The person and the glass tunnel

Among the things we see outside, there are other glass tunnels, which we suppose to contain other inner worlds, each doubly cut off from us as we are from them.

All this, we suppose, is real; but only the inner world of our glass tunnel seems real to us in the first degree. Out of philosophical laziness, perhaps, we are not solipsists, but solipsism seems quite credible to us. After all, all this outside could be a dream...

The tunnel there ends with our death. The world, we suppose, will still exist, but “our world”, the only one whose existence we feel assured of, will disappear forever. There will be a continuation, but we will not know it. There will be projects, sorrows and joys, but they will not be ours; so, we say of them, “What do I care? For me it will be the end of everything!”

We wish it were not so. So we think that there will be an “afterlife”, that is, a life of our own – the only one that really counts. So the best we can imagine is that our tunnel is endless.

This image of the glass tunnel comes to me from Derek Parfit. At the beginning of chapter 13 of Reasons and Persons1 , in an eloquent passage, he tells us how, when he believed in the reality of personal identity

[...] it was as if I were imprisoned within myself. My life seemed to me like a glass tunnel, through which I moved faster and faster as the years passed, and at the end of which there was darkness.

But once this notion was challenged:

[...] the walls of my glass tunnel disappeared. I now live in the open air.

Concerning his relationships with others:

[...] Other people are closer to me. I am less concerned with the rest of my own life, and more concerned with the lives of others.

Concerning his death:

Instead of saying, “I will be dead”, I should say, “there will be no future experiences that will be related in some way to these present experiences”. By putting the actual content of this fact back in front of me, this redescription makes it less depressing.

These profound changes in Parfit's attitude towards our innermost selves – towards life and death – are simply the result of his perception of a factual truth. This truth concerns the nature of our being, and, I would say, more generally that of sentient beings2. Parfit understood that there is no such thing as personal identity. Or, to use his formulation3 : that the personal identity that results from our relations of physical and psychological continuity is reduced to these relations; it is not a further fact. Parfit thus claims to be a reductionist about personal identity, in opposition to non-reductionism, which I will also call 'personalism' after a text I quote below.

Personal identity is the digital identity of the person, or subject, over time. This issue is in fact that of the very existence of the subject: if to each fact of sentience corresponded a new subject, this one would be confused with this fact of sentience and would have no real existence.

Reasons and Persons is meticulous in its reasoning and often difficult in its analytical style. Despite this, some presuppositions are not always made explicit, such as the identification made at the outset of rational choice with that concerning the interests of the individual alone. Some are also based simply on the author's intuitions4 . If the reader does not share them, he may be left wanting; in particular, while the “Repugnant Conclusion ”5 , which leads Parfit to reject utilitarianism itself, is not repugnant, some of the developments in the last part of the book, devoted to the search for a more impersonal ethics, seem less relevant. On the other hand, the question of personal identity is probably too close to us to ignore our intuitions about it. The Perfessor method consists rather in taking these seriously, as having a factual content, in order to confront them with thought experiments, the realities of neurology and the demands of logic. In this way, the critique also reaches our innermost self-evidence, which we thought could only be expressed in tautologies ('I am I').

Parfit is not the first to have questioned personal identity. In part, his conclusions are similar to those of Locke and Hume, to whom he refers. He tells us that the evidence of the Cartesian 'I' was challenged by Georg Christoph Lichtenberg:

Lichtenberg claimed that, in what he held to be the most certain, Descartes was mistaken. He should not have claimed that a thinker must be a separately existing entity. His famous cogito did not justify this belief. He should not have said: 'I think, therefore I am'. [...] [He] could have said instead: “It is thought: of thought takes place.” Or again: “This is a thought, therefore a thought at least is being thought.”6

Thus it seems that on the subject of personal identity there was an important debate in the seventeenth and eighteenth centuries, which is now largely forgotten. Utilitarianism, which was born at that time, appears to be the natural place for the criticism of personal identity, because it assigns to each moral agent the same goal, that of maximising the collective good, and because this good is defined on the basis of an impersonal aggregation of subjective facts. However, today it is mainly its opponents who criticise utilitarianism for being an impersonal theory and for considering individuals only as 'containers' for sensations, as Tom Regan puts it7 , even though utilitarians seem to have largely abandoned the challenge of personalism. I will return to R.M. Hare's universal prescriptivism, which is based on a personalist presupposition, as is this view expressed by Henry Sidgwick:

