. 6
( 9)


examples: “(B) that some people are unjust in a sense of ˜injustice™ that has
nothing to do with the legalistic de¬nition of justice” and “(C) As things
are, just actions are typically in accordance with (perhaps required by) the
laws of the state, and unjust actions are typically violations of those laws”
(his emphases).26
Chappell is certainly right that Thrasymachus never explains what he
believes constitutes an adequate de¬nition. It is true as well that Socrates™
demand for de¬nitions is not something an ordinary Athenian would be
familiar with, as the dialogues depict many times. But Thrasymachus is
not as ordinary and na¨ve about de¬nition as Chappell suggests. First, his
entrance into the conversation makes clear that he knows, or at least knows
about, Socrates.27 Thrasymachus has already told the others present that
Socrates would engage in his “usual irony” and that he would refuse to
answer questions (337a). Thrasymachus™ complaint about Socrates is in fact
that he has not said anything about justice “clearly and precisely” (336d3).
So, although Thrasymachus™ own responses will turn out not to be as “clear
and precise” as he may hope, these facts show that he is concerned with
providing a more determinate answer to what justice is than Socrates has by
simply calling it bene¬cial and so on.28 More importantly, however, Thrasy-
machus has obviously just witnessed Socrates™ conversation with Cephalus
and Polemarchus. And although Socrates never provides a description of
what he seeks in a de¬nition in Republic 1 as he does elsewhere (e.g., Eu.
6d“e, Meno 72b“73c), it is nevertheless clear that he seeks a statement of
what all just actions have in common; for this was why “returning what is

26 Chappell (2000), 105. Chappell™s point against Everson is that there is nothing contradictory or
incoherent about these two claims. Such an interpretation of Thrasymachus™ de¬nitions, as observa-
tions and descriptions of what justice is like (“dazzling one-liners”) and not as de¬nitions of justice
in a stricter sense, is in keeping with the views about Thrasymachus that Chappell defends in an
earlier article, Chappell (1993).
27 Has Thrasymachus met or talked to Socrates before? I don™t think the evidence is clear.
Socrates talks about Thrasymachus™ speaking ability in the Phaedrus (267c“d), and when Socrates
arrives at Polemarchus™ house he acknowledges Thrasymachus™ presence as someone familiar to
28 The list at 336d1“2 of unclear and imprecise terms that Thrasymachus wishes to disallow includes
more terms than Socrates has used up until that point, which might indicate Thrasymachus™ acquain-
tance with other Socratic accounts or conversations.
Socrates and Thrasymachus: Republic 1 177
owed and speaking the truth” failed as a de¬nition of justice (331c“d, 331e“
332a). This makes it overwhelmingly unlikely that Chappell™s (B) or (C),
which offer “for the most part” descriptions of justice, could be the cor-
rect interpretation of Thrasymachus™ claims. It is true that Thrasymachus
need not have in mind supplying necessary and suf¬cient conditions or
property-identities, but there is textual evidence that he understands (and
never disagrees with) Socrates™ demand that he state something common
to all instances of justice. The fact that Thrasymachus believes that he
has a much better answer to what justice is than “speaking the truth and
returning what is owed” only makes sense if we assume that he believes
that his answers will apply to all cases of just action and so will not be
simply “typically” true and therefore refutable by Socratic counterexam-
ples. This explains why he says things like “justice is nothing other than
the advantage of the superior” (338c2“3) and that it is “the same every-
where” (339a2“4). Thrasymachus is, at least at the start, willing and eager
to play the Socratic game as he understands it “ to state what it is that all
instances of justice have in common “ because he thinks he can beat Socrates
at it.
So, I think that we ought to expect from Thrasymachus™ answers prop-
erties that he takes to be common to all instances of justice. We should not
expect, however, a clear distinction between an essential property of justice
and mere “necessary accidents,” to borrow a phrase from Aristotle. In fact,
I think that it is most plausible to see Thrasymachus as supplying both
in his remarks so far. Following the rule of the established government,
which he takes to be the stronger, is what justice is. The sophisticated ide-
ological analysis offers us insight into a feature of all actions proclaimed by
the rulers to be just in a society: they are to the rulers™ advantage. When
Thrasymachus says that justice is the “advantage of the stronger” he is not,
at this point, offering this as a de¬nition of what makes an action just, but
supplying an allegedly sophisticated description of it. He does the same later
when he avoids calling justice a “vice” but only “very noble na¨vety” (p†nu
genna©an eɞqeian, 348c12). “Noble” is certainly sarcastic; Thrasymachus
is thinking of a person who falsely believes that he is acting nobly by acting
justly, when in reality he is simply the dupe of a ruler. It is not, however,
an action™s being to the advantage of the ruler/stronger that makes it just;
what makes it just is that it is lawful. Thrasymachus is simply pointing
out that laws are made for the advantage of the stronger, so that acting
justly (which is really acting according to laws) is also “noble na¨vety.”±
Thus, although according to Thrasymachus all just actions are both sim-
plistic foolishness and to the advantage of the ruler, these are, in the terms
178 Aiming at Virtue in Plato
of the Euthyphro (see 11a), simply “affections” (p†qh) of justice, not its
I have argued, then, that until 339d Thrasymachus™ remarks about justice,
that it is the advantage of the stronger, the advantage of the rulers, and that
it is the following of laws by the ruled, are all part of the same package. The
last, brought up by Socrates to complete the argument, is by Socratic lights
the account of justice; the other two are part of Thrasymachus™ allegedly
sly ideological insight into ordinary justice and its real bene¬ciaries.

5 .4 c leitoph on™s re comm en dat i on
Although Socrates says that he is in doubt not about whether justice is
a kind of advantage but about whether it is an advantage of the superior
(339b4“7), what he relies on to generate a con¬‚ict in Thrasymachus™ views
is the latter™s admission that a ruler might err about what is truly to his
advantage. If Thrasymachus believes both that justice is obedience to the
laws instituted by the rulers and that it is possible for the rulers to err about
what is really to their advantage, then someone who followed the laws might
be acting justly according to the conventionalist criterion for just action,
but also unjustly since she is not really acting in a way advantageous to the
superior/rulers. Socrates points out that therefore the same act could be just
insofar as it follows the rules or laws, but also unjust insofar as it is not really
to the advantage of the rulers (339d1“3, 339e1“8). So, Thrasymachus cannot
consistently simultaneously hold that (a) justice is obedience to the laws
instituted by the rulers, (b) justice is the advantage of the superior/rulers,
and (c) rulers sometimes make mistakes about what is to their advantage.
This argument triggers a brief interlude between Cleitophon and Pole-
marchus, before Thrasymachus returns to the argument (340a“c). Pole-
marchus compliments Socrates™ clarity in eliciting a contradiction in
Thrasymachus™ statements. Defending Thrasymachus, Cleitophon claims
that there is no inconsistency. For he maintains that Thrasymachus said

29 As an analogy we might think of someone asking, “What are religious acts?” and receiving the
response, “Silly superstitions.” Here the respondent takes a superior attitude, which he justi¬es by
an alleged insight into what underlies such acts. He is offering something he takes all religious
acts to have in common. But he is not thereby explaining what makes an act religious; nor does
he take himself to be doing so. In that respect he is not offering a Socratic de¬nition, but rather
an ideological analysis. Just as such a response would be provocative and potentially shocking to a
person of faith, Socrates is clearly a “believer” in justice, as he has already shown in his conversation
with Polemarchus. Thrasymachus intends to be shocking in the same way to a believer in the value
of justice, namely, to Socrates. Thrasymachus™ superior tone is explainable as well: it is one thing to
argue against someone™s belief, but another to claim to know why the person really holds that belief.
Socrates and Thrasymachus: Republic 1 179
that “the advantage of the superior is that which the superior person believes
[¡go±to] is to his advantage; this is what the weaker person must do,
and this is what he [Thrasymachus] posited as the just” (340b6“8). What
Cleitophon is offering Thrasymachus is to combine his conventionalism
with a subjectivism about advantage, where that is the view that what a
person thinks is to her advantage is to her advantage.
We recognize, as Cleitophon does, that being a subjectivist about harm
and bene¬t would render Thrasymachus™ statements consistent; he would
be able to maintain both that just actions are (believed to be) to the advan-
tage of the superior/ruler and that just actions are following the orders/laws
of the rulers. If he accepted Cleitophon™s suggestion, then justice would
be completely determined by the laws of the city, which (by the way)
the rulers institute for what they take to be their own advantage. Their
being liable to error would be neither here nor there, since justice is fol-
lowing whatever laws the rulers institute, believing them to be to their
advantage. Before we consider why Thrasymachus rejects this emendation
to his position, we should ask why Cleitophon proposes it, and why he
goes so far as to suggest that it is what Thrasymachus has been saying all
We recall that Cleitophon in the eponymous dialogue praises Socrates
for his ability to convince a person that one ought to aim at virtue, while
he criticizes his failure to determine what virtue is. Let us suppose that the
Cleitophon of the Republic and that of the Cleitophon hold the same posi-
tion. If this is so, we can see why Cleitophon is attracted to thoroughgoing
conventionalism combined with subjectivism about advantage, given his
frustration in the Cleitophon at Socrates™ inability to answer determining
questions. What makes such a conventionalism so attractive, relative to the
frustration Cleitophon expresses in the other work, is that it provides one
with an answer to what virtue (in this case, justice) is. If justice is following
whatever the rulers establish as laws, then, in order to know what the just
action is in any society, one simply looks at the laws.
Thrasymachus vehemently rejects Socrates™ offer to accept Cleitophon™s
subjectivism about advantage (340c6“7); he is an “objectivist” about advan-
tage and disadvantage (harm and bene¬t) (340e4“341a4). Thrasymachus
argues that a craftsman never errs qua craftsman. If a person errs, then, at

