You are here

6. Identity and Continuity of Experience

6. Identity and Continuity of Experience
I would like to begin here with a celebrated article by Mr Derek Parfit which has understandably been the catalyst for a spate of very lively recent discussions. It is entitled ‘Personal Identity’ and was published in the Philosophical Review 1971. Professor Penelhum claimed1 that this work ‘has largely transformed the discussion of self-identity in the last few years.’

In one respect I find Parfit’s position a little obscure in spite of the delightful clarity of his writing. His central theme is that the notion of identity does not have the importance normally ascribed to it in the concerns we have usually in mind questions of survival or moral responsibility for example when we give special philosophical attention to it. All that matters for the concerns in question is perfectly well assured in a proper understanding of the continuity and connectedness of experience. It is to these matters therefore that we should turn. We can forget about self-identity in the strict sense for ‘all that interests us — all that matters’ is provided for in some aspect of the continuity of experience.

The point which is not altogether clear to me is what after all we do say on Parfit’s view about personal identity. He is not in what he proceeds to say giving an account of such identity in terms of continuity of experience etc. Indeed it seems clear that he regards identity — ‘a one-one relation’ in his terms — to be certainly different from the relations he has in mind in what he tells us about continuity of experience. ‘Identity is all-or-nothing. Most of the relations which matter in survival are in fact relations of degree.2 Parfit is not therefore offering a reductionist alternative to the sort of view of identity I have been defending. But does he have any view at all about it? I suspect he would not agree with my view. But can he as a philosopher writing about these questions avoid having a view or at least saying explicitly that he was baffled? Even if it were true that all that matters in concern about survival and so forth is properly taken care of in the considerations Parfit has in mind the problem what exactly is self-identity or what do we mean by ‘personal identity’ remains. Its importance may be diminished in some respects but it cannot be wished away by just refusing to attend to it and diverting attention to other matters. Even if its ramifications are not as wide as has normally been thought and as I shall be maintaining that they are it remains an inherently important philosophical problem and as most would admit irrespective of questions of survival and accountability a central one. Was Kant right or wrong was Hume? I do not see how we can avoid tackling such questions or just decide to shut our eyes to their importance at least in philosophy.
The furthest that Parift is prepared to go in this article on the question of what the distinct ‘one-one relation’ of identity involves in the case of personal identity is to cast doubt on the assumption that in all cases ‘the question about identity must have an answer’ that ‘whatever happens between now and any future time either I shall still exist or I shall not’.3 He is not certain how to disprove this belief but he thinks he can make it seem implausible. His inclination is clearly to assimilate questions of personal identity to other identity questions — ‘Is it the same machine?’ ‘Is it the same nation?’ — here there is an arbitrary element and problematic cases at the edges. To this extent he is certainly giving an account of personal identity and taking sides on the question whether there is some finality in the view that I shall either exist or not exist at some future time. But on the whole he seems to leave the question open or pass it by according it at least enough distinctness as to declare it unimportant. What he is concerned to maintain is that whatever is to be made of the question of identity that has nothing to do with the important questions we ask about ‘survival memory and responsibility’.4 How then does he make that out? Let us turn to this question.
Parfit begins with a refinement upon the problem case envisaged by Professor Sydney Shoemaker namely that of a transplant of one person’s brain to the body of another or a switching of brains. Most people would argue Parfit holds correctly in my view that each person in this case would be the person whose memories and character he has in virtue of the way a brain makes this possible. But Professor David Wiggins had put forward a further variation based upon the relative independence of the two halves of one’s brain the supposition namely that the two halves of one person’s brain are separated and each housed in a different body. This can at least be imagined. But shall we then say that this is a genuine case of fission that one person has now become two? Could it also be thought possible that one of the new persons is the same as the original? But how could it be one and not the other especially if as is supposed in this example the two halves of the brain are exactly similar? Could the one original person be also both the new ones?
This latter supposition it is urged is not as absurd as might at first appear. For might we not have ‘two bodies and a divided mind?’ ‘We can imagine a man having two simultaneous experiences in having each of which he is unaware of having the other’.5 We have in fact actual cases of this as when the bridge between the two hemispheres of the brain is cut in the treatment of epilepsy. There are here ‘two separate spheres of consciousness’ but ‘what is experienced in each is presumably experienced by the patient’.6
Parfit himself takes this further by supposing that ‘the bridge between my hemispheres is brought under my voluntary control’7 so that I can divide my mind at will. This would have its uses for example by my setting my two spheres of consciousness to work simultaneously on a problem to be solved in a hurry and later uniting my mind and selecting the best of the solutions proffered as remembered in my unified state of mind.
Parfit thinks that this possibility is a perfectly proper one to entertain however improbable we may think it all in practice and this has a considerable bearing on the further presentation of his views and the variations upon similar themes in the further discussions which this set off. But at this point it seems to me we get on to very questionable ground. It seems plausible enough in one way to suppose that two ‘parts’ of my mind could go on functioning at the same time without the one obtruding upon the other. Is not this happening all the time? Could I not be so lost in thought that with that side of my consciousness I am not at all aware of driving my car — the latter operation continuing quite well on its own? This is how we are often apt to think and to speak. If I am right however in what has already been maintained this way of thinking is too simplistic. It is the one person who is deep in thought about some problem and who is driving the car. I do not drop my consciousness of the one in my concentration on the other. The mind just cannot be strictly divided in that way. If it happened as was noted earlier there would be a crash. What happens is that one matter occupies more the centre of my attention and demands more energetic pursuance. I do not become an automatic driver it is myself who is driving the car and thinking out the problems but the mental activity involved in the former is more subdued. There are in short no strictly separated spheres of consciousness. There is only the one consciousness diversified into various procedures undergone or enacted by the one person I am. There are no bits of my experience which can float away altogether from me. So if it were possible for me to think out two solutions of the same problem at the same time as I can listen to (or vaguely hear) the droning of a plane and compose this sentence I could not be completely oblivious in the one case that the other was happening also.
