Hume on Personal Identity

Hume Studies Volume XIX, Number 2, November 1993, pp. 289-299 Hume on Personal Identity DAVID PEARS The question that I discuss in this paper has often been raised and it has been answered in many different ways. "Why did Hume retract his theory of personal identity?" He puts it forward in the main text of the Treatise with his usual panache, and then takes it back in the Appendix. Why? Many different answers have been given to this question and what most of them have in common is the confidence with which they are given. Each interpreter sticks to his own interpretation in spite of the plausibility of its rivals, few of which are actually excluded by what Hume says. For this is an extreme case of the underdetermination of interpretation by evidence. I confess that in spite of all the recent work on this subject, I remain loyal to my own suggestion, which I first made in 1975.1 I will start with a sketch of my interpretation. Then I will quote Hume's recantation and develop an argument for my interpretation, based on that text and on others. After that, I will show how my suggestion is related to some of its rivals and I will say something about their comparative merits. Naturally, there is a considerable amount of overlap between the more plausible answers to the question, why Hume recanted. 1. My Interpretation In his Appendix, Hume confesses that his theory of personal identity seems to him to understate the strength of the connections between the David Pears is Emeritus Professor of Philosophy, Oxford University. 290 David Pears impressions and ideas that constitute a single person; or, as I would prefer to express his doubts, the impressions and ideas that constitute a single person's mind. He says that he would have been content if he could have attributed all those perceptions to a single substance; or alternatively, if he could have found some real connection between them. But all that he can find are connections based on the three natural relations: resemblance, contiguity, and causation. And those connections are, he implies, not enough. It is certain that he is thinking of a fundamental difficulty. For he evidently sees some reason for doing something which is simply not possible within his system, namely finding stronger connections between perceptions than those allowed by the doctrine of the main text of the Treatise. But what is his fundamental difficulty? The suggestion that I put forward was this: when an impression of sense occurs in my series of perceptions, it does not hold its place in the series by possessing any property, or by standing in any specific relation to any other perception in the series (except, of course, the relation of being co-experienced). Consequently, when I review the series, I do not find that the sensation is anchored by anything other than the fact of its contingent occurrence at the point where it occurred. Nevertheless, there is a strong modal statement that can be made about it: It, the very same sensation, could not conceivably have occurred in any other series. This modal statement is based on the special character of the identity-line of a particular perception: the identity-line of a particular perception starts at the point where it occurs in a particular series, and no sense can be attached to the speculation that it, the very same perception, might have occurred in some other series. However, in Hume's system, perceptions are treated like stars which form constellations, but only form them contingently, and it is obvious that the discovery of additional contingent relations between perceptions would never provide any support for the strong modal statement. Yet in Hume's system the only possible relations between perceptions are contingent relations. The other alternative, that the self is a substance, would, perhaps, do the trick, but for Hume it is unthinkable. It may seem fanciful to suggest that Hume's problem was the difficulty of accounting for the necessary ownership of perceptions within the framework of his system. But it is hard to see what other reason he could have had for requiring stronger relations...