PERSONAL SURVIVAL AND THE CLOSEST-CONTINUER THEORY  

International Journal for Philosophy of Religion, Vol. 41, 1997, pp. 13-23.

I.  A Logical Argument Against Personal Survival

                Philosophical sophisticates can easily trot out arguments against the likelihood of life after death.  Many of these arguments are quite compelling.  My purpose is neither to analyze nor refute all of these arguments.  Indeed, I think that a couple of them may provide us with exceedingly good evidence that western theism’s conception of personal survival of biological death is impossible.  One common argumentative strategy for establishing such a substantive philosophical and religious conclusion, however, is deeply flawed.

                Several contemporary philosophers have offered analyses of the concept of personal identity that suggest the following argument.  Personal survival requires strict personal identity between the pre-mortem individual whose survival is in question, and the post-mortem individual whose continued existence would seem to provide the answer.  Unfortunately, so this argument goes, the notion of personal identity that is at the heart of this story is logically incoherent.  A logically necessary condition for the required strict personal identity is strict bodily identity -- physico-spatio-temporal continuity.  Biological death is by definition an interruption of bodily continuity, whether or not the interruption is temporary in theories of bodily resurrection, or permanent in theories of disembodied existence.  Thus, philosophical analysis of what is meant by the concept of being the same person over time rules out any possibility of personal survival.

                If life after death is a logical impossibility, this has obvious theological significance.  My purpose in the present context is to offer one reason for being highly skeptical about the claim that strict bodily continuity is any kind of logical precondition for personal identity.  On methodological grounds I doubt that philosophical proof is ever possible for substantive theses like this.  I make no attempt to argue for this general methodological position here.  Rather, I will be content to show that we possess the conceptual resources to make sense of personal survival, and personal identity, in contexts where there is a clear lapse of bodily continuity.

                My strategy will be to briefly review the attractions of a rival view of personal identity based on the concepts of memory and psychological continuity.  Unfortunately, this theory has faced sustained philosophical criticism since its introduction into the literature with the work of Locke.  Analytic philosophers in the latter half of the twentieth century have expanded on these early criticisms of the psychological continuity theory with such force that many believe that the bodily continuity theory succeeds by logical default.  The way around this problem is to articulate a new perspective on identity, that of Robert Nozick,[1] that will allow us preserve the theological advantages of the psychological continuity theory.  We can, I believe, clearly articulate a schematic view of personal identity that is not dependant on strict bodily continuity, and hence, allow for a robust sort of personal survival of death.

II.  The Enduring Appeal of Locke's Thought Experiment

                The claim that bodily identity could be a necessary condition of personal identity, when approached in a certain way, strikes most of us as clearly false.  Situations where a person comes to inhabit a new physical body seem conceivable.

[S]hould the soul of a prince, carrying with it the consciousness of the prince's past life, enter and inform the body of a cobbler, as soon as deserted by his own soul, every one sees he would be the same person with the prince, accountable only for the prince's actions.[2]

 Philosophers in this literature have offered countless contemporary extensions of Locke's puzzle case of the prince and the cobbler.  Hollywood screenwriters have turned out several feature-length films that depend on audiences making sense of bodily transfer.  It goes without saying that we are never asked to buy into the biological possibility or the general likelihood of these accounts.  But it is argued that the mere fact that we can coherently imagine the circumstances described in these puzzle cases shows that bodily continuity is not any sort of logical precondition for personal identity.  These thought experiments are also taken to establish a more positive hypothesis about personal identity -- that memory or psychological continuity is a logically necessary, and perhaps sufficient condition, for personal identity. 

                Almost as soon as Locke first explicitly proposed the psychological continuity theory of personal identity, theoretical problems began to emerge.  There is the classic puzzle of the gallant officer who in his old age remembers his greatest military success as a young officer, but fails to remember being disciplined as a boy in school.  This is problematic because that younger officer clearly remembered the disciplinary act.  Thus, on the memory theory, the boy is identical to the officer, the officer is identical with the old man, but the boy fails to be identical with the old man.  I read this example not as establishing a logical paradox for memory based theories of identity, but simply as a reminder of how fragile human memory really is.  Real world puzzle cases involving amnesia, dementia, and Alzheimer's disease, illustrate that worries about remaining the same person over time are not simply intellectual games, but can have real personal significance.