30 Thrasymachus did not say exactly what Cleitophon says he did: that justice is what a ruler believes
is to his advantage, rather than what is. But I agree with Everson (1998a), 123 that 338c“339c initially
commits Thrasymachus to conventionalism insofar as he has agreed that following the laws of a city,
whatever they may be, is just, notwithstanding the fact that they were instituted by rulers explicitly
acknowledged to be fallible.
180 Aiming at Virtue in Plato
least while he errs, he is not exercising his craft. Thrasymachus distinguishes
between rulers and Rulers, and by implication laws and Laws.31 A Ruler,
unlike a ruler, always institutes Laws, where Laws are those laws that are
truly to the advantage of the ruler. So a ruler who errs cannot be, while he is
erring, a true Ruler, since a Ruler never makes mistakes about what is really
to his advantage (340d“341a). When Thrasymachus will not accept that
justice is simply what the rulers believe is in their interest, but what really is,
he rejects subjectivism about advantage. The nature of true advantage, and
so the nature of what is truly just, will not be exhausted by the attitudes
of the rulers manifested in the laws that they actually decree. It will rest
rather in the laws that they ought to decree. Thrasymachus has thus placed
restrictions on what constitutes a real ruler and what constitutes a real law.
Conventionalism may continue to be true in a sense (insofar as justice will
be following the Laws), but what are going to count as rules or laws and
who are going to count as rulers are now subject to further restrictions: only
those who are correct about which laws are to their advantage are actually
Rulers. And then we are back to the problem of determining content, which
so exasperates Cleitophon in the short dialogue. We recall that at the end of
the Cleitophon he threatens to abandon Socrates and go to Thrasymachus
instead, who offers an account of what virtue is (although Cleitophon does
not say what Thrasymachus™ account is). Here in the Republic perhaps we
can see what Cleitophon thought it was, and why it appeared to him to
be satisfying. We can imagine that Cleitophon, after Thrasymachus rejects
his suggestion, ends up frustrated by him as well. Indeed, we hear no more
from Cleitophon after this.32
Why does Thrasymachus reject Cleitophon™s suggestion so completely
and quickly? I think that what bothers Thrasymachus is similar to what
bothers Callicles, which underlies a deeper similarity between them and
Socrates. A thoroughgoing conventionalism, like an unrestrained hedo-
nism, is entirely subjective. The advantage of these positions is that they
make the questions “What is virtue?” and “What is justice?” relatively easily
answerable. But Callicles™ and Thrasymachus™ elitism is at odds with such a
thoroughly subjectivist position. They think that some sophisticated peo-
ple (and they clearly include themselves in this group) have, in some way,
“seen through” ordinary ethical judgements, seen them for what they really
are, and are truly better off than others for it. The problem they then have
is how to understand the nature of their own allegedly objective superiority.
31 To borrow a device from Reeve (1988), 11ff.
32 Conventionalism will come back in Book 2 when Glaucon and Adeimantus take up Thrasymachus™
challenge, and we shall see the role it plays there in the next chapter.
Socrates and Thrasymachus: Republic 1 181
What are the values they possess which are truly the right ones? In what
sense are they (or the ¬gures they aspire to be) objectively superior to ordi-
nary people? Once Socrates has moved Thrasymachus from thoroughgoing
conventionalism or Callicles from unconstrained hedonism, and so they
have agreed that what is pleasurable is not the same as what is excellent,
and what is just is not simply what seems right to rulers, they are thrown
back once again to the question of what excellence really is. One has an
answer if it is simply instituting apparently self-serving laws or ful¬lling
appetites, but this leads to the unpalatable examples of the deluded tyrant
(who loses his tyranny through instituting “mistaken” laws) and the insa-
tiable catamite; and such ¬gures are certainly not the “superior” types that
Callicles and Thrasymachus have in mind. Once these ways of determining
virtuous actions fall by the wayside, however, there seems to be nothing to
¬ll in the gap.33 What is the excellence that their superior types possess?
What is the true advantage that Thrasymachus™ Rulers hope to gain over
the ruled? We have seen that this is a central issue throughout the dialogues,
particularly in the Gorgias.34
The upshot of the argument so far is as follows. Thrasymachus begins
with a sly ideological insight into justice “ namely, that it is to the advantage
of the rulers “ but his account of what justice is consists in the ruled
following the rules/laws laid down by the rulers. So, Thrasymachus begins
as a conventionalist. Having conceded that rulers sometime err about what
is to their advantage, the only way to render his ideological insight consistent
with his conventionalism would be to accept subjectivism about advantage.
Since this is unacceptable, he lays down strictures on what will count as
rulers and laws. This in turn makes what began as an ideological insight
now play the role of a criterion for a law™s being truly just, namely, its
truly being to the advantage of the ruler. This then ends what I have called
“thoroughgoing” conventionalism, since it is no longer simply something™s
being a law that makes it just, but its being a law that is truly to the
advantage of the ruler. This is unsatisfying for Cleitophon because we are
thrown back to the normative question of what constitutes the advantage

33 This will be the position pushed to the limits by Glaucon and Adeimantus in Book 2; see chapter
34 Below we shall see that Thrasymachus™ account of the bene¬ts of injustice makes his position quite
close to Callicles™. The question of what constitutes true harm and bene¬t for an agent becomes
quite urgent in the arguments with Polus and Callicles. Socrates of course thinks of harm and bene¬t
as primarily harm and bene¬t to the soul as such, independently of what happens to the body or
one™s material possessions. We shall see in the next chapter that this idea of harm and bene¬t again
plays a central role in Glaucon™s and Adeimantus™ challenge in Republic 2 and in Socrates™ subsequent
182 Aiming at Virtue in Plato
of the ruler; in other words, how do we know when a law is really a Law
and so when the following of it is really just?

5.5 aiming a nd determini n g i n th e “t hra s ym ach us
episod e”
After Thrasymachus declares that a ruler qua craftsman never errs, and so
unerringly decrees what is best for himself, which is what his subjects must
do and what justice is (341a), Socrates responds that, on the contrary, no
craftsman (or craft) ever seeks his/its own advantage, but the advantage
for whatever it rules over. If Socrates were right then the Rulers, as true
craftsmen, would be seeking the advantage of the ruled “ since all true
craftsmen seek the advantage of the objects of their craft “ and would
thus institute Laws to that effect. Thrasymachus then would be wrong
that justice “ the following of Laws “ is to the advantage of the superior.
Socrates™ argument thus depends on what the true relationship is between
craftsman and object, not on any determination of what is truly bene¬cial
for ruler or ruled.
Thrasymachus is exasperated by this,35 and, after insulting Socrates,
launches into his “immoralist” speech (343b“344c), which defends the supe-
riority of injustice over justice. Thrasymachus then intends to leave, but is
restrained and persuaded by everyone, including Socrates, to stay (344d).
Socrates refers to the issue raised in Thrasymachus™ speech as what sort of
way of life makes living most worthwhile (344d7“e3). He claims that he
was not persuaded by Thrasymachus™ argument on behalf of the unjust life,
and he urges Thrasymachus to remain and persuade them “that we who
make more of justice than injustice [dikaios…nhn ˆdik©av perª ple©onov
poio…menoi] are not deliberating correctly [oÉk ½rq¤v bouleu»meqa]”
(345b3“4). In chapter 2 we saw the signi¬cance of this phrase concern-
ing what a person “makes more of” in the argument of the Crito. I argued
that what a person makes more of is what aim he takes to trump other
aims. The personi¬ed “Laws” tell Socrates and Crito that they ought not
to “make more of anything” above what is just (pr¼ to“ dika©ou, 54b2“5).
The last thing Socrates asks Thrasymachus, before he begins a new argu-
ment against his position, is for him to be honest and to declare openly
when he changes topics or positions (345b8“c1). When Socrates starts
the next argument, he says explicitly that he is considering the “things
from before” (345c1“2), and indeed his discussion ignores the points that

35 Perhaps not unreasonably; see Beversluis (2000), 234“5 for succinct criticism of the argument.
Socrates and Thrasymachus: Republic 1 183
Thrasymachus has just made in his “immoralist” speech. Instead it raises a
new argument against Thrasymachus™ earlier idea that no techne seeks its
own advantage. At the end of a somewhat lengthy ¬nal speech (347b6“e2)
he concludes that this is why he does not agree at all with Thrasymachus
that justice is the advantage of the superior. Then, without offering anyone
the chance to respond, Socrates continues:
But we will take this [discussion of whether justice is the advantage of the superior]
up at some other time; it seems to me that what Thrasymachus is saying now is a
much greater matter, when he claims that the life of the unjust is superior to the
life of the just. (347e2“4)
What is going on here? The aiming/determining distinction makes sense
of this complex ordering of arguments. Socrates is putting down the ques-
tion of determining which actions are just and whether they are to the
advantage of the superior person or not, and turning to the topic that
we shall see is central to Thrasymachus™ immoralist speech: what should
be the supreme aim of one™s life? He accuses Thrasymachus of changing
the topic: from determining what just actions are in general to whether
one ought to aim at justice or injustice above all. In the argument with
Polemarchus it was Socrates who shifted the discussion from a determin-
ing question to an aiming question, “Should the just person ever harm
(do wrong to) anyone?”;36 here it is Thrasymachus. Pressed by Socrates
on the determining question, which in all the other dialogues always
ends inconclusively, Thrasymachus shifts to an aiming question. Instead of
defending the idea that virtue is supreme, however, he of course defends
Until the immoralist speech, the argument has been about the nature of
just actions: are they to the advantage of the superior, and do true rulers,
who do not err, make laws for their own advantage or do real rulers make
laws for the bene¬t and advantage of the ruled? In order for this eventually
to yield a satisfactory answer to Socrates™ “What is F?” question, some
content would have to be given to the notion of “advantage.”38 But for

36 Recall from the discussion of the Crito in chapter two and of the argument with Polemarchus above
that “harm” does not mean do bodily harm to, but act unjustly or contrary to virtue towards,
someone. So the question “Should the just person ever harm anyone?” is equivalent to “Is it ever
right to do wrong?” A negative answer is, of course, the af¬rmation of SV.
37 Annas (1981), 38“9 ¬nds it somewhat odd that, while these two questions are separated in Book 1,
the rest of the Republic considers them together. I shall argue that, on the contrary, the distinction
is central to the main argument of the Republic. See chapters seven and eight.
38 Indeed Socrates complains about Thrasymachus™ answer that justice is “the advantage of the supe-
rior,” since Thrasymachus himself had just told Socrates to answer without using any unclear or
disputed terms like “advantage” (339a5).
184 Aiming at Virtue in Plato
most of the argument, Socrates only focuses on what he said he would
focus on: the “of the superior” part, since, as he himself says, he agrees
that justice is an advantage, only he is unsure about the “of the superior”
part (339b5“7).39 The particular notion of advantage and disadvantage,
however, knowledge of which would enable a person to pick out the just
and unjust actions, is not clear. And this is important with respect to the
¬nal evaluation of Socrates™ argument. As I have emphasized, his argument
thus far has been based on the idea that a craft and craftsman as such
act for the bene¬t of the object of the craft rather than their own. If he
could establish this he would cause a problem for Thrasymachus™ position,
which maintains that the proper ruler of a state is both a craftsman and
one who aims at his own advantage. But it would not go far towards
saying what justice is; the content of justice remains a problem. This makes
sense of the end of Book 1 (354a“b), when Socrates refers to himself as
having behaved like a “glutton” in moving to the question of whether the
just life is better than the unjust before he had adequately resolved the
question of what justice is. He is presumably referring back to this point
In understanding the immoralist speech, we should be clear that the
laws are always meant for the ruled to follow. The Rulers do whatever
they want; they are superlatively strong, free, and masterly (as Thrasy-
machus describes them, 344c5“6). Since they have these qualities they will
always simply act to their own advantage, primarily by instituting Laws for
others to follow which will be to their own advantage. Stephen Everson
claims that Thrasymachus™ praise of injustice raises additional problems
about the coherence of Thrasymachus™ overall position, for there is no way
to account for the injustice of the Ruler (tyrant) in the immoralist speech.
Everson argues that his injustice cannot be captured by the “legalistic” def-
inition that justice is following the Laws (since the injustice of the tyrant is
not explained with reference to the Laws), and must be understood instead
in terms of Thrasymachus™ new de¬nition that justice is “the advantage
of another” (343c3“4).40 The tyrant is unjust because he ignores the inter-
ests of others and acts only in his own interest. A further problem is that
39 We might wonder whether Socrates entirely rejects this de¬nition. After all, in the body of the
Republic, he argues that justice is to the advantage of the rulers, for it is to the advantage of everyone.
So, although this description of justice as it stands is incomplete and perhaps misleading, especially
insofar as Thrasymachus™ interpretation of it maintains that justice is to the advantage of the rulers
and to the disadvantage of the ruled, it is not simply false as stated. Furthermore, Socrates will also
end up endorsing the view that the superior should rule “ although he will have a much clearer
account of what makes a person superior (see 7.4).
40 Everson (1998a), 115, cf. also 124.
Socrates and Thrasymachus: Republic 1 185
Thrasymachus™ initial claim was that a just action is to the advantage of the
superior and the ruler is the superior. But when the tyrant acts unjustly he
is acting for his own advantage, and since he is the superior, he would also
be acting justly according to the earlier account.
This is all the more vexing since, at the end of his speech, Thrasymachus
claims that he has been saying the same thing all along (344c7). I think that
the solution to this puzzle depends on recognizing that for Thrasymachus
injustice is a real, objective feature of the world independent of human
agreements, conventions, or laws, but that justice is not. If this is true, then
Thrasymachus™ claims about injustice, particularly about the “complete
injustice” of the tyrant, ought not to be understood, contrary to Everson
and others, in light of his answers to the question, “What is justice?” To
give an account of justice is not ipso facto to give an account of injustice.
For Thrasymachus injustice is a matter of exercising the natural impulse
to pleonexia; it is worthwhile to remind ourselves that “to do injustice”
(ˆdike±n) is simply “to do wrong.” Justice, by contrast, involves some human
intervention to establish laws and, further, on his revised account, laws are
only true Laws when they are to the advantage of the superior rulers. We
typically assume that if one is a conventionalist about justice, one must
also be a conventionalist about injustice. It would then follow that prior to
the institution of the conventions that constitute justice (in, for example,
“the state of nature”) there would be neither justice nor injustice. My claim,
however, is that this is not the way that Thrasymachus thinks about these
matters. Injustice is a natural state of the world, in which individuals seek
their own advantage; by contrast, justice is an arti¬cial construct aimed at
constraining people™s natural desires. He has been asked, and has tried to
give an account of, what justice is. The essence of the arti¬ciality of justice,
according to Thrasymachus, is that it must be unnaturally imposed on
people by forcing them to act in a way that bene¬ts another; it is the height
of irrationality to do so.
We shall see in chapter six that Glaucon similarly takes injustice to exist
naturally and prior to the “invention” of justice.41 But let us note here, as
a parallel, what Protagoras says in his “Great Speech” in the eponymous
dialogue. Protagoras is telling a story about how justice was given to human
beings. He explains that at ¬rst Prometheus had given humans wisdom in
the crafts along with ¬re, but that humans did not yet have “political
knowledge” (politikž, 321d5). So, when they attempted to found cities to
live together in order to survive against wild animals, “they wronged/acted

