Vous êtes sur la page 1sur 24

UnvaveIIing TIvasnacIus' Avgunenls in TIe BepuIIic

AulIov|s) F. F. NicIoIson
Souvce FIvonesis, VoI. 19, No. 3 |1974), pp. 210-232
SlaIIe UBL http://www.jstor.org/stable/4181941 .
Accessed 24/04/2013 1257
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
BRILL is collaborating with JSTOR to digitize, preserve and extend access to Phronesis.
This content downloaded from on Wed, 24 Apr 2013 12:57:46 PM
All use subject to JSTOR Terms and Conditions
Unravelling Thraymachus' Arguments
in "The Republic"'
There has been much discussion recently of the encounter between
Thrasymachus and Socrates in Book I of Plato's Republic.
I am not here concerned with whether Thrasymachus' arguments
and Socrates' replies are valid, interesting and important though that
problem is, but with the fundamental problem of deciding what
exactly Thrasymachus is saying about so 81x%ctov, justice. Clearly,
this is a necessary preliminary to the raising of any other question
about Thrasymachus' arguments. Such an investigation may also
contribute to our understanding of The Republic as a whole: given
the structure of the dialogue, to know what Plato dissents from can
provide valuable clues to what he assents to.
In brief, my aim is to support Professor Kerferd's interpretation that
Thrasymachus' doctrine is that "justice is the advantage of another".2
Kerferd's view has not been generally accepted. I shall argue that the
standard view found in most commentaries, that Thrasymachus
thinks "justice is the advantage of the ruler(s)", is open to major
objections, and that Kerferd's interpretation is to be preferred be-
cause it avoids these objections. I shall argue further that some of
the implications of Kerferd's interpretation provide additional
I am grateful to John Gould, W. H. Greenleaf, Dale Hall, G. B. Kerferd,
Hugh Price, and J. C. Rees for their comments.
G. B. Kerferd "The Doctrine of Thrasymachus in Plato's Republic" Durham
University Journal n.s. 9 (1947) 19-27.
Since I shall be using the single word "advantage" throughout, it must be
stressed that Thrasymachus uses two words, speaking of sY -oi xpe'rrovo
kubtupkpov (e.g.
but of
X6-rptov &Tayokv (e.g.
C). "Advantage"
"interest" are the common translations of
although "good" has very
occasionally been used; "good" is the usual translation of &ya&6v. Now, in both
cases "advantage" seems to me the word best suited to Thrasymachus' meaning
and to his nuances, ambiguities, and word-plays, despite the risk of making his
position more uniform in appearance than it is. I shall therefore write of "the
advantage (vb
of the ruler" and of "the advantage (&yc&6v) of an-
This content downloaded from on Wed, 24 Apr 2013 12:57:46 PM
All use subject to JSTOR Terms and Conditions
reasons for thinking that it comes nearer the mark than any other.
With Kerferd as guide, Thrasymachus' arguments can be exhibited
as both consistent and coherent, once it is understood that he is
concerned to characterise and not define, that he prefers to do this
by his own method of argument, and that he presents his case in an
unexpected order.
According to most commentators in English, Thrasymachus sees
justice as a political relationship between rulers and subjects, and
asserts that justice is the advantage of the stronger, i.e., of the ruler(s).3
On the face of it, this means that Thrasymachus is only discussing
justice in the political sphere, since rulers' enactments cover a limited
range of the conduct which can be called just or unjust. However far
one stretched his terms like "rule"
or "law"
so that
they included not only legislation but also social regulation such as is
attributed to Lycurgus, they still would not encompass all moral
conduct. This interpretation, that Thrasymachus means that justice
is the advantage of the ruler(s), is open to two objections: (1) it seems
not to fit properly into the development of the argument of The
Republic; and (2) it is not a doctrine which Thrasymachus holds
(1) It is, to say the least, odd and surprising to find the discussion
* E.g. R. L. Nettleship Lectures on the Republic of Plato (London 1901)
E. Barker Greek Political Theory. Plato and his Predecessors (London 1918)
p. 180; P. Shorey What Plato Said (Chicago 1933) p. 210; G. M. A. Grube
Plato's Thought (London 1933) pp. 265-6; A. D. Lindsay The Republic of Plato
(London 1935) p. xxxi; M. B. Foster Masters of Political Thought.
Volume I Plato to Machiavelli (London 1942) p. 47; T. A. Sinclair A History of
Greek Political Thought (London 1951) p. 74; I. M. Crombie An Examination
of Plato's Doctrines. Volume I Plato on Man and Society (London 1962) p. 81;
G. F. Hourani "Thrasymachus' Definition of Justice in Plato's Republic"
Phronesis 7 (1962) 110-20; L. Strauss "Plato" in Strauss and J. Cropsey (eds)
History of Political Philosophy (Chicago 1963) pp. 11-12; R. S. Brumbaugh
Plato for the Modern Age (New York 1964) pp. 86-7; R. C. Cross and A.D.
Woozley Plato's Republic. A Philosophical Commentary (London 1964) pp.
2441; A. Bloom The Republic of Plato. Translated with Notes and an Interpretive
Essay (New York 1968) pp. 326-8; W. K. C. Guthrie A History of Greek Philo-
sophy. Volume III The Fifth Century Enlightenment (Cambridge 1969) pp. 88-90;
and K. Dorter "Socrates' Refutation of Thrasymachus and Treatment of
Virtue" Philosophy & Rhetoric 7 (1974) 25-46, especially section II.
This content downloaded from on Wed, 24 Apr 2013 12:57:46 PM
All use subject to JSTOR Terms and Conditions
of justice narrowed down to politics at this point. Very early on
(330 E ff) justice is spoken of as a general quality relevant to all
conduct and not simply to political activity; and the discussion con-
tinues in the same vein after the Thrasymachus debate until Socrates
makes a point of diverting it into the political channel (368 D-369 A).
It is true that the specifically political aspect is first raised by Thrasy-
machus, yet considering the dialogue's structure, it would be strange
if he were concerned exclusively with politics. Since the
subject of
The Republic is justice at large, it is natural to expect Thrasymachus,
like Cephalus, Polemarchus, Glaucon, and Adeimantus, to be intro-
duced in order to present a contrasting account of it. Moreover,
it can be argued that Thrasymachus is expressing a version of the
traditional values of the ordinary Greek. Adkins has elaborated this
kind of view, and writes for instance, "scratch Thrasymachus and
you find King Agamemnon", and "Thrasymachus is merely drawing
out what appear to him to be the logical consequences of Greek values".'
But as the traditional Greek conception of justice did not confine
itself to political obedience,' once more one would not expect Thrasy-
machus to limit himself to justice in a political context. In fact, when
he comes to give examples of just and unjust actions (343 D-344 B),
only some concern one's relations with rulers (e.g. paying taxes),
whilst others concern one's relations with fellow-subjects (e.g. business
dealings). And he sees the crucial issue as being which of the two lives,
the just or the unjust man's is superior (e.g. 347 E): meaning by
"life" a man's total imputable conduct not merely his political activity.
