Abstract

In this article, I argue that fregean approach to semantics, like that of modelling semantics, doesn’t have the conceptual tools to define the notions of truth and reference. The reason is that these programs ignore what Wittgenstein taught us, namely the link between meaning and use. More specifically, I intend to show how a theory of descriptive language can only be constructed by clarifying the connection between the notions of linguistic game, meaning and description. But how can this connection be clarified? Wittgenstein, in the second phase of his thought, had never been interested in a theory of this type, and all his reflections on use and meaning had the sole purpose of “curing” from that essentialist attitude that had characterized the Tractatus, and which still today characterizes almost the whole analytic philosophy. This means that it is not in Wittgenstein's thought that we can find such a theory, but in Moritz Schlick’s works, who attempted to build a theory of linguistic meaning based on the connection between the notions of image, use and verification.