be persuaded to be happy, that we ought to do the virtuous action above
all is another matter. This leads to the question of why doing the virtuous
thing should be our supreme aim, and not, for example, ¬nancial gain
or survival. We, like Socrates and his contemporaries, have beliefs about
what is right and wrong and we can act on them without having answers to
questions about their origin or their justi¬cation. In the context of an aiming
deliberation Socrates assumes that a person has, somehow, determined that
action A is the virtuous action but not ¬nancially pro¬table, while action
B is ¬nancially pro¬table but not virtuous. Given this, he asks, “Do you
think you should choose A or B?” And in fact for most of us, I think, the
answer is “A.” The moral skeptic, of course, says “B”; so the simple asking
of this question fails to move him.15 But for almost everyone else, this
14 I show in chapter two the damage that this confusion has done to our understanding of the Apology
and Crito. Since it appears that Socrates is taking into consideration his life, his children, his friends,
money, and so on in deciding what to do, commentators must ignore prominent parts of the Apology
and Crito (or else dismiss them as rhetoric, sophistry, irony or merely ad hominem argument). But
if we appreciate that he is not taking the good condition of these things as his aim, but only taking
them into consideration as factors in determining what the virtuous action is, we shall see that there
is no con¬‚ict, and we can make sense of all of his remarks as consistent.
15 An ancient skeptic would of course say “no more A than B.” But in this discussion by “skeptic”
I mean someone who in some way challenges or repudiates ordinary claims of morality. Thus
Thrasymachus and Callicles will count as skeptics, even though they clearly do not meet the ancient
de¬nition of a skeptic as someone who suspends judgement.
8 Aiming at Virtue in Plato
point has some teeth. Without a clear method for answering determining
questions, the call by itself to commit to SV will rely on people™s untutored
beliefs about which actions are virtuous. But even so, being consciously
committed to SV can have a substantial effect on action, and a potentially
good effect, if people paused in many situations for one additional thought
and asked themselves whether they were doing what they took to be the
virtuous action.
We should recognize that the agent committed to SV does not necessarily
have to think the thought “I ought to do what is virtuous” each time she acts,
even in cases where there is a virtuous action that must be done. The worry
here is related to a criticism by Bernard Williams. Williams maintains that
in a variety of cases the thought that one ought to do the virtuous action
would be “one thought too many.”16 If a loved one falls into a river, and an
agent begins her deliberation about what to do with the thought that she
ought to do what is virtuous above all, it does seem reasonable to think that
there is something amiss. A virtuous agent would not have such a thought,
regardless of its truth, before acting. And if she did, it would detract from
her virtue.
We can agree with the “one thought too many” point, but still quite
reasonably acknowledge that Socrates™ fellow Athenians frequently have
what we might call “one thought too few.” Ignoring entirely any question
of whether they are acting virtuously, they focus simply on the aims of
survival, wealth, reputation, and so on. It seems to me that the same holds
for us as well. Many of us might act better if we paused to ask whether we
were aiming at the excellent action, or simply aiming at what secured our
professional reputation, ¬nancial gain, pleasure, and so on.17 We can grant
the correctness of Williams™ point “ particularly in cases where a quick and
relatively straightforward decision must be made “ while still recognizing
that sometimes, indeed perhaps fairly often, an agent ought explicitly to
remind herself of her commitment to SV. I believe that this is a signi¬cant
part of the force of Socrates™ role as “gad¬‚y” of Athens (Ap. 30e): he accuses
his fellow citizens of typically having one thought too few. As we shall see
in chapter two, when Socrates must decide whether or not to escape from
prison in the Crito, he does think that it ought to be explicit that the aim
of his action is to do what is virtuous. For these reasons I shall retain the
expression “aiming at virtue,” with the understanding that it ought not to
16 Williams (1976), 214“15.
17 There is a related (and quite complex) political question about the deliberations of nations. Is the
military supremacy, economic health, or even the survival of a nation more important than its acting
Introduction 9
imply that the agent will necessarily have a particular thought in her head,
although sometimes she will and she ought to.
A different problem arises insofar as someone could easily, of course, be
too cavalier about his assumption that the action he is about to engage
in is in fact the truly virtuous one. This is just the sort of danger a ¬gure
like Euthyphro poses. While Euthyphro™s commitment to SV is secure,
he shows no interest in deliberating about determining what the virtuous
action is; indeed, I shall argue that Plato depicts him as even failing to
understand the importance of the question (see 4.3). Moral and religious
fundamentalists claim to devote themselves to what is right, and not to what
is pleasurable or ¬nancially pro¬table, but they then appear to think that
determining what is right is an entirely straightforward matter. They share
Socrates™ commitment to SV, while by contrast with him they are unques-
tioningly con¬dent that they know what virtue is; this can be a dangerous
and repellent combination. In examining the difference between aiming
and determining questions, we, unlike Euthyphro, must be careful not to
“moralize.” A person “moralizes” if she, like a fundamentalist, takes deter-
mining questions about virtue to be prematurely settled without adequate
justi¬cation. Thus a person moralizes if she assumes, without argument,
that, for example, telling a falsehood is always wrong or doing someone
physical harm is always wrong. One might agree with Socrates that one
must never do wrong, but then be unjusti¬ably con¬dent that one knows
which actions or action-types are wrong. This would be “to moralize” in
the sense I intend.
When Socrates claims that it is never right to do wrong, the question that
ought to follow is, “What is right and wrong?” That is, SV leaves us with
a puzzle about how to determine what virtue is, while accepting that virtue
ought to be our supreme aim. There are two questions here: one might
ask what the virtuous action is in the here and now, and one might ask
what virtuous actions are in general. The dialogues, as we shall see, address
both of these questions. In chapter two, we shall examine how Socrates
deals with determining what the virtuous action is in the here and now.
He offers an example of how to put SV to work in action. In the so-called
dialogues of de¬nition, considered in chapter four, we see Socrates and his
interlocutors try to determine what virtuous actions are in general by trying
to answer the Socratic “What is F?” question.
I have just discussed the force of SV on ¬gures other than the moral
skeptic. The distinction between SV and questions about what virtue is,
however, also results in a proliferation of skeptics. One could hold an ordi-
nary conception of virtue, but deny SV, that is, deny that one ought to
10 Aiming at Virtue in Plato
do the virtuous action above all; in chapter three we will see that Polus
in the Gorgias is such a ¬gure. Then again one might agree that virtue is
most important, but hold a radical and unconventional conception of what
virtue is, disagreeing both about particular token actions and more general
categorizations; Callicles is such a ¬gure. Finally one might be willing to
shift positions both on whether virtue is supreme and on what virtue is. I
shall argue in chapter ¬ve that this ¬ts Thrasymachus. So although I shall
begin in the opening chapters, following Socrates™ lead, by ignoring the
challenges of the moral skeptic, the conceptual structure that I see at work
in the dialogues actually generates not simply one skeptic, but a range of
skeptics. The dialogues work through these possibilities to reveal the partic-
ular requirements that arise in dealing with particular types. In chapters six
through eight I turn to Republic 2“10. I argue that the aiming/determining
distinction is crucial to understanding the central argument of that work
and its unity. We shall see that Republic 4™s notorious answer to the question
“What is justice?”“ that it consists in the harmony of the tripartite soul “
is part of the justi¬cation for SV; it tells us why we should be just. By
contrast the metaphysics of the middle books, which introduce the tran-
scendent Forms as the objects of knowledge for philosophers, explains how
the outstanding determining questions may be answered.

