Davidson, Wittgenstein and the computational semantics

Abstract

Famously, Davidson denied that there is such a thing as a language as many philosophers and linguists have thought of it: that the semantic competence of the speaker is learned in advance of the contexts of interpretation. On the contrary, Davidson sustains that we do not insert the utterance of a speaker into a meaning-generating computer and read off the output. In Wittgenstein’s treatment of such notions as understanding there is an attack on the conception of a person’s linguistic understanding as a mechanism, or calculus according to definite rules, which guides, and in that sense explains, the particular manifestations of her understanding. In the article, I make a free use of these ideas in order to undermine a general model of computational semantics.