We do not therefore in Parfit’s example begin to approach a situation of a strictly divided mind. Whatever we think of the possibilities of his example in itself it does not set us on a path where there are really two persons and not one or a genuine fission. The one subject must be however dimly in one respect the subject of awareness in both cases.
Parfit himself does not in fact want to say that in the example indicated one person becomes for some period properly two but only that this prepares the way for envisaging a situation when that would be the proper thing to say; and it is this which is helped by supposing it possible to imagine the same man ‘having two simultaneous experiences in having each of which he is unaware of having the other’.8 My submission is that in the strict sense this is an impossible contingency however much for rough and ready purposes we may speak in that way.
On the basis of the supposition just noted Parfit concludes that it is not obviously absurd to suppose that there could be two bodies and a divided mind and to this extent we might conclude that this was the proper account to give of the situation where different halves of the brain had been housed in different bodies. I would also admit that we could say something of this sort but only on the further vital qualification that it was the one mind which was continuing to function in these very different ways and that it could not fail to be aware of what went on in both the new contingencies.9
Before we go further with that let us see how Parfit himself proceeds. He maintains that while it would not be absurd to envisage two bodies and a divided mind in the sort of limited situation he envisages namely where the division is not protracted and ends in the reuniting of the two ‘streams’ of the mind yet it would be absurd to entertain this notion if the mind were permanently divided and the halves developed in their respective bodies in different ways — ‘they could live at opposite ends of the earth’ they might meet and ‘fail to recognise each other’.10
Here again the issue turns on whether one half of the divided mind could be totally unaware of what happens in the other. On my view this is not possible though I think it is at least conceivable that one mind could be involved with two bodies at the same time. This is not a likely contingency — indeed our first reaction is to rule it out altogether. There are so many adaptations and difficulties to be overcome before it becomes remotely intelligible to us how this contingency might come about due especially to the peculiarly close involvement with one particular body which is not only in point of fact the case but also a requirement of almost everything we do and undergo in an embodied condition. But there might be ways in which these difficulties could be surmounted and I have already discussed them and the complications to be considered in pages 106 and 113 of The Self and Immortality. The main point at the moment is that as I also stressed in the context just noted there would have to be awareness in the one ‘life’ of what goes on in the other although much would be dim and little heeded as already happens for many simultaneous experiences.
To put this again briefly as it is very central to most that I would want to maintain against Parfit it is on the view I have already presented impossible in the nature of what it is to have experience to have an experience without being aware however unreflectively of having it. If therefore in addition to all that goes on normally in my life there is another embodied existence I have perhaps far away as envisaged in Parfit’s embroidering of Wiggins’s example then I just cannot inherently cannot be unaware of this. It is only therefore in a very restricted sense that one could ‘have two bodies and a divided mind’. The mind could not be strictly divided but only have two very distinct spheres and modes of operation.
For Parfit envisaging as he does a strictly divided mind the question of identity becomes very acute in his case of a divided mind. We would have two persons and we could not say they were identical without changing the concept of a person. In the context envisaged ‘all the possible answers to the question about identity are highly implausible’.11 The course to be followed therefore it is maintained is to give up the assumption that there must be a true answer to any question about personal identity and having thus taken away much of the significance of the question of identity perhaps dispensed with it altogether though this is not made explicit then we prise the really important questions apart from any question of identity leaving the latter ‘with no further interest’.12
We come thus to the conclusion that ‘the relation of the original person to each of the resulting people contains all that interests us — all that matters — in any ordinary case of survival’.13 In terms of such relations it is not difficult to find a sense in which ‘one person can survive as two’. We set aside therefore the idea of identity as ‘all or nothing’ and turn to relations most of which are ‘matters of degree’. ‘The belief that identity is what matters is hard to overcome’ but once the breach has been made ‘the rest should be easier to remove’.14
All that matters in judgements of identity is thus found to be psychological continuity a principle that could operate even if there should be no bodily continuity. This is in sharp contrast with the move made by Bernard Williams in his discussion of someone seeming to be Guy Fawkes alive again having the appropriate ‘memories’ etc. Williams insists that sameness of the body is necessary for identity. For Parfit we can pass beyond this and regard the new Guy Fawkes as the old one having survived or alive again provided the appropriate psychological continuity is assured. But what is it that is appropriate here? If there is not a ‘one-one’ relation of identity what is to prevent the psychological continuity from being ‘incomplete and arbitrary’?15
To deal with this we have to invoke the idea of ‘psychological connectedness’. This consists in turn of ‘direct psychological relations’. The most important of these concerns memory. It is normally taken as a logical truth that we can only remember our own experiences but this does not debar us from framing a new concept ‘q-memory’. This is found when we have an experience in all respects like memory except that it refers to an experience which someone else has had my belief about this experience depending upon it in the same way as my ‘memory of an experience is dependent upon it’.16
This is very attractive but it begs the most important question of all. Suppose we did have some awareness of another person’s experience as direct and dependable as memory of my own experiences that is not established by independent evidence and observation. We would presumably regard this as paranormal but whether that is the appropriate term or not we would still not have knowledge of the other person’s experience in the same way as a memory is dependent on the experience remembered. For whatever else we say about memory in the strict sense it seems to be quite essential to it that my assurance about what happened involves its having happened to me I am able to recall it in the special way of memory just because I have lived through it myself. The alleged ‘q-memory’ is therefore not a variant of memories at all it could only be some peculiar way in which I become assured of what someone else’s experience was like somewhat like clairvoyance or telepathy; and these establish nothing about the relation of myself to other persons. I do not have the experience of another person in telepathy that phenomenon provides no justification whatsoever for an alleged merging of persons. In memory I recall the past experience as essentially an experience I have had and however the dependability of it is finally understood it would be quite off the mark to try to account for it regardless of what is most central namely my having had the experience myself.