More will be said about this aspect later. I leave it now for the second
objection to the standard interpretation of Thrasymachus.
(2) There is apparently a major inconsistency in Thrasymachus'
account of justice. Initially, he says that it is the advantage of the
stronger, i.e. the ruler(s) (338 C-341 A), but later that it is the advan-
tage of another (343 C-344 C). But to assert both propositions can
lead to contradiction, because only for the weaker is "the
' A. W. H. Adkins Merit and Responsibility (Oxford 1960) p. 238, and Moral
Values and Political Behaviour in Ancient Greece (London 1972)
119. Cf.
Bloom, op. cit. p. 336: Thrasymachus "gives voice to common opinions which
are usually kept quiet."
' As Barker puts it, 8LxaCXoa7 is an ethical as well as a legal term, and "has
the overtones and suggestions of our own word 'righteousness"'. The Politics
of Aristotle. Translated with an Introduction Notes and Appendices (Oxford 1946)
p. Lxx.
This content downloaded from on Wed, 24 Apr 2013 12:57:46 PM
All use subject to JSTOR Terms and Conditions
of the stronger" synonymous with "the advantage of another".
For instance, if a tyrant levies a tax to pay for his orgies, is he acting
justly or unjustly, according to Thrasymachus? Justly, if justice is
the ruler's advantage, unjustly if justice is another's advantage.
Those commentators who have paid attention to this problem have
reacted in a vanety of ways. Some accept that Thrasymachus is
inconsistent, offering a number of different explanations. Jowett
depicts Thrasymachus as a vain clown and "a mere child in argument",
implying that consistency was beyond him.6 Sidgwick does not think
are to be taken too seriously, believing he
defines justice "as a rhetorician, possessing the cultivated omniscience
to which ancient rhetoricians commonly laid claim, and so able to
knock off a definition of Justice, as of anything else. That 'Justice is
the interest of the stronger' is a plausible cynical paradox which a
cultivated person might naturally and prosperously maintain in a
casual conversation".7 Cross and Woozley conclude that Thrasymachus
"has advanced two different criteria of justice ... without appreciating
that they do not necessarily coincide".8 And Maguire contends that
only some of the arguments Thrasymachus states are his own, others
not consistent with them being falsely attributed to him by Plato.9
However, we should not accept any explanation of Thrasymachus'
inconsistency before asking whether his arguments will stand being
treated seriously, and whether they can be taken as a whole and
rendered internally consistent. Accordingly, I shall next consider three
different interpretations (without claiming that they exhaust the
possibilities) each alleging that Thrasymachus' two statements can be
made consistent, and this second objection met.
(i) It might be argued that "the advantage of another" is simply a
synonymous expression for "the advantage of the stronger, i.e. the
ruler(s)", for at 343 C Thrasymachus tells Socrates "you don't know
that justice and the just are literally the other fellow's good
the ad-
vantage of the stronger and the ruler, but a detriment that is all his
own of the subject who obeys and serves..."'1' However, it is clear
B. Jowett The Dialogues of Plato Translated into English with Analysis and
Introductions (Oxford 1871) vol. II p. 6.
H. Sidgwick The Philosophy of Kant and Other Lectures (London 1905) p. 370.
Op.cit. p. 41.
' J. P. Maguire "Thrasymachus ... or Plato?" Phronesis 16 (1971) 142-63.
The Loeb translation, by P. Shorey (London 1930). This is the translation
used throughout.
This content downloaded from on Wed, 24 Apr 2013 12:57:46 PM
All use subject to JSTOR Terms and Conditions
from the speech taken as a whole (i.e. to 344 C) that Thrasymachus
is not talking only of rulers and ruling but of all men and all social
relations. His examples of just and unjust acts, as was noted, include
other relations besides those between ruler and subject. Therefore
"another" cannot be equated with "stronger, i.e. ruler(s)", because
Thrasymachus' examples show that "another" can be a fellow-subject
as well as a ruler.
(ii) If "stronger, i.e. ruler(s)" and "another" are not synonymous, a
contradiction between them can be avoided by taking one as the pri-
mary idea and subordinating the other to it. The second interpretation
takes as primary Thrasymachus' first account of justice ("advantage
of the stronger, i.e. ruler(s)"), and subordinates to it his second
account ("advantage of another"). Barker, for instance, writes:
"if justice... consists in whatever is for the ruler's interest, it follows
that, for everybody other than the ruler, justice may be further de-
fined, according to a popular definition, as 'another's good'. To be
'just', in the popular sense, is to be a means to the satisfaction of the
ruler: to be 'unjust', in the popular sense, is to act for the satisfaction
of oneself".1' But this is to gain consistency at a heavy price, for it
entails that Thrasymachus applies his doctrine only to some people,
namely, the subjects. Two main difficulties impede our acceptance of
that conclusion.'2
(a) Thrasymachus gives an example, "business dealings", in which
what the just man does is to the advantage of someone other than the
ruler(s) (343 D). The example cannot be ignored: yet it is hard to
account for it within this line of interpretation without introducing
some further hypothesis, with difficulties of its own. For instance,
it might be argued that although in private transactions what is to the
advantage of another is not to the advantage of the ruler(s) directly,
indirectly it is. It would not be necessary to suggest, with Lee, that like
Marx, Thrasymachus thinks that all morality is somehow determined
Op. cit. p. 180. Cf. Cross and Woozley op. cit. pp. 3941.
One might add, as a further difficulty, that it is an odd doctrine of justice
which applies only to some people, that this is a point which Socrates should
have seized upon, and that since he does not it cannot be what Thrasymachus
means. But we have no grounds for assuming from Socrates' absence ot reaction
that Thrasymachus has not made a point. If we did, it would wreak havoc
with much of the interpretation of Thrasymachus and other thinkers treated in
Plato, including that offered in this paper.
This content downloaded from on Wed, 24 Apr 2013 12:57:46 PM
All use subject to JSTOR Terms and Conditions
by the ruler(s) to their advantage ;13 it might simply be said that Thra-
symachus believes that the observance of moral rules, however
have been determined, always works out to the advantage of the
ruler(s). In this way, every just act is
to the
advantage of
the ruler(s), though it may also be to the advantage of someone else.14
But the difficulty with this interpretation is to see how the text can
support it. It involves sophisticated and contentious ideas, yet Thra-
symachus provides no special explanation of them such as he gives
with other unusual ideas, for example that of the ruler who makes no
mistakes. On the contrary, what he does say runs counter to this
interpretation, for whereas many moral rules are the same under
different constitutions, what Thrasymachus stresses is the diversity
of the ruler(s)' enactments under different constitutions. His emphasis
is upon the limited number of rules which vary correspondingly with
the different forms of government (338 D, E). Even if we accepted that
Thrasymachus means, although he does not say, that all just acts
are sooner or later to the advantage of the ruler(s), the difficulty
remains that an act could then be to the advantage both of another
subject, directly, and of the ruler(s), indirectly; whereas Thrasy-
machus conceives of advantage in terms of private goods which can
be enjoyed only by one person exclusively (343 D-344 B).