                A second early problem for Locke's theory involved a reliable means for distinguishing genuine memories from pseudo-memories -- those experiences sincerely felt to be memories, but which turn out to have false or illusory content.  It would seem that the way that I would confirm my memory of a heroic moment in little league is by having others verify my being there and having performed the heroic deed.  Similarly, I could discover that my sincerely held pseudo-memory of a kindness I performed as a young boy, really involved my brother, and simply resulted from my egotistically internalizing an oft told family anecdote.  It seems that we need some prior criterion for personal identity, perhaps bodily continuity, as a precondition for genuine memories.

But though consciousness of what is past does thus ascertain our personal identity to ourselves, yet, to say that it makes personal identity, or is necessary to our being the same persons, is to say, that a person has not existed a single moment, nor done one action, but what he can remember; indeed none but what he reflects on.  And one should really think it self-evident, that consciousness of personal identity presupposes, and therefore cannot constitute, personal identity, any more than knowledge, in any other case, can constitute truth, which it presupposes.[3]

                I do not wish to minimize the importance of these problems for psychological continuity theories of personal identity.  At the very least they show us that a naive memory criterion is likely to fail.  Still, much of the argument to follow amounts to endorsing the intuitive force of Locke's thought experiment.  I think most of us would agree with David Lewis about what is important in personal survival.

[W]hat matters most in survival is mental continuity and connectedness.  When I consider various cases in between commonplace survival and commonplace death, I find that what I mostly want in wanting survival is that my mental life should flow on.  My present experiences, thoughts, beliefs, desires, and traits of character should have appropriate successors.  My total present mental state should be one momentary stage in a continuing succession of mental states.[4]

                Psychological continuity theories seem to many to be the most plausible theoretical accounts of personal identity.  They also appear to offer the most attractive models of personal survival of death.  Unfortunately, they have encountered a new species of counter-example that is wholey unanticipated in the classic literature inspired by Locke.

III.  The Reduplication Puzzle Case

It can be fairly said that after the classic initial discussion of the problem of personal identity by John Locke, and the responses by Butler, Reid and Hume . . . the shape of the controversy was fixed for the next 200 years. . . . [It] remained so until as late as 1956. . . . [Here Bernard Williams] put his famous Reduplication Argument.  This argument transformed subsequent discussion of the problem and led philosophers to formulation of positions which were wholly new.[5]

                I wake up one morning in a new body.  I know it's me, my memories tell me this.  The problem is that whatever the means of transfer is -- brain transplant, use of the Enterprise teleporter, down-loading the contents of my brain, soul transfer, or whatever -- it seems possible that there will be a second me produced by the same process.  If at t0 there is one Johnson0, but at tn there are two bodily people, Johnson1 and Johnson2, each having inherited precisely the same memories and other psychological characteristics by whatever means allowed the prince's soul to inhabit the cobbler's body, Johnson0 seems to have fissioned into Johnson1 and Johnson2.  But now we have a serious identity puzzle.  Which present Johnson is identical with the past Johnson?  Johnson1?  Johnson2?  Both?  Neither?  Things seemed fine when the bodily transfer started with one person and one body, and ended in the same single person in a single different body.  But when the end result was two, or twenty-six, things pretty much went to hell.

