Are Natural Languages Necessary?

Abstract

Against Davidson's criticism of the usual notion of a natural language, Dummett and most philosophers of language have argued that such a notion is necessary to account for the normativity of meaning and to avoid declaring meaningless much of our everyday talk on languages. This paper tries to show that both worries are unjustified by arguing that: 1) It is possible to talk of linguistic mistakes without commitment to natural languages in the usual sense; 2) The rejection of natural languages in the usual sense does not entail the possibility of non-trivially private languages; 3) The rejection of natural languages in the usual sense does not entail that conventionalized languages do not exist, but only that their existence is not necessary for linguistic communication.