Ten or fifteen years ago, the complaint against moral philosophy was that it did not address practical problems, but concentrated on meta-ethics: that is to say, on questions about the status, meaning, objectivity and so forth of ethical thought. That complaint is now out of date. For a decade, analytical philosophy has been conspicuously concerned to display its credentials for being of use in helping us to think about concrete problems.
In doing that, it has escaped the charge of evasiveness, but has slipped back into the line of fire of other accusations. One is that it has disconnected itself from other speculative, critical or, indeed, philosophical thought. Philosophers have tended to turn to ethical theory, an enterprise that tries to resolve practical dilemmas by appealing to a structure of moral principles, a systematic framework which philosophical ingenuity can hope to apply to concrete issues. This raises the question why a set of ideas should be thought to have any special authority over our sentiments and our lives because it has the structure of a theory. Besides having this very basic problem of what might be called theoretical authority, ethical theory has sometimes been impoverished because it has cultivated too much the autonomy of ethics, and neglected other areas of philosophy, and (with the exception of some philosophers such as John Rawls) other disciplines.
Derek Parfit has written a brilliantly clever and imaginative book which treats in a very original way a wide range of ethical questions. It spends virtually no time on meta-ethics (perhaps too little), but it avoids many of the deformations that sometimes afflict first-order ethical philosophy. It makes contact with other subjects, such as welfare economics. It is deeply involved with some other parts of philosophy, in particular with questions of personal identity and of what a person is. It also starts the subject, rightly, not within the sphere of morality but in the wider area of practical reason, setting out from the question ‘what have we most reason to do?’ rather than from any distinctively ‘moral’ question.
Within ethical thought, Parfit does not start off with any ethical system. Nor does he hope to conjure one out of nothing at all. He concentrates on questions of consistency, asking us, over and over again, in different connections, what is implied by our ethical judgments, and whether what is implied hangs together with other implications to which, equally, we seem to be committed. That is not his only method. He uses many methods of ethical argument, more than moral philosophers often acknowledge. It is only when in his concluding chapter he quietly displays a few of them, that one realises how naturally they have been deployed. In these ways he goes some way to meet the problem of theoretical authority – though not, I believe, far enough.
In starting with practical reason, and in some of his methods of argument, Parfit agrees with the Victorian moral philosopher Sidgwick, whom he greatly admires. Keynes thought that Sidgwick lacked intensity and was suffocated by respectability. Parfit would deny these charges against Sidgwick, but whether he is right in that or not, the charges certainly do not apply to this strange and excitingly intense book. It is in four parts. In the first, Parfit considers what it is for a theory of rational action to be, in any of various ways, self-defeating. He deals, very subtly, with such problems as this: if one believes that one’s aim should be to produce the best outcomes all round, it is very unlikely that the best way to do this is to consider, on each occasion, how one can bring about the best outcome. The best outcomes are more likely to be produced if each person acts from motives which do not involve thinking directly about the outcome. This has been thought to be a problem for consequentialist theories of this kind. Parfit insists that it is not, and that this result does nothing to refute the theory that we should produce the best outcomes all round. It merely tells us how to produce them, by cultivating in ourselves other dispositions. In other cases, however, theories can be damagingly self-defeating, by enjoining on each of us courses of action which, when we all pursue them, collectively defeat the objectives at which the theory was aiming in the first place (which is not so, Parfit claims, with the innocuously self-defeating consequentialist theories).
In these connections, Parfit has a lot to say about problems that have concerned decision theorists, such as the famous Prisoner’s Dilemma, which makes it distressingly clear how courses of action that are individually rational can be jointly damaging, and also that when the parties know that fact, they may still have good reason to follow them. These issues, and many others of the same kind which he discusses, have a good deal to do with politics. Parfit makes it clear that they do, but he does not for the most part discuss them as though they did. The discussion is detailed, quite hard, and very revealing, but in social or political terms it is rather airless. He does not consider what institutions would be needed, for instance, or what forms of social understanding, in order to do what he, like Sidgwick, recommends us to do, which is to induce in ourselves dispositions of action which serve the ends of an underlying ethical theory while not revealing its content.
It is perhaps a pity that this rather daunting section has to come before the winningly ingenious discussion of rationality and time that forms the second part. In this Parfit asks whether we should be more concerned with what will happen tomorrow than with what will happen years from now, and if so, why. That is only the most familiar of such questions. He also wants to ask, for instance, why we should be more concerned with what will (or rather, may still) happen than we are with what has happened. Why is it good news that the nasty operation has already happened? If one is disposed to think that this issue, at least, is perfectly obvious, Parfit, with a very light touch, can turn one round to see that it is not.