(b) Even if it is felt that the first obstacle can be overcome, the
problem remains of what justice is for the ruler himself, if Thrasy-
machus is only talking about "everybody other than the ruler"
(Barker, above). Barker's interpretation implies that if justice is the
ruler's advantage, then for the ruler, justice must consist in pursuing
his own advantage. As Kerferd points out, "nowhere does Thrasy-
machus say, hint or imply that this was his view."15 In fact, Thrasy-
machus says the reverse: that the ruler who pursues his own advantage
is unjust (344 A-C). And this last passage also shows that Thrasy-
machus is not thinking solely of what is just for subjects, but for rulers
(iii) This is the converse of the last interpretation; it takes as pri-
mary Thrasymachus' second account of justice ("advantage of an-
other"), and subordinates to it his first account ("advantage of the
stronger, i.e. ruler(s)"). This is in effect what Kerferd does, and,
H. D. P. Lee Plato. The Republic Translated with an Introduction (Harmonds-
worth 1955)
This view was put to me by Dale Hall.
Op. cit. p. 22.
This content downloaded from on Wed, 24 Apr 2013 12:57:46 PM
All use subject to JSTOR Terms and Conditions
with some variations, Sparshott16 and Herderson.17 Kerferd argues
that 343 C to 344 C, where Thrasymachus says justice is the advantage
of another, is "the fullest account of his views on justice and injustice
which we have" and "has hardly received the attention which it merits,
and ... has often been dismissed as hardly worthy of serious study".118
In this passage, Thrasymachus has a single conception of justice
("another's advantage") and injustice ("one's own advantage").
Until this passage, he has been talking about what is just for the
subject, namely the advantage of the stronger, the ruler. Yet that is
but one instance of the general doctrine that justice is the advantage
of another; in this case, of the ruler. What then is justice for the ruler?
Since justice is the advantage of another, justice for the ruler must be
the subject's advantage. It is because this is so, that Thrasymachus
always prefers injustice to justice, i.e. prefers the promotion of one's
own advantage. Kerferd's interpretation, it seems to me, provides
an account of Thrasymachus which is consistent both internally and
in relation to the rest of The Republic, and which avoids both of the
objections to the standard interpretation. In what follows I shall
argue that the best sense can be made of Thrasymachus' arguments
by abandoning the standard interpretation for Kerferd's, that is,
first, by conceiving Thrasymachus to be discussing all just relation-
ships and not simply those between subject and ruler, and second
by making central the idea that justice is the advantage of another,
rather than the idea that it is the advantage of the stronger, i.e.
However, I must make it clear that I only subscribe to that part of
Kerferd's article in which he discusses what Thrasymachus says.
He also discusses what we should take Thrasymachus
to mean by
what he says, and here I cannot agree with him. He classifies Thrasy-
machus' doctrine as one of Natural Right ("moral obligation has real
independent existence, and arises from the nature of man") rather
than of Ethical Nihilism, Legalism, or Psychological Egoism."
agree that Thrasymachus thinks injustice is supenor to justice as
an aim in life, but I am not sure that he asserts a moral obligation to
F. E. Sparshott "Socrates and Thrasymachus" Monist 50 (1966) 421-59,
especially section V.
T. Y. Henderson "In Defense of Thrasymachus" American Philosophical
Quarterly 7 (1970) 218-28.
Op. cit. p. 22.
Op. cit. p. 19.
This content downloaded from on Wed, 24 Apr 2013 12:57:46 PM
All use subject to JSTOR Terms and Conditions
be unjust, as Kerferd claims. The issue is complicated by our holding
different conceptions of "moral obligation" from the Greeks; in partic-
ular, it has been suggested that they did not make a clear distinction,
as we do, between morality and prudence.20 But even by Greek
standards, it seems doubtful whether Thrasymachus was thinking
terms of morality at all.2' He calls injustice not a virtue but sound,
"good", judgement, zu43ovka. (348 C, D). EU'ouALoc is used in Thucydides
of expedient judgements.22 Likewise, iuicpgpov and cyo m6 are words
often connected with prudence, efficiency, success, and the production
of desired results; they are, writes Adkins, part of "a system of values
based on calculation".23 Nor is it clear whether Plato introduces
Thrasymachus in order to contrast his own moral theory with a dia-
metrically opposed moral theory or a completely non-moral theory.
However, there is no need for my present purposes to settle whether
Thrasymachus' ideal of injustice is a moral theory or not. Since
Kerferd's classification of Thrasymachus' doctrine in terms of moral
theories is quite distinct from his interpretation of what Thrasymachus
says, one could reject the first whilst accepting the second. I consider
Kerferd's interpretation of what Thrasymachus says to be an important
contribution to our understanding of both Thrasymachus and The
Republic. Only Hourani has confronted Kerferd directly and dissented
from his interpretation: I shall not discuss Hourani, because I regard
Kerferd's reply as conclusive.24 Instead, I aim to show how, with
Kerferd's interpretation in our hands as the vital clue, it is possible
to unravel Thrasymachus' arguments, and reach a better under-
standing of the whole course of his debate with Socrates.
'? E.g. E. F. Carritt "An Ambiguity of the Word 'Good"' Proceedings o/ the
British Academy 23 (1937) 51-80, H. A. Prichard Duty and Interest (Oxford 1929)
and Moral Obligation (Oxford 1949) chs 1, 3, and 5, and Adkins Merit and
Responsibility especially ch. xi. Cf. R. G. CollingwoodAn Autobiography (Oxford
634, and chs IV, V, and VII passim.
21 Cf. Maguire op. cit. pp. 158-9, who is sure Thrasymachus does not assert a
moral obligation.
si E.g. Peloponnesian War iii 40 and 42-8. It is used in this way by Kleon and
Merit and Responsibility p. 223.
24 Hourani loc. cit.; Kerferd "Thrasymachus and Justice: a Reply" Phronesis 9
(1964) 12-16. One of Kerferd's points, that Thrasymachus is not a legalist,
is further supported by D. J. Hadgopoulos "Thrasymachus and Legalism"
Phronesis 18 (1973) 204-8.
This content downloaded from on Wed, 24 Apr 2013 12:57:46 PM
All use subject to JSTOR Terms and Conditions
A corollary of Kerferd's interpretation is that Thrasymachus has
presented his case backwards. For he begins with the particular appli-
cation to political inferiors of a general doctrine which he presents
later. Did the real Thrasymachus adopt this procedure? If so, why?
Or is this another instance of Plato's "manipulation" of Thrasy-
machus?26 If so, why? These are awkward questions, but they must
be raised. I shall argue that we can come partially to terms with them
if we pay attention to the dramatic setting. This entails examining
the first steps in the debate (336 B-338), which are usually ignored.26
Indeed, I suggest that whether or not my particular use of this passage
is acceptable, this section of the debate is of great importance and has
been neglected. Plato is a philosopher who thinks he has tamed art
and put it to work as philosophy's handmaiden, and consequently it is
always necessary to consider whether a dramatic interlude or by-play
may not be the means of making a philosophically relevant point.