41 We should expect this if Glaucon is, as he claims to be, genuinely restating Thrasymachus™ position.
186 Aiming at Virtue in Plato
unjustly towards one another [ d©koun ˆllžlouv]” (322b7) until Zeus,
fearing that humans would be destroyed, told Hermes to give all people
a sense of shame and justice. Protagoras™ account, although very different
from Thrasymachus™, shares with it the idea that people can “wrong” and
“do injustice to” one another in a state of nature prior to the presence of
What is most signi¬cant about the immoralist speech for our purposes,
however, is that Thrasymachus has shifted topics in precisely the way
Socrates describes:42 he has moved from talking about what determines
whether some actions are just (the answer to which was whether they are in
accord with Laws) to the question of whether a person is better or worse off
by acting justly. In other words, he has moved from a determining question
to an aiming question.
In Thrasymachus™ immoralist speech he sticks to many of his previous
themes, to which he now adds another insight: that just actions are in reality
the good of another.43
And you [Socrates] are so far out there [kaª oÌtw p»rrw e²]44 concerning the just
and justice and the unjust and injustice that you are ignorant that justice and the
just are in reality the good of another, an advantage of both the superior and the
ruler, but the peculiar, special [o«ke©a] harm of the one obeying and serving. But
injustice is the contrary and rules over those who are in truth simple-minded and
just, and who, being ruled over, do what is to the advantage of that one “ the one
who is superior “ and they make him happy by serving him, but [do not make]
themselves [happy] in any way whatsoever. (343c1“d1)
This description of just action deepens the ideological analysis of how
justice really works “ an analysis that “far out” Socrates, who is so out
of touch with the way things really are, misses. Thrasymachus begins
to supply the evidence for his view in the next line by saying, “Con-
sider, o most simple-minded Socrates [å eÉhq”state SÛkratev] . . .”
(343d2), clearly echoing his remark that Socrates is out of touch by not
understanding what justice is and as such is one of the “truly simple-
minded,” whom he referred to in the passage above, as well as foreshadowing

42 Beversluis (2000), 233 maintains that the shift Socrates refers to has to do with an argument about
craft. But in his speech following Thrasymachus™ immoralist speech Socrates clearly distinguishes
the topic that Thrasymachus is discussing as being about which life is better, the just or the unjust;
and he equally explicitly returns to respond to what Thrasymachus was saying “before” (345c1).
43 As I argued above, all of Thrasymachus™ “accounts” of justice are supposed, minimally, to state
something common to all cases of just action and so to avoid the refutation that Cephalus™ earlier
de¬nition suffered.
44 Following Shorey (1930) note ad loc., and not Adam (1902).
Socrates and Thrasymachus: Republic 1 187
his later remark that justice is only “very noble simple-mindedness”
The key to appreciating Thrasymachus™ immoralism is quite similar to
what we saw was critical to Callicles™ account. Thrasymachus™ conception
of harm and bene¬t is eminently ordinary and consists of harm and ben-
e¬t to one™s body and one™s possessions. In these terms it is clear that
the unjust person is better off than the just one.46 As evidence Thrasy-
machus cites how the just person gets the short end of the stick in all of
his transactions, public and private. He pays more in taxes, and receives
less in refunds. When he holds public of¬ce his own personal property
falls apart and his friends and household despise him for not using his
power to bene¬t them (343d“e). Although, unlike Callicles, Thrasymachus
never explicitly endorses hedonism, it would seem to be a fairly natural
extension of his view. For he extols the ability and willingness of the unjust
person “to have more than his share” (pleonekte±n, 344a1) in the example
of the tyrant who succeeds in taking the “property” of the citizens and
enslaving them. The advantages of being a tyrant are taken to be obvious.
He will be supremely wealthy, and so also happy (343c8, 344b7). Why,
according to Thrasymachus, is the capable, unjust person so well off? It
would be natural to say that, in addition to the material and bodily gains,
he also bene¬ts because he can have whatever he wants; his appetites are
unconstrained and grati¬ed. In this way, like the Calliclean ideal man, the
Thrasymachean tyrant gains appetitive satisfaction. Callicles too, prior to
presenting his hedonism, extolled “having more than one™s share” as part of
excellence and the happy life (Gor. 490a ff.). Callicles™ hedonism would ¬t
in quite well as a further explanation of the tyrant™s happiness, and we shall
see in the next chapter that appetite grati¬cation makes an explicit appear-
ance in Book 2 when Glaucon and Adeimantus restate Thrasymachus™

5.6 soc rates ™ def ens e of s v i n r e p u b l i c 1
Socrates moves without pause from his second argument, which attempts
to show that a craft seeks the advantage of what it rules over (345c“347d), to
consider the “much greater” question raised by Thrasymachus™ immoralist
speech: is the unjust life superior to the just life? (347e). Socrates has already

45 Socrates will pointedly offer his own revision of what constitutes “simple-mindedness” in Book 3,
46 This point will be developed in detail by Glaucon and Adeimantus in Book 2; see next chapter.
188 Aiming at Virtue in Plato
declared that Thrasymachus™ speech did not persuade him (345a“b) that
the unjust life is better than the just one, and in the present passage he
turns and asks what Glaucon thinks. Glaucon replies:
I at any rate think that the life of the just is more pro¬table [lusitel”steron].
But, I [Socrates] said, didn™t you hear how many good things Thrasymachus just
went through that belong to the life of the unjust?
I heard, he [Glaucon] said, but I was not persuaded [¢kousa, ›fh, ˆll ¬ oÉ
pe©qomai].47 (347e7“348a3)
This passage repeats the theme of persuasion, recalling the opening of
the dialogue, when Polemarchus asks Glaucon whether he and Socrates
could persuade someone who does not listen (327c12). Glaucon represents
an example of someone who “listened to and heard” the alleged good
things that accrue to the unjust person, but who was not persuaded by
them to believe that the unjust life is superior.48 Thrasymachus as well
seems to be someone who will listen, but not be persuaded, at least not by
the argument of Book 1. Furthermore, this passage serves to place Glaucon
¬rmly on Socrates™ side in this argument, a place he will continue to occupy
for the rest of the work.
Getting nowhere with the determining question about justice, Socrates
happily turns to the “much greater” question of whether the just life is
better than the unjust. But he runs into trouble. Since Thrasymachus has
said that injustice is more pro¬table than justice, Socrates looks for quick
agreement that justice is a virtue and injustice a vice. It is clear where he
is headed: to repeat the argumentative move employed with Polemarchus

47 The lines that follow this interchange, 348a7“b7, are also quite striking. When Glaucon says that
he would like to try to persuade Thrasymachus that he is wrong, Socrates then raises the issue of
their approach. One method would be to oppose Thrasymachus™ logos (presumably the “immoralist
speech”) with a logos of their own that presented the good things that accrue to the just life. Then
Thrasymachus would respond again, and then they again. After this they would need a jury to
add up and measure all the good things on each side to determine which was best. Alternatively,
they could simply continue with the question-and-answer way. Glaucon opts for the latter. But, as
Shorey alone notes, in Book 2 (358d ff.) it seems that they adopt the former method. It is tempting
to see here a contrast of approaches from the perspective of the outer frame. The results of this
method turn out to be dissatisfying even for Glaucon himself. If this is right we might wonder who
the “jury” is who counts and adds up the advantages of the just and unjust lives; perhaps it is the
audience of the outer frame, we readers.
48 Book 2 opens with Glaucon once again raising the issue of persuasion: does Socrates want simply to
seem to have persuaded them that justice is in every way better than injustice, or to have really done
so (357a5“b2)? For Glaucon points out that he has not yet really done so, and appears to talk for
all of the people present. He and Adeimantus will emphasize repeatedly, however, that they are not
persuaded by Thrasymachus despite the fact that they will reformulate his argument. See chapter
Socrates and Thrasymachus: Republic 1 189
(335b“d). In that interlude they moved from consideration of a determining
question, “What is justice?”, to an aiming question, “Would a just person
ever harm anyone?”, where Socrates carefully and quietly, if uncritically,
imported the idea of a harm to the soul “ that is, harm to a person™s virtue.
But when Thrasymachus denies that justice is virtue and that injustice is a
vice, Socrates™ path is blocked, and he must try a new tack. He is forced to
supply a new argument from 349a to 350d (“the pleonexia argument”). This
argument relies on a very questionable use of the notion of techne.49 What
is most striking for our purposes is simply that Socrates tries to establish
that justice is a virtue without considering the effect of just actions on the
soul. There is no discussion of the harm or bene¬t that justice does to the
soul. The effect of just actions on the soul will, of course, be center stage
in Book 2.
Once the pleonexia argument is completed Socrates reaches the point
from which he tried to begin, concluding, “justice is virtue and wisdom,
and injustice is vice and ignorance; let™s take this as set” (350d4“6). There
is clear textual evidence that the argumentative relationship with Thrasy-
machus is breaking down at this point. At the beginning of this argument,
Socrates declares that they will have to push on, despite Thrasymachus™
radical stance, adding that Thrasymachus really seems to him to be saying
what he believes (349a). Thrasymachus retorts that it makes no difference
whether he believes what he is saying or not; it is the account Socrates ought
to refute.50 Socrates quickly concedes, and the pleonexia argument com-
mences. After the pleonexia argument, Socrates narrates in present time that
Thrasymachus fought against him throughout, sweating profusely and not
easily conceding Socrates™ points, contrary to how Socrates has presented
matters (350c“d). Then Socrates says that Thrasymachus blushed (350d3).
Why does Thrasymachus blush? In the Euthydemus (297a), Dionysodorus
blushes as well. In that case his brother has just pointed out to him that
he has jeopardized the argument; Socrates will win. When Dionysodorus
realizes his error, he blushes. What is at work here is the sense of shame
mentioned in chapter three. Philosophical argument is agonistic. And

