Thrasymachus' Definition of Justice in Plato's Republic

Abstract

T HE PROBLEM of interpreting Thrasymachus' theory of (tb 8LxoLov) in Republic i, 338c-347e, is well known and can be stated simply. He makes two assertions about the nature of or right action, each of which appears at first glance as a definition: i. Justice is serving the interest of the stronger.' 2. Just action is obedience to the laws of one's state. But, as Socrates quickly points out (3 39b-e), these two assertions cannot both be true as definitions, because their two predicates conflict in certain instances; for obedience to the laws by the subjects is occasionally not in the interest of the rulers. Thus the only way in which Thrasymachus' position can be maintained without inconsistency is if one of his assertions is not seriously meant by him as a definition. Assuming, then, as a working hypothesis that he has a consistent position, we must look in the text for answers to the questions, which assertion does he mean seriously as a definition? and what is the logical character of the other assertion? We shall then be able to see the relation of the two assertions to each other in his argument. The obvious answer to the first question, and the one that generally remains in our memory of Thrasymachus, is that he defines justice (a loose word for just action) as doing what is in the interest of the stronger. This answer also seems to be supported by Thrasymachus' own deliberate choice in face of the contradiction shown by Socrates, for he then says that is the real interest of the stronger, not what they think to be their interest as shown in the laws they make (34oc34ia). Nevertheless I believe there are good reasons for the other view, that Thrasymachus' intended definition of is obedience to law (conventionalism or legalism). This view is not new it has been held by Grote, Gomperz, Lindsay, Bosanquet, Winspear and others but it needs to be justified by a closer examination of the text than has yet been made for this purpose. The present article attempts to do this, and to answer objections raised by Professor G. B. Kerferd in an article publish-