There is of course the danger of reading too much into Plato: but this
is less of a danger than not considering all possible readings.
But first, a distinction must be drawn. Thrasymachus is joining in a
discussion on "what is justice".
There are, broadly speaking, two ways
of dealing with such a question: (1) by "delimitation" - showing
which people and acts are counted just; and (2) by "characterisation"
giving some characteristic or essential feature common to some or all
of the uses of the word. These are logically distinct, and it is possible
to be able do one and not the other. Now, in The Republic everyone,
including Thrasymachus, seems to be agreed upon which people and
acts are just, for they all hold that certain acts, which are both illegal
and normally called unjust, are what a just man does not do (331 A,
333 C, 343 D-344 B, and 360 B, C). Examples would be dishonesty,
tax evasion, corrupt practices, sacrilege, murder and adultery. It is
significant that when Socrates later illustrates the "commonplace and
vulgar" conception of justice (442 E-443 B), he repeats the others'
examples but also greatly expands the category of just conduct which
is not covered by the law, adding the examples of loyalty to friends
The word "manipulation" is E. L. Harrison's: see "Plato's Manipulation of
Thrasymachus" Phoenix 21 (1967) 27-39. It is approved by Maguire op. cit.
p. 142 n. 3.
Though not by T. G. Tucker The Proem to the Ideal Commonwealth of Plato
(London 1900) pp. xxxviii-xlv, or Cross and Woozley op. cit. pp. 23-6.
This content downloaded from on Wed, 24 Apr 2013 12:57:46 PM
All use subject to JSTOR Terms and Conditions
and keeping promises. None of the other participants can object to this
extension, since it follows their own principle that to be just is -to
obey social rules. Whether these are also legal rules or not is surely
secondary, and so Socrates is right to add these further examples.
We may note in passing that if we assume, as seems legitimate, that
Thrasymachus too would accept Socrates' list of just actions, then his
doctrine cannot possibly be reduced to "justice is the advantage of the
ruler" since some of Socrates' additional examples have nothing at all
to do with the ruler. Be that as it may, clearly in their debate Thrasy-
machus and Socrates agree on which acts are to count as just.27
The disagreement between them concerns the characterisation of
justice, in particular, whether the just man's acts are advantageous
to him. Once again, although they differ in their answers, the partic-
ipants all agree that this question is central. Cephalus has said that
the chief service of wealth is that it enables him to act justly, and
Glaucon and Adeimantus require Socrates to show that justice "pays".
What we must expect Thrasymachus to do, therefore, is to give us
such an account of the characteristics of justice as will enable us to see
whether it "pays" to be just.28
Now to turn to the text itself. Thrasymachus' first words are mainly
about the method of argument that has been adopted until then (336
C, D). First be condemns Socrates, and the others, not only for talking
nonsense, but also for using the question and answer method: "why
do you Simple Simons truckle and give way to one another?" He next
demands that Socrates give his own answer, and finally tries to pre-
scribe the kind of answer admissible: "and don't you be telling me that
[justice] is that which ought to be, or the beneficial or the profitable
or the gainful or the advantageous, but express clearly and precisely
whatever you say". What exactly is Thrasymachus telling Socrates to
do: and what therefore is he implying he will himself do instead?
It has been variously suggested that Thrasymachus is trying to rule
out "stale and barren platitudes",29 "synonyms",30 and "slogans...
The same view is taken by D. Sachs "A Fallacy in Plato's Republic" Philo-
sophical Review 72 (1963)
142-3, (reprinted in A. Sesonske [ed.] Plato's
"Republic": Interpretation and Criticism [Belmont 1966] and in G. Vlastos
[ed.] Plato. A Collection
Essays [New
Henderson op. cit. p. 219.
A similar conclusion is reached by Sparshott op. cit. pp. 422-3.
29 J. Adam The Republic of Plato. Edited with Critical Notes, Commentary and
Appendices (Cambridge 1902) vol. I. p. 24.
Shorev in his translation. p. 39.
This content downloaded from on Wed, 24 Apr 2013 12:57:46 PM
All use subject to JSTOR Terms and Conditions
never less ambiguous than the terms they replace."'3' This is part of it,
but I think there is a further point. These suggestions all presuppose
that Thrasymachus is looking for a different answer, whereas what he
seeks is a different method of answering. What we have here is a dis-
pute about the method of argument to be employed.32 Thrasymachus
objects to Socrates' elenctic method because, not understanding its
purpose, he regards it as merely a device by which Socrates gets out of
giving his own account of justice. So he sees Socrates' evasiveness as
the consequence of his method of discussion.
The alternative method of discussion which Thrasymachus thinks is
more satisfactory is that of
where one speaks at length,
develops an argument without interruption, and deals properly with
an issue. A rhetorician, he is himself at his best in giving set speeches
of some length, whilst on the other hand, as the sequel reveals, not
very skilled at the cut and thrust of the Socratic question and answer.
Indeed, the Sophists of Thrasymachus' generation, as suited their
profession, preferred long set speeches, and Plato describes clashes
over the two methods between Protagoras and Socrates and between
Polus and Socrates.33 Sidgwick emphasises the proficiency of Prota-
goras, Polus, Hippias, Gorgias
and Thrasymachus at j&xpoXoyta, and
their ineptitude at Socratic discourse, noting that Thrasymachus
is the least inept; and indeed claims that Socrates
method of question and answer, was at that time its only practitioner,
and naturally could beat everyone else at it.34 So it is not surprising
to find Thrasymachus in The Republic trying to switch the course of
the discussion into the channel he prefers, forbidding Socrates to come
up WIth a bare formula which is unclear and imprecise, and demanding
a full account, i.e. a ,uocxpoAoyEo, for this opens the way to Thrasymachus
31 Sparshott op. cit. p. 456. Cf. Cross and Woozley op. cit. pp. 24-5.
It might be objected that Thrasymachus is concerned with the content of the
argument as well as its form, since he tries to rule out several possible answers
(336 D). But that here too he is primarily making a point about method becomes
clear when we consider that he himself uses openly (339 A) one of the terms he
forbade Socrates
the difference being that Thrasymachus specifies
that justice is the advantage of another and that he embeds this in a whole
theory about justice and injustice. That is to say, the different method by
which Thrasymachus describes justice in terms of advantage makes his account
radically different in content too. Questions of content, then, are secondary
to those of method.
3" Protagoras 328 D-329 B and 333 E-338 E; Gorgias 461 D-462 B.
a' Op. cit. pp. 333 and 338 ff.
This content downloaded from on Wed, 24 Apr 2013 12:57:46 PM
All use subject to JSTOR Terms and Conditions
giving his own
What is surprising, is that he does not im-
mediately launch into his
in which we find later that he is at
his clearest and most forceful, but instead expresses his view partially
and very briefly, and, moreover, at once allows Socrates to cross-
question him about it. What has happened? Is Thrasymachus so angry
that he has lost his head, and thrown away all his advantages? On
the contrary, his action is not impetuous but deliberate: he does not
plunge in, but pauses to arrange his fee, and to repeat his complaints.