                The fact that we can imagine the prince and the cobbler exchanging bodies, so Locke argues, tells us something about personal identity.  The fact that we can imagine reduplication seems also to tell us something.  One strain in the literature embraces bizarre possibilities of fission, fusion, and the like.  It abandons personal identity as the central normative concept when people consider their futures.  As Parfit puts it:

[c]retain important questions do presuppose a question about personal identity.  But they can be freed of this presupposition.  And when they are, the question about identity has no importance.[6]

           

A very different response to these sorts of thought experiments argues that the way to avoid reduplication, and also the way to preserve "a distinction between identity and exact similarity,"[7] is to insist on bodily continuity as a logical criterion for personal identity.  This, of course, is precisely the move that plays into the hands of those skeptics of life after death who seek to establish their theory on logical and linguistic grounds.  If the possibility of puzzle cases like reduplication force us, as a matter of  preserving logical coherence, to recognize the principle that "bodily continuity is always a necessary condition of personal identity,"[8] it will become impossible for the very same pre-mortem person to survive his or her bodily death.

IV.  A No Man's Land of Indeterminacy

                The professional literature in personal identity is dominated by increasingly creative and complex thought experiments.  A number of philosophers have expressed worry that our imaginations have outstripped our linguistic resources.  Perhaps we are debating situations where there is no clearly correct answers.  As Anthony Flew reminds us:

[m]ost words referring to physical objects are vague in some direction, somewhere there is an undemarcated frontier; somewhere there is a no man's land of indeterminacy; often there is a complete encircling penumbra of perplexity.[9]

Contemporary discussions of personal identity have consistently mentioned the possibility of borderline cases.  Although much lip service has been given, there a consistent reticence to take seriously the possibility that:

[w]e . . . can describe cases in which, though we know the answer to every other question, we have no idea how to answer a question about personal identity.  These cases are not covered by the criteria of personal identity that we actually use.[10]

                What should we do when we are confronted with imagined or actual cases that do not admit to clear linguistic categorization?   The answer is straightforward, I believe.  Admit the fact; don't try to stipulate to avoid it; and don't worry about it.

                We know about real world situations that rival any philosopher’s ability to construct personal identity puzzles.[11]  Amnesia, split-brains, and multiple personalities all strongly hint that our idea of the same person is perhaps inadequate to current biological reality.  Alzheimer's disease presents a particularly poignant case.  When I consider the scary possibility of my body and central nervous system existing in a severely demented state, I have real problems seeing that future person as identical to myself.[12]  At the same time, it seems arbitrary to deny that this person counts as me; indeed, it is precisely because he does in a way seem to be me that this future state is so frightening.  When these kinds of biological possibilities are supplemented with imagined cases,  it is hardly surprising that circumstances can be conceived for which we have no idea what to say.  Flew diagnoses this problem as going back to Locke and the very beginnings of the personal identity literature.

A . . . source of Locke's unhappy analysis of personal identity lies in his un-Lockean assumption that we can find a definition such that, granted we are provided with all the relevant factual data, we shall be able to say in every actual or imaginable case whether or not the expression "same person" can be correctly applied.  This assumption is mistaken. . . . [It] overlooks the possibilities of vagueness,  of marginal cases in which we do not quite know where to draw the line.[13]

                Of course there are borderline cases of personal identity.  Baroque examples of fission and fusion will certainly result in situations where have no clear idea who a person is, and consequently, whether he or she is identical to some earlier person.  This should not distress us.  And it should not tempt us to solve the problem by philosophical fiat.  We need not over-react.

                Just because a situation is new, unexpected, or non-standard does not mean that our current linguistic resources are inadequate.  Not every puzzle case is a borderline case.  Perhaps Locke was right and our shared intuitions tell us that the person in the prince's body at t1 is the very same person as the one in the cobbler's body at t2.  But perhaps not.  All I am insisting on at present is that the recognition of vagueness in some of the personal identity puzzle cases should not commit us to treating all of them in this way.  Perhaps there will remain some clear cases of personal identity in spite of surprising things that have happened.  Perhaps the theistic notion of life after death is one of those puzzle cases where clear personal identity is preserved.

V.  The Closest Continuer Theory

                It can be argued with some plausibility that the only model of personal identity that avoids the paradoxes of fission, fusion, and the like, is one of strict bodily continuity.  Such a view has many virtues.  It is simple, empirically straightforward, and thoroughly materialistic.  It is defended on conceptual grounds, not religious.  Nevertheless, it has immediate, and obvious, theological implications.  Indeed, if no better theoretical structure can be found we may be forced to concede that survival of death is conceptually impossible on the grounds that the strict bodily identity required for personal identity is by definity interrupted.  Fortunately, a workable alternative theory has recently been defended.