✓ It would be contrary to Common Sense to deny that the distinction between any one individual and any other is real and fundamental, and that consequently “I” am concerned with the quality of my existence as an individual in a sense, fundamentally important, in which I am not concerned with the quality of the existence of other individuals: and this being so, I do not see how it can be proved that this distinction is not to be taken as fundamental in determining the ultimate end of rational action for an individual. [...] [A man] may still hold that his own happiness is an end which it is irrational for him to sacrifice to any other; and that therefore a harmony between the maxim of Prudence and the maxim of Rational Benevolence must be somehow demonstrated, if morality is to be made completely rational. This latter view [...] is that which I myself hold.8

Reasons and Persons deals with important topics other than that of personal identity, which is, however, its core and to which I shall confine myself. Parfit's great merit seems to me to be that he has posed the debate from the fundamentally ethical angle of what matters9 . For beyond the descriptive discourse on the existence of an “I” entity persisting in time, it is the plane of the prescriptive that constitutes the crux of the matter. Despite Parfit's deep admiration for Sidgwick,10 his position is the opposite of the theses stated in the above passage. If Sidgwick is wrong and personal identity is not 'real and fundamental', it is not what matters, and it is on the basis of a false belief that we care about the quality of our existence in a radically different way from the way we care about the existence of others. It is on the basis of a false belief that we fear death so much. It is on the basis of a false belief that we consider prudence to be unproblematic and rationally founded, while despairing of finding any basis for ethics other than our own best interests. It is also on the basis of a false belief that we give a central place in ethics to personalist notions such as rights, distributive justice, reward and punishment, etc.

The term “reductionism” that Parfit claims for personal identity sometimes comes as a shock, suggesting a petty desire to deny what is great. However, one can be a reductionist about different things; we are generally, for example, reductionists about Father Christmas, believing that he is reduced to the more mundane causes of gifts appearing in children's shoes. So there is no reductionism 'in general'; each reductionism is to be judged on its own merits.

Parfit's use of the term may reflect a desire to maintain a distinction between the reductionist and the non-existenceistential thesis; that is, between the assertion that X is reduced to Y, and the assertion that X does not exist, the phenomena attributed to X being in fact due to Y. This distinction, however, seems to be justified only if one seeks to preserve, in the first assertion, the idea that X can nevertheless be more than Y, i.e., can be seen as an “emergent phenomenon”; otherwise, X is only a short name for the fact(s) Y that underlie X. But Parfit does tell us that personal identity is nothing more than the facts of physical and psychological continuity. The distinction between reduction and assertion of non-existence is therefore hardly justified; Parfit's thesis is, in essence, that of the non-existence of personal identity, and is opposed to the opposite thesis: that of its existence, as an entity irreducible to relations of physical and psychological continuity.

It is essential to distinguish the issue of personal identity, or the subject, from that of sentience, or subjectivity. Parfit's thesis precisely challenges the apparent evidence that there can be no subjectivity without a subject. However, there is also a reductionism of sentience, and many of the arguments that claim to show the non-existence of sentience, for example in Daniel Dennett11 , are in fact only valid against the existence of the subject. Parfit's reductionism, on the other hand, far from aiming to deny sentience, allows us to apprehend the events of sentience as existing in the world, rather than as 'private' entities existing only from the point of view of a subject.

Because the non-reductionist thesis merges with common sense and is rarely stated as such, I will cite a particularly clear expression of it:

At the heart of Pope John Paul II's personalism (his philosophy of the person) is the recognition that it is the concrete individual person who is the subject of consciousness. [...] This subject may exist before consciousness (as in the case of the human embryo) or during interruptions of consciousness (as during sleep or coma). But the existing subject is not to be identified with consciousness itself, which is an operation or activity of the subject12.

This subject is an object of a very specific nature, “mysterious and unique” according to the Catechism of the Catholic Church:

✓ In the kingdom, the mysterious and unique character of each person marked with God's name will shine forth in splendor. "To him who conquers . . . I will give a white stone, with a new name written on the stone which no one knows except him who receives it.” (Rev 2:17)13.