In this section, too, he makes some important moves in a campaign which runs throughout the book and helps to unify it – the war against the Self-Interest Theory, which holds that the rational thing to do is to be concerned with one’s own aims and interests as viewed, so far as possible, over one’s whole life. This war Parfit conducts on two fronts, as he puts it. On one side, the theory is harassed by Morality, which says that we should be concerned with more than ourselves – for instance, with everybody. On the near side, it is undercut by the Present Aim Theory, which says that what it is rational for one to do now is what one wants now. For this view of things, or rather for a slightly more respectable version of it, Parfit makes a very good case against the Self-Interest Theory. One of his main objectives is to show that prudence does not have the special priority in rational behaviour that is often given to it. This is a good objective, but feckless readers who hoped to be liberated by it will find their enthusiasm dampened when they learn later that there is something wrong with imprudence after all: it is not irrational but immoral.
The reason for this is that our later selves are properly to be seen as rather like other people. Parfit is trying to get us to see that in practical reasoning ‘when?’ is much the same sort of question as ‘who?’ We should get rid of the picture that dominates us, or most of us, that there is some special identity that one has, some underlying item which is really me. We should get rid of the very compelling idea that there must always be a fully determinate answer to such questions as: ‘Will that person who will be in pain in ten years’ time be me or not?’ On the true view of things, according to Parfit, there may be simply no answer to that question. We should realise that, as Hume believed, a person is no more than a collection of experiences held together by certain relations, such as those of memory and continuity of character. When we see that, we shall understand that it is misguided to draw a sharp ethical or prudential line between ourselves and others.
These are the subjects of the third part of the book. In the final part, Parfit turns to problems raised by our concern for future generations, in particular by population policy and the question of how many people there should desirably be. As with personal identity, he has already published articles on this subject, and has made notable contributions to it – for instance, in discovering what he calls the Identity Problem. This lies in the fact that when we discuss whether future people will be better-off or not as a result of our policies, we cannot suppose that the same people will be there to be affected by one or another of our policies, since our actions will radically affect what individual people will come to exist. Parfit shows how arguments that may seem plausible in this area can lead to undesirable results, such as the Repugnant Conclusion, as he calls it, according to which an indefinitely large population of people whose lives were just worth living would be morally preferable to a smaller population of people who were a lot better-off. Parfit tries to find a theory that will avoid this result and at the same time certain other paradoxes. In the end, despite much ingenuity and refinement of argument, he confesses failure: but he can claim credit for identifying some remarkable problems along the way, which will undoubtedly generate discussion for a long time to come.
The intensity displayed by the book is in good part argumentative. Short, sharply-defined sentences are loosed at one in compact formations; the effect, at times, is of one who will not let you go. But there is an imaginative intensity as well, displayed above all in the examples, often simple, carefully designed, each presented with a title – a device that could have been arch if used with less skill. Many of these examples are fanciful, particularly in the personal identity section, where teletransportation, bodily fission and other fantasies are introduced to construct cases that challenge our everyday assurance that we know what would and what would not count as the same person. Such fanciful cases have often been used by the philosophers who over the past decades have helped to set the agenda of Parfit’s discussion. Others reject them, saying that our concepts have developed to deal with the actual, not with worlds extensively different from ours, and there is no reason to expect those concepts to be able to breathe that alien atmosphere. To this line, Parfit has several sophisticated replies. One is that this idea could explain why in certain unlikely cases we might not know what to say, but it can hardly explain why, with other equally unlikely cases, we do seem to know what we would say. In some matters, again, and personal identity is one of them, the whole idea of not being able to give an answer is something that our common notions seem to exclude, and is a basic part of the problem.
They are good replies, it seems to me, when these are regarded simply as metaphysical issues. But it is less clear why they are adequate when we are concerned, as Parfit is, with supposed ethical consequences of metaphysical positions. To put it another way, it is not always clear why metaphysical positions, arrived at in this way, have ethical consequences at all. Parfit is encouraged by his metaphysics of the merely agglomerated self to accept an ethical outlook which abstracts from self-interest and sees other people, and stages of oneself, as more like one another than we normally suppose. He thinks that philosophy should move us to a more impersonal outlook. But the extent to which it should do that must surely depend on what the world is actually like. If the experiences which constitute one person are powerfully related to one another, and give their owner (as Parfit, rather riskily, allows us to call that person) a strong sense of his or her own identity and of difference from others, why should a metaphysical belief, that he or she is really a fuzzy set of experiences, provide a reason for feeling and acting in some altered way?
