1 Fraser, A. C. ed., Essay Concerning Human Understanding, (Dover, 1959),Google Scholar Bk. II, Ch. XXVII, sect. 10, p. 450.

2 The interpretation I will offer derives in part from a few of Perry's, John remarks in “The Importance of Being Identical,” in The Identities of Persons, ed. Rorty, Amelie (University of California Press, 1976).Google Scholar However, Perry's way of developing his suggestive observations is quite different from mine.

3 Compare Perry's, discussion in his introduction to Personal Identity, (University of California Press, 1975) pp. 11–12Google Scholar.

4 I intend the phrase the “stages of each person” here and throughout to be ambiguous between “all and only stages which are stages of the same person” and “virtually all and only stages which are stages of the same person.” The reason behind this peculiar usage is two-fold. First, I simply believe that theoretical generalizations can withstand some theoretically unimportant exceptions, so I think it is more accurate to state such generalizations in this way. Secondly, I would not want to be taken as suggesting that our whole theory of persons could possibly depend on our not finding a single case of a person whose stages failed to stand in a particular relation. The suggestion is, rather, that our various views about persons may rest on a core of beliefs about what persons are like, which includes a belief about a particular relation among the stages of single persons.

5 The first interpretation of Locke's question given in the text is suggested by his heavy use of examples in the section of the Essay on personal identity, op. cit., Bk. II, Chap. XXVII, sections 11-29. A general consideration supports the second interpretation. The section on personal identity appears in the book devoted to (general) ideas and their origins. More specifically, the section “Of Identity and Diversity” begins: “Another occasion the mind often takes of comparing the very being of things, when, considering anything as existing at any determined time and place, we compare it with itself existing at another time, and thereon form the ideas of identity and diversity” (p. 439, Locke's emphasis). I interpret this passage as Locke's account of the basis and origin of the idea of identity (individuation) in general. This opening makes plausible the suggestion that the subsequent sections of the chapter are concerned with the grounds of ideas of more particular kinds of identity.

6 That is, most contemporary philosophers believe that psychological continuity, rather than just memory continuity, is the criterion of personal identity. I will define psychological continuity more precisely in section 3.

7 A causal analysis would substitute for the second condition on direct memory continuity, for example, something like “and (2) the experiences of the earlier stage and his beliefs about various facts are causally operative in bringing about the recollections and beliefs of the later stage.” Martin, C. B. and Deutscher, Max offer a causal analysis of memory in “Remembering,” Philosophical Review 75 (1966), pp. 161-96.CrossRefGoogle Scholar

8 Since we are looking for the mainstay of our current theory of persons, we need consider only what we can and cannot believe given our present views on the relevant matters.

9 Quinton, presents the original case in “The Soul,” journal of Philosophy 59 (1962), pp. 393–409CrossRefGoogle Scholar. In his version, memories and character traits are switched. I do not know how to make the description of the alleged “memory switch” more detailed. For example, it would be controversial to suggest that part of the brain is duplicated. I think the difficulty with describing how to switch just memory suggests that we do not have much understanding of memory as an isolated phenomenon.

10 In “The Reality of the Past,” in Philosophical Analysis, ed. Black, Max (Cornell University Press, 1950), pp. 38–59Google Scholar.

11 Martin and Deutscher offer this argument in “Remembering,” op. cit. Their remarks about Hume and others on p. 167 suggest that their goal may be to establish only the possibility of memory in the absence of conscious belief.

12 There is an obvious objection at this point. Given a case like Quinton's, the continuity of belief could be disturbed if the pre-operation stages had incompatible beliefs. The Scot after the experiment might remember the Pole's Christmas dinner of 1955, but because he has incompatible beliefs reject the belief that he enjoyed roast duck on that occasion. Thus although the Scot would actually recollect experiences of the Pole and facts known by him, he would come to regard the recollections as somehow delusory. I doubt whether this sort of thing should be regarded as memory. In any case, however, the latter stage's belief structure — his inclinations to believe as well as his actual beliefs — would still be counterfactually dependent on the belief structure of the earlier stage.

13 The argument is somewhat more complex if we construe memory continuity and desire continuity causally. On this interpretation, my previously desiring chocolate ice cream would cause me to recall that I desired it. Anyone who believes that previous desires, beliefs and experiences are causally responsible for later beliefs and desires would presumably believe that my recollecting that I desire chocolate ice cream is not only a prima facie reason for my now desiring it, but is actually causally efficacious in bringing about a present desire for it. Thus, my present system of desires would be an effect of my previous desire structure. (See fn. 12.)

14 For example, we are easily able to understand Shoemaker's, Sydney description of “Brownson” in Self-Knowledge and Self-Identity, (Cornell, 1963), pp. 23–25Google Scholar.

15 Cf. “Personal Identity,” Philosophical Review 80 (1971), pp. 3–27.

16 E.g. David Lewis in “Survival and Identity,” op. cit.

17 Cf. Essay, op. cit. Bk. II, Chap. XXVII, sect. 10, p. 443.

18 These arguments meet two important necessary conditions on explanation: the proposed explanantia are both testable and relevant to their explananda. (For a discussion of the importance of these conditions, see Hempel, C. G. Philosophy of Natural Science, [Prentice-Hall, 1966Google Scholar].) Needless to say, I cannot take up the enormously difficult problem of providing necessary and sufficient conditions for genuine explanation.

19 Here, the claim is just that, in fact, we never see human stages related by one of memory continuity, belief continuity, desire continuity and character continuity and not the others. The argument of section 2 is that we could not see such a case, given our present beliefs.

20 Essay, op. cit. Bk. II, Chap. XXVII, cf. sects 13, 16, 19, 20 and especially 22.

21 Ibid. Bk. II, Chap. XXVII, sect. 76, p. 467.

22 I am temporarily ignoring the fact that we cannot believe person-stages to be memory continuous alone.

23 I will leave open the ticklish question of whether this explanation also justifies our treatment.

24 We are construing “psychological continuity” as including any amount of continuity manifested by an uncontroversial case of the same person. This construal again seems appropriate, because it is precisely this range of individuals who are more or less susceptible to influence by praise or blame. Since our explanandum includes a range of cases, our explanans must at the outset, at least, also cover a variety of cases.

25 I am grateful to Philip Kitcher, David Lewis and George Pitcher for helping me to clarify my ideas about personal identity. I have also benefitted from the published writings of John Perry.