This is not a matter of words of what we decide to call ‘memory’. It is a case of what the phenomenon we do call ‘memory’ is like in itself. If we leave out what is vital no other resemblances will help. This is what Parfit wholly repudiates. He does so quite expressly. ‘When’ he says ‘I seem to remember an experience I do indeed seem to remember having it. But it cannot be a part of what I seem to remember about this experience that I the person who now seems to remember it am the person who had this experience’.17 I only make the latter assumption ‘because I do not in fact have q-memories of other people’s experiences.’18
Indeed it is urged that all memories are initially ‘q-memories’. It only happens that I have them as memories of my own experiences. If I did begin to have ‘q’-memories’ of other people’s experiences I would then be assured that someone had the remembered experience and ‘have to work out who it was’.19 ‘Memories are simply q-memories of one’s own experiences’.20 In Wiggins’s case of the two halves of a brain being housed in new bodies we could ask of one of ‘the resulting people’ whether he had ‘heard this music before’ and he might reply ‘I am not sure whether it was I who heard it or the original person’.21 Indeed if only we started having q-memories of other persons’ experiences we could drop the concept of memory altogether.22
For this purpose however we would have to assume that some memories at least could ‘come to me simply as beliefs about my past’.23 It is this that seems to me to be a wholly unwarranted assumption. It is not in accord at all with what we find memory to be which is not just some peculiar assurance about my past but the recalling of what my past was like in some respect because it comes as something through which I have lived myself. If it were not this it would be something radically different.
Let us turn now to the second major example of the relations in which connectedness consists. This is ‘q-intentions’. ‘It may be a logical truth that we can intend to perform only our own actions’.24 But why not again ‘redescribe’ intentions as q-intentions which could allow us to ‘q-intend to perform another person’s actions’.25 This seems to me a very daring line. One wonders to what excesses this convenient policy of redescription could lead. In the present case the supposition proposed falls down irretrievably from the very start in respect of one matter which philosophers seem peculiarly apt to overlook today namely that we just cannot intend to act in the strict and proper sense independently of or ahead of actual performance. We do for rough and ready purposes speak of forming intentions to do something in the future. I may for example say that I have formed a firm intention or made up my mind or decided to have my next holiday in Scotland. I was earlier very undecided being much attracted to the idea of going to Italy. But I have now decided my choice is for Scotland that is where I will go. We are in fact forming intentions ahead in this way all the time all our planning involves it. I intend to give a certain lecture in the autumn to visit Emory University next year to call to see a friend tomorrow. We would achieve little without planning or intending ahead in this way.
And yet what do such intentions involve? In my intending to have my holiday in Scotland I have firmly put out of my thoughts all consideration of going elsewhere. I have stopped thinking about it; my mind is made up. I also take the necessary steps book my hotel cancel other engagements etc. I do all that is necessary to do ahead. I make it such that going to Scotland will be the obvious and natural thing to do when the time comes but I do not actually set myself to go until I do go. Circumstances could prevent me or I could change my mind. Hell we have all been told is paved with good intentions and New Year resolutions have a notorious habit of not being fulfilled. My dear father intended to give up smoking many times but never did. It seems inherently impossible to do anything ahead of doing it and though we might think it malicious of someone to harbour thoughts of murdering his rival or in fact intending in the present sense to do so we would not take this to be as bad as actually doing it unless it were just his being prevented. He might change his mind on further thought or ‘when it comes to it’ and we would certainly modify our moral condemnation accordingly.
The example that Parfit offers is again the Wiggins case of one person becoming two. As the original person he may intend to perform a certain action as one of the ‘resulting people’. But if I am right he cannot intend in this way even in his own case and if there is any sense in speaking in that way seriously at all it could only be in the form of making such dispositions including the set of his own thoughts as would strongly (irresistibly?) induce one of his future selves to do something. I do not think that the notion of vicarious intending is a possible one and the only way in which it would begin to be plausible would be if in the Wiggins case we took these to be one person with two bodies not properly two resulting people.
Undaunted however Parfit anticipates that the vicarious operations could be extended to the situations where we ‘q-recognise’ ‘be a q-witness’ of what one has never seen ‘have q-ambitions make q-promises and be q-responsible for’.26 Indeed this could pave the way to thinking experiences ‘in a wholly “impersonal” way’.27 That seems to me the appalling consummation of it and in a much more disastrous way for moral accountability and religion than in the normal idealist elimination of the ultimate distinctness of persons. One wonders how far Parfit and those of like mind are prepared to take this and whether they take the full impact of the consequence.