Parfit, on the other hand, takes a physicalist perspective, where sentience is what happens in our brain, and the book concludes with a call for the development of an atheistic ethic:

'Belief in God, or in a plurality of gods, has prevented the free development of moral reasoning. Non-belief in God, openly admitted by a majority, is a recent and still unfinished event. For this reason, non-religious ethics is still in its infancy14.

Yet it is among the declared rationalists that one finds some of the most virulent personalist proclamations, rationalism and materialism being interpreted as disqualifying ethics and 'therefore' as justifying 'every man for himself' against a background of exaltation of the individual and his heroic autonomy15. Without going so far, even the irreligious among us readily subscribe to the Sidgwickian belief in a “self” that separates us, in a fundamental sense, from any other “self”. Criticism of this conception seems to me to be one of the founding elements of a non-religious ethic.

Relationships of continuity

Our personal identity, one might think, is simply the identity of our body over time. This, however, is not really grounded, and especially is not really what we have in mind when we believe in personal identity.

We are well identified with our bodies in our daily interactions. This is operationally possible because each body describes a continuous spatial trajectory over time, which theoretically allows it to be tracked at all times and thus unambiguously assigned an invariant identity.

The material composition of our body changes over time, but again only continuously. Between two close instants, this composition having changed little, we find negligible the error committed in affirming that it is the same material body. However, after a few years our matter is almost entirely renewed. Can we then say that it is the same body, in the material sense?

This question has been asked at least since ancient times:

The ship on which Theseus embarked [...] was a thirty-oared galley, which the Athenians kept until the time of Demetrios of Phalerus. They removed the old parts as they spoiled, and replaced them with new ones, which they joined firmly to the old ones. Also the philosophers [...] maintain some that it was always the same [vessel], others that it was a different vessel16.

Starting from a vessel A, we obtain, by successive replacement of all its parts, a vessel B which, if it had been built directly from these new parts, no one would have thought of saying that it was the same vessel as A. Now the truth value of the proposition A = B depends on what A and B are, and not on the intermediate states. This value must therefore be the same, whether B was created by gradually replacing the parts of A or by directly assembling the same new parts. Since in the latter case we would not say that A = B, we cannot say so in the former. The same applies to our body: if after a certain time all its molecules have been replaced, we cannot say that it is the same body.

Above all, physical continuity does not account for our real conception of our identity. Indeed, whether we believe in heaven or not, we attach a meaning, true or false but intelligible, to the idea of going there. We believe that I is an object that may be factually, but not logically, impossible to exist without our bodies. And what matters to us is this object: it matters to us whether “we” go to heaven or not, not whether our body goes there. Our identity, this object whose future we care so much about, and in a way so different from the way we care about the future of others, is not constituted by our physical continuity. This only serves to identify it on a daily basis.

This 'self' is psychological. According to John Locke, personal identity is defined by memory:

'For as soon as an intelligent being can repeat the idea of any past action with the same consciousness as he had of it originally, and with the same consciousness as he has of any present action, to that extent it is the same personal self.'17

Parfit adds to this criterion other psychological continuities, such as the relationship between an intention to act and its subsequent execution. But he also notes that these relationships must be completed by transitivity: we rarely remember a distant past day, but every day we remember (much of) our experiences of the previous day. The relationship is considered to be established if it is close to each other18.

Parfit seems to accept that personal identity can validly be defined on the basis of such continuity relations, while arguing that this identity is not what matters. I find it preferable to say that such an identity cannot be one. Indeed, continuity relations, whether they concern the trajectory of bodies in space, their material composition or our memories, always suffer from the same defect: they make the identity A = B depend not on what A and B are, but on the intermediate states between A and B.

Because these continuity relations almost always coincide with our idea of our identity, however, it is difficult to highlight what we are in fact generally non-reductionist about. Parfit turns to thought experiments in which this coincidence is broken.

Teleportation

The idea of teleportation, as it appears in science fiction and is taken up by Parfit19 , involves the transfer not of the matter of our body, but of information about its instantaneous physical state. On Earth, the traveller enters a cabin and presses a button. A 'scanner' then records the position and state of every molecule in the body, including the brain, and sends this information via radio waves to another cabin on Mars, where it is used to create an identical copy from locally available material. The scanner also destroys the original body20 . The copy has all the memories of the original, will carry out all its projects and perceives itself as the same person.

