Knowledge of Meaning

Abstract

Can we explain the meaning of a sentence by appealing to the notion of the truth conditions of that sentence? The thought that we can is characteristic of what will be called here 'classical' theories of meaning. Davidson's approach to the construction of a theory of meaning is one recent attempt to embody this sort of connection between meaning and truth conditions in a precise way.1 For some time now the classical theory has been under attack from Michael Dummett, according to whom the problems associated with this program are not so much difficulties which arise in the details of its undertaking, but in the very principles on which it is based. In what follows we first of all try to isolate the essential features of Dummett's argument. Our business is then to show that these objections of principle can be met, but our aim is not to defend any particular version of the classical theory. 1. According to Dummett a theory of meaning is to be conceived of as a theory which explains a speaker's ability to understand and use his language. But it does not do this by identifying and describing physiological or psychological mechanisms possessed by the speaker. The explanation is to be agnostic with respect to the question of what actual processes go on in a human language user. It is to be a theory which would function equally well as an explanation of the linguistic competence of a martian or a robot, if either of those beings spoke English, or whatever language it is we are interested in providing a theory of meaning for. We assume that the speaker knows how to speak his language: the theory of meaning translates his 'knowing how' into 'knowing that.' It is a specification of the theoretical knowledge which would represent the practical knowledge which a speaker has and which constitutes his linguistic competence ([10]: 70). A question which immediately presents itself with respect to such a theory is this: under what conditions is such a theory to be regarded as