He does have a pocxpoXoyEa. (i.e. 343-344 C), but he apparently wants to
make a stir by presenting its essential point (justice is the advantage
of another, so it is better to act unjustly) in its most striking applica-
tion, the conduct of the ruler.35 We may suppose that he feels free to
abandon his normal method of presentation temporarily because he is
confident that he can handle Socrates' questions. Socrates tied Pole-
marchus into knots, reducing him to
that he no longer knew
what he meant (334 B), and twisted his words to suit himself; but
then, from Thrasymachus' point of view, Polemarchus suffered two
grave disadvantages: he did not know what he was talking about, and
he was prepared to defer to Socrates. Whereas Thrasymachus thinks
that he knows the essential character of justice, and that he is able to
prevent Socrates confusing or tricking him
he is not a "Simple Simon"
(341 A-C).36 He succeeds in keeping his end up remarkably well, but
seems to feel that Socrates is steering the course of the discussion too
much in his own direction and away from Thrasymachus' thesis (i.e.
in the discussions 341 C-342 E), and this is why he shows reluctance in
answering, and at last breaks up the discussion (343 A), abuses
Socrates, and opens his
That completed, he is about to
depart, since he does not think there is anything else he can do to
convince Socrates (345 B). He cannot be said to participate properly
when Socrates questions him after this: perhaps he sees that they can
never agree. When Socrates later justifies his use of question and
answer instead of the set speech which might have been expected as
a counter to Thrasymachus', it is noticeable that the point is settled by
Socrates and Glaucon without reference to Thrasymachus himself
(348 A, B). In the dialogue that follows, Thrasymachus is made to
*6 There is a measure of agreement that Thrasymachus is out to impress:
Kerferd "The Doctrine of Thrasymachus" p. 26, Hourani op. cit. p. 112.
I6 We note here that Thrasymachus uses the same word, -6s?uz, for the simpli-
city of those who accept and succumb to Socrates' elenchus (336 C) and for the
simplicity of the just man (348 C). He implies that he has seen through both.
This content downloaded from on Wed, 24 Apr 2013 12:57:46 PM
All use subject to JSTOR Terms and Conditions
feel that he cannot handle Socrates' questions so well as he had thought:
he is trapped, knows it, and blushes (350 D). At this point Thrasy-
machus not unnaturally wants to revert to
but says he knows
he will not be allowed to (350 D, E): after all, it was Glaucon, who had
taken the lead in
Thrasymachus' payment (337 D), who had
just agreed with Socrates that paxpoXoyxa be dropped (348 B). De-
prived of his chosen weapon of argument, Thrasymachus in effect
ceases to oppose Socrates (e.g. 351 C, 351 D) but agrees to all he says,
making clear that he is simply allowing Socrates to elucidate his ideas
and is not committing himself to holding them (351 A-354 A). In
other words, because the conditions of debate set by the audience are
not those which Thrasymachus sought, he feels unable to continue;
but he remains unconvinced by Socrates' arguments. Thrasymachus is
shown to us as a man who is rather truculent,37 a little vain, and over-
confident. No doubt such qualities were important in his profession.
Yet he is also shown to have considerable self-control and
for he extricates himself rather well from a situation where everything
is weighted against him. He is not portrayed as either a foolish or an
unsympathetic character, though perhaps a little reckless. So it is
plain to all that in his encounter with Socrates, his argument suffers
no real defeat. He has made a strategic withdrawal, not a retreat.
However we account for the order in which Thrasymachus presents
his argument, it remains that he expresses a clear preference for
FompoMoyEa as the method of argument. If we take that preference
seriously, then we shall hesitate to leap in as soon as Thrasymachus
begins to speak and say that he means by justice "the advantage of
the ruler". Rather, we shall refrain from trying to interpret his argu-
ment until he has finished giving it, and then we shall see that he
means justice is "the advantage of another". This enables him to give a
"characterisation" of justice relevant to the discussion he has joined.
I have argued that it is no objection to Kerferd's interpretation that
Thrasymachus means justice is the advantage of another, to argue
that this is not what Thrasymachus begins by saying, for a plausible
explanation can be provided to show that Thrasymachus presented
"... is
sharp tongue
are almost the only other things inde-
pendently recorded of him." Guthrie op. cit. p. 297.
This content downloaded from on Wed, 24 Apr 2013 12:57:46 PM
All use subject to JSTOR Terms and Conditions
his case in reverse. Next, I shall say something about the connection
between the earlier and later parts of
Thrasymachus' account of
justice, stating my understanding of his main thesis and
how the opening formula follows from it. This may help to explain
why Thrasymachus places so much emphasis on that formula.
According to Thrasymachus'
(343 B-344 C), men are in
competition for scarce resources (all his examples concern money in
one way or another). There are times when one is in a position of trust
and could take advantage of another person to one's own profit, e.g.
in a business partnership, or when holding public office. The rules
that make up justice forbid one to do that. But if you do not do what
is to your advantage, your inaction must be to someone else's ad-
vantage; your loss is his gain. Therefore justice is another's advantage
and your own detriment, and injustice, being the contrary, is your
own advantage and another's detriment. The rules of justice thus
favour those who break them. Consequently, the unjust man has the
advantage over the just, for the latter does the work but the former
takes the profit. The simple-minded just man's efforts make happy
not himself (as he thinks) but the unjust man. The unjust man is thus
in control of (&px&) the just man. He is using the just man, who is
unaware of it. (Thrasymachus seems to imply that men are just only
unwittingly: anyone who can consider the choice will see the advantages
of being unjust.) Here the question of "strength" arises. If we consider
not, say, a random or occasional thief, but a professional criminal or
persistently dishonest shop-keeper, we see that they are living off
their just fellows through a policy of injustice. This latter seems to be
the kind of situation of which Thrasymachus says that "injustice
on a sufficiently large scale is a stronger, freer, and more masterful
thing than justice" (344 C). The consistently unjust man comes off
better than the consistently just man: and he is therefore called
"stronger", referring to both his strength of mind (i.e. his knowledge)
and the strength of his consequent position (i.e. the advantages he
gains). Physical strength does not come into it (338 D). Every "ruler",
i.e. man in control, whether or not in a political context, is "stronger":
presumably because control implies knowledge. But we should note
that "stronger" is not a synonym for "another". Justice is always
the advantage of another, but it is only the advantage of the stronger
when someone has the strength to refuse to be just and seek his own
advantage instead. Thrasymachus is justified in speaking loosely
(though it does seem to be a bad habit of his, e.g. 340 C-E) as if the
This content downloaded from on Wed, 24 Apr 2013 12:57:46 PM
All use subject to JSTOR Terms and Conditions
two were synonymous because in this speech he has limited himself
to cases where there are only two parties, one just and the other unjust,
and here for the just man the advantage of another is always the
advantage of the stronger. Though even in these cases, justice for the
unjust man is not the advantage of the stronger but only of another,
i.e. the just man, the weaker. So "advantage of another" is the uni-
versal characteristic of justice. It is the "advantage of the stronger"
only in those cases where the stronger takes advantage. In short,
Thrasymachus has two propositions: (1) to be just is always to another
person's advantage and your disadvantage, whilst to be unjust is
always the opposite; and (2) if you pursue your own advantage on a
sufficiently large scale you will be strong and happy. The final step in
Thrasymachus' argument adds a third proposition which at first glance
is not closely connected. In order to see how much more advantageous
it is to be unjust than just, Socrates is invited to consider the extreme
case of the man who does the greatest wrong and is the most happy
- the tyrant. Thrasymachus' examples make clear that by a tyrant
he means not a usurper but a despot, someone using his position and
power to take everything he wants, committing "complete and entire
injustice" (344 C). This gives another proposition: (3) to rule is to
seek one's own advantage, i.e. to act unjustly, so that the tyrant is the
perfect ruler.