4 soc rate s a nd pl ato on v irtuou s acti ons a n d v i rtuou s
c h a racters : a sta nd a rd account
It may be useful here in the Introduction to sketch brie¬‚y what I take
to be a common understanding of Plato™s ethics and then by contrast to
explain how this book™s focus on virtue and on the aiming/determining
distinction affects it. In the dialogues typically called “early” or “Socratic,”18
the character Socrates believes that virtue is the most important thing, but
he also disavows knowledge of what it is. This is one of the “paradoxical”
features of Socrates. At the same time, a familiar account proceeds, Socrates
believes that virtue is knowledge. This consists of two claims: (1) that
knowledge is necessary for virtue; and (2) that knowledge is suf¬cient for
virtue. While the ¬rst is plausible enough, the second is far-fetched, for
it denies the possibility of incontinence. The more mature Plato corrects
for this. In the Republic, for example, Plato introduces a tripartite division
within the soul, which allows for the possibility of intra-psychic con¬‚ict. He

18 See Vlastos (1991) and Irwin (1995).
Introduction 11
thus rejects Socratic intellectualism by allowing that there are non-rational
motivations in a human being that can con¬‚ict with, and sometimes win
out over, a person™s reason, leading him to act incontinently. Not all wrong
action, then, is a matter of ignorance. Further, the “Socrates” of the Republic
(i.e. Plato) provides positive accounts of the nature of the virtues, which
had eluded the “Socrates” of the earlier, aporetic dialogues.
On such an account we begin with a Socrates who does not know what
all virtuous actions have in common; he cannot ¬nd the Socratic form
which would enable him knowledgeably to identify one action or another
as being virtuous. This is a serious epistemological problem: how do we
know what the virtuous action, either in the here and now, or in general, is?
But Socrates also, on the usual reading, maintains something about what
it is to be virtuous. To possess a virtue is to possess knowledge. At this
point, commentators rapidly proceed to discuss the implausibility of the
idea that being virtuous consists entirely in the possession of knowledge.
The discussion has moved from a question in moral epistemology “ how
do we know what the right action is? “ to a related, but distinct, question in
moral psychology: is possession of knowledge by itself suf¬cient for an agent
to act on that knowledge? What gets dropped in the shift to a discussion of
the possibility of incontinence is the question of how to identify virtuous
actions in the ¬rst place. Once we are discussing the merits or defects of
the idea that virtue is knowledge, we are concerned with the question of
what it is to be a virtuous person and have stopped considering how to
determine which actions are virtuous.
Of course the standard reading of Socrates and Plato explains this. Plato
has moved from an “act-centered” account of virtue to an “agent-centered”
one. Accordingly, when Plato de¬nes the virtues in the Republic, he de¬nes
what it is to be just, courageous, and so on, but he does not tell us what
all just or courageous actions have in common. This is because Plato has
allegedly given up trying to give an account of just action, having seen the
futility in that,19 and will instead say what it is to be a just person. And
once there is an account of this, then the just actions will be those actions
that are done by a just person. Presumably the fact that these actions issue
from someone whose soul is harmonious ipso facto makes them just.
I am not denying the importance of questions about incontinence. But
we should be clear that if Socrates believes that virtue is knowledge, and,
presumably, knowledge of goods and evils, then what the virtuous person
19 As perhaps Socrates also had when he turns in the “dialogues of de¬nition” to questions about, for
example, what it is to be courageous rather than questions about what courageous action is.
12 Aiming at Virtue in Plato
will know is how to determine which actions are good and which are
evil. He will have some knowledgeable way of correctly identifying what
is objectively the right or wrong action in any situation (to take examples
that will be important, whether one should escape from prison or prosecute
one™s father for murder). Whether knowing this will be suf¬cient to get him
to act in accordance with that knowledge is a wonderful and interesting
problem, but we shouldn™t allow it to eclipse the problems of what the
knowledge is that enables one to determine what is just or virtuous, of how
one should act given that one does not possess this knowledge (as Socrates
claims is his own position), and of how such knowledge might be obtained.
I argue in this book that the problem of determining what the virtu-
ous action is in concrete circumstances, perhaps via knowing what virtu-
ous actions are in general, is front and center throughout the “early” and
“middle” dialogues. Socrates never claims to know which actions (or action-
types) are virtuous, neither in the “early” dialogues nor in the Republic.
He consistently disavows throughout the entire corpus knowledge of how
to determine whether some action is virtuous or not.20 When Socrates
“de¬nes” the virtues in Republic 4 he is not answering the question that was
unanswered in the earlier dialogues (see 8.1“2). In those dialogues Socrates
is in part searching for a way of determining which action is virtuous, and
accounts of what it is to be virtuous go no distance whatsoever towards solv-
ing that problem (see 4.2). If Euthyphro is attempting to determine what
the right action is with respect to his father, knowing that justice is a har-
mony in the tripartite soul or that virtue is knowledge of good and evil does
not help at all. He needs, as Socrates tells him, some way of determining
whether prosecuting his father for murder is the right action or not. I shall
argue that the Republic does not replace an act-centered account of virtue
with an agent-centered one. Rather Socrates tells us where knowledge that
would settle outstanding determining questions might be found (although
he consistently continues to disavow having it): in knowledge of the Forms.
The answer to Socrates™ “What is F?” question in the Republic is not the
account of virtues in Book 4, but the Forms. What makes actions, people,
and everything that is just just is participation in the Form of Justice. The
tripartite division of the soul with the corresponding accounts of what it is
to be courageous, wise, and so on, by itself leaves the determining questions

20 His divine sign is an exception to this, but Socrates does not consider the divine sign as providing
him with knowledge or understanding. The divine sign, as a form of revelation, simply tells Socrates
not to do a certain token action. See 2.2.
Introduction 13
This book, then, focuses on moral epistemology in the sense that it
argues about Plato™s concern with the problem of how to determine what
the virtuous action is. It claims that this question is central to the dialogues™
discussion of virtue and is not supplanted or resolved by switching to
questions about what it is to be a virtuous person. If one is committed to
SV, she will be committed to being virtuous, since, by de¬nition, a virtuous
person does virtuous actions. One who is committed to SV is motivated
to act virtuously above all. This commitment also, I argue, focuses the
agent on the proper problem: how to determine what the virtuous action
is. Accounts of what it is to be virtuous are important for defending SV. A
person might wonder why she would want to be virtuous. If Socrates can
explain how being virtuous bene¬ts a person, then she would see why she
ought to be committed to doing virtuous actions above all and thus to being
virtuous. We shall see that Socrates supplies consistent, but progressively
more sophisticated, defenses of why one should commit to SV in the Crito,
Gorgias, and Republic.
To avoid misunderstanding, I should say that I do not have an origi-
nal contribution to make about the nature of Socratic ignorance; I simply
agree with the standard response in the literature that Socrates denies hav-
ing knowledge about what virtue is (which amounts to a denial that he
can successfully answer the “What is F?” question about virtue: see 1.3.2),
but that he doesn™t deny that he may have beliefs and indeed true beliefs
about what virtue is. I am thus not especially concerned with and never
discuss in any detail the paradox about Socratic inquiry as expressed in the
Meno (80d ff.).21 I do not maintain that Socrates is “completely ignorant”
about the content of virtue. As is well known, Socrates often takes ordinary
examples of virtuous actions for granted in the course of his arguments.
The distinctiveness of my argument will concern what Socrates™ lack of
knowledge is about “ namely the answer to determining questions as to
what the virtuous action is, either in the here and now or in general “ and,
by contrast, what he claims to have knowledge about “ that he ought to be
committed to doing the virtuous action above all (SV) (see chapter one).
Socrates™ view, and my interpretation of it, would indeed be absurd if it
claimed that he adheres to, and even, as I argue, claims to know SV, but
then is completely ignorant (that is, does not even have any beliefs about)
which actions or action-types are virtuous. For commitment to SV to be at
all meaningful of course we must, as indeed we do, have some beliefs and
ideas about what is right and virtuous. As will be clear in chapters one and