49 Reeve (1988), 20 calls it “grossly fallacious.” Book 1 frequently plays with the concept of techne
in strange and dif¬cult ways. See also, e.g., Irwin (1995), ch. 11; Parry (1996), ch. 1, and Roochnik
50 Beversluis (2000), 237 claims that this moment is unique in the “early” dialogues. But in the
Protagoras Socrates offers the abandonment of the “say what you believe” constraint after things
start getting dif¬cult between him and Protagoras. He tells Protagoras that he does not care whether
Protagoras believes the answers, so long as Protagoras answers; it is chie¬‚y the “account” he is testing
(333c). Vlastos (1983/1994) emphasizes the importance of the “say what you believe” constraint; see
Irwin (1994) for a discussion of Vlastos™ view.
190 Aiming at Virtue in Plato
even if, as is surely the case, Plato wants to distinguish Socrates™ type of
argumentation and his own from the sort of eristic practiced by a Euthyde-
mus and Dionysodorus, that does not mean that there are not winners and
losers in the more noble type of argument practiced by Socrates. Callicles
and Thrasymachus, with their esteem of the “superior” type, who conquers
and “has more” than ordinary people, are keenly aware of the shame of
being bested publicly by Socrates. Once Socrates has, apparently at least,
established his key premise “ that justice is virtue and injustice vice “ and
so has put the argument back on a Socratic playing ¬eld, Thrasymachus
has lost.
But Plato has given us reason to temper such an assessment in this case.
After all, Socrates is the narrator, so he is in control of the descriptive details,
and, from the perspective of the outer frame, Plato could expect a careful
reader to be aware of this. Socrates says that Thrasymachus did not con-
cede these points easily, but he does not tell us how Thrasymachus objected.
Socrates hides from us some of Thrasymachus™ argument. We can see some
further justi¬cation for his own ¬nal description of himself as a glutton
insofar as he keeps the precise details of what happened in the argument
for himself, away from the audience of the outer frame. We know that, at
the beginning of Book 2, Socrates believes that the discussion is over. Obvi-
ously, however, he is wrong; even Glaucon and Adeimantus, who explicitly
agree with Socrates about the value of justice, want to go through the argu-
ment again. We might suspect then that Thrasymachus™ responses made a
bit more of an effect on the audience of the inner frame than Socrates, in his
narration, lets us in on. (This is not to deny that Glaucon and Adeimantus
also take the argument up again in Book 2 because of a dissatisfaction with
Socrates™ positive arguments in Book 1.) Moreover, Socrates does report that
Thrasymachus himself says he is not satis¬ed with Socrates™ argument and
that he has quite a bit to say about these matters, but that he knows that if he
did Socrates would accuse him of demagoguery (dhmhgore±n) (350d9“e1).
Thrasymachus then agrees simply to nod agreement and disagreement
“just as one does to old women telling stories” (350e2“3). Socrates, rather
lamely, tells him not to answer contrary to his own opinion, but when
Thrasymachus says that he is doing it to please Socrates since he won™t
let him make a speech, Socrates happily concedes. From here to the end
of Book 1 Thrasymachus has ceased, like Callicles in the Gorgias after
510, to be a real interlocutor. Furthermore, an attentive reader, as in the
Gorgias, is not fooled into thinking that Socrates is persuading anybody
Socrates and Thrasymachus: Republic 1 191
The book ends with one of Socrates™ most sweeping disavowals in the
corpus (see 1.3):
I have not feasted well, however “ but it is my own fault, not yours. Just as gluttons
always snatch at every dish that is passed and taste it before they have appropriately
enjoyed the previous one, so it seems to me that I too, before ¬nding the ¬rst object
of our inquiry “ what justice is “ let that go and hastened off to consider something
about it, namely, whether it is vice and ignorance or wisdom and virtue; and again
later when the account broke in that injustice is more pro¬table than justice I
could not refrain from turning to that from the other subject. So that for me the
present outcome of the discussion is that I know nothing. For when I don™t know
what the just is, I shall hardly know whether it is a virtue or not, and whether the
one who has it is or is not happy. (354a13“c3)
Socrates here draws out the metaphor of a “feast of words,” begun by
Thrasymachus in the midst of his declaration that he will no longer try to
oppose Socrates “ he may have his feast (352b).51 As I emphasized in chapter
one, Socrates™ disavowals ought to be considered in context. This disavowal
comes at the end of a summary of the argument that has preceded, and I
think we can see that this summary is not wholly accurate. For one thing,
the order is incorrect. Socrates and Thrasymachus were in the midst of
arguing about a determining question whether justice was the advantage of
the superior “ when Thrasymachus embarked on his “immoralist speech,”
exasperated by Socrates™ claim that the craft of sheep-herding aims at the
welfare of the sheep. In the immoralist speech Thrasymachus addresses an
aiming question, arguing that the unjust person is far better off than the
just person. It is then Socrates who, after saying that he was not persuaded,
asks Thrasymachus not to switch topics, and who goes back to offer one
more argument about what justice is. Then, at 347e, Socrates does “rush
headlong” to the next topic, but it is the aiming question ¬rst raised by
Thrasymachus; it is not the topic that he mentions above, whether justice is
a virtue and wisdom or a vice and ignorance, but one concerning whether
the just life is better than the unjust life, which he is eager to discuss. It
is only because Thrasymachus rejects the idea that justice is a virtue that
Socrates is forced to embark on the pleonexia argument and to consider the
more radical question of whether justice is a virtue in the ¬rst place.
In the next three chapters we shall see that the aiming/determining
distinction at play here in Book 1 proves pivotal to understanding the
argument of the body of the Republic.

51 See Worman (2008), esp. ch. 4, on the broader signi¬cance of this type of imagery, in Plato as well
as in other prose and in poetry in the ¬fth and fourth centuries.
c h ap t e r 6

The bene¬ts of injustice

6 .1 d e fining j us tice a nd t h e proj ect of th e r e p u b l i c
Whatever may have been the case in previous dialogues, and even in
Republic 1, most commentators believe that once Glaucon and Adeimantus
take over the argument they “identify being just with a property possessed
primarily by psyches.”1 This point is put in more general terms by saying
that Plato moves in the Republic from an act-centered account of justice
or virtue to an agent-centered account.2 The explanation of this is that
once Plato has despaired of being able to provide an act-centered account
of what justice is, he begins in Book 2 to present an agent-centered one,
which culminates in his Book 4 account of “Platonic” justice as harmony
in the tripartite soul.3 The attractiveness of this interpretation is allegedly
buttressed by the plausibility of the ethical theory behind it. Plato, one of
the original virtue ethicists, rejects the idea that one can specify virtuous
actions ¬rst and then de¬ne the virtuous agent as the one who performs
such actions and performs them in the right way. Matters will instead be the
other way round. De¬ning what it is to be a virtuous person ¬rst, virtuous
actions will then be the actions done by that sort of person.4

1 Reeve (1988), 35; cf. 24“5. Irwin (1995), 378, n. 5 agrees.
2 While I shall often speak for obvious reasons about justice and injustice in the context of the Republic,
we should recall that, as in the discussions of the Apology, Crito, and Gorgias, the just/unjust action is
understood to be synonymous with the virtuous/vicious, the right/wrong, and the ¬ne/ignoble action.
Thus “doing injustice” is synonymous with “wrongdoing.” Under a suitably broad understanding
of “moral,” introductory philosophy textbooks are right in saying that the central question of the
Republic is “Why be moral?”
3 This view can be found in Annas (1981), 23“4, who sums up the force of Socrates™ objections to
Cephalus and Polemarchus as follows: “They show up the inadequacy of the notion that one can say
what justice is by specifying kinds of action at all. Later in the Republic, when we ¬nd what Plato™s
own account is, we shall see that he moves away entirely from the doing of certain actions, and instead
characterizes justice as a state of the agent. The primary questions will turn out to be those about the
just person, and questions about which actions are just actions will be in an important way secondary.”
(my emphases).
4 See too discussion in 4.2.

The bene¬ts of injustice 193
If the argument of this book is correct, however, this will not be the best
way of understanding how the central argument of the Republic develops.
Appeal to the state of the virtuous person™s soul is used, as it has been used,
to address an aiming question: why should one do the virtuous action above
all? Socrates™ ¬nal speeches in the Gorgias describe passionately and vividly
the ill effects of unjust action on that most important part of ourselves,
our soul. Although this is not a way of addressing a determining question
at all, we should not conclude that the determining question becomes
“secondary,” if that means either that it can somehow be answered by
de¬ning justice in terms of a state of soul, or else that it can simply be
dropped. A harmonious state of soul and a property J which is common to
all and only just actions and explains why they are just could both be answers
to a question of the form “What is justice?” But they are not answers to the
same question. As I argued in chapter four, an account of what it is to be
virtuous does not answer the determining question.
Furthermore, the habituation principle, explicit in the Crito and elab-
orated in the Gorgias, is repeated in much more detail in the Republic. It
provides an important link between virtuous actions and virtuous charac-
ter.5 The fact that the habituation principle is never abandoned by Plato, but
rather developed with increasing sophistication, is a strong reason not to see
the difference between discussion of justice as a state of soul and as a prop-
erty of actions as two con¬‚icting ways of discussing the nature of justice.
Rather, when you explain the effect of justice on the soul, you explain why
a person ought to act justly above all: the question at issue, as we shall see,
in Glaucon™s and Adeimantus™ challenge. But, far from relegating questions
about actions to a “secondary” position, it makes them very urgent. We still
need some way of determining which actions are the just ones and which are
not. The question that has been plaguing Socrates has not gone away and
it is not diminished in its centrality merely because there is a newer, more
complex, account of the effect of just actions on a person™s soul. In fact,
insofar as the good effects of just action have been persuasively described in
detail, to that extent the need to discover an answer to determining ques-
tions becomes that much more intense. An account of the just soul may be
very useful for persuading an agent of the truth and signi¬cance of SV, but
this simply magni¬es the problem of Cleitophon with which we began.
Moreover, we need to keep in mind a distinction of the type in the Euthy-
phro (see 10a ff.): although by de¬nition a virtuous person does virtuous

5 In 7.2“3 we shall see that Plato™s elaborate discussion of proper education is predicated on the principle
that repeated performance of the right sort of actions generates a soul of a corresponding type.
194 Aiming at Virtue in Plato
actions, he does them because they are virtuous, they are not virtuous
because he does them. The point of an “agent-centered” or virtue-based
account of the type that Plato and Aristotle adopt is that a person who does
not have the “perspective”6 of the virtuous person cannot be counted on to
determine correctly what the virtuous action is in all circumstances. The
virtuous person by contrast has a grasp of objective moral truth: she knows
without error what is truly virtuous and ¬ne. So although we non-virtuous
types might seek help when we deliberate about what to do by looking at
what the virtuous person does, or by thinking about what a virtuous agent
would do, the simple but important fact remains that virtuous actions are
not virtuous because a virtuous person does them, rather, the virtuous per-
son is virtuous because, at a minimum, she correctly discerns what is truly
virtuous and unfailingly acts accordingly.
Readers, particularly those thinking of Aristotle, might object that this
cannot be an adequate account of the virtuous person and of virtuous
action. For the virtuous action itself must be characterized with reference to
the agent™s motivation. In Aristotle™s language it is not enough to do what
the virtuous person would do, one must do it in the way the virtuous person
would do it (NE 2.4, 1105a28“33).7 But, even if we agree with Aristotle, I
have emphasized that an agent must have doing the virtuous action as her
supreme aim in order to determine correctly what the virtuous action is
(see 1.6). For, if her overall aim is, for example, to maximize her ¬nancial
gain, this may sometimes lead her to do the virtuous action and some-
times not, depending on varying circumstances. Thus, as we might put it,
the issue is not the “purity” of her motive, but whether she has some fur-
ther end (pleasure, money) beyond the doing of the truly virtuous action.
Commitment to SV secures that she does not, and thus secures that, by
being committed to SV, one aims at virtue above all. What this does in turn
is to ensure that the agent is solving the right problem: namely, determining
what the virtuous action is (and not what the most pleasurable or ¬nancially
pro¬table action is). Thus being committed to SV is a necessary condition
for correctly solving determining questions about virtuous action.

