The verification of linguistic characterisations

Abstract

How do I know the sorts of things about language that I claim to know? Even assuming that I do not need to back my intuitions by appeal to criteria of certain sorts, still if they are to be shown to be valid must they not be backed by something? What sorts of explanation, or account, or justification could I offer for the claim that such and such a string of words is a sentence or that ‘oculist’ means eye doctor or that it is analytically true that women are females? How, in short, are such claims to be verified? These questions acquire a particular urgency if they are taken as expressions of the following underlying question: ‘It is not the case that all such knowledge, if really valid, must be based on an empirical scrutiny of human linguistic behaviour?’ How could one know such things unless one had done a really exhaustive statistical survey of the verbal behaviour of English speakers and thus discovered how they in fact used words? Pending such a survey, is not all such talk mere pre-scientific speculation?