We are now approaching, in its correct order, the point from which
Thrasymachus began. From the point of view of the tyrant's subjects,
justice is the advantage of the ruler,
i.e. the
tyrant. And, more gen-
erally, justice for the subject is the advantage of the rulers. However,
there is still one matter to clarify. Thrasymachus begins by talking
of rulers in a very general way (338 D-339 A). However, before long
he has elevated the discussion from the case of actual rulers, who are
trying to be unjust but do not completely succeed, to the case of the
perfect ruler who always succeeds in his injustice (339 A-341 A);
and thereafter, I think, he is talking only about the perfect ruler, i.e.
the tyrant. Many commentators think that there is something wrong
with Thrasymachus' adoption of the position that "the ruler in so far
forth as ruler does not err" (341 A).
It has been said that Thrasy-
machus blunders here by abandoning a realistic empiricism for an
idealism which makes him an easy victim for Socrates,38 that he should
"S E.g. Adam op. cit. vol. I p. 33, Shorey What Plato Said pp. 211-12, H. W. B.
Joseph Essays in Ancient and Modern Philosophy (Oxford 1935)
18, Lee
cit. p. 63, and Harrison op. cit. p. 31.
This content downloaded from on Wed, 24 Apr 2013 12:57:46 PM
All use subject to JSTOR Terms and Conditions
have slipped through the loop-hole which Cleitophon had provided
and that Plato is again falsely attributing a position to him.'0
None of this is acceptable to those who deny that "the advantage of
the ruler(s)" is Thrasymachus' main view of justice.41 And they are
right to reject the traditional interpretation of his response to Clei-
It can be argued that Thrasymachus' is not an
empirical thesis, that he is not refuted by Socrates, that Cleitophon's
formula is quite different from and indeed in several ways opposed to
Thrasymachus', and that, given the context of his
there is
no reason to think that he is saying anything other than what he wants
to." Above all, there is no need to assume that Thrasymachus is
trapped or "manipulated" into a false move. On the contrary, the
perfect ruler is a crucial and logical deliberate step in the unfolding
of his ideas. By taking it, Thrasymachus has raised the whole argument
to a higher level.
Having reached the culmination of Thrasymachus' view in his
and having considered his progress towards it, let us turn
back to his original statement. Some of those who interpret his position
as being that justice is the advantage of the ruler, and especially those
who see him as a Legalist ("justice is obedience to the law"), are apt
in my view to miss part of what he is saying here.
for instance
interprets Thrasymachus' position very narrowly: (a) only the subject
can be either just or unjust; and (b) in relation to some laws only,
viz. those ensuring the stability of the particular regime in power;
from which he concludes that Thrasymachus "is not, at the outset,
talking about moral rules at all, but only about political arrange-
ments".'3 If this were so, then there would indeed be a disjunction
between Thrasymachus' two discussions of justice. However, if we
read the
first, and then consider Thrasymachus' opening
remarks about rulers as an application of its doctrine, we see that his
initial views are not Legalist. It is true that at first he explains him-
self by speaking of the rulers malding v6suOL (338 E), but rulers do far
ag E.g. Joseph op. cit. p. 17, Sinclair op. cit. p. 75, Cross and Woozley op. cit.
p. 46, Bloom op. cit. p. 329, Guthrie op. cit. pp. 95-6, and Maguire hc. cit.
Hourani loc. cit. p. 115, Harrison loc. cit. pp. 30-2, and Maguire loc. cit. p. 146.
See Kerferd "The Doctrine of Thrasymachus" pp. 20-1, Sparshott hc. cit.
section III, and Henderson loc. cit.
There is not the space to develop these points here.
Loc. cit. pp. 146-7. Cf. Sparshott's opinion that the law which
equates with the ruler's advantage is only consitutional law, loc. cit. pp. 426-7.
This content downloaded from on Wed, 24 Apr 2013 12:57:46 PM
All use subject to JSTOR Terms and Conditions
more besides, and it is clear from the p.axpoXoyEtx that we are to con-
sider all their acts and not simply their
I suggest that Thrasy-
machus has in mind not only what we call constitutional law, but as
elsewhere, anything to do with money. Taxation is the obvious
example of a situation which provides opportunities for just or unjust
action by both rulers and subjects.44 A democracy like Athens raised
money for some public purposes by imposing trierarchies on the rich.
Thrasymachus would see this as something where justice (paying up
honestly) was to the advantage of the democrats and to the dis-
advantage of the nrch, so that "justice is the advantage of the rulers".
In an oligarchy money was more likely to be raised regressively than
progressively, e.g. by a poll tax, which is to the advantage of the rich
since they pay a smaller proportion of their wealth, and again Thrasy-
machus would say that justice was to the advantage of the rulers.
Finally, a tyrant levies taxes without paying any, so clearly justice is
to his advantage. At the same time, justice for the rulers would be to
use the money raised from taxes to the advantage of the taxpayers,
and injustice would be to use it directly or indirectly for their own
advantage, e.g. for their own pleasure or to pay men to defend their
regime. (Of course, Thrasymachus is not committed to saying that
rulers never do what is for the advantage of their subjects. His thesis
is that whatever rulers who know their own advantage do, they do
for their advantage. This could include acts which do benefit the
subjects in the short term, in the same way that the profit-seeking
shepherd does care for his sheep. That is, when these rulers do benefit
their subjects, when they keep law and order, defend the polis, promote
trade, and so on, they do it for their own advantage, and in fact it is
ultimately for their own advantage and not their subjects'. Healthy
sheep give the best fleeces and best meat, and likewise orderly,
prosperous and contented subjects are less troublesome and more
profitable to the rulers who "fleece" them. It may look as if the subjects
benefit, but at the last count every advantage finds its way into
the ruler's coffers.) So I think we must assume that Thrasymachus is
concerned with what is just in relation to all the rulers' actions, not
simply their legislative acts in any strict (and modern) sense. After
all, this is clearly the case with the tyrant in the FaxpoAoyEx (344 A).