6.2 th e c l a ss if ic ati on of g oods
In the opening Stephanus page of Book 2, Glaucon lays out a notori-
ous tripartite classi¬cation of goods: those desirable only for themselves

6 Whatever combination of cognitive and desiderative states this amounts to.
7 See Whiting (2002).
The bene¬ts of injustice 195
independently of their consequences, those desirable for their own sake
and for their consequences, and those which are desirable only for their
consequences. The debate about the proper interpretation of the classi¬-
cation of goods, both on its own terms and in terms of how it ¬ts with
the subsequent arguments, has traditionally focused on whether Glaucon™s
notion of a good in itself does or does not include causal consequences of
the good, including, notoriously, happiness.8 Many commentators agree
that Glaucon™s category of “good in itself” includes things that are, in some
sense, consequences or effects, and so what puts them in the category of
being a good in itself is their having a particular sort of effect or conse-
quence. The challenge then becomes specifying what type of causal effect
still counts as being part of a thing™s being good in itself.9 For reasons I
shall offer below, I think it is uncontroversial that the category of goods in
themselves must include things that we consider “causal consequences” in
the modern sense. The debate that unfolds is one about whether a person
is better off performing just or unjust actions and so better off being just or
unjust. Most scholars interpret this as part and parcel of the eudaimonist
outlook of the Republic as a whole.10
While I do not dispute this, I shall focus on how being just or unjust,
and acting justly or unjustly, is better or worse for a person™s soul; this is
the speci¬c challenge of the Republic.11 Throughout the dialogues we have
seen the signi¬cant presence of the distinction between goods of the soul,
goods of the body, and material possessions.12 I have shown that the concept
of a good for the soul is crucial to Socrates™ argument for SV, especially
in the Gorgias. These points are important for understanding Glaucon™s
classi¬cation of goods and its connection to his and Adeimantus™ challenge.
I shall argue that when Glaucon speaks of a “good itself by itself” he means
something that bene¬ts the soul qua soul. When he talks about something
being chosen (or avoided) for its good (or bad) consequences, he is referring
to the bene¬ts gained to the body or to one™s material possessions. Therefore

8 The modern debate begins in earnest with the articles of Foster (1937) and Mabbott (1937/1971).
9 See, e.g., Sachs (1963/1998), 209, Kirwan (1965), and White (1984). Irwin (1995), §135 argues that
Plato maintains that justice is not a causal means to happiness, but a dominant part or component
of it.
10 For example, see Irwin (1995), ch. 12; Kraut (1992b); Butler (2002).
11 I do not deny that having more bene¬ts for one™s soul is necessary for, or perhaps a component of,
being happy. But I think that it is not the concept of happiness in any very robust sense that is doing
the argumentative work in the Republic. Rather the focus is more speci¬cally on what being just
and virtuous (which we shall see is a causal consequence of acting justly and virtuously) does for the
soul; is being just somehow good for the soul, and if so, how?
12 Recall that I intend the phrase “material possessions” to cover a broad category including not only
one™s possessions and property, but also one™s reputation and luck.
196 Aiming at Virtue in Plato
I shall argue that the category of a good in itself for Plato is not concerned
with the modern notion of being independent of any causal consequences
(some goods in themselves are, some are not), but with the very Platonic
idea of something being good for the soul independently of any effects
it may have on one™s body or the state of one™s material possessions. On
this reading, when Glaucon wishes to hear justice praised “itself by itself ”
(aÉt¼ kaq¬ aËt», 358d2“3), his question amounts to: how is being just
good for the soul? We shall see that Glaucon is concerned with just actions
and that he holds that one causal effect of acting justly is that it causes the
soul to become just.13 Someone might worry that in order for the doing of
just actions to yield a just soul, the person must also do the actions with
the right motive. Therefore simply doing just actions (for example, paying
one™s fair share) is not suf¬cient to make a person just. While this is an
issue that does come up to a degree later in the Republic, I deny that there
is any concern in Glaucon™s and Adeimantus™ argument with the motive of
the just person. Whenever they discuss characters who seem to be just but
are not, such as the possessor of the ring of Gyges or Adeimantus™ fellow
who keeps a “facade of virtue” around him (365c3“6), it is clear that these
are people who are actively engaged in wrongdoing and injustice, but who
succeed at keeping their wrongdoing hidden (and so they appear just “
that is, appear not to be doing anything wrong). Indeed, a dif¬cult but very
important goal for the successful unjust person is to keep his actions hidden
(365c“d). Glaucon and Adeimantus never consider a character whom we
would describe as engaging in all and only just actions, but for the wrong
I shall speak, then, of essential properties of just actions (which would
include both their intrinsic features and their causal consequence of making
a person just), as distinct from merely accidental properties which justice
may have in particular contexts. So Glaucon is indeed asking a question
about causal consequences, but of a select type: the causal effects of acting
justly on the soul. And this is why he moves easily from almost ubiquitous
talk of “practicing justice” to questions about the justice of the soul. Taking
it for granted that acting justly (not merely seeming to act justly) makes

13 Heinaman (2002), 324 clearly recognizes this. Citing 588b“590a, he writes: “It is clear that psychic
justice is a causal consequence of just action and that psychic injustice is a causal consequence of
unjust action.” I entirely agree with Heinaman™s claim that some things which we would call causal
consequences are part of what counts towards something being a “good in itself” in Plato™s sense.
Heinaman, however, retains the language of “intrinsic good” as a gloss for Plato™s idea of a good
“for itself,” which can be confusing since something being an intrinsic good typically means that its
goodness is independent of any of its causal consequences.
The bene¬ts of injustice 197
one just, he then presses the question: “Why would one want to be that?”
The ¬nal answer to why we ought to be just will be in terms of the value of
the state of the soul that is effected by acting justly.
In the rest of this section, I shall show that such an interpretation is
compatible with the examples Glaucon offers of goods in the three classes.
In the later sections of the chapter I shall provide evidence that it best
explains the argument we actually get from Glaucon and Adeimantus on
behalf of the many.14
The ¬rst class of goods are those that are desired not for their conse-
quences (t¤n ˆpobain»ntwn) but are each welcomed “itself for the sake
of itself ” (aÉt¼ aËto“ ™neka); for example, “enjoyment” (t¼ ca©rein) and
those pleasures (a¬ ¡dona©) which are harmless and from which nothing
other arises later on account of them other than having enjoyment (ca©rein
›conta) (357b4“8). The discussion about this ¬rst class has focused on
whether Plato already presents the idea of a “good in itself” as being good
in itself because of a causal consequence insofar as the passage seems to
suggest that what makes pleasures goods in themselves is that they have
the causal consequence of enjoyment and no other (bad) consequences.15
I do not think that this is the central issue, however. We might instead
think about this ¬rst class of goods in terms of Glaucon™s and Adeimantus™
upcoming concern with how being just could be good for the soul. Pleasure
and having enjoyment are desirable states of soul for their own sakes. As
we saw in chapter three, in the Gorgias Callicles maintains that appetite
grati¬cation is the only good in itself: excellence at appetite grati¬cation
(gratifying the most and most extensive appetites) is what excellence and
happiness are. Although the examples from this ¬rst class, by contrast, stop
quite short of hedonism of any stripe “ pleasures and enjoyment are goods,
so long as they are “not harmful [ˆblabe±v] and nothing arises on account
of these at a later time other than having enjoyment” (357b7“8) “ never-
theless what makes them goods in themselves are the pleasures that they
are in the soul.
The second class contains goods that we value both for themselves and
for the sake of their consequences. The examples are: thinking, seeing, and
being healthy. There is no explanation of how or why these examples fall
14 In chapters seven and eight I shall argue that it makes good sense of the Republic™s answer as well.
15 See Heinaman (2002) 325, n. 32 for a defense of this reading, and against Irwin™s (1977), 325, n. 8
attempt to justify taking ca©rein and ¡dona© as equivalent. Heinaman rightly does not rely on this
by itself, however, to conclude that Plato counts a certain type of causal consequence as being part of
valuing something for its own sake. For, if ¡dona© are activities, then it is possible that the activity is
identical to the enjoying: watching a TV show I like is identical to enjoying myself; the enjoyment
is not a causal consequence of the watching.
198 Aiming at Virtue in Plato
under this class. Why are these “goods in themselves”? We might simply
explain their status as essential goods in terms of the ¬rst class: thinking,
seeing, and being healthy are enjoyable, regardless of what else they might
lead to: it is simply enjoyable to think, see, or be healthy, at least in many
cases.16 On this reading, what makes them essentially valuable is identical
to what makes enjoyment and pleasures essentially valuable. They are going
to differ from items in the ¬rst class insofar as enjoyment itself (we assume)
has no further positive or negative consequences, while thinking, seeing,
and being healthy clearly can also lead to all sorts of other goods of the
body and material possessions.
We might be tempted to resist this idea because we think that these
three examples are goods in themselves in a way that is different from how
the goods of the ¬rst class are goods in themselves; the essential goodness
of thinking is surely different from the essential value of enjoyment. But
succumbing to this temptation would lead us astray. While I think that Plato
ultimately believes that all of these examples have goodness in themselves
in a way that is not reducible to their harmless enjoyment,17 that cannot be
relied on at this stage of the argument. For there is a dialectical requirement
at work here. The classes of goods that Glaucon distinguishes, together with
the examples, ought to be uncontroversially acceptable to “the many” for
the purposes of his argument. It is generally unremarked that Glaucon
and Adeimantus do not just take up Thrasymachus™ position, but from
the beginning Glaucon identi¬es it as a position of “the many” (358a4). If
Glaucon is really to be defending Thrasymachus™ and the many™s position,
he must begin from premises they would accept. The many acknowledge
only one kind of good for the soul, one kind of essential good: enjoyment
(ca©rein). To take enjoyment is to satisfy one™s desires; it is inherently
enjoyable to have one™s desires satis¬ed and inherently painful to have
them frustrated. On my reading, Glaucon™s challenge will thus ask: what
does it mean for something to be good for the soul, other than being a case
of enjoyment? Does it mean anything at all? The many would say “no.”
To hold that some other mysterious, unmentioned, concept of goodness is
smuggled in here other than enjoyment is not warranted by the text, and
would violate the dialectical requirement of the argument.18