Parfit moves to the closing stages of his argument by setting out patterns of psychological continuity in terms of greater or lesser degrees of strength and closeness ranging from those where the relation is strongest namely the psychological connectedness of the ‘direct’ q-relations to where it is so remote as to be of no account. In a case of fusion the one person who results can for example q-remember the lives of the original two and it would make sense therefore to speak of two persons now surviving as one. Alternatively we are asked to think of beings who are like ourselves except that they reproduce by a process of natural division. These would all have some psychological continuity but there will also be psychological connectedness which can ‘vary in degree within a single life’28 and in terms of the strength and closeness of this and of other relations which are closest to it we can distinguish cases where for the purposes that matter we could say for example ‘it was I who did that’ and those where we could not do so. This gives us in essentials all that seriously counts when we are concerned about such matters as survival and responsibility. It is the closeness or degree of relation that really matters.
These possibilities are deployed with great ingenuity. But for me at least they produce little conviction. Even if we grant the central and most direct q-relations and I have already urged that this is very questionable it is by no means obvious that I should feel the most concern and affinity with those with whom I stand in the closest relations on Parfit’s account. Suppose someone does have some peculiar direct awareness of what my experiences or some of them were like does it follow that I would be more interested in him than others — members of my own family and circle of friends? The relation could be just a freak one and I might well resent it. Again if someone could q-intended that I do something would not the natural reaction be to hold him solely responsible? How can these be properly his intentions and mine respecting the same actions short of being strictly the one person and not two?
Suppose again one were assured that something pleasant (or unpleasant) was to happen to someone directly related to me in the present senses and were excited or dismayed as the case may be this is not the same interest as I would have in some properly future experience of my own. I may if I love someone deeply and am nobly motivated prefer some disagreeable fate to happen to me rather than to him but this does not begin to be tantamount to the concern we have about some bad experience we may be likely to suffer ourselves. The latter turns on its being strictly my experience just as my accountability turns on an action being strictly my own.
The distinctness of persons is not eliminated in the sort of relationships Parfit envisages we are only diverted to arbitrary substitutes for it and it seems to me that none of these however close and ingeniously conceived begins to do duty in matters of survival and morals for the way each person is uniquely and finally the person he is. To these implications of personal identity in the strict sense I shall return in the sequel to this book. In the meantime let us look at some of the ramifications of Parfit’s procedures in the work of other writers who took their start from them.
Professor David Lewis maintains very ingeniously that for ordinary purposes that is excluding the notorious problem cases there is nothing seriously at issue between the common-sense insistence on personal identity as the crucial concern and the sort of connectedness described by Parfit. What matters he insists when it comes to the point is ‘identity between the I who exists now and the surviving I who will I hope still exist then’.29 This is not Parfit’s view he relies on the mental continuity and connectedness. ‘One question two answers’. But on David Lewis’s view it all comes to the same thing. ‘We need not choose the answers are compatible and both are right’.30 The discrepancy is only ‘formal’. ‘What matters is that one and the same continuant person should have stages both now and later. Identity among stages has nothing to do with it since stages are momentary. Even if you survive your present stage is not identical to any future stage. You know that your present stage will not survive the battle — that is not disconcerting — but will you survive’?31
The answer is that ‘identity among continuant persons induces a relation among stages: the relation that holds between the several stages of a single continuant person’.32 This is ‘the I-relation’ and it is this that we must compare with the R-relation defined as ‘some definite relation of mental continuity and connectedness among person-stages’.33 ‘You wonder whether the continuant person that includes your present stage is identical with any of the continuant persons that continue beyond the battle’.34 It is a question in fact of ‘whether any of the stages that will exist afterward is I-related to — belongs to the same person as — your present stage’.35 But Lewis claims that ‘any stage is I-related and R-related to exactly the same stages’.36 The ‘I-relation is the R-relation’. For ‘if we left out any stages that were I-related to one another and to all the stages we included then what we would have would not be a whole continuant person but only part of one. For short: a person is a maximal I-inter-related aggregate’.37 ‘I cannot tolerate any discrepancy in formal character between the I-relation and the R-relation for I have claimed that these relations are one and the same …although the I-relation is not identity’.38
This comes extremely close to what I have myself been maintaining in somewhat different terms. The experience ‘the stage’ as it is put here belongs to the continuant person who is not himself the stage — the ‘present stage will not survive the battle’39 and I do not see how anyone could expect that. The continuant will and that is what matters. This is why David Lewis holds that ‘the I-relation does inherit much of the formal character of identity. But ordinarily the R-relation also is well behaved’.40 The difficulty arises in the problem cases. How does that come about?