                I want to briefly outline a theory of identity, and personal identity, articulated by Robert Nozick.  It is not without philosophical critics,[14] but it remains in my judgment the most plausible and useful guide through the maze of puzzle cases and thought experiments with which we have concerned ourselves.  If it is an accurate conceptual model of what we mean by the same x, as I believe it is, the closest continuer theory provides us with theoretical framework for surviving dramatic changes to, and exchanges with, a person's body.  It will also make perfect sense of personal identity being preserved between a pre-mortem individual and a post-mortem individual.

                According to the closest continuer theory there are three logical necessary conditions for identity through time. 

                                X at time t1  is identical with Y at time t2 just in case:

                                                  i.            Y is a continuer of X.

                                                 ii.            Y is a close enough continuer X.

                                                iii.            Y is the closest continuer of X.

Each of these criteria will require explication, and in one case some correction.

                Suppose a faculty intramural softball team at a small college in the mid-west in the early 1980s called the Null Set.  Another faculty team in the northwest is currently playing intramural softball and calling itself the Null Set.  The common name is a product of nothing other than the fact that the team is composed of academics who desire to appear witty and at least one of whom knows a little about set theory.  The one team is not a continuer of the other and is, therefore not even a candidate for being the same team.  Nozick insists that:

[t]o say something is a continuer of x is not merely to say its properties are qualitively the same as x's, or resemble them.  Rather it is to say they grow out of x's properties, are causally produced by them, are to be explained by x's earlier having had its properties, and so forth.[15]

                We can easily imagine the second Null Set team being a continuer of the first.  Suppose I played on both teams and was so struck with the name of that first team that I badgered my colleagues into adopting it when I came to my present appointment.  Identity questions now focus on the relationship of being a close enough continuer of the original Null Set.

How close something must be to x to be x, it appears, depends on the kind of entity x is, as do the dimensions along which closeness is measured. . . . Closeness, here, represents not merely the degree of causal connection, but also the qualitative closeness of what is connected, as this is judged by some weighting of dimensions and features in a similarity metric.[16]

It is doubtful that the common membership on both teams of one philosopher would be sufficient to make the second Null Set even a candidate to be identical with the first.

                But what about a situation like the following?  The original Null Set contained three or four key players in terms of organizing ability, enthusiasm, and athletic skill.  The chair of the history department, in particular, was the driving force behind the original Null Set.  In 1984 he accepted his current position in Oregon as academic dean.  Over the next couple of years he managed to bring three of us from that original team to this same institution.  [Since this is a thought experiment, we need not worry about Affirmative Action guidelines or the like.]  Once we are all on the faculty we decide to revive the Null Set, even to the point of using our old jerseys as patterns for our new ones.  Obviously there are many new players on our team, but there is always some turnover in personnel on a team from one year to the next.  Suppose that you are open minded to the possibility that the new Null Set is a continuer, and a close enough continuer, of the old team.  We might now wonder if it is the same team, transported to a new academic institution.

                Any openness to considering the current Null Set as a serious candidate for being identical with the earlier team is undercut when we learn the following additional information.  When my colleagues/teammates left the mid-west we egotistically assumed that our old intramural team would disband -‑ we were, after all, the driving force behind the team.  We were wrong.  Our former colleagues discovered that they enjoyed competing with students and celebrating an occasional win with a few pints at the local pub.  They recruited some new younger colleagues to take our place.  The original Null Set has in fact continued to compete every year since its inception.  They have, of course, needed to gradually replace personnel as age, injury, lack of interest, and academic mobility have taken their inevitable toll -- just like any faculty intramural team.

                Situations like that of the two Null Sets are designed to show that the two criteria of being a continuer, and being a close enough continuer, are not always sufficient to establish identity.  The Oregon Null Set arguably satisfied these conditions, but failed to be identical with the original Null Set because later instantiations of the mid-west Null Set counted as an even closer continuer.  According to Nozick's model identity is reserved for the single closest continuer.