We have now completed the outline of Thrasymachus' views.
"4 Thrasymachus could also instance the distribution of benefits such as grain,
land, or citizenship.
This content downloaded from on Wed, 24 Apr 2013 12:57:46 PM
All use subject to JSTOR Terms and Conditions
These do not conflict but, as Kerferd claims, form a consistent whole.
Though Thrasymachus' doctnrne is formulated in different ways in
different contexts: (1) if we consider just acts as a single class, then
their common characteristic is that they are "the advantage of an-
other"; (2) if we consider the sub-class where the just man's act is
to the advantage of the consistently unjust man, justice is "the ad-
vantage of the stronger"; and (3) if we consider the sub-class of (2)
where the just man's act is to the advantage of the stronger who is
his ruler(s), justice is "the advantage of the ruler(s)". This perhaps
brings out the crucial importance of getting Thrasymachus' argument
in its right order. As it has now been set out, it is possible to move
from (1) to (3), but one cannot move from (3) to (1), which is what
we have to try to do if the text is taken in the order in which it stands.
If we take (1) as "nothing but an expansion" of (3),4f we encounter
difficulties which never arise if we take (3) for what it is, an instance
of (1). Again, if we start from (3), we may find it inconsistent with (1);
thus Maguire, nrghtly thinking that between (3) and (1) the meanings
of the key terms change, tries to save Thrasymachus from inconsistency
by arguing that the incompatible (1) is not his but Plato's."6 Whereas
the principle I have followed is that Thrasymachus is not being
hounded from one position to another, without thinking out the
implications of each, but that before he speaks he already has a
consistent doctrine which, for reasons at which we can only guess,
he chooses to reveal piecemeal.
Why then does Thrasymachus place so much emphasis on the pro-
position that justice is the advantage of the ruler? We have mentioned
the supposition that he wants to present a striking paradox. But there
may be more to it, surely? Sparshott thinks that "the sphere of govern-
mental arrangements is indeed that in which justice becomes most
problematic and questions of justice most pressing: if it is Thrasy-
machus' chief interest, it is also Plato's (as the rest of the Republic
shows) and Aristotle's and ours ... What Thrasymachus at first
provides can only be, at best, a special case of a general theory. No
one challenges him in these terms because the special case that he
chooses is the most conspicuous and the one most usually debated...""'
I should go further still: Thrasymachus begins with the case of justice
between subject and ruler because, besides suiting the course of the
4" Hourani oc. cit. p. 117.
Loc. cit. pp. 148-53 and 155-63.
47Loc. cit. p. 427.
This content downloaded from on Wed, 24 Apr 2013 12:57:46 PM
All use subject to JSTOR Terms and Conditions
discussion, it expresses the core of his doctnrne that justice is the
advantage of another, and includes its key exemplification, the ex-
treme case of the tyrant. We have noted that his third proposition
is that to rule is to seek one's own interest. He insists on this (340 C-
341 A). Yet to postulate a necessary connection between ruling and
acting unjustly seems if anything even more implausible than Socrates'
assertion of a necessary connection between ruling and acting justly.
Moreover, this proposition might seem to be superfluous to Thrasy-
machus' account; why cannot he simply say that rulers will be happiest
if they act unjustly and seek their own advantage, and ignore the
question whether a true ruler would be just or unjust?
The answer lies in the connection which Thrasymachus sees between
knowledge, strength, ruling, and injustice. The unjust man is aware
that injustice is to his advantage, and he knows how to use the just
man to his own ends. The unjust man, being thus the "ruler" of the
just man, is "stronger" than him in terms both of this knowledge
and his control. But a government is a clear case of men who are in
control, being able in the last resort to coerce their subjects by some
means, so if strength goes with injustice in one instance, it must do in
the other. Otherwise, Thrasymachus' theory will be destroyed by a
powerful counter-example. Further, on his premises, it is inconceivable
that the cleverest and most unjust man, the tyrant, should not be the
perfect ruler. For to succeed at injustice requires a great deal of
knowledge: you have to know what is to your advantage and how to
get it. So the tyrant must know all about ruling, or he would not
succeed in his injustice. If Socrates were right, and to rule were to
govern for the advantage of the subjects, there could be no such thing
as an unjust ruler and in exactly the same way, according to Thrasy-
machus there cannot be a just ruler, only a simpleton in a position of
power which he does not know how to use - he is not a ruler because
he is just. Thrasymachus must assert that the essence of ruling is to
be unjust, and Socrates that it is to be just, given the general view of
justice of each. There is a further connection between ruling and in-
justice for Thrasymachus. He realises, as his examples show, that
laws are one way in which the ordinary rules of justice are formulated.
Possibly he believes, like Plato and Aristotle, that the rulers' laws are
in the long run the main determiners of the moral climate. Certainly,
Thrasymachus and Socrates agree that laws tell us at least in part
which acts are just, that the rulers are strong and the subjects weak,
and that it is just to rule for the advantage of the subjects (344 A,
This content downloaded from on Wed, 24 Apr 2013 12:57:46 PM
All use subject to JSTOR Terms and Conditions
342 E). But if Thrasymachus says justice is the advantage of another,
and that the strong act unjustly, and that rulers are strong, it follows
that he must say that ruling is acting unjustly. Socrates is likewise
committed to the opposite.
The particular cases where justice is the advantage of the ruler
are therefore not simply the most striking and most interesting, they
also reveal the heart of Thrasyrnachus' general doctrine, and this is
perhaps why he begins with them before he has stated the framework
in which they lie embedded. Thrasymachus' account of justice is not
only consistent, it is coherent too.
I now hope to show that Kerferd's interpretation of Thrasymachus,
as well as revealing a consistent and coherent doctrine, fits in with
the rest of the dialogue. The first way to test this hypothesis is to ask
whether the details of the debate between Socrates and Thrasymachus,
and in particular Socrates' questions and positions, accord better
with the standard interpretation or with Kerferd's. However, I have
not the space for this task. I suggest only that Socrates' discussion
moves from "justice is the advantage of the ruler" to "justice is the
advantage of another" to "injustice is preferable to justice", that this
is a development from the least to the most important question, and
that thus Kerferd's interpretation is the more plausible. Furthermore,
it is in his discussion of Thrasymachus' characterisation of justice as a
disadvantage to be avoided, that Socrates first introduces some of
the central ideas of the rest of the book
limit (349 B-350 C), harmony
(351 C-352 A), and function (352 E-354 A), which again indicates
that Socrates takes Thrasymachus to have a thesis about justice as a
general moral concept and not a narrowly political one.