It comes about in the case of fission for example because the same stage may belong to two or more continuant persons and these very strangely may be persons to a ‘diminished degree’.41 There may even be such a case that there is no such thing ‘as the person of whom S is a stage’.42 The I-relation may not have the formal character of identity in the problem cases it may not be so if there is not ‘some one continuant person of whom both S1 and S2 are stages’.43 It might be possible to deal with this by supposing that there could be a ‘partial overlap’44 of continuant persons. This would preserve the I-relation and its transitivity. Some stages would be shared. But this is not possible for there could then be more continuants than those that have stages at one time. In the case of fission for instance there would be two persons all along. A common ‘initial segment’ is ruled out because ‘on the day before the fission only one person entered the duplication centre’.45 The way out is to count not by a method that agrees with the result of counting stages but by ‘a weaker relation’ of ‘tensed identity’. This is not an identity among continuants but ‘an identity-at-a-time’. ‘So counting by identity-yesterday there was only one (person). Counting by identity-today there were two’.46 ‘There is a way of counting on which there are two all along; but there is another way on which there are first one and then two. The latter has obvious practical advantages. It should be no surprise if it is the way we prefer’.47
The justification for this is that in the problem cases ‘we cannot consistently say quite all the things we feel inclined to. We must strike the best compromise among our conflicting initial opinions’.48 In the case of fusion we can vary this by counting ‘at t by the relation of identity-at-all-times-up-to-t’.49
The final conclusion is: ‘if the R-relation is the I-relation and in particular if continuant persons are maximal R-interrelated aggregates of person-stages then cases of fission and fusion must be treated as cases of stage-sharing between different partially overlapping continuant persons’.50
This idea of ‘stage-sharing’ and of ‘partially overlapping continuant persons’ seems to me to do the greatest violence to all we find ourselves normally disposed to think about ourselves. I am not a creature who can have a pain let us say by proxy. I have it while I have it or not at all. I cannot have it by sharing my existence with someone whose experience at some time I did not have at all. No amount of ingenuity will identify me with a person whose stage of being yesterday was not my own. The kind of identity posited here is entirely artificial. I am the same person as I was yesterday because I had the experiences I had yesterday. This is not a matter of convention but of fact.
This is why it seems to me out of the question for me to become two or more persons or for two persons to become me. It is not a question of ingeniously finding a way of representing this. I may be obliterated and two or more persons take my place or use my brain or neurological system — or a part of it. But if it is to be me it has to be wholly me as an integrated irreducible being however changed or confused or many-sided. If another who has part of my brain shares my memories he has to be me. I do not see how it can be otherwise in the nature of things and I fail to see how David Lewis has provided any advance in the problem cases on the insurmountable difficulties they provided for Parfit.
This relates to the celebrated example of Methuselah alleged to have lived nine hundred and sixty-nine years. The mental connectedness which matters most for Parfit fades with the passage of years. We are asked to suppose that they fade out completely in a matter of around 137 years. By the time this has been repeated nothing of the relationship that matters remains and so for all that matters Methuselah by this time will be another person. This may be approached at least by very sharp changes of personality brain damage amnesia conversions senility. But quite independently of this Methuselah will have been many persons by the time he dies in no way confined to the specific periods from which we begin to count for 137 years. ‘There are infinitely many different 137-year segments that include all of Methuselah’s stages on his 300th birthday’.51 We may get out of this however according to David Lewis by counting in terms of ‘tensed identity’. This leaves us only one person for ‘all the continuum many nonidentical continuant persons are identical-at-the-time-in-question’.52 It follows however that while identity cannot be a matter of degree the t-relation may be so; and in this way we may come into line with Parfit’s position. In cases of fusion for instance there may be relatedness to a reduced degree and in the case of longevity while we have to hold on to identity in the strict sense the t-relation can become less significant or weaker and could for all that matters fade altogether. At a certain point it matters as little to Methuselah now what his lot will be centuries hence as if he were not strictly the identical person he is. ‘In this way personal identity can be just as much a matter of degree as the mental continuity of connectedness that matters in survival’.53
In effect the two positions are so similar as to make no difference — or at most a merely formal one. They invite the same comment. If I am assured that I shall live as long as Methuselah and shall spend the last two years of it in agony I am certainly not so upset as I would be if threatened with the start of the same torture next week or even ten years hence. But I still know that it will be me and am disturbed by this fact just as much in David Lewis’s account of things as in those of Parfit. Remoteness makes a difference but as the centuries creep on my trepidation will increase because it will be me I the unique being that I am who will be the one that is going to suffer (or have a marvellous time) a thousand years hence.
In his subsequent comment on David Lewis’s view54 Parfit points out that it all depends on the way different people might share stages when the alleged t-relation between them holds. But this he points out is not possible after fission when there really are two persons ‘two bodies and a divided mind’.55 That the two minds seem to remember the same things will not save the situation they must really remember and be the same person. Otherwise it is not identity that matters but something more akin to Parfit’s own connectedness etc. David Lewis’s case gains its plausibility by subtly importing into his crucial arguments the assumption I have been defending that I really am strictly identical in all stages whether I remember them or not if they are stages of me and therefore that it is this strict identity that matters. I may be peculiarly interested in someone very like me who seems to have my memories etc. but he will not be me on that token alone.
What happens on Parfit’s view has degrees. Psychologically this may be so we have a ‘discount rate’ in respect to our distant future. But this does nothing to alter the facts of the case. Myself as Methuselah dying under torture will still in the fullest sense be me however little it may trouble me now.
Professor Georges Rey has a somewhat different attack on the problem though maintaining like Parfit that ‘what matters to us is not identity over time’56 and a one-one relation. She even suggests that more extensive duplication might be a more exciting prospect and that we may look forward to surviving as several different persons. In this way she agrees with Parfit that identity is not what matters but rather our ‘serious concern’ with survival and what this involves. Identity may only come in in a secondary way and she even goes so far as to say in a moral instance that moral blame could not be escaped by a person contriving his own fission. This is an incidental point or relatively so; but I would certainly maintain that proper moral estimation not always identical with what is more expedient in treatment for social and like purposes would be very seriously affected if fission strictly meant that we had now two distinct persons. I would be disposed to say that there could be no moral disapprobation affecting these two persons of what had been done by the person we had before fission. He just would not be the same person any more and one or more persons cannot be morally blamed on my view for what someone else had done. There would be nobody remaining accountable for what we could blame morally — it ended with the fission.