21 See Fine (1992) and (2004) for a clear account.
14 Aiming at Virtue in Plato
two, Socrates must rely on his beliefs about examples of virtuous behavior
just like anyone else; he is different only insofar as he never claims to know
which actions are virtuous.
Furthermore, if all one has are beliefs about which actions are virtuous,
but doesn™t have knowledge, one might worry that perhaps one™s beliefs
about which actions are virtuous may be entirely off the mark. The rec-
ollection theory, about which I say little in this book, might be a way of
mitigating this sort of worry. If it is true that our souls existed prior to our
birth and were in cognitive “contact” with the Forms, knowledge of which
would enable us to know which sensible actions were just and unjust, then
we might have some reason to think that our beliefs about virtue are not
entirely false (especially those about which there is signi¬cant agreement)
insofar as they may be partially recovered by our being reminded of Forms
that we once knew but forgot (see 8.4).
A second familiar topic in the literature on the dialogues I discuss, but one
which receives little direct treatment in this book, is the Socratic elenchus
and the so-called “problem of the elenchus.”22 I cannot begin to treat
this topic properly here nor do I have any original view about how the
elenchus works. I do, however, discuss closely related topics such as the
Socratic “What is F?” question (1.4 and 4.2), which involves universally
failed attempts to say what all virtuous actions have in common, and the
priority of de¬nition. In the terms I am using, if Socrates (or his interlocu-
tor) could successfully answer the “What is F?” question about virtue (which
they never can), then they would know what all virtuous actions have in
common and they would have answered the determining question for virtue
in general; this knowledge, in turn, would enable them to say of any token
action whether it is virtuous or not (see 4.2). As everyone agrees, the elenchus
is (at least) used to elicit contradictions in an interlocutor™s beliefs during his
attempt to answer a Socratic “What is F?” question. All that needs to be con-
cluded from this, as far as I am concerned, is that the interlocutor does not
know what all virtuous actions have in common, since he has inconsistent
beliefs; neither he nor Socrates need know which belief is necessarily the false
one. More importantly, the principle that I shall argue that Socrates claims
to know, SV, is not established via the elenchus but via an argument about
the effect of virtuous actions on the soul, elaborated with increasing sophis-
tication in the Crito, Gorgias, and Republic (see 2.4, 3.8, 7.2“3, and 8.1).23
22 See Vlastos (1983/1994) and Benson (2000), chs. 2“4, for discussion and additional references. See
too Nehamas (1998), ch. 3.
23 A reader may be concerned that if, as I am arguing, Socrates claims to know that virtue is supreme
(SV) without knowing what virtue is (i.e. without knowing how to answer the “What is F?” question
Introduction 15
I hold, then, a rather de¬‚ationary view of the elenchus and do not think
it establishes ethical truths. I should add, however, that I most emphatically
do not think there is nothing “positive” about the method. Indeed, insofar as
it exposes the lack of knowledge of an interlocutor about how to determine
what the virtuous action is, it helps (or ought to help) avoid the serious
danger of moralizing (see section 3 above, 5 below, and 4.3).
A ¬nal issue that is often treated in discussions of Platonic and Socratic
ethics, but that I do not discuss in any detail, is the set of puzzles related to
“the unity of virtue.” As should be clear from my discussion of SV here and
in the opening chapters, the virtuous action for Socrates is the action that is
the excellent, right action in some circumstance (e.g., in Euthyphro™s prose-
cution of his father or Socrates™ deliberation about escaping from prison). I
focus on how Plato argues that one ought to do the virtuous action above all
and how he addresses the problem of determining what the virtuous action
is. The differences between the particular virtues (temperance, piety, and
so on) and the ways in which they relate to one another do not concern me

5 a b rief overview of some ce nt ra l pri nci ple s
Let me describe a couple of concepts and principles that will be critical to
this study. Some of these will be uncontroversial, others not. Here I am
simply stating them, and in a very general way showing some connections
among them, and between them and the aiming/determining distinction.
I start with a point that will be familiar to anyone who has experience
with Greek philosophy. For the Greeks, to be alive is to have a soul; using
words with Latin roots, having a soul marks the difference between animate
and inanimate objects. Thus to deny that things have souls in the Greek
sense would be as absurd as to deny that there are living things. Human
beings thus have both a body and a soul. In a human being the soul includes
and is responsible for (at least) one™s mind and thoughts, one™s emotions,
likes and dislikes, character traits, and so on. In Greek thought, the concept
of soul by itself does not carry with it any necessary metaphysical or physical
account of what it is to have a soul. A debate about whether people have
souls in the Greek sense would be ridiculous. To deny it would be to deny
that people have minds, thoughts, emotions, attitudes, character traits, and