It is as if, instead of being destroyed, the traveller has simply travelled. However, we can also describe the event as involving, on the one hand, the death of the original, and on the other hand, the creation on Mars of a copy that mistakenly believes itself to be the original. The criterion of physical continuity is broken, and if psychological continuity is maintained, it is not by the usual means. Memories, for example, are transmitted from the original to the copy by a causal chain involving the scanner, radio waves, and so on. If we accept such a causal chain as being able to constitute a relationship of continuity that founds personal identity, why should we not accept in the same way another causal chain, the one that occurs when we tell our memories to others, and which passes through sound waves, etc.? Where do we draw the line? It seems impossible in these circumstances to say whether personal identity, defined by this continuity, is maintained or not.

The question posed by Parfit is whether, at the moment of pressing the start button, the subject is justified in believing that he or she is about to die – and entertain the corresponding dark thoughts – or to travel. From the outside as well as the inside, there is no difference between teleportation and travel. I leave in the morning for Mars, and return in the evening by the same means. I – or the copy of my copy – have as vivid a memory of my breakfast on Earth as I do of the afternoon on Mars and the return journey in the evening. The copy of my copy will testify that everything went well; and who could, or ever could, contradict it? Parfit suggests that the repetition and trivialisation of the experience might even lead us to forget our initial fears. Are we right or wrong?

We have here a special kind of question, the status of which I think it is important to consider. It is not a question of what is (a descriptive question), nor what we should do (a prescriptive question), but of what we are entitled to be sad or happy about in advance. Anticipatory fear or pleasure may, of course, serve us when they prompt us to act; but the question here is what these feelings are in the absence of such utility. It is primarily this domain, which I will call anticipatory sympathy, that is covered by the Perfidian question of what matters: am I justified, when I press the teleport button, in feeling sad about my death or in looking forward to the beautiful afternoon that “I” will spend on Mars?

We can ask ourselves to what extent anticipatory sympathy is accessible to reason and fact. We could say that no reasoning that would conclude that I should or should not experience it can touch it, because it is not a matter of duty. If I am going to be tortured tomorrow, should I be anxious about it? It is expected that I will be; but if, while being fully aware of what will happen to me, I retain my good humour, so much the better! Tomorrow I will suffer, but why add present suffering? To what extent, then, can one ask, as Parfit does, whether I am justified in rejoicing, or in being sad, when I enter the teleporter's cabin? Should we not simply note what I am in fact doing?

This would be a mistake, and Parfit's question is justified. In one respect, anticipatory sympathy is not immune to reason and fact. If I am unhappy because I believe that tomorrow something will happen to me that in fact will not, I am wrongly unhappy. It is not that my emotions should be urged to change: it is that they will in fact change. Whether my fear of flying tomorrow is rational or not, or of undecidable rationality, it will not remain if I learn that tomorrow I will not fly. If it remains even though I have learned a truth that deprives it of its factual basis, it is because I have not managed to convince myself intimately of this truth (to 'realise' it).

Anticipatory sympathy is thus accessible to the cognitive. This in turn is accessible to reason: it can be reason that allows me to determine the facts.

If I feel sad about one situation, but happy about another, there must be a substantial difference between these situations, i.e. not just a difference in description. Parfit shows that in the case of teleportation, as in other cases he examines, it is possible to make a complete impersonal description of events, including their subjective aspects. Teleportation can be described as follows:

— in the departure cabin, it is thought: “I am on Earth and will press the button”;

— then it is thought, on Mars: “I am on Mars and think I remember pressing the button”.

Such a description is complete (describes the facts both materially and subjectively), and yet says nothing about whether the person on Mars is, or is not, the original traveller. There is therefore no substantial difference between the following two descriptions:

(1) The traveller is killed, and a new sentient being, qualitatively but not numerically identical, is created on Mars.

(2) The traveller is transported to Mars.

If there is no substantial difference between (1) and (2), then personal identity cannot be what matters. If I have intimately integrated the reasoning that leads to this conclusion, my anticipatory sympathy cannot be saddened by (1) while rejoicing at (2).

