Aristotle on Language and Universal Proof

Abstract

This paper shows that Aristotle’s conception of language is incompatible with a divide between syntax and semantics. Language for Aristotle is not reducible to a formal syntax of names and sentences, since these are linguistic entities if and only if they have linguistic meanings by convention. Nor is there an abstract semantics, in so far as Aristotle’s De Interpretatione resorts to a distinction between two kinds of meaning. On the one hand, linguistic meanings are by convention, so that names, verbs, and sentences are meaningful spoken sounds, and Aristotle concludes that linguistic meanings cannot be the same for all. On the other hand, non-linguistic thoughts are about mental contents, and when they are related to actual things, they are the same for all, because actual things are the same for all. For instance, the non-linguistic thought of snow is the same for all, but the linguistic meaning of the name ‘snow’ is not the same for all. While linguistic meanings rely on the learning of linguistic conventions, non-linguistic meanings are mental contents, derived from perceptions. In that respect, a mental conception of meaning is very different from a semantic theory of meaning. Nevertheless, the absence of syntax and semantics does not prevent Aristotle from developing a theory of deduction based on universal proofs, so that scientific explanations are made possible in the context of demonstrations. Accordingly, demonstrative knowledge is about explanatory middle terms, whose formalization allows Aristotle to establish a universal discourse for science.