The second way to test the hypothesis is to ask how Thrasymachus'
doctrine has been understood by Glaucon and Adeimantus. Plato
represents Glaucon from the first (357 A, B) as taking Thrasymachus'
main point to have been that all injustice is preferable to justice,
that is, Glaucon does not limit the discussion to rulers. Adeimantus
formulates Thrasymachus' doctrine in his own words as "the opinion
that justice is the other man's good, the advantage of the stronger,
and that injustice is advantageous and profitable to oneself but
disadvantageous to the inferior" (367 C). Clearly, Adeimantus takes
the doctrine to be that justice is the advantage of another, though he
This content downloaded from on Wed, 24 Apr 2013 12:57:46 PM
All use subject to JSTOR Terms and Conditions
speaks loosely and equates "another" with "stronger", which we saw
is strictly speaking incorrect, but there is no indication that he means by
"stronger" merely "ruler" (in the political sense). Neither Glaucon
nor Adeimantus simply repeats Thrasymachus' attack on justice,
they renew it, expanding some points and adding others. For instance,
Glaucon adds an account of the origin of justice (358 E-359 B), and
expands Thrasymachus' point (344 C) that men revile injustice only
because they are afraid of being unjust,
arguing that men are just
unwillingly because they cannot be unjust (359 B-360 D)."8 As an-
other instance, Adeimantus adds a whole new dimension concerning
the teaching of attitudes towards justice (362 E-367 E). My point is
that though Glaucon and Adeimantus do not claim to be urging what
Thrasymachus would have said, they do claim it is what he could
have said (e.g. 367 A). They think their arguments are consistent with
his. And what their arguments fit, is Thrasymachus' statement that
justice is the advantage of another.
The third way to test the hypothesis is to ask with which of Thrasy-
machus' doctrines does the remainder of the book deal. And again,
the answer is that The Republic deals with the doctrine that justice is
the advantage of another, including the idea that justice for subjects
is the advantage of the ruler, and not the latter solely or even mainly.
No one is satisfied with Socrates' replies to Thrasymachus, not even
Socrates (354 A-C). Thrasymachus himself, as we saw, simply falls
silent, obviously unconvinced. Glaucon and Adeimantus demand a
further defence of justice. When Socrates sets out to reply to their
demands in the remainder of The Republic, he is also making his reply
to Thrasymachus, and making it by a method that Thrasymachus
cannot ignore, that of
Plato, unlike Socrates, seems to
agree with Thrasymachus over method. He knows that he cannot
"prove" Thrasymachus wrong
Sachs' mistake is to suppose that is
what Plato wants to do
and that to rebut his characterisation of
justice he must resort to Thrasymachus' methods and produce a rival
Sparshott, loc. cit. p. 431, claims that "Glaucon and Adeimantus plainly misin-
terpret Thrasymachus on the issue of the conventionality of justice, for they take
him to hold that the just man would be unjust if he dared (360 C), whereas
in fact he had attributed justice to 'an honest simplicity"' in 348 C. If Sparshott
is right about Glaucon, then there is an inconsistency in Thrasymachus himself,
since he plainly states that some are not unjust only from fear of the conse-
quences (344 C). But there is no misunderstanding by Glaucon, and no in-
consistency in Thrasymachus, if we take Thrasymachus to mean that some men
act justly out of folly (as at 348 C) and others out of fear (as at 344 C).
This content downloaded from on Wed, 24 Apr 2013 12:57:46 PM
All use subject to JSTOR Terms and Conditions
and more appealing characterisation. In their debate, Socrates and
Thrasymachus in effect swap definitions of the key terms (art, ruling,
wise, strong, happy, advantage, etc). Neither can be confuted provided
that his own set of definitions is adhered to. As a quick illustration
of this, consider "art" ('ixvy). Thrasymachus claims that the essence
of ruling is to take advantage of the subjects, Socrates that its essence
is to look to the advantage of the subjects. They agree that ruling is
an art, but each is defining both "the art of ruling" and "art" in his
own way; and the ordinary usage of "art" is sufficiently vague to lend
itself to both definitions, and to others too.49 Socrates could have
repeated the move he makes in the Gorgias (462 ff.) and dismissed as
pseudo-arts those which do not fit his definition; but the difficulty
would have remained, namely, why accept his definition and not that
of Thrasymachus? The existence of the rest of The Republic is Plato's
tacit admission that more is required. Only when the idea of the Form
of the Good is broached do we finally understand why Plato sees a
necessary connection between arts and justice, via his conception of
The importance of the debate with Thrasymachus is that it sets many
themes for the book as a whole (of course still others are set by Glaucon
and Adeimantus). Socrates and Thrasymachus agree that ruling is
an art, that an art involves knowledge, that the just ruler governs
for the advantage of his subjects, that rulers are stronger, and that
there must be a comparison made between the lives of the perfectly
just man and the perfectly unjust man to see which is happier.
Thrasymachus, in fact, has set up an ideal which is the mirror image
of Plato's (a procedure pursued in the Gorgias through the opposition
between Socrates and Callicles). That is to say, their ideals are often
the same yet turned back to front at the vital point. For instance,
for both, ruling is an art, but for one it is the art of being unjust, for
the other, of being just. Again, they agree that the man with knowledge
will act on it, but for one this means never being just wittingly, and
for the other, never being unjust wittingly. Overall, Thrasyrnachus'
tyrant is the mirror image of Socrates' Philosopher Ruler. And if this
is so, since Plato is concerned with justice in a general sense and not
4" Varying examples of "arts", to show how some fit Socrates' case but not
Thrasymachus', and others the converse, are cited by Tucker op. cit. pp. 1-li,
Joseph op. cit. pp. 224, Cross and Woozley op. cit. p. 50, Sparshott loc. cit.
section IV, and Henderson loc. cit. pp. 226-7. Cf. L. Wittgenstein on "game",
Philosophical Investigations (Oxford 1953) I 66.
This content downloaded from on Wed, 24 Apr 2013 12:57:46 PM
All use subject to JSTOR Terms and Conditions
a specifically political sense, so too must Thrasymachus be, as Kerferd
The burden of the last section was that there is nothing elsewhere in
The Republic which leads us to abandon the line of interpretation of
Thrasymachus' arguments which, following Kerferd, has been offered
here. Neither, I would argue, does The Laws 714B-D, nor what is known
independently of the real Thrasymachus, require any modification of
the interpretation.
This is not to say that this interpretation eliminates every problem.
Indeed, as it has been developed here, it may be felt to leave us with
at least three new and difficult questions unanswered: (a) if Thrasy-
machus' fundamental premise is that justice is the advantage of an-
other, why does he not spell it out at greater length; (b) why does he
present his case backwards, making it hard to follow; and (c) why does
he speak of
4uzqkpov ("the advantage") of the stronger but of &y OCv
("the advantage") of another, instead of using one word. These
difficulties, and their importance, must not be overlooked. On the other
hand, the interpretation offered has the merit of making a coherent
whole of all Thrasymachus' arguments and examples. On balance, I
suggest that his arguments are in such disarray that no interpretation
can be established beyond all doubt, but that whilst the present inter-
pretation rests in part upon conjecture, and leaves some matters in the
air, it is yet the most plausible reading of the text.
University CoUege of Swansea
This content downloaded from on Wed, 24 Apr 2013 12:57:46 PM
All use subject to JSTOR Terms and Conditions