Parfit describes several other thought experiments, sometimes close to the present technological possibility, in which a description in terms of personal identity is impossible or arbitrary, whereas an impersonal description, which notes the mental states present in the world, locates them in a specific body, but does not attribute them to a subject persisting in time, is possible and complete. One example is the split-brain experiments, which give rise to two streams of consciousness21. Parfit imagines that the division can be reversed at will. The subjective description before the division – a single stream of consciousness -, during the time of division – two streams, each possessing the memories of the previous single stream – and after the reunification – a single stream, which possesses the memories of the original stream and of each of the two separate streams – is not problematic, whereas it would be hard to say which of the two streams, during the time of division, “is really” the original person.

The dissolution of the tunnels

We have seen that the descriptions (1) and (2) above are, in substance, identical. Let us see what makes (1) identical to (2), but different from the following proposition:

(3) The traveller is killed, and no copy is created.

The answer is: the creation of the copy. Thus, if I am in the position of the traveller who is about to press the button, what is going to happen far away concerns me, as if it were to happen to me; it is even insubstantial to say that it will happen to me or not. It is insubstantial to say whether it will happen in my inner world, or in the world of others. In short: what happens elsewhere in the world concerns me, just as much as what happens 'to me'. There is not an inner world, which necessarily concerns me, and an outer world, which only concerns me if I want it to. My glass tunnel is an illusion. The soft solipsism that marks the personalist attitude is not tenable.

It is the realisation of this fact that has produced the effects described by Parfit in this passage already quoted:

'Instead of saying, “I will be dead”, I should say, “there will be no future experiences that will be related in some way to these present experiences”. By putting the actual content of this fact in front of me, this redescription makes it less depressing.

If the “I” does not exist, and we are intimately convinced of this, it is impossible for us to fear its disappearance. We understand that there is indeed an afterlife: that of others, which is like our own and just as real.

Yet Parfit only says that death depresses him less. He wonders whether it is possible for us to really believe in reductionism:

Nagel has argued that, even if the reductionist thesis is true, psychologically it remains impossible for us to believe in it. So I will briefly review the arguments I have provided. I will then ask myself whether I can honestly say that I believe my conclusions. If so, I will assume that I am not the only one [...]22.

His answer is mixed:

'No matter how much I review my argument, I fear that I shall never be able to dispel my doubts entirely. At the reflexive, or intellectual level, I will remain convinced of the truth of the reductionist thesis. But there will always remain in me at some lower level the tendency to believe that there must necessarily be a real difference between a certain person being me, and her being someone else. A similar thing happens when I am on top of a skyscraper. I know there is no danger. But, looking down from that dizzying height, I am afraid. It's a similar irrational fear that I would have if I were about to press the [teleporter] button.

[...]

It is hard to believe that personal identity is not what matters. If tomorrow a person is to experience great suffering, it is hard to believe that the question of whether I will be that person is unanswerable. And it is hard to believe that, if I am about to pass out, there can be no definite answer to the question “Am I about to die?”

However, even a partial belief is enough to make an idea operative. I may never quite believe, when I see a plane on the ground, that such a thing can fly, but that does not stop me from getting on it. Like Parfit, I find it 'hard to believe' in the reductionist thesis entirely, but I feel powerfully helped by thinking about it to fear death less and to feel closer to others.

Parfit tells us that the walls of his glass tunnel have disappeared and that he now lives in the open air. The exclusive relationship that personal identity established 'from me to me' – between successive moments of my existence – cut me off from others and from the physical world itself. In reality, the relationship of me to me is always of the same nature as that of me to others. It is the causality of physics that ensures the structural continuity of our body and brain. It is this same physical causality that could, in a teleporter, ensure the structural continuity between the original and the copy. It is again physical causality that links our experiences and ideas to those of others, when we tell them about them. The epistemological cut-off between the observer and the external world is an illusion: the relationship of me to me, as well as that of me to others, passes through the world. Our own survival depends on its laws. If I really abolished all confidence in the world – as in the radical Cartesian doubt – I could not believe in my own existence in even a second.

Some consequences

The consequences of the critique of personal identity are many and profound. I will sketch some of them.