16 For thinking and seeing it is easy to imagine things we would rather not see or think about. But
presumably the point is that these activities are themselves enjoyable as such, even if some particular
case of thinking about x may not be enjoyable.
17 See 8.5 for discussion of the arguments about pleasure in Book 9.
18 As far as I know, no commentator discusses why thinking, being healthy, and so on would be counted
as goods in themselves. But, if we are not explicit that the only kind of intrinsic goodness that would
The bene¬ts of injustice 199
The third class of goods reinforces this interpretation. Pursued only for
their consequences, things in this class are onerous or painful in themselves
(–p©pona). The etymology of –p©pona carries the implication that such
things are attended with pain, which highlights the marked contrast with
the ¬rst class. Robert Heinaman has persuasively and importantly argued
that what belongs in this class are what the many regard as evils in them-
selves.19 Following the above interpretation of what makes things essentially
good, what makes these things bad in themselves is the fact that they are
themselves painful or have pain as their causal consequence for the soul.
The examples are: exercising, medical treatment for a sick person, medicine
itself and other ways of making money. It may be good for your body to
undergo a certain treatment, but, insofar as you do not want to, your soul
experiences the opposite of pleasure in the frustration of its desire. Just as
the essential goodness of the previous examples consists in their satisfying
desires, what makes these examples painful is that they frustrate desire; it
may be “good” for your wealth to go to work and make money, but inso-
far as you don™t want to go, your soul experiences pain, for its desire is
Moreover, examples from the third category also provide insight into
what will count as “good consequences”: beauty (the result of exercise),
physical health (the result of receiving medical treatment), money (the
result of employment). What is crucial to these examples is that they are
all bene¬ts to the body or to one™s possessions.
When Glaucon lists the three classes, he surely already knows which
class Socrates will place justice in: Socrates will hardly deny that justice
is a good for its own sake, and could not plausibly deny that it is “good
for its consequences.” The idea that justice, under ordinary circumstances,
is valuable for its consequences is taken as obvious in the argument, as
is the claim that injustice, under ordinary circumstances, has bad conse-
quences.20 After Socrates has declared that justice ought to be placed in
the second class of goods, those that are good both in themselves and for
their consequences, the “¬nest” class (358a1“3), Glaucon states that the
many, like Thrasymachus,21 do not agree, placing it instead in the third
class that “must be practiced” (–pithdeut”on) only for its consequences,
be uncontroversially accepted by Thrasymachus and the many is enjoyment, we risk missing the
importance of the upcoming argument.
19 Heinaman (2002), 311“15 argues that the class of things that are good only for their consequences
does not contain anything that is in itself indifferent or intermediate.
20 See Heinaman (2002), 315ff.
21 Irwin (1995), 181“2 says that this description of Thrasymachus™ position is “rather surprising” since
he does not seem to consider justice a good at all, but calls it simply foolishness (348c2“e4). But it is
200 Aiming at Virtue in Plato
but avoided “itself by itself” as something harsh (calep»n) (358a4“6).22
The challenge that Glaucon raises here on behalf of the many is clearly
a question about just action. What they “practice” for consequences and
what they “¬‚ee” (feukt”on) in itself are doing just actions. According to
the many the reason to act justly is “for the sake of rewards and for the
sake of popularity on account of reputation” (358a5). This refers to the
bene¬ts accrued to one™s body and material possessions. Although acting
justly involves the frustration of appetite (and so is a bad thing in itself and
painful [–p©pona]), in ordinary circumstances it avoids the physical harm
of punishment, secures you employment for wages, and gains you further
material goods on account of your good reputation (bene¬cial marriages,
and so on).23
My interpretation then of the distinction between goods in themselves
and goods for their consequences is as follows: something is good (or bad)
in itself if it is bene¬cial (or harmful) to the soul itself; something is good
(or bad) for its consequences if it leads to bene¬ts (or harms) to one™s body
or to one™s material possessions. Furthermore, I have claimed that the only
candidate acknowledged by the many as a good in itself is enjoyment; they

clear that, when Thrasymachus and even Glaucon himself (359a7“b1) speak of justice as not a good,
they are not denying that it can belong in the third class of “goods.” Heinaman (2002) maintains
that Glaucon and Adeimantus argue that justice is intrinsic evil and injustice an intrinsic good “
that is, that the third class of goods are intrinsic evils. I agree, but avoid the language of “intrinsic.”
See Heinaman, 322“3, n. 28 for discussion of Irwin.
22 As we shall see, there are frequent references to acting. The continual use of “practice” (–pitždeuma
and cognates) highlights the importance of action. Practice has to do with engaging in actions of a
certain sort. See Vasiliou (2002b), §3.1 on the importance of this word to the Gorgias and in other
dialogues. Acting rightly remains the core issue, since, if being just is shown to be essentially valuable
(as it will by the end of Republic 4) and thus shown to be desirable in itself, we will need to engage
in truly just actions in order to be just. See below and, especially, chapters seven and eight.
23 Heinaman (2002), 327ff. argues that “in every passage where Plato talks about excluding the conse-
quences of justice from consideration of its intrinsic worth, it is the reputation for justice and the
consequences following on that reputation which are at issue” (his emphasis). I disagree. In his dis-
cussion of 358a, Heinaman must construe both misq¤n (“rewards”) and eÉdokimžsewn as explained
by di‡ d»xan. Following Adam (1902) ad loc., I take di‡ d»xan to go with eÉdokimžsewn alone.
I think it is implausible that misqo© refers to rewards received on account of one™s reputation. The
primary meaning of misq»v and its compounds has to do with wages or pay for work, although it can
mean more broadly “rewards.” But strong evidence that it should be understood as meaning “wages”
is that it is so used a couple of paragraphs earlier (357d1) in the description of the consequences
for the sake of which people consider as goods items in the third class. The examples there are
explicitly “other ways of making money” (357c6“7). The concluding phrase foreshadows the one at
358a5: we do not choose things in the third class for themselves, “but for the sake of wages and other
things, however many arise from them” (t¤n d• misq¤n te c†rin kaª t¤n Šllwn ‚sa g©gnetai ˆp ¬
aÉt¤n, 357d1“2). Here there is no mention of “reputation” at all (this locution, with no mention of
reputation, is repeated again at 358b6“7, where Glaucon explains what he wants Socrates to argue).
Reputation is added at 358a5 as one of the “other things” (referred to at 357d1“2) that arise in addition
to wages from such onerous activities as acting justly. See also Rep. 1, 345e ff., where misq»v is used
repeatedly to refer to wages. I shall discuss other passages brought up by Heinaman below.
The bene¬ts of injustice 201
conceive of nothing else as a good for the soul. In this light, let us consider
Glaucon™s following request:
I desire to hear both what each [justice and injustice] is and what power it has itself
by itself when it is in the soul [t ¬ ›stin —k†teron kaª t©na ›cei d…namin aÉt¼ kaq¬
aËt¼ –n¼n –n t¦€ yuc¦], but leaving off rewards and the consequences from these
[toÆv d• misqoÆv kaª t‡ gign»mena ˆp ¬ aÉt¤n –†sai ca©rein]. (358b4“7)
When Glaucon asks for an explanation of the value of justice itself by itself,
he is asking: what value does justice have for the soul itself, aside from the
consequences of having such a state of soul for one™s bodily and material
well-being? We shall see in chapters seven and eight that, in the rest of the
Republic until Book 10, 612b ff., this is precisely what Socrates addresses in
the course of defending SV.
So far I have only shown that such an interpretation is compatible with
the initial classi¬cation and the examples it offers. In what follows I shall
argue that this interpretation is not only possible but also most plausible
by showing how it makes neat sense of the arguments we actually get from
Glaucon and Adeimantus.24 Moreover, it shows a continuity and coherence
of concerns throughout the dialogues and makes sense of what we ¬nd in
the rest of the Republic, including the problematic arguments about pleasure
in Book 9 (8.5).

6 .3 u nders ta nding gl au con™s e xa m pl e
Before turning to Glaucon™s and Adeimantus™ argument we need to consider
a methodological point that greatly affects how we understand it. Glaucon
repeats that he wishes to hear justice praised “itself by itself” (358d3) and
then explains the rationale behind his upcoming argument:
For this reason I will speak by lengthening out my praise for the unjust life, but by
speaking [thus] I will show you the way in which I in turn also wish to listen to you
on the one hand disparage injustice, and, on the other, praise justice. (358d5“7)
Commentators have read this passage so as to restrict drastically
the exegetical options for understanding Glaucon™s upcoming argument.
Robert Heinaman concludes from the passage:

24 This has been the sticking point in the literature. It should be clear too that I agree with two very
important claims by Heinaman (2002): (1) that some causal consequences of a thing count towards
whether it is a good in itself in Plato™s sense; and (2) that Plato™s class of things that are good only for
their consequences are evils in themselves. My understanding, however, of what makes something
an essential or accidental (in Heinaman™s language: an “intrinsic” or “instrumental”) good according
to Plato is quite different from his.
202 Aiming at Virtue in Plato
Glaucon says that he wants to hear Socrates praise justice in the same way he is
about to praise injustice. He wants Socrates to praise justice by showing that it
is good in itself. Therefore, the reasons Glaucon goes on to give for preferring
injustice to justice are reasons why injustice is supposed to be good “in itself,” in
the sense of the phrase that Plato has in mind.25
Heinaman™s discussion of this passage is part of his argument that Glau-
con and Adeimantus aim to show that injustice is a good in itself. I am
happy to accept this as part of their goal, but Heinaman™s reading of the
passage understands it as requiring that it be their sole aim. For, if Glaucon
is offering an argument that is the same as the one he wants Socrates to offer,
then, since he wants Socrates to praise justice in itself, whatever considera-
tions Glaucon brings to bear in arguing the case for injustice must, on pain
of contradiction, be aimed at establishing that it is valuable in itself.26 By
reading the passage as Heinaman does, one must go on to say that everything
that Glaucon appeals to in his argument “ all of the advantages accrued to
the person acting unjustly with Gyges™ ring “ is part of what makes injustice
good in itself according to the many.27 But I think that, if even the material
goods gained by an unjust act do not count as “consequences” of the act,
and thus are not, as the saying goes, to be counted among “the wages of
sin,” but instead are to count as part of injustice itself, then we, and Plato,
are entirely at sea about what a good in itself is.28
Fortunately, 358d3“6 does not need to be interpreted so as to necessitate
this conclusion. When Glaucon says that he will show Socrates how he
wants him to praise justice by lengthening out his own praise of injustice,
he need not mean that Socrates is supposed to give a speech that cites the
very same sort of considerations. What he could also mean, and what makes
more sense, is that, by painstakingly examining all of the consequences of
injustice and justice, as well as the evil in itself of the pain of acting justly
when one could get away with injustice and the good in itself of the pleasure
25 Heinaman (2002), 318“19. Reeve (1988) 25, 28 also uses this methodological remark to conclude that
reputations and rewards are consequences of justice itself according to Glaucon, but not consequences
in the same sense of injustice itself. Such an asymmetry seems to me an excellent reason to rethink
the interpretation of 358d5“7 that appears to necessitate it.
26 Part of Heinaman™s larger goal is to establish that Plato allows some causal consequences of a thing
to count towards whether it is a good in itself or not. As I have said, this issue is of secondary
importance to me, since I do not think that it helps us to explain Plato™s notion of a good in itself.
Something is essentially good for Plato if it either is a bene¬cial state of soul or causes a bene¬cial
state of soul “ so both justice as a property of a soul and just actions will count as essential goods.
27 Heinaman (2002), 319“20.
28 Heinaman includes such causal consequences as part of what makes something (in this case, injustice)
a putative “intrinsic” good, which leads him ([2002], 327) to doubt that any interpretation could
make consistent sense of which causal consequences are “consequences” of a thing and which part
of the thing itself.
The bene¬ts of injustice 203
of acting unjustly when one can get away with it,29 he will leave Socrates
no conceptual room to maneuver other than to praise justice itself by itself,
if he can. Thus the way Glaucon shows Socrates how to praise justice is not
by providing an example to imitate, but by excluding all other options for
praise other than praising justice for itself, that is, for its “effect” on the soul.
Therefore in his and Adeimantus™ speeches we will see them referring both
to the supposed good consequences of unjust action (understood on my
interpretation as bene¬ts to body, possessions, and reputation) and to the
alleged essential positive value of acting unjustly (the unfettered satisfaction
of appetite).
After all, what is distinctive about Glaucon™s and Adeimantus™ argument
is not that they establish some points that are in dispute by ordinary people.
The advantage of stealing when one can get away with it and the pleasure of
doing whatever you feel like are both taken to be obvious. What is special
about their argument is that they cover the case so thoroughly and in such
detail that they leave the question of the value of justice by itself entirely
isolated and “stripped” (see 361c4). But this is the result of the cumulative
effect of all of their examples, not of any controversial claim about what is
to a person™s advantage in some particular instance. And this is how they
“show” Socrates how they want to see justice praised. Finally, this should
not be surprising, for they are claiming to be presenting the view of the
many; the phenomena they appeal to are plain for all to see.