                Nozick argues that many of the classic puzzles regarding identity result from the fact that there exists more than one plausible candidate for being a close enough continuer.  This seems exactly the problem in the reduplication thought experiment.  Johnson1 is a continuer of Johnson0, and a close enough continuer.  All seems well until we learn of Johnson2 with an equal claim to be a close enough continuer of Johnson0.  Neither counts as the closest continuer, and therefore, neither is identical with the original.

                It may seem that the closest continuer theory promises more than it delivers, since we are never given explicit criteria for the relationships of continuation, close enough continuation, and closest continuation.  But the analysis was never intended as an analysis of being a team, or being a person.  It is not even a complete theory of what is required for being the same team, or the same person.  The closest continuer theory provides us some insight into the deeper logic of identity, on which more specific theories of personhood and personal identity can be superimposed.  As long as the theory allows us to diagnose our confusion about personal identity and navigate through the puzzle cases it will have served its function.

                I want to register two quibbles with the closest continuer theory, though neither will affect its application to life after death.  Nozick claims that the schema provides us with necessary and sufficient conditions for one thing being identical with another thing.  All is well when we are concerned with ships, teams, and people.  Nozick's last condition, the one that gives the theory its name, seems too strong, however, when the "thing" in question is a story, melody, or argument.  A muzack version of a classic rock tune can count as the same melody, even if a closer continuer is being performed at a local club.  Of much more direct relevance to issues of personal identity, Nozick's theory resolves borderline cases of identity into clear cases of non-identity.  In the reduplication scenario, Johnson1 turned out to be non-identical with Johnson0 because of the existence of Johnson2.  The precision here seems artificial.  It is not so much that we are confident about the non-identity relationship, but rather we come to realize that our concept of identity is indeterminate with respect to this highly unusual (and almost certainly biologically impossible) situation.  Neither of these problems show fundamental problems in the closest continuer framework, though they do indicate to me that there is further work to do.  Fortunately, the theory as it stands now provides us with a powerful tool for articulating a candidate for meaningful personal survival, one that completely rejects the notion of strict boidily continuity.

VI.  Personal Identity Across Death

                Annabellealive lives a long and prosperous life from her birth at tb to her death at td.  At some later time, tb+n, a post-mortem individual calling herself Annabelle  dead claims to be identical with Annabellealive.  You may picture  whatever sort of post-mortem existence your imagination, your theology, and your culture permits.  For most of the contributors to the analytic literature on personal identity, however vague the details of post-mortem existence are, they would nevertheless include some kind of continuing consciousness.  I will assume, therefore, that Annabelledead is conscious, has memories, and has some recognizable personality.  The obvious question is whether Annabelledead is identical with Annabellealive.  Has Annabelle survived her death?

                We should pause here to notice how odd this thought experiment has become.  Who should we ask to answer these questions?  It is far from clear that any pre-mortem person would have the needed epistemological perspective.  Annabelle, herself, is an attractive candidate.  Unfortunately, if she does this at td+n she will be unable to distinguish genuine memories of Annabellealive from realistic seeming pseudo-memories.  Presumably the thought experiment is for Annabellealive to conduct by imagining a post-mortem Annabelledead.  If she does this by utilizing the closest continuer theory I think she gets surprising clear answers, ones in direct opposition to the bodily continuity theorists.

                Annabelle can easily imagine that Annabelledead is a continuer, a close enough continuer, and the single closest continuer of Annabellealive.  I would guess that Annabelle will need a good deal of help from her culture and religious heritage in order to carry out this thought experiment in any detail.  If she is a member of a theistic interpretive community she will be supplied with a number of helpful ideas.  If she is a metaphysical dualist, the concept of an immortal soul will provide a potential mechanism for the preservation of those psychological traits that she determined satisfied the relationship of close enough continuation.  If she is a materialist, things will be a little more difficult, but not impossible.  The key, of course, will be the concept of an omnipotent and morally perfect God.  Omnipotence is required for the creation of a mode of post-mortem existence, including individually existing post-mortem people.  Moral perfection is required to rule out any sort divine mischief making.  The transference from pre-mortem existence to post-mortem existence seems to allow for a version of the reduplication problem.  A God who was inclined to metaphysical practical jokes, or was simply careless, would make survival problematic. 