The relationship of me-now to me-future is of the same nature as that of me to others. Ethics is no more problematic in its foundation than prudence; or, put differently, prudence is no less problematic than ethics. We can no longer say that our future fate must logically concern us simply because it is our own. Prudence is as difficult to establish as ethics, and for the same reasons; it is no longer possible to make it a separate domain. The result is a situation of aporia: I seem to have no valid reason to do anything, since any act can only benefit a future being, which is not me-now. This aporia is, I think, fruitful, forcing us to redefine our relationship to the world24.

Prudence and ethics merge into a single domain: that of the prescriptive, or ethics in the broad sense. We find here a consequence of classical, hedonistic utilitarianism, which commands us to maximise the pleasure of the world, which includes our own. Recklessness can be seen as morally reprehensible, as can harm to others25. 25 Conversely, one can conclude that ethics is 'demoralised': harm to others is, like recklessness, a mistake, due to the fact that, trapped in our glass tunnel, we do not see the pleasures and pains of others as fully existent.

While classical utilitarianism is impersonal, more modern versions such as R.M. Hare's universal prescriptivism26 and Peter Singer's preference utilitarianism27 tend to reintroduce a large measure of personalism. This trend seems to me to be untenable, at least at the critical level28. I will sketch out the reason for this by taking the example of Hare. Universal prescriptivism does express moral duties in impersonal terms, but is characterised by the maintenance of an essential residue of arbitrariness in the individual's determination of his or her own good:

'Some preferences, even in the prudential realm, may well be more rational than others [...]; there remains nonetheless an irreducible large minimum of purely autonomous preferences which rational thought can only accept for what they are, or will be.'29

Any self-regarding preference can only concern one's future self; if the relation of now-me to future-me is of the same nature as that of now-other to future-me, Hare's privileging of now-me, but not now-other, in the arbitrary determination of preferences concerning future-me is unjustifiable. If a residue of arbitrariness is to remain, it should allow everyone to choose each other's preferences; but then we would have a multiplicity of arbitrators for them, which is absurd. This element of arbitrariness cannot therefore remain.

If ethical propositions no longer have any arbitrariness, they are likely to have a truth value. The critique of personal identity thus promotes ethical realism, according to which ethical truths are, in a certain sense, part of the world. It also leads us to conceive sentience itself as part of the world. As long as subjective experience was conceived as inherently linked to a subject, it existed only for that subject: it was 'private'. If others took my pleasures and pains into account, it was by an act of heroic faith, or by the hope of reciprocity. The critique of personal identity leads us to see that others' experiences exist in the world in their own right as our own future experiences. This realism of sentience seems to me to be a necessary condition for solving the 'hard problem of consciousness' (as David Chalmers puts it), that is, the very existence of sentience within the physical world. On the contrary, the subject, being neither subjectively nor objectively perceived, is “mysterious and unique” and outside the world. As long as the existence of subjectivity is conceived as implying the existence of the subject, the “hard problem” is a hopeless problem.

By taking sentience seriously, the critique of personal identity leads us to take into account any sentience, however different from our own; and this as if it were our own, and therefore equal to our own. However, the notion of equality, as it is formulated in the purely human framework, is mainly relative to the person, in a non-reductionist sense. We speak of the equality of value or dignity of persons. Death appears as an absolute evil, since it is the destruction of the person; homicide is the ultimate crime. In contrast, since non-human animals are not seen as persons – they live only in the moment, it is said – their lives are worth almost nothing, less, for example, than the pleasure we take in eating their bodies. The notion of animal equality thus seems to me to be unformulable in a personalistic framework; the immediate answer would be: the life of a fly cannot be worth that of a human.

Utilitarianism, which advocates equality not in terms of the value of persons, but in terms of taking their interests into account, allows for a clear advance in this area. However, the version proposed by Peter Singer following Hare, preference utilitarianism, reintroduces the notion of the person30 , and makes it difficult to take serious account of the vast majority of sentient beings. It seems to me that only by freeing ourselves from the notion of personhood, through the critique of personal identity, can we arrive at a coherent ethical vision, at least at the critical level, that can take into account sentient beings beyond the boundaries of our species.

This critique has many other political and economic consequences, especially in relation to private property, the notion of contract, merit, etc. It relativises other aspects of the social and economic life of the individual. It relativises other notions of identity, such as national identity. It has the potential to bring about profound changes in the whole of our thinking, in our vision of ourselves, in our individual and collective inter-human relations, in our relations with other sentients and in our understanding of our position in the world.