about virtue), then he might be guilty of violating the “priority of de¬nition.” I address this objection
in 1.4. Again, when I say Socrates doesn™t know what virtue is, I mean that he cannot answer the
“What is F?” question about virtue; I do not mean that he does not have beliefs and even true beliefs
about which actions are virtuous.
16 Aiming at Virtue in Plato
so on. I shall simply translate yucž as “soul” throughout, understanding
it as I have explained.
The belief that human beings have a body and soul is, as I have said,
common to both popular and philosophical Greek thought. But Socrates
crucially highlights the principle that the soul is something that itself can
have a good or bad condition, independently of the condition of one™s body
or one™s possessions. I shall express this idea by saying that the soul is an
independent locus of harm and bene¬t.24 Thus there are not only goods
of the soul like useful knowledge, which any ordinary person would agree
with, but also things that are good for the soul. That the body may be
harmed or bene¬ted is obvious, but the idea that one™s character or soul
might be in a state that is intrinsically better or worse is a distinct and more
subtle claim. By “intrinsically better or worse” I mean that better or worse
states of soul are not determined by considering their effects on the body
or on one™s possessions. Cleverness, for example, might be a good thing
insofar as it enables me to acquire lots of money, but that is to say nothing
whatsoever about whether cleverness is in itself a bene¬cial or harmful state
for the soul. It is necessary simply for understanding Plato™s ethical outlook
(never mind whether one agrees with it or not) that one grasps that the
soul itself is an independent locus of harm and bene¬t. We are thoroughly
familiar with the contention that our characters are the most important
part of ourselves. From earliest childhood we have been exposed to the idea
that our characters, our real selves, are distinct from what we look like, what
we own, how physically healthy we are, and so on; it is a concept central
to the entire Judaeo-Christian-Islamic tradition. But it was a much newer
idea for the Greeks than it is for us.25
If we accept that a human being has both a soul and a body, and that
the soul is an independent locus of harm and bene¬t, we may add a third
principle: the condition of the soul is of a value that is incomparable to the
value of the condition of the body or of one™s material possessions. As we
shall see, it is dif¬cult to determine with certainty whether the intended
24 The tripartite division of goods into goods of the soul, goods of the body, and possessions is common
throughout Greek thought. “Material possessions” or “material goods” ought to be understood
broadly so as to contain not only wealth and property, but good fortune, noble birth, and excellent
25 Scholars have argued that the character traits of a person were thought of as identical to what one
looked like, what one had, and what one™s visible actions were. Someone who was a king, and
therefore excellent and virtuous by birth, appeared excellent; one could see his excellence by looking
at his body, his chariots, his house, and his dress. The character of Socrates, as well as the sophists,
destabilizes this easy reading of real character from surface appearance. Plato presents Socrates as
the ugly surface appearance which masks an “inner” incomparable beauty (see Alcibiades™ speech in
the Symposium [215b ff.]).
Introduction 17
superior value of the state of one™s soul is quantitative, qualitative, or both.
A person might well ask, “Is no amount of physical well-being or gain
in material circumstances worth even the slightest harm to one™s soul?”
Throughout the book I argue that such a question is misguided according
to the ethics of Plato. The key is to appreciate the distinction between
aiming and determining, and to avoid what I called above “moralizing.”
Moralizing takes for granted what is right or wrong, virtuous or vicious; that
is, moralizing preemptively takes the determining question to be settled.
Thus, for example, acting destructively towards the state or bribing guards
is thought to be obviously wrong. Then one can easily wonder, “What if I
act in a way that is only slightly destructive of the state, but preserves my
life, which is unjustly in jeopardy?” This is typically taken as a challenge
to SV, to Socrates™ principle that it is never right to do wrong. On my
view, however, Socrates and Plato do not consider any actions described in
purely non-evaluative terms as virtuous or vicious in themselves. As long
as one™s aim is to do the virtuous action (or to not do what is contrary to
virtue) and not, say, to maximize one™s wealth, it may turn out in some set
of circumstances, for example, that it is virtuous to tell a falsehood in order
to acquire some possessions.
The fourth principle that I argue runs through Plato, with varying degrees
of explicitness, I label “the habituation principle.” According to the habit-
uation principle, engaging in actions of a certain ethical type contributes to
the formation and maintenance of a character of the corresponding type.
The habituation principle is often expressed in conjunction with examples
from health or beauty. As certain types of exercise and physical activity con-
tribute to the formation and maintenance of beautiful and healthy bodies,
so performance of certain types of actions, for example, virtuous actions,
contributes to the formation and maintenance of a virtuous character. This
principle holds for both positive and negative types of habituation.
The habituation principle explains why the type of actions one engages
in has such importance. Each action is not only the action that it is, but also,
at the same time, contributes to making a person the type of person she is.
There is no possibility of engaging in an action in such a way that it does
not “mark” one™s soul and does not contribute to the formation of one™s
character, either for better or worse. The habituation principle is a well-
known component of Aristotle™s ethics. In Nicomachean Ethics 2.1, the locus
classicus for this concept, Aristotle relies on it to explain how we acquire the
so-called “virtues of character.” How habituation works in Aristotle, and
its role in the formation of the fully virtuous character, have received much
attention in recent scholarship. But the extent to which the habituation
18 Aiming at Virtue in Plato
principle is present in Plato as well and the critical role it plays there have
not been adequately recognized. We should note here that the habituation
principle underscores the signi¬cance of the distinction between aiming
and determining questions. Given the truth of the habituation principle, it
is essential to determine correctly which actions are virtuous and which are
not. Without the ability to determine which actions are correct, one will
not be able to acquire the proper state of soul, so the incomparably most
valuable part of oneself will not be bene¬ted.
The last principle that I shall mention is that which actions are virtuous
and which are not is an objective fact, not dependent on humans™ or gods™
attitudes about what is virtuous. This is important, and conditions Plato™s
arguments against hedonism and conventionalism. It makes determining
questions urgent: what is true virtue and vice? If the good is pleasure or the
right is established by convention, then it is relatively simple to determine
what is good or right in some circumstances by determining what grati¬es
one™s appetite or what the laws of a city command. The virtuous action
would be the action that maximizes one™s pleasure or the one that does not
break any societal agreements. We shall see this issue arise in chapters three,
four, ¬ve, and eight. If hedonism or conventionalism is rejected, as it is by
Plato, we are deprived of a simple way of settling determining questions
and are thrown back to Socrates™ nagging question: what is virtue after all?

6 a not e on rea ding pl ato (i) : th e s ig ni f i ca n ce of th e
dia log ue f orm
As readers of this book are well aware, the contemporary study of Plato
employs a wide variety of methodological approaches. The reasons for
this are many and complex, involving not only features of Plato™s writing
and the time in which he wrote, but also developments in contemporary
philosophy and even literary theory. One clear reason, however, that there
are so many more widely divergent approaches to (and not simply differing
interpretations of ) Plato than, say, to Aristotle, Aquinas, or Leibniz, is
doubtless the unvarying way that Plato writes philosophy: in dialogue form
and without ever including himself as a speaking character.26

26 Excepting, if genuine, any letters. This makes Plato rare, but not unique, among philosophers in
the Western tradition. Nietzsche and Wittgenstein are two notorious examples of philosophers
whose way of writing philosophy provokes not only differing interpretations (as any philosophi-
cal writing will) but profoundly differing approaches to the writing itself, frequently breaking the
canonical disciplinary boundaries. If you do not include occasional borrowings of theory or con-
ceptual distinctions from philosophy (such as conclusions from twentieth-century philosophy of
Introduction 19
Arguments about the proper approach to Plato often center around the
question, “Why did Plato write dialogues?” I am not sure, however, that this
is a particularly useful question to focus on, even if it is an obvious one to
ask,27 in part because I am skeptical that there is any interesting substantive
answer to it. I agree that most of the explanations typically offered are at
least part of the story about why Plato chose to write philosophy in this
way: to some largely unknowable extent, it imitates the way the historical
Socrates actually conducted philosophical discussion, and the historical
Socrates had an enormous in¬‚uence on Plato; it enables Plato to distance
himself from any conclusions reached by the interlocutors and he does
this because of a hesitation to present some view as the ¬nal truth of the
matter; Plato believed that in some way philosophical truth could not be
communicated fully or directly (whatever that means exactly) in writing;
“the Socratic dialogue” was a genre that was popular at the time Plato
wrote, engaged in by many who had close contact with Socrates. Most
probably all of these explanations of why Plato wrote dialogues, as well
as others, are at least partly accurate. But I remain unpersuaded by more
detailed accounts or elaborations of these ideas insofar as they seek to ¬nd
some key that would make the “problem” of the dialogue form dissolve,
typically either by implausibly diminishing or implausibly augmenting its
signi¬cance. No one can deny that the dialogue form creates a certain
distance between author and reader that is not present in a treatise. In that
distance questions arise about irony, the author™s real intent, whether he
identi¬es with any character, and if so, which, and so on. But is there any
way, at more than a speculative level, to say that Plato de¬nitely did not
hold the views expressed by an interlocutor at some point, or any way to
say that he de¬nitely did? To a signi¬cant degree these are questions about
Plato™s intentions, and, even if we are interested in answering them, the
dialogue form simply precludes our knowing them.
I think we may make more headway if we begin from the manifest fact
that Plato did write dialogues, whatever his reason(s) for doing so, and
consider the consequences of that. The effect of writing a dialogue is to
create a distinction between what I call the “inner” and the “outer” frames
of the dialogue.28 The outer frame refers to the relationship between the
text and the reader. To be conscious of the outer frame is to remember and
re¬‚ect on the fact that every dialogue, and every speech of every character