                Theism, however, implicitly addresses all of the worries above.  An omnipotent, omniscient, and morally perfect God would seem to have the ability to do the following.  He could create a distinct person, Annabellealive.  After Annabelle's death He could bring it about that a post-mortem individual, existing in some appropriate post-mortem realm, Annabelledead, stands in the following very specific relationship to that first individual.

                  i.            Annabelledead is a continuer of Annabellealive.

                 ii.            Annabelledead is a close enough continuer of Annabellealive.

                iii.            Annabelledead is the single closest continuer of Annabellealive.

According to Nozick's schema the two are identical and survival of death has taken place.  More modestly, the thought experiment has led to the conclusion that meaningful personal survival is conceivable.

                I want to close pretty much where I started.  None of the considerations above say anything at all about the truth of theism.  I fear that the evidence here is disappointing.[17]  It also says nothing about whether the relationship of personal close enough continuation can ever be manifested on entities not possessing biological central nervous systems.  The evidence that it might doesn't look that promising either.[18]  Still, when materialist philosophers claim that personal survival is a conceptual impossibility based on thought experiments like reduplication, those extravagant claims need to be countered.[19]


ENDNOTES

 1.            Robert Nozick, Philosophical Explanations (Cambridge, Mass.: Harvard University Press, 1981).

 2.            John Locke, Essay Concerning Human Understanding, 1690.  His chapter, "Of Identity and Diversity," is reprinted in John Perry, Personal Identity (Berkeley: University of California Press, 1975), p. 44.

 3.            Joseph Butler, The Analogy of Religion, 1736.  His appendix, "Of Personal Identity," is reprinted in Perry, op. cit., p. 100.

 4.            David Lewis, "Survival and Identity," in Amelie Rorty, editor, The Identity of Persons (Berkeley: University of California Press, 1976), p. 273.

 5.            Harold W. Noonan, Personal Identity (London: Routledge, 1989), p. 149.

 6.            Derek Parfit, "Personal Identity," Philosophical Review 80 (1971).  Reprinted in Perry, op. cit., p. 200.

 7.            B. A. O. Williams, Problems of the Self (Cambridge, U.K.: Cambridge University Press, 1973), p. 9.

 8.            Ibid, p. 1.

 9.            Anthony Flew, "Locke and the Problem of Personal Identity," Philosophy 26 (1951).  Reprinted in Baruch A. Brody, editor, Readings in the Philosophy of Religion (Inglewood Cliffs: Prentice-Hall, 1974), p. 636.

10.           Parfit, op. cit., p. 199.          

11.           See, Kathleen Wilkes, Real People (Oxford: Oxford University Press, 1988).

12.           For a poignant and insightful discussion of the personal, moral, and legal implications of Alzheimer's disease see Ronald Dworkin, Life's Dominion (New York: Alfred K. Knoff, 1993).

13.           Flew, op. cit.

14.           See, for example, Noonan, op. cit., and Derek Parfit Reason and Persons (Oxford: Oxford University Press, 1984).

15.           Nozick, op. cit., p. 216.

16.           Ibid, p. 218.

17.        See my discussion of the strength of the evidence that pain and suffering provides                 for the atheistic hypothesis that God does not exist, "Inference to the Best                                 Explanation and the Problem of Evil," The Journal of Religion 64, 1984.

18.           See, for example, Daniel C. Dennett, Consciousness Explained (New York: Little, Brown, 1991), and Paul M. Churchland, The Engine of Reason, the Seat of the Soul (Cambridge, Mass.: The MIT Press, 1995).

19.           My sincere thanks to Charles Coate, the editor of this journal, and an anonymous               referee, for help in clarify my argument.