Abstract

This chapter criticizes the standard truth–theoretic model of implicit definition whereby we stipulate that a word is to have whatever meaning will make true a certain set of sentences containing it. The alternative model proposed here is that, in such cases, the word derives its meaning from our way of using it, from our regarding those sentences as true—and so it acquires that meaning even if they are not true. It is argued, on this basis, that there is no route from meanings, so constituted, to our a priori knowledge in logic, arithmetic, or geometry.