language borrowed by some types of literary theory), most philosophers are written on only by other
philosophers; this is not so for Plato, Nietzsche, or Wittgenstein.
27 And of course needs to be addressed in any introductory course or work on Plato.
28 I ¬rst discussed this distinction in Vasiliou (1999a).
20 Aiming at Virtue in Plato
within a dialogue, is written by Plato. We should remember that there is no
evidence that Plato ever circulated his dialogues anonymously. Everyone
who read (or heard) the Gorgias was well aware that it was written by
Plato, the man who ran a well-known school in Athens for some decades.
The inner frame refers to the relationship between interlocutors within the
dialogue.29 Rather than attempting to say in general why Plato wrote dia-
logues, I shall consider the text from the perspectives of both the inner and
the outer frames. It is not the case, however, that each of these perspectives
is always equally important. At times in Plato the signi¬cance of the outer
frame may be limited to the weak claims I listed above about why Plato
wrote dialogues at all “ perhaps, for example, in the Laws. Other times, in
the Symposium, Laches, Gorgias or Protagoras, re¬‚ection on the outer frame
may be extremely important exegetically. In yet other works, the fruitful-
ness of re¬‚ecting on the outer frame may be moderate. I have very little
to say a priori about the signi¬cance of the distinction without showing
its value in particular cases; and I believe it does not always have the same
value in all dialogues.
Thus I do not always raise the distinction between the inner and outer
frames. I sometimes go into considerable detail about the dramatic context
and focus on the way interlocutors argue, spending less time on the analysis
of arguments themselves. In other places I focus in great detail on an argu-
ment, while ignoring elements of the drama and style. What dictates these
choices is my goal of explicating and elucidating what I shall argue is the
central role of the distinction between aiming and determining questions.
This is what drives the study, not some a priori belief about the relative
importance of the “philosophical” as opposed to the “dramatic” elements
of the dialogues.

7 a note on rea ding pl ato ( ii) : doct ri ne s a nd
developm en ta li sm
I shall also say something brief about Plato, doctrines, and philosophical
development. Like the general question about why Plato wrote dialogues,
I also do not ¬nd the question of whether Plato held doctrines especially
useful. It is clear that views are put forward and defended in the dialogues.
It is clear too that many of these views are complex, defended by long
and intricate arguments, and approached from different angles in more

29 Of course, sometimes there are nested inner frames; the Symposium is a quite complex example of
nested inner frames (four).
Introduction 21
than one dialogue. Regardless even of what Plato himself might have said
about his own epistemic relationship to these views “ whether he would
regard himself as knowing them, believing them, entertaining them, or
whatever “ it is clear at least that the dialogues take certain views very
seriously. We ¬nd elaborate views defended in detail, typically by Socrates,
but sometimes by other interlocutors, in the dialogues. As with the question
about the dialogue form, there are various generalizations that I would ¬nd
it dif¬cult to dispute, at a suf¬cient level of generality: Plato held some
views that bore a close relationship to those held by the main interlocutor
in many dialogues; his views underwent some development or change in
the course of more than forty years of philosophical work; there is some
moving away from the ¬gure (and in¬‚uence) of Socrates in his latest works,
exempli¬ed by the unpublished Laws, which we know Plato was working
on at the end of his life and which alone does not contain a character
While I do believe that all of these claims contain elements of truth,
I do not think that one can put much ¬‚esh on these bones with any
degree of certainty. All we can do is interpret the dialogues individually
and in comparison with one another, asking questions about both the
inner and outer frames where appropriate, and applying the principle of
charity with as much self-consciousness about our own philosophical views
as we can. I maintain that particular conclusions about development or
changes in views over the course of the dialogues can only be the product
of the detailed study of them and that we must be careful not to employ
an argument that begins from the assumption that a dialogue belongs to
a certain period and so therefore must be presenting certain views. This
book is unitarian in spirit insofar as it argues for an elaboration of the
same concepts and distinctions throughout the dialogues discussed.30 I
do not deny that on some other topics there may be clearly distinct and
con¬‚icting views in different dialogues, which may plausibly suggest some
sort of development. I shall argue that the defense of SV, and the distinction
between it and determining what virtue is, is developed most intricately
and thoroughly in the Republic. As a device of convenience I shall refer
to the other dialogues I discuss as “earlier.” I do not, however, intend this
to imply anything about the chronological date in which a dialogue was
written nor do I claim that the order of my discussion matches an order in
Plato™s own philosophical development.

30 Annas (1999), ch. 1, emphasizes the relative recentness of developmentalist readings of Plato.
ch a p t er 1

Socrates and the supremacy of virtue

1 .1 in trod uct i on
Even a casual reader of the Apology understands that Socrates believes that
virtue is more important than anything else, even his own life. What has
not been recognized, or at least not accorded any signi¬cance, is that for
Socrates virtue is an aim. He believes that you should never aim sim-
ply at saving your life at the expense of aiming at what is virtuous; in
other words, your life counts for nothing as an aim when compared with
virtue as an aim. But “the supremacy of virtue” does not imply the quite
implausible view that many readers apparently attribute to Socrates, that
one™s loss of life is not relevant to the deliberation about what is in fact
the virtuous action in some circumstances; that is, the view that one™s
life counts for nothing in a different sense. It is not as though one could
determine what the virtuous action is independently of considerations of
life, death, pleasure, pain, or material loss or gain. Socrates is not saying
that we should ignore these things absolutely. Any fact described in non-
evaluative terms may in principle be relevant in a deliberation that seeks
to determine what the virtuous action is here and now.1 Certainly factors
like pleasure and pain, life and death, wealth, and the welfare of friends
and family will be most relevant to such deliberations. We must ignore
such things, however, as aims of action when they con¬‚ict with what virtue
I shall argue that while Socrates believes that one ought always to adhere
to SV, it may well be that in some cases material bene¬ts gained or lost is a
relevant factor in the determination of what the excellent action is here and
now. To deny this would be to deny the relevance of, for example, a person™s

1 Although this sounds like a particularist position, it is not necessarily so. It could be the case that
there are universal principles that could be made concrete enough to cover any possible case; or,
perhaps more plausibly, there may be “prima facie” Rossian-type generalizations.

Socrates and the supremacy of virtue 23
losing all of her property in an assessment of what is right. It is true that in
Socratic deliberation the simple fact that an action involves killing someone
or depriving them of property does not, by virtue of that fact alone, render
it wrong or vicious.2 Moreover, one™s aim cannot be simply to save lives;
one™s actions must always be regulated by SV. But whether someone will be
killed or not is certainly relevant in assessing whether an action is virtuous
or not. Socrates is not depicted as intuiting through some magical insight
what the right thing to do is, without considering the mundane features
of life that we all ordinarily consider.3 What is distinctive about Socrates™
deliberation is not that he ignores such factors, but that he always takes
them into consideration regulated by SV either as his explicit aim or as a
limiting condition.4 This argument will not be complete until the end of
chapter two.
Here in chapter one I shall begin with the Apology, and show that both the
way Socrates states SV and the argument within the speech itself indicates
that SV ought to be understood as an aiming principle, which operates
in action both as an explicit aim and limiting condition. Appreciating
the signi¬cance of the concept of an aim will lead to new readings of some
familiar texts, and a new understanding of Socrates™ avowals and disavowals
of knowledge. Scholars have not noticed that, except for the ¬nal argument
in the Crito (50a ff.) and the way he conducts himself at trial, Socrates never
argues that anything he has done, or is doing, is in fact virtuous.5 Rather he
argues simply that, whenever virtue was at issue, he has always unfailingly
aimed at doing the virtuous thing or avoiding the vicious action. Whether
the actions he has engaged in “ for example, ¬ghting in various battles or
refusing to help to bring in Leon for prosecution “ were in fact virtuous he
does not argue one way or the other. Of course he believes they were, but
he provides no argument for this. Without exception his point is that on
all such occasions he has acted in accord with SV.

