Plato’s Last Word on Naturalism vs. Conventionalism in the Cratylus. II

The second part of the article, dealing with the question of Plato’s own position in the discussion of naturalism and conventionalism in the Cratylus (for part I, see Hyperboreus 29: 2 [2022] 196–233), continues with an analysis of a key point in the discussion between Socrates and Cratylus (434 a – 435 c). Cratylus argues that when the descriptive properties of a word confl ict with each other and thus make it impossible to establish which “thing” such a word denotes, a competent native speaker capable of analyzing such properties is forced to follow linguistic habit like ordinary speakers. Socrates points out that following the linguistic habit with which a competent native speaker is forced to “agree” is nothing but understanding a word that is dissimilar to its referent. This in turn entails, in accordance with the dilemma of the whole discussion, that understanding in such cases can be based only on a “contract”, or “agreement” stipulating the meaning of a word, with the competent speaker negotiating it with himself. Cratylus’ forced agreement that contract plays a role in linguistic communication is usually understood by scholars as a partial concession to conventionalism on the part of Plato himself, or even as evidence that Plato fully supports the arguments of conventionalism and sees no need for words that have similarities with their referents. The article substantiates a diff erent understanding of this part of the dialogue. It argues that Socrates agrees with Cratylus in treating linguistic habit as an independent factor in linguistic communication: understanding a word on the basis of habit does not need a word to resemble its referent through the descriptive and “mimetic” capacities of words; however it does not follow that the meaning of such words derives from the arbitrary “agreement” by which any arbitrary name can be assigned to any thing and at any moment change its name, as according to the theory of Hermogenes. The “agreement” of the competent native speaker with the meaning that a word possesses according to habit applies, first, only to a limited category of words that have no resemblance to their referents (in this the author of the article agrees with David Sedley’s understanding). Plato does not mean that understanding according to habit should make the similarity of a word to its referent superfl uous: the highest purpose of words is their philosophical purpose as “instruments” for distinguishing the essential properties of things, while linguistic habit provides only knowledge of what a word refers to, but not of the properties of this referent. Second, the “agreement with oneself” by which a competent native speaker is forced to agree to habit in the course of communication diff ers signifi cantly from the arbitrary contract of assigning meaning to words in Hermogenes’ theory. Socrates’ argument does not assume that linguistic habit arises from such an establishment of meaning that makes any people, even the most mediocre ones, creators of language. On the contrary, his reasoning about the word σκληρότης is based on the premise that this word was created by one of the wise creators of language who strove to create words similar to the things they designate. This implies that the similarity was either not achieved, through error, or was lost in the course of the long history of language; the habit has nevertheless preserved the word’s connection to the thing to which the word was assigned by a “lawgiver” of language (apparently by virtue of his high authority), although the sound composition of the word does not allow us to defi ne this thing by virtue of its intrinsic properties. Following the linguistic habit should thus be understood as an imperfect kind of linguistic communication, a forced retreat from the principles of naturalism, but not as a concession to conventionalism or even as Plato’s acknowledgment of its victory.