The matter would of course be entirely different if as I have also envisaged the one original person continues functioning with different bodies. He would certainly retain his responsibility undiminished. But this is not the situation which Rey envisages. She does not think that it is identity that matters here any more than in the case of survival.
On the other hand she certainly thinks that something does matter — ‘what conditions underwrite our usual personal concern’?57 We cannot regard this as purely arbitrary a matter eventually of decision or persuasion as might be the case she supposes in determining what is to count as humiliating or honourable. Even these matters are not wholly capricious; and certainly ‘not all decisions are arbitrary’.58 ‘There must be some fairly well defined or definable basis rooted in some general view upon which we may systematically apply or withhold our concern’.59 It is such a basis we must find in place of identity.
There is therefore no reason not to ‘take seriously the kind of concern each person normally exhibits about all and only the person with whom she is identical’.60 But there must be ‘a basis’ for this and that is what must be seriously sought. This is closely related to ‘the relation that obtains between a person and usually only her own present experiences’.61 However we may be gladdened or pained by other people’s impending pleasures or pains only each one can be personally concerned about his or her impending experience — ‘only that person about only hers’.62 This may be due to the fact that each person feels or has some unique privileged access to his own experiences ‘his present conscious states’63 and this as must be evident is what I would maintain myself. But Rey will have none of this. We may be in error about our own experiences and we have thus in all cases to ask she claims how we each enjoy a personally privileged epistemological stance what natural conditions an entity must satisfy to enjoy64 this epistemological stance. This is quite at odds with my own view that we are aware of these experiences in just having them whatever their causes. She also claims that if there were telepathy we should be having the same ‘privileged stance’ to the conscious states of other persons or even most mysteriously to experiences a pain for instance which that other person does not have. There could indeed be ‘inculcated’ pains induced by extreme concern or sympathy or in the bizarre case of a person’s own writhings giving him the pains his wife is taken to have in labour. But this is quite artificial and gets us nowhere near actually having or undergoing the pains of another person. On the view I am defending it is just inherently impossible to have another person’s experience in this way or to enter into it as it is for the person himself.
But what account then does Professor Rey offer of what others of us would regard as knowing an experience in having it? To deal with this she takes her cue from the well-known position of H. P. Grice who says that a person has a particular experience notably perception if it is caused in some specific appropriate manner. If this means that the experience and my identity as the being that has it can be accounted for in all essential terms in terms of the causal conditions fully considered it seems to me as will be plain from what has been said already that it is totally unacceptable. The experience just is what it is or what I find it to be and my having it is also just that ultimate fact. But let us leave this aside and see how Rey proceeds. She claims that ‘perception and privileged belief …must be the products of particular kinds of processes’.65 If there seems to be a break-down of this it must be explained as in the case of possible telepathy or of a person’s having or seeming to have his wife’s labour pains in terms of peculiar causal conditions.
It is the causal condition that is thus really ultimate.
Certain behaviour is pain behaviour because it is typically caused in a specific way by a person’s experiences of pains which are also typically the causes of her abilities to relate and compare them as of her privileged beliefs. Typically that is to say the evidence converges. Indeed I do not think it would be far wrong to conceive of the network formed by these typical and appropriate kinds of causal chains or mechanisms as largely constitutive of the experiential relation: without any of the chains — without the appropriate causes for the (dispositions toward) privileged reports the abilities to compare the behaviour typical of the particular states — it would seem extravagant to suppose that there was nonetheless some full conscious experience. At least one would expect ‘the experience’ would be diminished. It is phenomena such as these and especially their interrelations that are the material insofar as there be any material of conscious life. Given then that this experiential relation is the object of our present tense personal concern it would seem that the basis for that concern at least at any particular time consists in the normal functioning of that causal network.66
The functioning of that network entails personal identity ‘at the time’ ‘and across time’. There might be difficulties where this leads to extensive division but the key to solving them would be the experiential network. That is what matters for survival. Personal qualities character beliefs even memory will not suffice. For all these change and even memory may be just ‘seeming memory’. There may indeed be many persons at the present time some in remote parts of space hallucinating or otherwise having our conscious experience. But we would not be seriously personally concerned about such psychological duplicates. I should myself question this if the memory is not just a ‘seeming memory’ for reasons made clear already. But on the Gricean model followed by Rey there is no proper remembrance unless it is brought about in the appropriate way. Indeed she goes so far as to declare that ‘it would certainly be surprising if the appropriate kind of causal chain could not stretch between persons whom for independent reasons we might want to distinguish’.67 What matters is ‘that the experience has been produced in a very particular sort of way’.68
In support of this view reference is also made to a possible objection to the ‘mnemonic and psychological criterion’ of identity namely that we may dread or eagerly anticipate what may happen to us in changes like ‘dreams drugs love affairs meditations religious conversions’69 and so on notwithstanding that we ‘know full well we invariably will forget who we are during the dream’ etc. The point is I take it that if we do so forget our concern cannot turn on a psychological criterion. We must therefore look elsewhere to the way experiences are produced.