2 “Vicious” is used as the antonym to “virtuous,” not in the colloquial sense.
3 With the exception of his divine sign, which I discuss further below and in chapter two.
4 See Introduction, section 1, for discussion of the terms “explicit aim” and “limiting condition.” As
we shall see in chapter two, in the argument of “the Laws” in the Crito (50a“54d), Socrates considers
the effects of possible exile, his age, the prospects of future care of his children, and the welfare of his
friends in his deliberation about whether to escape from prison.
5 Lane (1998), 313 ff., recognizes a related point that Socrates in the Crito engages in what she calls
“deliberation”: that is, attempts to determine what to do in a particular practical situation. Contrary
to Lane, however, I do not think that the Crito is the only place Socrates faces such a situation,
although it is the most prominent. Both his decision to engage in certain conversations (although
this is arguably a different sort of situation) and also his decision to conduct his trial speech in the
manner he does are additional examples. See chapter two.
24 Aiming at Virtue in Plato

1 .2 th e su premacy of vi rtu e i n th e a p o l o g y
In the Apology Socrates provides no explicit argument for the claim that
people should follow SV, but a substantial portion of the speech does consist
of an argument to show that he himself has unwaveringly adhered to it. If
Socrates is a persuasive advocate of SV in this setting, it is not because of
any argument he puts forward, but because he stands as an inspirational
example of commitment to it for a listener who already believes in it.
Let™s consider Socrates™ ¬rst statement of SV:
Perhaps someone might say: “Aren™t you ashamed that you have pursued the sort
of pursuit on account of which you are now likely to be put to death?” But I
would reply to this with a just statement [d©kaion l»gon], “You are not right, sir,
if you think that a man who is worth even some little bit ought to take under
consideration the risk of living or dying and not instead look to this alone when
he acts: whether he is doing just or unjust things, the deeds of a good or a bad
man.” (28b5“9)
This is an aiming principle in the clearest sense: Socrates says that a man
should “look alone” (m»non skope±n) at whether he is acting virtuously or
not. A person™s goal ought to be to realize just actions, the deeds of a good
man.6 Here Socrates refers to SV as an explicit aim. He tells us that, in
deciding what to do, there is one goal in action that necessarily trumps all
others: virtue. It ought to override any other value one might be inclined
to offer as the aim of one™s action, such as survival, wealth, pleasure, good
reputation, and so on. SV does not help an agent, however, to determine
what to do or not do in any particular situation. In the passage the imaginary
interlocutor supposes that a life that results in a premature death at the hands
of others must be a life led in the wrong way.7 SV denies this by claiming that
acting virtuously trumps any other aims one might have in action, including
staying alive. We should be clear that it is not offering an all-inclusive
account that says that we should always act aiming at virtue. Rather, the
context of the conversation with the imaginary interlocutor makes clear
that the principle SV applies only in certain situations: whenever there is
a virtuous action to be done (or a vicious action to be avoided), then one
ought not to consider anything else as providing a competing aim for one™s
action. It reasonably leaves open that in some, perhaps many, situations
where an action™s being virtuous or not is not at issue, one may pursue

6 The virtuous action, the ¬ne or noble action, the just action, and the good action are all synonymous.
7 We shall see in 3.6 that Callicles in the Gorgias is an interlocutor who raises just such an objection to
Socrates, in strikingly similar language.
Socrates and the supremacy of virtue 25
some other aim. This is prohibited only in situations in which another aim
runs contrary to virtue.
Socrates follows this passage with an example from the Iliad. Achilles
made his decision to remain (and so to die) at Troy rather than to return
home solely on the basis (according to Socrates anyway) of what Achilles
thought was the excellent thing to do. It is critical to notice that the point
of his example is not that what Achilles actually did was right “ it may or
may not have been, and he does not argue that substantive question one
way or the other. The important issue for Socrates is that Achilles had only
one aim in his deliberation: to do the virtuous action. When doing the
virtuous thing was at issue, Achilles gave no thought to “death and danger”
(qan†tou kaª kind…nou) (28c9, 28d4“5); he feared only “living as a bad
man” (de©sav t¼ z¦n kak¼v ßn, 28d1). Socrates claims, then, that Achilles
shared his commitment to the aiming principle SV, but he does not af¬rm
that what Achilles actually did was correct.8 One might argue that what
Achilles took to constitute excellence in his situation was incorrect; perhaps
he should have given up the false glories of the Homeric hero, and instead
returned home and worked for the less fortunate. Socrates™ praise of Achilles
leaves this question entirely open, and the claim that Achilles adhered to
SV does not affect it.
Consider another passage:
For in truth, men of Athens, things are this way: where someone positions oneself,
believing that it is best [b”ltiston], or having been placed there by his commander
[Ëp¬ Šrcontov], he must [de±] remain there, as it seems to me, and face danger,
not considering either death or anything else in place of the disgraceful [pr¼ to“
a«scro“]. (28d6“10)
“The disgraceful” here is clearly meant to be contrasted with “the noble”
(t¼ kal»n) (cf. Cr. 47c9“10). Again Socrates states that avoiding death
is nothing as an aim for action when contrasted with excellence and its
opposite. He goes on to illustrate his adherence to SV by citing his actions
at the battles of Potidaea, Amphipolis, and Delium. As above, the point of
the examples is not that these speci¬c actions actually were the right ones,
but that he acted in the way he thought required by virtue even in the face
of death; he made his decisions solely on the basis of whether the action
was right or wrong. A paci¬st might take issue with Socrates™ participation
in these battles, and try to argue that he was in fact acting wrongly. As I
understand him, Socrates would have welcomed such an argument and, as
we shall see below, would have acknowledged that he might have actually
8 Weiss (1998), 8“9, has a different reading.
26 Aiming at Virtue in Plato
done wrong. In order to settle such an issue, however, he and his interlocutor
would ¬rst have to know what virtue is.9 So Socrates may have in fact done
wrong, while never wavering from his commitment to SV.10
One more passage:
To do wrong [ˆdike±n] and to disobey one™s better [belt©oni], whether man or
god, this I know [o²da] is bad and disgraceful [kak¼n kaª a«scr»n]. (29b6“7)
This statement emphasizes SV™s role as a limiting condition. For Socrates, it
is clear that one™s better must be one™s moral superior “ not someone who is
wealthier, or holds higher political of¬ce, or is better at shoemaking.11 This
formulation of SV, which is so central to the Crito (cf. 49a“e), expresses
its ubiquitous role as a limiting condition. Although one may have many
aims in action other than virtue, those aims must always be limited by
the condition that one™s aim is not contrary to virtue, does not constitute
“doing wrong.”
Socrates proceeds to argue, as before, that he himself has always adhered
to SV by citing his experience on the Council (32a ff.) and his refusal during
the rule of the Thirty to bring in Leon of Salamis (32c). It is again no part of
Socrates™ argument here that what he did was right; a particular Athenian
might take issue with that, and it would need to be argued about separately.
It is not, surely, that Socrates thinks that what he did was wrong; he thinks
it was right, and expects his audience to agree. But his point in discussing
these events is not about this question one way or the other. Rather, Socrates
emphasizes that the sole aim he allowed to guide his deliberations was
whether he was avoiding wrongdoing:
[I was ordered to bring in Leon], however I showed again at that time, not in word
but in deed [oÉ l»gwƒ ˆll ¬ ›rgwƒ], that I consider death (if it were not too wild
to say) to be nothing whatsoever, but not doing anything unjust or impious, this
I consider to be everything [t¼ pŽn]. (32c8“d3)
9 See chapter four.
10 We shall see below how this affects our understanding of Socrates and incontinence.
11 See Kraut (1984), 23, n. 38. Reeve (1989), 110“12, disagrees. He claims that Socrates includes his actual
commanders in the battles of Potidaea, Amphipolis, and Delium as “his superiors in the requisite
sense,” despite the fact that Socrates would hardly consider them his moral superiors. Reeve argues
that for a person assigned to a post it is reasonable to believe that “[. . .] his commander is in
the relevant sense his better. For unless there is strong evidence to the contrary, X™s [the person™s]
commander must be presumed to be in a better position to know what it is best for him to do on the
battle¬eld or its analogue, the Athenian polis, than X is himself. But . . . the only contrary evidence
X has is that staying at his post puts him at risk of dying. And that . . . has insuf¬cient weight to
justify disobedience. Thus (2) [the above passage, 29b6“7] explains why disobeying a commander
is bad, wrong, and shameful.” (112). But when Reeve says that “unless there is strong evidence to the
contrary, X™s commander must be presumed to be in a better position” he almost begs the question:
if the commander is in a better position to see what virtue truly requires, then he is morally superior
at least in that context. The qualifying clause about strong evidence leaves open the possibility of
justi¬ed disobedience. Thus Socrates does not obey his of¬cial commander simply qua commander.
Socrates and the supremacy of virtue 27
Socrates focuses again on how SV governs all of his deliberations. The delib-
erations themselves we are not privy to, but he boldly states his adherence
to SV. His argument implies that if he has never let staying alive count
as a competing aim against doing what is virtuous in the past, a fortiori
he would never allow aims such as health, wealth, or reputation (all the
things he has complained the Athenians falsely take to be the most impor-
tant things [29d ff.]) to override the aim of acting virtuously and being
sure never to act viciously. As we shall see, I am not attributing to Socrates
the view that factors like wealth, life, health, reputation, and so on might
not fairly count as one™s aim under certain circumstances, nor that such
factors certainly might be relevant in the substantive deliberation involved
in determining what is here and now the right thing to do.12 SV is simply
the claim that when a question of right or wrong is at issue (something not
the case in every action in one™s life) a virtuous action is the supreme and
explicit aim an agent should have, and that in all other actions one must
be sure they are never contrary to virtue.