6 .4 t he origin of j u stic e accord i ng to th e m a n y
Glaucon sets out to show three things. He lists his ¬rst topic as explaining
“what sort of thing they [the many] say that justice is and from where it
arises” (dikaios…nhn o³on e²na© fasin kaª ‚qen gegon”nai, 358c1“2). It is
striking that, every time Glaucon discusses what justice is, he conjoins it
with a question about its origin.30 By contrast, when in propria persona he
simply asks Socrates to tell him what justice is and what capacity it has
when it is present in the soul (358b4“7), he does not ask about its origin.
The many™s argument, however, includes an account both of what justice
is and of its origin.
Glaucon maintains (on behalf of the many) that while everyone knows
that doing injustice is good and suffering injustice is bad, people discov-
ered that the badness of suffering injustice without recourse exceeded the

29 Both to be explained, as above, in terms of satisfaction or frustration of desire.
30 In addition to the quoted line, see 358e2, 359a5, 359b4“5.
204 Aiming at Virtue in Plato
goodness of doing injustice with impunity. Since most people are too weak
to act unjustly without running signi¬cant risk of also suffering injustice,
they decide it pro¬ts them to make agreements with one another neither
to do nor to suffer injustice. To abide by such agreements is to be what is
called lawful and just. Given this account of the genesis of justice we can
see that the “being” of justice is an intermediate. Agreeing neither to do
nor to suffer injustice is worse than doing injustice with impunity, but it is
better than suffering injustice without recourse (359a). Although Glaucon
says that doing injustice is “by nature good” and suffering it “by nature
bad” (358e4“5), he does not say how or why this is so. He clearly relies on
ordinary ideas, available to the many, about how this works “ some details
of which will be forthcoming in the story of the ring of Gyges. But there is
no hint here that the natural goodness of injustice is to be restricted to its
essential goodness as opposed to the goods that result from it. What is nat-
urally good about acting unjustly without punishment is that one bene¬ts
from its consequences (that is, one accrues material and bodily bene¬ts “
wealth, good food, and so on) and that one gains bene¬t to one™s soul by
satisfying, not frustrating, one™s desires.
The simplest and most straightforward way of reading Glaucon™s ¬rst
point, then, is to see him offering a conventionalist account of justice:
justice is a matter of agreements that people make between one another
neither to do nor to suffer injustice. This way of putting the point, however,
seems to run into an immediate obstacle.31 If Glaucon were really giving
a conventionalist account of the origin of justice it would make no sense
to say that “doing injustice” is by nature good. Doing injustice would be a
matter of breaking the laws and agreements that people form with one
another; prior to the formation of such covenants, in the state of nature,
there would be no such thing as justice or injustice, only people acting
in their own self-interest. Likewise, independently of agreements between
people, there is no such thing as justice. For this reason, to say, as Glaucon
does (359a2), that people form agreements with one another “neither to do
nor to suffer injustice” makes no sense at all, since it is the agreements that
themselves constitute what is just and unjust in the ¬rst place.
In response we can apply the point made in chapter ¬ve about the inter-
pretation of Thrasymachus™ “immoralist speech” (5.5). Glaucon is explicit
that he is providing an account of the nature and origin of justice, not
of justice and injustice. If, like Thrasymachus, Glaucon (as spokesperson
for the many) believes that there is injustice in the state of nature prior

31 Recognized by Gauthier (1986), 309ff.
The bene¬ts of injustice 205
to the “invention” of justice, then we would expect that Glaucon would
speak of people “doing injustice” to one another even though there is, as
yet, no such thing as justice. So criticisms, by Gauthier and others, who
notice that Glaucon speaks of people “doing and suffering injustice” prior
to the formation of the agreements which constitute justice, would be off
the mark.32 This will enable us to appreciate a deep connection between
Thrasymachus™ and Glaucon™s accounts, as we should expect, given that
Glaucon is explicitly restating Thrasymachus™ position.
Why does Glaucon include this account of the origin and being of justice?
Why not skip right to the ring of Gyges story and the arguments that jus-
tice is practiced only as something necessary and that the unjust person
is happier and better off than the just person? The latter two arguments
focus on the question, “Why ought a person to act justly/morally?”,33 where
justice is understood to consist in just actions as ordinarily conceived. We
might think that these arguments themselves would be suf¬cient to raise
the important challenge to justice. What the account of the origin and
nature of justice shows, however, is that part of the many™s view is that
justice is not, after all, something “real”; it is simply the product of an
agreement or convention.34 What is real are bene¬ts to the self, and the self
can be bene¬ted in three ways: in its soul, its body, and its possessions. As
we saw in the Gorgias, the conception of harm and bene¬t that emerges for
Callicles consists of ful¬lling appetites. The advantage of such a view is that
one can determine what bene¬ts oneself in a relatively straightforward way:
simply determine what one has an appetite for. If a person is strong enough,
he can get it, and thus obtain the good in itself of appetite satisfaction as well
as the good consequences of bodily and material bene¬ts. This goes hand in
hand with the idea of justice as something that is not really real, something
that does not have an objective, independent, nature of its own. Unlike the
next couple of arguments, the conventionalist account appears to supply an
answer to Socrates™ “What is F?” question: one can determine whether an
act is just or not by looking to see whether it is forbidden or allowed by the
32 This would make moves like Irwin™s (1995), 183 unnecessary: “Glaucon does not claim that just and
unjust action are impossible outside a state . . . Glaucon claims that the fact that the laws require just
action explains how justice came into being (359c5); this is not because there was no such thing as
just action before there was a law, but because the existence of law promotes the growth of justice.”
Although this interpretation might successfully counter Gauthier™s criticism, it does not seem to me
to be very plausible that Glaucon would then be giving an account of the “origin and nature” of
33 That is, why ought one to refrain from “doing wrong”?
Glaucon says at 359a that he will supply the g”nesiv kaª oÉs©a of justice. In the middle books of the
Republic we shall learn that the real oÉs©a of justice is the Form of Justice; and Forms, of course,
have no genesis.
206 Aiming at Virtue in Plato
agreements made.35 If we ask why a person should act justly the upcoming
arguments claim that in fact, if a person is strong, he would not and should
not. But the conventionalist position adds a further point that is indepen-
dent of its being able to resolve disputes: when a person violates justice,
he is not violating some objective standard of true goodness; he is simply
going against the agreements of people. Slightly further on Glaucon says:
For they [the many] will say that in reality [t¤€ Ànti] the unjust man, insofar as he is
pursuing a matter of the truth [Œte –pithde…onta prŽgma ˆlhqe©av –c»menon],
and not living by opinion [oÉ pr¼v d»xan z¤nta], desires not to seem to be unjust
but to be . . . (362a4“7)
This is a striking passage that draws on the resources of the convention-
alist argument. The unjust person is not only unabashedly pursuing his
self-interest, but also acting in a manner consonant with reality and the
truth. To act justly, by contrast, not only does not bene¬t one, but it is to
act under a false idol, to be living by mere opinion.36 If conventionalism
were false, and the ordinary conception of justice were objectively true,
then, the unjust person would not be in as strong a position, insofar as
he could not simply dismiss justice as an arti¬cial ¬ction, as something
unreal. He might still, of course, query why he ought to act in accordance
with justice “ how would it be in his interest to act justly (or be just)?
But he could no longer oppose the reality of the bene¬t accrued to him
by injustice with the merely conventional nature of the justice he violates.
Rather he would have to concede that he acts against something that is real
and true. In the Euthyphro (10d), Socrates and Euthyphro agree that what
is pious and right has a nature that is independent of anyone™s (in particular
of the gods™) beliefs about or attitudes towards it. In the Phaedo, and later
in the Republic, justice will turn out to be an independently existing eternal
entity: the Form of Justice. These three dialogues have in common the
belief that justice is something objective and without an origin: what is just
is just because of properties it has and not because of what some people or
gods agree to, or think about it.

