語・意味・対象 : 『クラテュロス』におけるプラトンの言語哲学

Abstract

The Cratylus is the only dialogue in which Plato thematically elaborates his view on names (òvóματα, which include not only proper names, but also other kinds of word such as general names, adjectives, and verbs). In recent years several Plato scholars have put forward an interpretation that in this dialogue Plato treats names as disguised descriptive predicates and that he is committed to the view, which I call the 'description theory of reference', that names refer through their descriptive contents. In this paper I challenge this interpretation and try to show that Plato's own theory of reference is something quite different from the description theory as characterized above. Two theories of names are taken up and examined in the Cratylus----that of Hermogenes and that of Cratylus. Hermogenes' view, which is sometimes called 'conventionalism', is that a user of a language may employ on different occasions qualitatively different sounds or inscriptions as names of one and the same thing. Through examining passages of the dialogue in which Socrates criticizes this view, I bring out a realist and pragmatic vein of Plato's thought on language: i.e., he holds that the correctness or rather appropriateness of names consists in their capability of successfully discerning real differences among extralinguistic objects and of exhaustively expressing the truths about the real world. On Plato's view, however, this discernment is carried out by the dialectician (ó òταλεκτικóς), not by the name-giver (ó νουοθέτης). In this way Hermogenes' conventionalism is rejected. On the other hand, Cratylus' view essentially involves the description theory in the above sense. Many of those who attribute this theory to Plato make much of the fact of his treating proper names in parallel with other kinds of referring expression. But this parallelism depends on an etymological account of the descriptive content of a name, which account, however, Socrates rejects by seeking the name's descriptive content in the name-giver's opinion (òóξα). Thus a closer reading of relevant passages of the Cratylus reveals that Plato is not sympathetic with the description theory (or rather theory-schema) of reference. On the contrary, as I see it, his theory of names shows a certain similarity to a direct theory of reference on general terms, recently advanced by, for example, Kripke and Putnam. Finally, I briefly explore some epistemological implications of Plato's thoery of names and indicate how it can and must be linked up with his well-known doctrine of recollection.