Connections between metaphysical and ethical issues are central to this work, but it is not always made clear how they run. In at least one case, one which Parfit touches only very briefly, they do not run at all. He says that if, as some metaphysicians have claimed, the passage of time is an illusion, it cannot be irrational in practical thought to have no preference for one time over another, such as a preference for the near over the far. But this does not follow. If time’s passage is an illusion, so is the flow of time apparently involved in action and deliberation themselves; relative to the metaphysical truth of the matter, the whole enterprise of practical deliberation, and all the various principles that might be brought to it, would alike have to be bracketed. If time’s passage is an illusion, we live that illusion, and finding out that it was an illusion would not provide us with a reason for deliberating in one way rather than another within it.
Parfit can convert the metaphysical into the practical so easily, I suspect, because the view that he takes of the practical, and of experience in general, is throughout the book so radically external. Philosophically speaking – it is not true of his literary allusions – he sees everything from the outside. In dealing with personal identity, this conceals from him one of the main reasons why people think that it must be a determinate question whether some future experience will be theirs or not: that if it will be theirs, they can, as well as expecting that it will happen, also expect it, in the sense of imaginatively anticipating having it; and there seems to be no room for the idea that it is simply indeterminate whether I can appropriately do that or not. If Parfit had discussed that particular point, it would not necessarily have harmed his case, and it might even have helped to reconcile us to it. But in other ways his neglect of the first-personal view, in the theory of personal identity as in his earlier discussion of one’s need to induce certain dispositions in oneself, leaves a gap. When we think how the argument is to be understood and applied, a dimension is missing.
In one respect, Parfit leaves it unclear whether he has adequately applied his metaphysical conclusions to his own argument. In the last part of the book, where population policy is in question, the idea that people are only aggregates of experiences seems to have been left behind. The whole discussion rests on a notion which seems uneasily related to that idea, the notion of ‘a life worth living’. All Parfit’s paradoxes involve the question whether the people in various populations have lives which are, or are not, worth living. But the discussions of personal identity and of prudence have earlier led us to distrust the ethical importance of a life at all. Perhaps a life worth living need not be taken to mean a life which as a whole will have been worth living. Perhaps it just means some living which, at any given time, is worth living. But Parfit cannot, as things stand, simply contract it to that. Almost the only clue that he gives to what is meant by saying that a life is not worth living is that people who had a life very much not worth living would kill themselves if they could. But he cannot use that notion without reference to the identity of the life that such a person would be ending. On his own view, that involves the question of the lives which suicide would be preventing: meaning by that, not the children that the agent would not have, but the selves that he would not become. Parfit cannot use the willingness to commit suicide as a neutral test of how a person values his or her own life. If imprudence is, as Parfit says, immorality, then suicide is murder.
There is another question raised by the section on population policy, besides those that come from the metaphysics of persons. That section tests more severely than any other part of the book the reliability of our ethical reactions when we are confronted with extreme and very abstractly presented possibilities. Correspondingly, it is the part that most calls in question Parfit’s refusal to raise questions of meta-ethics. Asked by him to say whether it would be better if there were two large populations, not connected with each other, each consisting of people whose life was just worth living, rather than one of those populations with a standard of life rather higher, or some yet more complex question of the same kind, I may wonder what I am being invited to do. What real substance can such judgments possess?
The problem presses all the more when I have, for once, a belief on these questions that seems very solid, but it turns out that theoretical argument may lay it aside. Very many of us believe in what Parfit calls ‘The Asymmetry’. If any child that I had now would (very probably) have a miserable life, that in itself would be some reason against my having a child now. On the other hand, if any child I have is likely to have quite a happy life, that fact in itself is no reason for having a child rather than not. We do not think in terms of doing the child a good turn by bringing him or her into existence. Parfit argues that we should probably think in those terms. To me, I must confess, it seems that ‘The Asymmetry’ is as clearly valid as anything is in this area, and while we certainly need a philosophical account of that impression, I do not see how theory acquires the power to cancel it. If moral philosophy is to do as much as Parfit hopes, by his very abstract means, it badly needs an account of the authority of theory.
However, here as elsewhere, the conflicts that Parfit has discovered are entirely real, and his imaginative and powerful arguments have uncovered deep questions which have in most cases never been explored so thoroughly, while, in other cases, they have barely been thought about at all. They are important questions, for practice as well as for philosophy, and in a moving last chapter, Parfit makes it clear how important he takes them to be. This ingenious, unusual, compelling book fully meets the importance of its questions.
Send Letters To:
London Review of Books,
28 Little Russell Street
London, WC1A 2HN
Please include name, address, and a telephone number.