1 .3 s oc rates a nd m ora l know l edg e
One striking feature of Socrates™ statement of SV in the Apology is that it is
something he explicitly claims to know. Gregory Vlastos has made the pas-
sage from 29b (quoted above) famous as an example of an explicit avowal of
ethical knowledge.13 Crucially, however, Socrates avows not just any ethical
knowledge, but speci¬cally knowledge of an aiming principle, knowledge
of SV. Appreciating the distinction between SV and the ethical knowledge
that would be constituted by correct answers to Socrates™ “What is F?”
questions (that is, knowledge that would actually determine what virtue
is) suggests a new way of understanding Socrates™ disavowal of knowledge.
Commentators have offered a variety of explanations of how the disavowal
can be seriously meant.14 An adequate account must explain the apparent
contradiction between a Socrates who most often claims not to know, and
one who also sometimes claims to know.15
12 13 See his (1985/1994), 43 ff.
See chapter two.
14 The literature on Socrates™ disavowal is vast, and I cannot comprehensively survey it here. But, in
addition to discussion below, see Irwin (1977); Lesher (1987); Vlastos (1985/1994); Brickhouse and
Smith (1994), ch. 2; Irwin (1995), §§16“18; Nozick (1995), and the reply by Fine (1996); Stokes
(1997), 17“21; Benson (2000); Wolfsdorf (2004).
15 Vlastos (1985/1994) proposes that Socrates must be working with two types of knowledge: elenctic
knowledge and certain knowledge. For criticisms of Vlastos™ view see Lesher (1987); Irwin (1992);
Brickhouse and Smith (1993); Nehamas (1998), ch. 3. Both Irwin (1995), §17, and Benson (2000),
ch. 10, esp. 236“8, attempt to downplay the signi¬cance of the avowal passages. Irwin believes that
Socrates maintains only that he has true belief, not knowledge. Wolfsdorf (2004), esp. 124, 132,
rejects the idea that a consistent reading ought to be found.
28 Aiming at Virtue in Plato
Rather than focusing, as most accounts do, on differences in the cognitive
state of Socrates as a way of understanding the disavowal, I shall look at the
objects of Socrates™ knowledge and lack of knowledge.16 There is no puzzle
if Socrates disavows knowledge of x, but avows knowledge of y; attempting
to distinguish the cognitive states of Socrates is only necessary if we think
that Socrates is avowing and disavowing knowledge of the same thing (at
the same time). But if we look at the passages where Socrates avows and
disavows knowledge, and the contexts of those passages, we ¬nd that he
consistently avows knowledge of SV, but consistently disavows knowledge
of the answers to the “What is F?” question about virtue. An adequate
answer to the “What is F?” question is the criterion that Socrates demands
for claiming to know what F is, and it is this he does not have.17 But that
does not stop him from knowing SV, and therefore being unwaveringly
con¬dent in living according to it.18

1.3.1 Socrates™ disavowals of knowledge
I shall take the ¬rst appearance of the disavowal in the Apology as a paradigm
for how to understand it elsewhere. At 19d8 Socrates defends himself against
a charge of “making the worse argument appear to be the better” levied
by the “earlier accusers” which, in effect, accuses him of being a sophist.
Socrates denies that he undertakes to teach anyone or that he charges any
money for it. He ironically praises the sophists “since it seems to me to be
noble (kal»n), if someone is able to educate (paide…ein) people, just like
Gorgias of Leontini, and Prodicus of Ceos, and Hippias of Elis.”19 Socrates

16 Lesher (1987) is an exception. He believes that Socrates disavows knowledge of “essential natures”
(286) of the virtues, but that he avows knowledge of “the moral qualities of speci¬c actions” (285) and
that he “con¬dently identif[ies] the goods and evils of daily life” (287). Lesher further believes that
“identifying the goods and evils of ordinary life was in short as philosophically uncontroversial as it
was uninteresting” (287). Alcibiades expresses such a view on behalf of the Athenians in Alcibiades
1 (113d), but I see no evidence that it is Socrates™ or Plato™s. I think that Socrates does certainly
disavow knowledge of the “essential natures” of virtue, but that he therefore also disavows knowing
which token actions are virtuous or not (barring aid from his divine sign). Indeed I am arguing
throughout that the puzzle of how to resolve “determining questions” knowledgeably is one of the
primary concerns of the dialogues.
17 We shall consider Socrates™ “What is F?” question further in chapter four.
18 Another obstacle is the con¬‚ict between Socrates™ claiming knowledge of any sort and his alleged
belief in the “priority of de¬nition,” which maintains that if one does not know what F is, then
neither can one know anything about F. After I discuss the textual evidence that Socrates does indeed
avow knowledge of SV, while disavowing knowledge of what virtue is, I shall address how this can
be reconciled. See 1.4.
19 This is a case of what I call “conditional irony”; see Vasiliou (1999a). In conditional irony Socrates
literally means the conditional as a whole. The reader, however, has good reason to think that
Socrates does not believe that the condition stated in the antecedent actually obtains, and so good
Socrates and the supremacy of virtue 29
then describes how he asked Callias whether there is someone who is a
“knower” (–pistžmwn) of human excellence to whom he can send his sons
to become “¬ne and good,” on analogy with an expert horse-breeder who
makes horses excellent. Callias tells him that Evenus is such a person and
Socrates responds with the ¬rst disavowal:
And I considered Evenus blessed, if he has this skill [t”cnh] and teaches it for such
a reasonable amount.20 I at any rate would also preen and pride myself if I knew
these things; but I do not know [them], men of Athens. (20c1“3)
Here the knowledge that is attributed to Evenus is explicitly techne-
knowledge21 of virtue, that is, expert knowledge of what virtue is. This
is the knowledge that Socrates so frequently tests for in other dialogues “
for example, in the Charmides, Euthyphro, Laches, Meno, and Protagoras “
and this is the knowledge he disavows. As Socrates says at Meno 71c and as
all of these dialogues show, the inability of anyone to answer the “What is
F?” question indicates that no one actually has expert knowledge of virtue.
As I shall discuss further in chapter four, if someone had this knowledge,
then he would of course have a foolproof way of determining which actions
are virtuous and which are not.
After hearing that the oracle claimed that no one was wiser than he,
Socrates explains the source of his puzzlement:
For I am conscious of not being wise in anything great or small (oÎte m”ga oÎte
smikr¼n s…noida) (21b4“5).
In the context of Socrates™ account of his examination of the politicians,
poets, and craftsmen that follows, it becomes clear that “great or small”
refers to the upcoming contrast between the craftsmen™s knowledge of
their crafts and their subsequent misplaced conceit that they also know
“other most important things [˜greatest,™ m”gista]” (22d6“e1).22 The theme

