Uses of language and philosophical problems

Abstract

Men such as Nowell-Smith and Hart, although they have focused attention on important alternatives to the kinds of solutions traditionally offered to certain philosophical problems, have not justified our adopting the alternatives they propose. The obvious question to ask at this point is what they must do to justify their conclusions. I shall not attempt to answer that question here. However, I do wish to indicate what this would seem to involve. They must justify a premise something like one of the form “‘P’ is used descriptively if and only if ‘P’ functions in such a way that it has logical features A, B, C...” This brings up two related problems. First, there is the problem of showing why an expression used descriptively must have any special logical features. This, as I have tried to show, involves justifying the claim that there are in the logic of a language symptoms of the kinds of things the expressions of the language refer to, a claim, incidentally, like the one Wittgenstein made in 1916 when he said, “The way in which language signifies is mirrored in its use.” Second, if the first problem is solved, there is the problem of showing just what those special features are. Using the language of the Wittgenstein of the Tractatus, we can say that these followers of the Wittgenstein of the Investigations must show, first, that some logical features of descriptive expressions are essential to their referring; and, second, just which features are essential rather than accidental. Can such claims be justified? About this I can here offer only one clue. They cannot be justified if the early Wittgenstein is correct because to do so would involve us in the relationship between language and reality about which, according to Wittgenstein, we can say nothing.