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

Abstract

The paper discusses the results of scholarly debates on Plato’s own position on the issue of naturalism and conventionalism in the Cratylus and attempts to contribute to solving some problems. The author argues that there is no reason to suppose that Plato’s position differs from the one Socrates stands for in the dialogue: it is a naturalism of a definite kind, as argued for in the first part of the dialogue devoted to the refutation of Hermogenes’ conventionalism. Hermogenes, who treats a simple picking up of a referent by a name as sufficient for a full-fledged communication, holds the view that the connection between a name and a referent rests on the arbitrary and changeable agreement of ordinary language-speakers. As it is argued, he one-sidedly stresses the moment of imposition and re-imposition of names, without consideration of how the assigned meanings of names are transmitted beyond the participants of an agreement and are preserved through generations of language-speakers. Socrates opposes to him the theory of a name-instrument, that is a name that in its highest function should be employed successfully in dialectical enquiry, and thus should be made to be appropriate for properties of its referent. The creator of such names thus cannot be an ordinary language-speaker, but must be a competent lawgiver, and he should be supervised by a philosopher-dialectician who would use the products of his name-giving.

This general view is further explicated and illustrated in Socrates’ etymologizing and his hypothesis of mimetic capacities of mimetic sounds, which demonstrate that practically all names for various referents – from human proper names to the names of gods and physical, moral, and epistemological concepts – turn out to be meaningful descriptions of their referents. Although caveats are warranted by the text – the procedure of etymologizing is not entirely reliable and the opinions of name-givers are marred by a proto-Heraclitean teaching that all is in permanent motion, – this section demonstrates that the larger part of the philosophically relevant vocabulary consists of descriptive names that convey non-trivial, although not necessarily true judgments of their referents.

This result that Cratylus and Hermogenes applaud can be treated as the ultimate victory of naturalism. However, Socrates is not satisfied by his own reasoning and calls for its reexamination. In spite of this, he does not return to his own discourse, but turns to refuting Cratylus, who defends a more radical version of naturalism than that of Socrates. Some scholars treat this most debatable part of the dialogue as Socrates’ partial yielding to conventionalism, but other scholars see it as a complete victory of conventionalism. Among these latter, some find in the text itself evidence for this victory, while others believe that, although Socrates explicitly maintains that agreement plays only a complementary role in naming, Plato steers the course of the discussion to a full victory. The author argues in the paper against both kinds of proponents of the latter view that naturalism ultimately wins both according to the text and to the character of Socrates’ argument. Socrates assigns to agreement a certain role only in the communication, not in the assignment of names to their referents: in some cases, like that of σκληρότης, ‘hardness’, the resemblance of a name to its referent conveyed by a combination of σ and ρ is blocked by λ that conveys the opposite idea of ‘softness’. In such cases, a competent language-speaker who normally understands the meaning of names due to their resemblance to referents has no option but to appeal to linguistic habit, ‘to agree’ with it, that is to follow those meanings that are habitual from childhood. Socrates’ argument does not maintain that such meanings are arbitrary and based themselves on agreement, as according to Hermogenes. Rather it is implied that they correspond to the will of an ancient name-giver whose purpose was to make a name that resembles its referent, the resemblance however not having been attained, either because of some initial mistake or because of later distortion. Anyway, Socrates’ yielding to agreement in this sense thus does not amount to acceptance of Hermogenes’ conventionalism even for these particular cases.