Abstract

In lieu of an abstract, here is a brief excerpt of the content:45. HUME'S DIFFICULTIES WITH THE SELF One of the more baffling and apparently inconclusive parts of the Treatise is the section on personal identity. Hume himself, when he takes a backward glance at it in those notorious passages in the Appendix, singles it out as representing an unresolved problem in his philosophy. It is a matter of fairly general agreement among recent writers on the subject that one of Hume's chief difficulties -- though not one he recognizes — is his assigning to the mind (or, sometimes, the "imagination") certain activities, which he claims are instrumental in generating belief in, and awareness of, the very mind or self whose activities they are said to be. Hume talks about customs and habits of the mind and the expectations arising out of these, of the acts of observing, noticing and associating, of believing, of feigning, and so on, acts which the mind is required to perform in order for us to be able to explain along acceptable empiricist lines how the idea of an abiding self arises. Such talk, it seems, already presupposes the existence of a self and our possession of the idea thereof. So does, of course, the talk involved in Hume's accounts of the external world and of causality. Insofar as these involve his theory of belief, they, too, seem to presuppose a self which is more than a fiction, more than a product of the imagination. Yet the account he gives of the self seems to yield precisely a fiction in just the sense in which our ideas of the external world and of its causal connections are ideas of fictions. Quite generally, then, if Hume's theory of belief presupposes a continually existing self, what sense can be made of his arguments to show that our belief in such a self is the belief in a fiction? I want to argue that Hume can be defended against such criticisms and that the appearance of circularity here is just that: an appearance. It results largely from 46. two features of the Treatise. The first is that since at different points in the Treatise Hume is obviously concerned with different problems, his language reflects this. It is for this reason only that at times he talks in ways which make it seem that a continually existing self is being presupposed. The second feature responsible for the appearance of circularity is that while Hume draws a distinction between "perfect" and 'imperfect" identity, he fails to make that distinction explicit enough, so that one may easily miss the all-important point that it is only the former sort of identity which is being denied of the self, and the latter sort which is being explained and accounted for along Humean empiricist lines. If we keep this in mind, we shall see that the references in his account of the identity of the self to the activities of the self are not questionbegging, and that the attribution of these activities to the very same "fiction" whose existence is being accounted for carries no absurdity. II One way to bring out the differences between the kind of self which is alleged to be presupposed by Hume's account and that which, as I shall argue, is really involved in that account, is to put pressure on Hume's doctrine of 2 association. It is frequently urged by commentators that the notion of association which is so fundamental for much of Hume's philosophy is ambiguous, showing as it does the influence of two very different models, deriving from Hutcheson and Newton respectively. On the first model, a continuing self is presupposed as the agent of the various sorts of associative acts he invokes; on the second it is the items associated (perceptions) which are themselves the entities doing the associating. Clearly, the question whether Hume's theory of belief involves the assumption that a continuing self exists (harmless in the uses of that 47. theory in other contexts, but possibly fatal in its use in a theory of personal identity) may be seen to resolve into the question, which model of association is fundamental for Hume? It is, of course, an undeniable...