Philosophical discoveries

Abstract

There are two groups of philosophers in the world at present who often get across one another. I will call them respectively ‘analysts’ and ‘metaphysicians’, though this is strictly speaking inaccurate—for the analysts are in fact often studying the same old problems of metaphysics in their own way and with sharper tools, and the metaphysicians of an older style have no exclusive or proprietary right to the inheritance of Plato and Aristotle who started the business. Now metaphysicians often complain of analysts that, instead of doing ontology, studying being qua being (or, for that matter, qua anything else), they study only words. My purpose in this paper is to diagnose one (though only one) of the uneasinesses which lie at the back of this common complaint (a complaint which analysts of all kinds, and not only those of the ‘ordinary-language’ variety, have to answer). The source of the uneasiness seems to be this: there are some things in philosophy of which we want to say that we know that they are so—or even that we can discover or come to know that they are so—as contrasted with merely deciding arbitrarily that they are to be so; and yet we do not seem to know that these things are so by any observation of empirical fact. I refer to such things as that an object cannot both have and not have the same quality. These things used to be described as metaphysical truths; now it is more customary, at any rate among analysts, to express them meta- linguistically, for example by saying that propositions of the form ‘p and not p’ are analytically false. An analyst who says this is bound to go on to say what he means by such expressions as ‘analytically false’; and the account which he gives will usually be of the following general sort: to say that a proposition is analytically false is to say that it is false in virtue of the meaning or use which we give to the words used to express it, and of nothing else. But this way of speaking is not likely to mollify the metaphysician; indeed, he might be pardoned if he said that it made matters worse.

Sections II–V and VII of this paper appeared in the Journal of Philosophy, liv (1957) in a symposium with Professors Paul Henle and S. Kömer entitled ‘The Nature of Analysis’. The whole paper could not be printed there for reasons of space, and I am grateful to the editors of the Journal for permission to include in this revised version of the complete paper the extract already printed.