6.5 th e b enef i ts of i n j ust i ce
My concern in looking at the second two parts of Glaucon™s argument “
intended to establish that those who “practice” (o¬ –pithde…ontev) justice

35 This picks up on the discussion of conventionalism in Book 1; see 5.2.
36 Recall chapter ¬ve where we saw that Thrasymachus calls Socrates foolish and simple-minded. The
foolishness at issue is that of an adult who believes in Santa Claus or the tooth fairy.
The bene¬ts of injustice 207
practice it “unwillingly” and that this is reasonable because the unjust life is
much better than the just one “ is to add to the evidence that the distinction
between an essential good and an accidental good is based on one between
goods that bene¬t the soul and those that bene¬t the body or one™s material
possessions. Thus, when Glaucon argues that, given the ring of Gyges, the
just and unjust person would behave in the same way, he is establishing that
acting unjustly provides bene¬ts to soul, body, and possessions. Injustice
turns out to be a good in itself and for its consequences.
Recall that Glaucon™s goal is for Socrates to consider justice itself by
itself. He will show him how he wants justice praised by stripping it of all
its consequences. To strip acting justly of all its consequences is to consider
its effect on the soul alone. Glaucon™s claim, on behalf of the many, is that,
limited to this dimension of assessment, justice is painful and “harsh.” But
harsh in what sense? Take the example of Gyges™ ring. If I have the ring, and
so will suffer no harm to my body or to my possessions by acting unjustly
(that is, I run no risk of punishment), and I desire something that it is
not just for me to take, it will be painful for me to forgo it. Furthermore,
the pain of acting justly with the ring is precisely not bodily pain: it is the
pain of frustrated appetite “ not getting something you want. Acting justly
is thus in this sense bad; it is not simply indifferent. For the many, there
is no other interpretation of good or bad for the soul; they have no other
conception of the “health” or well-being of the soul. By parallel reasoning,
acting unjustly with the ring “ the obvious material and bodily advantages
aside “ is enjoyable insofar as it results in appetitive satisfaction as opposed
to frustration.37 To this extent, then, Glaucon holds that injustice is a good
in itself: insofar as injustice results in “enjoyment” (and not in frustration),
it is desirable in itself.
When Glaucon sets up the ring example, his language is quite speci¬c.
He says that he will give the power to the just and the unjust person “to do
whatever he would wish” (poie±n ‚ti ‹n bo…lhtai) (359c1“3) and then see
where desire (–piqum©a) leads each one. The appeal made for the value of
injustice by itself is centered on the notion of doing “whatever one wants”
“ a person™s desire not being frustrated. When Glaucon compares the two
types, assuming that each has a ring (360b4 ff.), he argues that no one would
be able to resist taking things that are not rightly his (the good consequences
of injustice), nor would they be able to resist the opportunities that the
ring affords its bearer of doing “whatever he wishes”: he can take from

37 We shall see that Plato takes this seriously and addresses the alleged value of satisfying appetite in
his discussion of pleasure in Book 9. See 8.5.
208 Aiming at Virtue in Plato
the market fearlessly “whatever he wishes” (‚ti bo…loito, 360b8), have
sex “with anyone he wishes” (‚twƒ bo…loito, 360c1) and kill or release
from bonds “whomever he wishes” (oÌstinav bo…loito, 360c2).38 Finally,
anyone who didn™t “ever wish” (pote –q”loi, 360d4) to do any injustice
would be thought a fool. The repeated emphasis on the satisfaction of
desires shows the essential goodness of injustice and the essential badness
of justice: satisfying one™s desire is “enjoyment” (ca©rein), to act justly is to
frustrate this desire, a state which is painful and harsh. As regards pleasure
and pain, then, injustice is good and justice bad for the soul.
The ¬nal stage of Glaucon™s argument shows how he wants Socrates to
argue. In order to judge properly between the just and unjust lives to deter-
mine which is better, he must separate the most just and the most unjust
person and make each perfect in his own “practice” (–pitždeuma). When
Glaucon “strips” the just man of everything except his justice, he deprives
him of goods of every type. The most extreme case is to be just without
seeming to be just. Such a person is deprived of the ordinary good conse-
quences of justice “ ordinary pay for work, security, good reputation, and
so on. He will gain, of course, no material bene¬ts, and then he will suffer
the most outrageous bodily harms (361e). Moreover, given the evidence
of the Gyges part of the argument, and its claims about human psychology,
there will also be no bene¬t to be had from one™s justice by itself insofar as
one™s desires are entirely and completely frustrated.
The completely unjust person, by contrast, not only pro¬ts and gains the
rewards from the fact that he has no dif¬culty doing injustice (362b4“5),
that is, he reaps material and bodily advantages (wealth, physical safety,
not being impaled, and so on), but he also has the good for his soul of
desire satisfaction. Glaucon again releases a ¬‚urry of “whatever he wishes”
(‹n bo…lhtai, 362b3 (twice), 362c4; ‹n –q”lh, 362b4). Like the position
of Callicles, Glaucon™s unjust man gains the good in itself of unrestrained
desire satisfaction. Since this is the only candidate considered as a good for
the soul, it is entirely unconstrained by any notion of a moral good. I argued
above that this was part of the point of the conventionalist argument: justice
is not a real, objective thing. In a passage from this section that I quoted
above (362a4“7), when the completely unjust person pursues his desires,
whatever they are for and wherever they lead, he not only has a good in
itself, but his actions have reality and truth on their side, as opposed to the
allegedly ¬ctional idea that the desire frustration involved in acting justly
has some real value of its own.

38 This is foreshadowed in Republic 1 when Socrates denies that a person is better off being unjust even
if he can do “whatever he wants” (Œ bo…letai) unhindered (345a4).
The bene¬ts of injustice 209
Glaucon™s argument is, then, as follows. Something is good either essen-
tially, or accidentally for its consequences, or both. If it is good essentially
that means it provides some good for the soul, and the only sort of good
for the soul recognized by the many is enjoyment or appetite satisfaction.
The many recognize, however, that sometimes appetite satisfaction must
be curtailed in order to acquire goods of the body or material possessions: a
person must go to work to make money, must submit to medical treatment
to be healthy. Similarly, in the ordinary world, a person must submit to
justice, although it is painful insofar as it is not what one would “want” to
do, as part of a societal agreement in order not to suffer an even greater
deprivation of goods by suffering injustice without recourse.
The point of Glaucon™s argument is that justice itself by itself is worth
absolutely nothing. It gains no material goods, no bene¬t for the body,
and no bene¬t for the soul. None of these claims is taken by itself to be
controversial: the conception of harm and bene¬t at work is the many™s. The
harm of being punched or of not getting what one wants and the bene¬t
of money or of desire satisfaction are supposed to be plain to everyone.
Glaucon™s argument can seem to leave no room for some other conception
of harm and bene¬t according to which justice is a good in itself after
all. Glaucon has shown Socrates how he wants him to argue, then, not by
himself arguing merely that injustice is a good in itself, but by leaving,
apparently, no conceptual space to advocate for justice. All of the goods,
of all types, accrue to the perfectly unjust person. Since the just person is
left with no goods whatsoever, there is no reason to be just, only to seem
It may be true that Adeimantus™ argument does not add much of philo-
sophical interest to a contemporary reader.39 What Adeimantus™ argument
emphasizes more vividly for my purposes, however, is that the good con-
sequences of being just are the same as the good consequences of seeming
just.40 Therefore, as Glaucon has already argued, if one does not seem just,
but only is just, there will be no good consequences. This is clearly not the
case for injustice. If one seems unjust, he suffers bad consequences, but if
a person can manage to be unjust without seeming unjust (either, say, by

39 Stokes (1987) argues that Adeimantus holds a distinct position which abandons the premise that
justice is a virtue; for discussion see Inwood (1987). Annas (1981), 65 remarks that the length of
his speech is not warranted given the signi¬cance of its impact on the argument. I think that this
impression is exacerbated by our distance from ancient Greek poetry and religion. Since the religious
beliefs and worries associated with them are rather remote from our own, it certainly adds to the
impression that Adeimantus is not contributing anything new to the argument.
40 We should note, however, that this does not deprive truly being just of having, in ordinary cir-
cumstances, the same good consequences; see 366a. After 612b in Book 10, Socrates takes up these
210 Aiming at Virtue in Plato
having a Gyges™ ring, or by diligence and effort [365c“d]) he will reap the
good consequences of injustice as well as the satisfaction of appetite ful¬ll-
ment. By contrast it is painful to be just; since it involves the frustration of
a natural desire for pleonexia,41 being just and acting justly have no essential
value either; so, there is really no reason left to be just. Adeimantus seconds
these points. At 364a ff., he adds that unjust actions are generally more
pro¬table than just ones, gaining wealth for the agent, “and other types
of power.” These are clearly consequential goods gained from unpunished
unjust actions: goods of the body and material possessions. With such
wealth, one can buy exemption from the gods from any punishment they
might otherwise in¬‚ict upon the unjust. It is clear that justice bears the con-
sequence of avoiding divine punishment (366a),42 but the unjust person
can obtain this through bribery, plus have all of the material advantages,
desire satisfaction, and physical well-being that comes from injustice.
The most important aspect of Adeimantus™ argument for my purposes
is his clear emphasis on the effect of acting justly on the soul. I have been
arguing that Glaucon™s and Adeimantus™ speeches work to isolate the goods
of injustice and justice, both those that are essentially and those that are
accidentally gained, and leave Socrates with the apparently impossible task
of explaining how being just could be a good in any sense. As mentioned,
in this way they “show” Socrates how he should argue; namely, by leaving
him no conceptual space to argue in any other way. Consider the following
[Among everyone past and present] no one has ever disparaged injustice or com-
mended justice otherwise than in respect of the repute, the honors, and the gifts
that arise from each [d»xav te kaª tim‡v kaª dwre‡v t‡v ˆp¬ aÉt¤n gignom”nav].
But what each one of them is in itself, by its own power, when it is within the soul
of the one who has it and escapes the notice of both gods and men, no one has
ever adequately set out in poetry or prose “ namely, the argument that the one
[injustice] is the greatest of all evils that the soul contains in itself, while justice is
the greatest good. (366e3“367a1)
Here the contrast is clearly between the consequences of just and unjust
action, on the one hand, and their effect on the soul on the other. As I have
41 See 359c.
42 Presumably even for the just person on the rack “ Glaucon™s extreme example “ for the gods will
know who is really just and who unjust. Now the unjust person will get around this by successfully
bribing the gods with his unjust gains. Adeimantus does not consider a more diabolical scenario
whereby an unjust person would bribe the gods to punish a just person in the afterlife, and so, in
effect, make the just person “seem” unjust even after death. Plato has a clear concern to rebut this
conception of the gods and to disrupt the role that traditional forms of poetry play in promulgating
it. See 7.2“3 for some discussion.
The bene¬ts of injustice 211
been arguing, this is precisely the contrast between being a good accidentally
and being a good essentially. The ¬rst sentence claims that no one has
criticized injustice in itself: that is, criticized its effect on the soul. On my
account, since the many think of acting unjustly as satisfying the natural
appetitive desire for pleonexia, and of acting justly as frustrating it, of course
they have never condemned the one and praised the other in themselves.
But, as the argument has shown, once a person has conceded this much,
it is clear that the consequences of appearing just are the same as the
consequences of being just (that is, the bene¬ts accrued to one™s bodily well-
being and to the well-being of one™s possessions), and the consequences of
really being unjust while seeming just outstrip by far the good consequences
of being just. On top of that the agent gains the bene¬t of the desire
satisfaction involved in acting unjustly. Thus the challenge to Socrates
clearly emerges: explain how justice has some effect by itself on the soul
that makes it worth being just; that it is the greatest good for the soul to
have justice. How this could be the case is, I submit, entirely unclear at this
point in the argument. Justice must be of some value to the soul that has
not yet been considered.
In the following two chapters I shall consider Plato™s central argument
that justice is a good for the soul. It is the same type of argument for SV
that we have seen throughout the dialogues: the soul is a locus of harm
and bene¬t independently of the body and is of supreme importance.
By de¬nition, an excellent soul is virtuous, and a terrible, ruined, soul is
vicious. What causes a soul to become excellent, or vicious, is the repeated
performance of the corresponding type of actions. This much is clear in
the Gorgias. What the argument of the Republic adds are crucial details.
What does virtue do to the soul that makes virtue a good thing? The
tripartite account of soul will provide a much more elaborate answer than
the “account” Socrates provides in the Gorgias. Furthermore, in order for
habituation to be effective, there must be some way of determining correctly


. 6
( 9)