reason to think that Socrates does not really believe the consequent either. Of course, this does not
follow logically; one cannot validly conclude “not-q” from “if p, then q,” and “not-p.” Nevertheless
in ordinary conversation when a person says, “If you f, I™ll y,” there is the implication that if you
don™t f, then I won™t y. The frequency of this sort of example in conversation is surely part of the
reason why people are so liable to commit the fallacy.
20 An example of conditional irony once again.
I will frequently leave t”cnh simply transliterated as “techne” (without long marks). It is typically
translated “art,” “skill,” or “craft” and can refer to the skill or craft itself (e.g. shoemaking) and also
to the expert knowledge that the craftsman possesses.
22 Vlastos (1985/1994), 43, n. 12 (see also his [1991], 238) claims that the “clear import” of 21b4“5 is
that Socrates knows “absolutely nothing.” To the extent that this suggests that Socrates claims not
to know that his name is “Socrates,” how to get to Piraeus, or that two plus two equals four and so
on, it is nowhere warranted in the dialogues. See Brickhouse and Smith (1994), 34“5, and Wolfsdorf
(2004), 129, for more tempered assessments.
30 Aiming at Virtue in Plato
of the “most important things” is repeated at 29e“30b, and refers to the
best possible state of one™s soul, that is, virtue. By contrast, knowledge
of something “small” or “unimportant” would be the knowledge of an
ordinary techne, which Socrates says that the craftsmen, as opposed to
the poets and politicians, actually do have (22d1“4). In this passage, then,
Socrates disavows knowledge of any techne, and knowledge of virtue. He
has met people, the craftsmen, who possess knowledge of their crafts, but
has never met anyone with techne-knowledge of virtue.23
In every other passage in which Socrates expresses his disavowal of
knowledge, it is always clear that what he is disavowing knowledge of
is knowledge of what virtue is. Let us consider some examples, beginning
with the Laches:
I say ¬rst about myself, Lysimachos and Melisias, that I have not had a teacher
in this [in making a person™s soul best], although I desired such a thing from the
beginning of my youth. But I did not have the money for sophists, who alone
claimed to be able to make me ¬ne and noble; but I myself am unable to discover
[this] craft [t”cnh] even now . . . [I urge you, Lysimachos, not to let Laches or
Nicias go but to question them] saying that Socrates denies that he knows [–pa¹ein]
about this matter nor is he capable of determining which of you speaks the truth “
for he has been neither a discoverer nor a student of anything about such things . . .
(186b8“c5 . . . 186d8“e2)
Laches and Nicias are willing to speak about what would be best for Lysi-
machos™ and Melesias™ sons, thereby implying that they have knowledge
about what is best for them.24 By contrast, Socrates denies having any
knowledge of a techne of virtue just as he did in the Apology. Again at the
end of the Laches (200e2“5) Socrates refuses the role of teacher, since to
teach one must ¬rst have knowledge, and insists instead that they con-
tinue the search together and not remain as they are. Likewise in the
Charmides, Socrates disavows having knowledge of what temperance is
(165b4“c2) and suggests, when Critias becomes annoyed, that Socrates is
refuting Critias more for Socrates™ own sake than for Critias™, in order to be
sure that he does not end up thinking he knows something he does not know
In the Protagoras Socrates sums up his long and complex discussion with
Protagoras as follows:
23 Although this is the knowledge that Socrates claims that the sophists believe they have: see, e.g., Ap.
20d9“e2 and La. 186c2“4.
24 See Vasiliou (1999a), §6, for the conceit of knowledge by Laches and Nicias, and Socrates™ use of
conditional irony to speak to them.
Socrates and the supremacy of virtue 31
I ask about all these things for no other reason than my wishing to inquire into
[sk”yasqai] how in the world matters concerning virtue stand [p¤v pot ¬ ›cei
t‡ perª t¦v ˆret¦v] and what in the world virtue itself is [t© pot¬ –stªn aÉt»].
For I know [o²da] that if this became apparent that [issue] about which each of
us has said so much would become perfectly clear, I maintaining that virtue is not
teachable and you that it is. (360e6“361a3)

Socrates here insists that his confusion is again about what virtue is. We
should note that he explicitly avows knowledge about the connection, itself
quite uncontroversial, between knowing what F is and knowing what F is
like.25 This stands as an example of an “ordinary” avowal; it is not about
ethics as such, nor is it the avowal of any techne-knowledge. It is simply
the claim that, if we were to know what virtue is, we would then know
whether it was teachable. This shows further that Socrates™ disavowal of
things “great and small” in the Apology (21b4“5) should not be understood
as a disavowal of all knowledge.
The pattern of disavowing the answer to the “What is F?” question when
posed about virtue or a virtue continues in every other passage containing
a disavowal: Socrates disavows knowing what piety is at Euthyphro 5a7“c5,
15c12, and 15e“16a; he denies knowing what the ¬ne is at Hippias Major
286d“e2 and 304d“e; what virtue is at Meno 71a“c and 80d; what justice is at
Republic 1, 337e.26 Every time Socrates disavows knowledge he is disavowing
knowledge of what virtue or a virtue is, that is, he disavows being able to
answer the “What is F?” question for virtue or a virtue. There remains
one important disavowal passage “ Gorgias 509a4“6 “ that I have not yet
considered. Before we can understand it properly, however, we need to
look at Socrates™ avowals of ethical knowledge, where he appears to avow
knowledge about the “most important” things inconsistently with his many
disavowals. I shall argue that this con¬‚ict is merely apparent, for we shall
see that what Socrates avows knowledge of “ the supremacy of virtue “ is
quite different from what he has disavowed knowledge of “ namely the

25 Note that what Socrates says in this passage is that, if they know what virtue is, then they will know
whether it is teachable. This is a signi¬cantly weaker claim than the priority of de¬nition, which
maintains that if one doesn™t know what F is, then one can™t know what F is like. This passage thus
doesn™t preclude the possibility of knowing something about what F is like without knowing what
F is.
26 This pattern holds also for the midwifery passage in the Theaetetus (148e“151d), where Socrates
disavows knowledge of what knowledge itself is, and claims to be able only to test the ideas of
Theaetetus. As we shall see in 7.4 and 8.4, it extends as well into the Republic when Socrates

. 1
( 9)