I have maintained however that even if we do forget in the dream for example what our previous situation was like or at least find our circumstances vastly changed we do not in the least lose the basic sense of who we are and the reason why I specially dread my having a nightmare or its like is that whatever its precise course it will be very unpleasant for me. In addition I do not think that even in very wild dreams we cut adrift altogether from awareness of the normal course of our lives. But on Rey’s view any psychological criterion comes down eventually to a ‘causal chain’ and this leads to our bodies. Causation requires some substance to sustain it. ‘Dispositions require some underlying continuous matter to be so disposed. Persons their capacities need embodiment’.70
This seems to me to be an unwarranted assumption and I do not see why along the lines of my chapter on ‘The World of Thoughts Alone’ in my The Self and Immortality we might not have beliefs characters etc. not sustained by any particular stuff at all. But for Rey this is inconceivable. ‘Failure in the functioning of the mechanisms of our personal embodiment is failure of ourselves to be embodied: we die even if the matter of which we are composed endures’.71
The most important ingredient in the required network are normally ‘our brains and nervous systems’ — not just flesh and bones and Sperry’s work is thus much more disturbing philosophically than that of Christiaan Barnard. But we are not quite dependent even on the crucial stuff of our bodies not even the continued identity of the brain for the latter as in Sperry’s examples can be dissipated but the causal network must survive. ‘The basis of what matters to us in our personal survival seems to consist in the not necessarily identical continuity of our functioning personal embodiment’.72 The underlying stuff could even be other than material substance though this seems not to be taken very seriously in the slighting tone of the reference to ‘rare ectoplasms distilled in the laboratories at Duke or in the jungles of Brazil’. The point is ‘not that we must be made of the stuff science has so far found; but that we must be made of something’.73 On this view reincarnation is not to be strictly ruled out and Catholics are thought to have an advantage over Protestants presumably because they are supposed to attach more importance to the resurrection of the body.
What matters is in any case the continuity of the causal network whatever the stuff that sustains it. But why should this concern us so much? Does it not matter much more that I should be myself whether sustained by some personal network or not? We know very little about the network it is a matter for specialists but we all have ideas of certain kinds of experience and their being enjoyed by us. This is what strictly concerns us and it is quite peripheral for normal concern and expectations how these come about. Would it give me any cause for anxiety or enjoyable anticipation apart from the expected experience to know that the network would persist? If the same experiences were to be produced in some other way as might well be in another mode of existence would this be cause for any serious concern? Professor Rey concludes with the suggestion that with the evanescence ‘in our concerns and accompanying differences in the distinctions we find useful to draw’ we may some day find it quite natural to be ‘indifferent to one’s own as opposed to anyone else’s personal fate’.74 The reverse seems to me to be the case.
An interesting and ingenious variation on these themes is provided by Professor John Perry. He inclines to the view though not wholly committed to it that personal existence depends on physical continuity but he also maintains that what matters in our concern about survival and our own future states is different. Bodily identity though not ‘a necessary nor a sufficient condition for a personal identity’ is ‘nevertheless importantly involved in our concept of personal identity …The stable relation between persons and bodies is in this sense not an accident although purely contingent; if the P-relation’ which explains the relations (H-relations) which are ‘stages of a single human being’ ‘was not generally accompanied by and causally related to the H-relation we would not have the concept of a person we do have’.75Having the same brain is at least a promising candidate for the P-relation’.76 But this does not explain the special concern we have about what happens to us in the future. If by pushing a button I can prevent someone being in great pain tomorrow I shall normally be very keen to do so. ‘But intuitively if the person is me I will have more reason or perhaps special reasons for pushing it’.77 What can be the reason for this?
The reason we are told is in terms of my projects for tomorrow. If I am in great pain I shall not be able to accomplish the projects I have in mind for tomorrow. ‘If I am not in pain tomorrow I will contribute to the “success” of many of my projects: I will work on this article help feed my children and so forth. If I am in great pain I will not do some of these things’.78 I should myself have thought there was a great deal more than this to it. It is not just my projects that will suffer I shall myself have to endure great pain. But let us keep to the importance of the continuity of our projects. This is something which at least at one level we have in mind in thinking of the continuity of a person. Our interests do sometimes undergo dramatic changes but generally we retain the idea of a person having ‘much the same desires goals loves hates’79 most of the time. It was stressed earlier that this is one important sense of being the same person. Moreover no one is as likely as I am to have the same concern for these projects ‘love my children vote for my candidates pay my bills and honour my promises as me’.80
But this does not wholly cover the case. It is not impossible for someone else to do all the work in question just as well. Indeed someone else may even do it better. I would still be more struck with horror at the knowledge that I rather than the sustainer of my projects will be hit by a truck tomorrow — because I would think of the dreadful ‘event as happening to me’.81 Why should this be?
The answer is given in terms of ‘identification’. ‘A person identifies with the participant in a past future or imaginary event when he imagines perceiving the event from the perspective of the participant; that is when he imagines seeing hearing smelling tasting feeling thinking remembering and so on what the person to whom the event happened did (or will or might) see hear smell taste feel think remember and the like as the event occurs’.82 I can imagine Napoleon losing the battle of Waterloo but not with any of the sorts of details mentioned.
Indeed we may not identify in the sense in question with many things in our own past or futures. When we do ‘the memory or expectation’ has more ‘impact’ upon us. I keep my equilibrium while I just know of the pain that is going to happen to me tomorrow and am not terrified till I start to imagine it. When I do so this does not give me more reason for pushing the button that will spare me more pain but it makes it more likely that I will do so. Identity is not a condition of identification. Nor do I have to imagine myself the same as the participant with whom I identify. There are in fact no limits to our ability to identify with various participants. But we are much more likely to identify with ourselves than others and with our own ‘future selves’. One of the reasons for this is that the participant with whom we identify may be significantly different from us (members of another species perhaps in some remote part of space) and secondly that we ourselves are in a better position to watch out for ourselves than others. Some may want a more metaphysical reason for wanting to get to the bottom of our concern for our future self. But for Professor Perry the empirical account he has offered provides an adequate ‘evolutionary derivation of the facts explained’ and generally meets the demand for explanation of why we have ‘a preponderance of reasons for actng in our own behalf and strong motivation for acting on them’.83
There still seems for Perry something incomplete about this. The reasons he has mentioned might be equally good for sparing someone else the pain. But there must be ‘special’ reasons in my own case. The main reason is in terms of projects that I have which no one is likely to complete as I wish them done. This would be strongest if it included ‘my being in a certain state’ (a ‘private project’) for that clearly requires that I be alive myself and not a substitute. We have many such ‘private projects’ and also some ‘ego projects’ which simply require my existence. In these ways it becomes important ‘not merely that this article be completed but that it be completed by me’. Why should this matter on Perry’s theory? It matters partly because I am not certain that it will not be ‘ill-completed and partly because it may be connected with many of my other non-private projects’. But suppose these doubts are set at rest. Suppose that everything I will do ‘for the rest of my life will be done and done as well and done in just the same way by someone else. Still I want that I complete it and not this benign impostor’.84
The answer is in terms of ‘habit’. ‘Usually surviving is the only way to achieve a good part of what we want done’85 and we thus naturally want to survive. Also we identify better with our own doings than with those of others. This ‘habit’ too is ‘ingrained in us by the demands of evolution and its utility for achieving our purposes in ordinary circumstances’86 and it explains in part at least the more metaphysical concern we are still inclined to have. Perry denies outright that more than this is involved in ‘my desire that it be I who finish the article’ ‘not only in the normal circumstances but even in the metaphysical’.87 He affirms that the only justification for my concern for a private project is ‘a derivative one’ — I do not really expect others to succeed as well. The real explanation is ‘the result of habit as it is and the demands of evolution’.88
If we remain reluctant to accept this the explanation is that it is peculiarly hard to believe that my projects will be carried out as I want them by another or to find adequate justification for this. A supporting reason is that so much that I wish to accomplish and complete depends on things in my past history including things that only I can remember. My duplicate will be lying or deceiving himself if he declares or implies that he remembers things which I did. The only way of avoiding this is of his having ‘been produced’ from me which is only meaningful through ‘a reliable causal relation’.89 If a team of scientists invent some way of disposing of me (because I have an incurable disease perhaps) and creating in secret a duplicate that ‘will simply take over my life’ then ‘he would not be me’.90 ‘But on my (Perry’s) account I would have the very same legitimate reasons to act now so as to secure for him future benefits as I would if he were me’.91 This is not a defect to be charged to Perry’s account ‘but an insight to be gained from it’.92
If after amnesia I were to change and become a very disagreeable or reprehensible person I would still care about this person and at moments perhaps be ashamed of him (certainly if I were told beforehand what would happen) but this it seems is to be explained entirely in terms of habit. The supposed ‘“ineffableness” of the fellow who would survive’ ‘when it is not the remnant of a bad theory of personal identity is simply the shadow of the enormous contribution that we are in the habit of expecting ourselves to make to the projects we have’.93
The importance that we do actually ascribe to identity is thus derivative. I will have many relations with people with whom I will identify (in Perry’s special sense). ‘But it is incredibly unlikely that I should have all of them (the relationships to others) I will have to myself to anyone else’.94 If however the ‘many relations’ or more strictly all of them are guaranteed if Smith (in the brain rejuvenation case) ‘can expect to have as tight a web of special relationships to the survivor of the operation as personal identity in its purest form could provide’ this ‘is all that need matter to him’.95
Now there can be little doubt that Professor Perry’s view is sound in important respects. Our projects do matter to us and in some cases (the statesman the reformer the artist the man of letters the scholar for example) these projects are of immense importance and such that in many cases the instigator of them would sacrifice his own life to ensure their fulfilment. If an author is told that he will be dead within a few days one of the matters that will grieve him most will be that the novel he is halfway through and which is going well will never be completed and certainly not completed as he would have it. But this I submit is not of the same order as my concern about my total extinction. In some cases of acute suffering extinction may be the preferable course. But short of some calamity of this kind being in prospect most of us would wish to survive as ourselves however ruined some of our major projects may be. In a long prison sentence or sustained illness I may accomplish little of what I had previously been expecting to achieve. But I shall be having some experiences and not all of them unpleasant or frustrating to the point of being unendurable. Most convicts would prefer a long or life sentence to execution. We are concerned that we should be having some experiences and that they should be had by us — by me this same individual that I am now.
It would be comforting to know that someone would complete my novel and do it superbly well that if other people survive in a future existence they will enjoy loving and co-operative relations with the friends of whom death has now deprived me. I am concerned that my parents will be happy and that their sanctified postmortem experience will continue to expand and be enriched. But this is not of the same order at all as that I shall meet them that I shall enjoy again their manifestations of their love for me. This is the sort of expectation which religious people have normally had and I think they are right and certainly interpreting the New Testament aright. Projects short-term or long-term may have the greatest importance. Life would be dreary to the extent of pointlessness without them. But projects are one thing my own continuation as the distinct person I am (and also in distinctive relations to other persons) is quite another. In hoping to survive in some form after my demise in this world I am hoping that it will be me not some copy or duplicate but the same consciousness the same strict identity as I have now even if I have only dim memories of what I am now or pass beyond the importance they may have for me now — a contingency which I however think wholly unlikely in the matters that are of greatest moment in my present existence.

From the book: