Death and Utility

1.

In Practical Ethics Peter Singer describes his moral position as “broadly Utilitarian.” Though he mentions, with some skepticism, a theory of rights and Kantian principles of autonomy as providing support for some of his practical conclusions, the main theoretical interest of this book, and also some of its difficulties, lie in the striking combination of two different forms of utilitarianism on which he principally relies. In addition to the merits of many of its arguments his book has the virtues of the best utilitarian manner in moral philosophy. For Singer, as for Bentham and Mill, moral philosophy is primarily a confrontation of problems of current concern and controversy, and a critique of current assumptions and conventional prejudices.

Though the book is short, its range is wide and it includes discussions of the moral basis and the scope of principles of equality, of the human use and misuse of other animals, of euthanasia and abortion, of the obligations of the world’s rich to the world’s poor, and of violence and obedience to law. These topics are not brought in merely as examples to illustrate abstract principles, but are discussed in vivid and disturbing detail as the main business of moral philosophy, and of this book. Throughout the tone is temperate and fair to opposed views, but Singer is plainly inspired by a passionate wish to extend the range of our moral concern for the sufferings of all sentient beings beyond the limits which much conventional morality assigns to it.

There is some repetition in this book of some points made in Singer’s earlier well-known Animal Liberation,^* but the two books differ not only in scope but in the form of the underlying utilitarian principles. The earlier book, mainly concerned to argue the case for vegetarianism, assembled in convincing and horrifying detail the evidence of the hideous suffering annually inflicted on millions of animals, both in factory farming for the production of food for human consumption and also in laboratory experiments, some of which are pointless or carried out for no more important end than the testing of cosmetics, shampoos, food coloring, or floor polish.

Singer showed how “speciesism” is a self-interested human prejudice which treats mere membership of the species Homo sapiens as the cut-off point for moral concern for suffering. This ancient prejudice, buttressed by religious and philosophical myths such as that animals were appointed by God or Nature for the unrestrained dominion of Man, or that they were incapable of feeling pain, has had a profound effect on modern society. Hence modern liberals, though devoted to the causes of racial and sexual equality, could feel no inconsistency in remaining indifferent to the infliction on animals of suffering which they would think monstrous to inflict on the most retarded or defective human beings, human only in the sense that they are members of the species Homo sapiens.

It was enough for the purposes of the earlier book to oppose to the speciesism that permitted these cruelties a principle of minimizing suffering, which gives equal consideration to the equal suffering of all sentient beings. Singer did not discuss the precise relationship of this principle of equal consideration to general utilitarian doctrine in Animal Liberation, and with the exception of one short comment he avoided any discussion of the ethics of killing animals as distinct from their maltreatment. But the single comment was important since here Singer has had tentative second thoughts; it was made to dismiss the idea, first ventilated by Leslie Stephen, that if the humane rearing and painless killing of animals for human food could be secured, and the animals killed were replaced by others whose lives were made equally pleasant, there could be no objection to such food production and the moral case for vegetarianism would collapse.

“Of all the arguments for Vegetarianism,” wrote Stephen, “none is so weak as the argument from humanity. The pig has a stronger interest than anyone in the demand for bacon. If all the world were Jewish, there would be no pigs at all.”

In the present book in which the taking of life, human and nonhuman, is a central topic, Singer now treats Stephen’s idea with respect but draws a distinction between those sentient beings who are not persons and those who are. The former category includes, besides most nonhuman animals, human fetuses and newborn babies; while among “persons” Singer includes not only normal adults but some higher nonhuman animals as well. Merely sentient beings are replaceable for Singer, but persons are not. To sustain this result Singer now draws upon two forms of utilitarianism—“Classical” and “Preference Utilitarianism”—each of which may take one of two forms (“Total View” and “Prior Existence”). Unfortunately the argument concerning the taking of life that Singer develops with this more elaborate apparatus (which I explain below) falls at certain points below the high standards of clarity and cogency of the rest of the book. But if there are mistakes here they are of an illuminating kind, prompting new questions, and it is rewarding to inquire how the different strands of utilitarianism fit together in Singer’s thought.

2.

Classical Utilitarianism proposes as the sole standard of both personal and public morality a maximizing and collective principle: all that matters morally is that human conduct should maximize the aggregate happiness and diminish the aggregate misery of all those affected by it. This doctrine has an egalitarian aspect; it is, in the benign sense of the expression, “no respecter of persons.” This is so because if the only elements of value and disvalue are pleasures and pains (or as some utilitarians prefer to say, the experiences of desire-satisfaction and desire-frustration) then in the determination of what would most advance aggregate happiness, the equal pleasures and pains, satisfactions and frustrations of different persons must be given equal weight. Differences of status, race, sex, religion, age, and intelligence are morally irrelevant except when they affect, as they may sometimes do, the amount or intensity of pain or pleasure caused by human conduct.

Suffering is suffering no matter whose it is; and if suffering is equal, it is equally bad to inflict it on blacks or whites, women or men, Jews or Christians, the stupid and intelligent. “Everybody is to count for one and nobody for more than one.” This Benthamite maxim is the source and gives the essence of the principle which Singer calls “The Principle of Equal Consideration of Interests” and which he uses with telling effect in many chapters of this book. Bentham himself urged the extension of this principle to all beings capable of suffering including nonhuman animals. His words, in effect identifying speciesism as a vice as objectionable as racism, are memorable.

The French have already discovered that the blackness of the skin is no reason why a human being should be abandoned without redress to the caprice of a tormentor. It may one day come to be recognized that the number of the legs, the villosity of the skin, or the termination of the os sacrum, are reasons equally insufficient for abandoning a sensitive being to the same fate. What else is it that should trace the insuperable line? Is it the faculty of reason, or perhaps the faculty of discourse? But a full-grown horse or dog is beyond comparison a more rational, as well as a more conversable animal, than an infant of a day, or a week, or even a month, old. But suppose they were otherwise, what would it avail? The question is not, Can they reason? nor Can they talk? but, Can they suffer?

This egalitarian aspect of utilitarianism is the first of two main features of the doctrine extending the range of our practical moral concern beyond conventional limits. The second is the utilitarian’s refusal to endorse the common attribution of moral significance to the distinction between acts and omissions. If all that matters morally is the outcome of conduct then it cannot matter whether a bad outcome is brought about by our deliberate active intervention, or by passively failing to intervene and allowing things to take the course which we could have prevented. If I desire a blind man’s death, allowing him to walk over a cliff when I could have prevented this by crying “Stop” seems no better morally than pushing him over, even if the law does not recognize this as murder or any sort of crime. In this simple case the motivation of a positive act and the passive omission are the same, and the moral hollowness of the distinction seems evident. In other cases, as Singer illuminatingly shows, this may be disguised by differences in motivation, among other things. Failure to prevent harm owing to indifference, laziness, or meanness may be less alarming and less blameworthy than active intervention for the purpose of inflicting harm. But the difference is attributable to the difference in motivation, not to that between act and omission.

Singer interestingly exploits the utilitarian insistence on the moral equivalence of acts and omissions in two diametrically opposed ways. He uses it first to awaken the conscience of Western society to the moral gravity of allowing millions of the world’s poor to suffer or die through poverty, when many of them could be saved by aid which would still leave wealthy countries with a much higher standard of living. Here we allow to die or suffer where we could save, and at a cost trivial in comparison with the evil that that cost would avert. Conversely, much medical practice, which accepts passive euthanasia as if this was morally distinguishable from direct killing, allows incurable human beings to die slowly and painfully though they could be killed quickly and painlessly, and would want to be. In such cases we withhold from suffering human beings the mercy killing we would extend to a suffering nonhuman animal.

These two features of Classical Utilitarianism extending the boundaries of moral concern—its egalitarianism and its refusal to distinguish morally between acts and omissions—may in part account for the fact that utilitarianism was so long regarded as the main source of progressive social thought, and under the aegis of Bentham and his followers the inspiration of great legal reforms. But things have changed, and Singer’s emphasis on the humane aspects of utilitarianism now contrasts strongly with the insistence by many contemporary moral philosophers that utilitarianism has a sinister side, since it licenses the imposition of sacrifices on individuals whenever this can be shown to advance aggregate welfare. Such sacrifices may be licensed and even required by utilitarianism because the egalitarianism embodied in the maxim “everybody to count for one, and nobody for more than one” is only a weighting principle, to be used in calculating what will maximize aggregate happiness. It is not a principle requiring the equal treatment of different persons, and it may yield grossly inegalitarian results, though in certain circumstances it will tend to favor the equal distribution of economic resources. (This is because when resources are distributed unequally, any transfer from richer to poorer is likely to benefit the poorer person more than it harms the richer, and therefore the effect is to increase aggregate welfare.) Individual persons and the level of an individual’s happiness are only of instrumental not intrinsic importance for the utilitarian. Persons are merely the “receptacles” for the experiences which will increase or diminish aggregate welfare. So utilitarianism is “no respecter of persons” in a sinister as well as a benign sense of that expression.

Faced with this uncongenial aspect of his doctrine the utilitarian may argue that such sacrifice of individuals to advance the general welfare is only an abstract possibility and that in practice the “side effects” or more remote consequences of imposing sacrifices on innocent persons will outweigh any good that they may do. Such side effects include the grief of relatives and friends and also, when such practices become known, the general alarm and insecurity awakened by the thought of their repetition. So the aggregate welfare is not increased after all.

Singer realizes that the argument from side effects, powerful as it often is, will not always apply, and whether it applies or not he believes that there are deeper moral objections which a more sophisticated form of utilitarianism—“Preference Utilitarianism”—can urge against the sacrifice of individuals for the sake of increasing aggregate good. Preference Utilitarianism, so Singer contends, does not regard persons as if they were mere replaceable or interchangeable receptacles for contributions to such an aggregate.

3.

These deeper objections and the principles required to support them emerge progressively as Singer discusses various cases of the taking of life, starting with the killing of nonhuman animals. The present painful rearing and killing of animals for human consumption is plainly condemned by the principles of Classical Utilitarianism since, as Singer convincingly argues, the vast suffering involved is plainly not outweighed by the meat-eater’s pleasures in his diet or by any other advantage which that diet secures. But what if the rearing of animals is humane enough to secure them pleasant lives and the killing is painless? Is killing then wrong and if so, why?

Singer’s present though tentative answer to this question reflects his second thoughts. He now thinks it at least arguable that such killing would not be wrong if new animals for whom an equally pleasant life would be secured are bred to replace the animals we kill. But killing without such replacement would be simply to cut short a pleasant life and so to diminish the aggregate happiness in the world, and would, for Classical Utilitarianism, be wrong unless it were necessary to avoid some greater evil, as it might be if there were no other sources of human food. But to argue in this way is to bring to bear on the question a specific form of Classical Utilitarianism—which Singer calls the “Total View”—that conceives of aggregate happiness as something to be maximized not merely by our increasing the happiness of beings that already exist or would anyhow come into existence, but also by our creating new happy individuals for the sake of increasing total happiness. On this view our duty is not merely to make individuals (human and nonhuman) happier or less miserable but to create happy individuals.

Opposed to this Total View which permits killing with replacement and so affords a moral escape route for the meat-eater from vegetarianism—when adequately humane rearing and painless killing can be secured—is another version of Classical Utilitarianism. This is the “Prior Existence View” which does not recognize that there is any value in increasing aggregate happiness by creating additional beings. Because an increase so procured has on this view no value, it cannot make good the lost happiness of a being whom we kill. Hence there can on this view be no “replacement” of a happy animal killed by an equally happy new animal created, and hence no escape route for the meat-eater.

Singer is not of course concerned to find an escape route from vegetarianism and he stresses how narrow the escape provided by the Total View would be, for the argument could neither justify factory farming where animals do not have pleasant lives nor justify the killing of wild animals since that does not lead to their replacement by others. In any case Singer does not embrace the Total View with any enthusiasm. It is rather something to which he feels driven by the argument that since it is plain that we would do something bad if we were knowingly to conceive a child who would certainly lead a miserable life—e.g., owing to an inherited defect—it must follow that we would do something good when we bring a happy being into existence.

Singer thinks it is at least arguable that Total View Classical Utilitarianism, which allows killing and replacement, is the appropriate standard determining the morality of killing all sentient beings who are not “persons,” as he defines them. His resultant views on the morality of human abortion and infanticide challenge many accepted views. But even in its application to lower animals, the Total View has some strange implications. These are not, I think, made sufficiently clear by Singer because his attention is concentrated on the idea of replacement. The Total View, however, does not merely authorize or require replacement of a happy animal killed, but except in a situation where the world’s population of happy animals is at its optimum it would require us to breed additional happy animals of every sort, and not merely animals of use to human beings.

Of course in determining whether the optimum number is reached, attention would have to be paid both to the financial and other costs of breeding additional animals and to the strain on resources which might cause aggregate welfare to fall. No doubt men would in fact only breed new animals on any large scale for profit or if immunized against costs. But the fact, as Singer remarks in another context, that men would follow self-interest rather than do what this form of utilitarianism requires of them does not mean that their conduct would not be wrong according to its canons.

4.

Before considering Singer’s argument about abortion and infanticide it is necessary to take account of Singer’s conception of persons and of the Preference Utilitarianism which he thinks is applicable to our dealings with them. Instead of the speciesist’s distinction between sentient beings who are members of our species and those who are not—which he considers devoid of all moral significance—Singer stresses the great importance of the distinction between those who are persons and those who are not, which cuts across the speciesist’s division of the animal world into human and nonhuman.

Persons (including some higher animals like chimpanzees and dolphins) are distinguished from other sentient beings (like the lower nonhuman animals, the human fetus or newborn baby) by the possession of self-consciousness and rationality. A person is aware of itself as a distinct entity with a future and is capable of desires for that future, including the desire to live. Killing a person is therefore not merely cutting short a sentient being’s life, which could be “directly” wrong, i.e., apart from side effects, only if its future life is likely to be on balance pleasant. Killing a person is wrong just because it thwarts a person’s desire to live and his plans for his future, though of course he will not live to experience their frustration. Killing a person therefore is wrong for a reason which Classical Utilitarianism cannot recognize since the only constituent of what it seeks to maximize are conscious experiences of happiness. Preference Utilitarianism regards the frustration of desires or preferences as wrong in itself, other things being equal, since what it seeks to maximize is just the overall satisfaction of preferences.

5.

Singer therefore offers as a moral guide a combination of Preference and Classical Utilitarianism. Preference Utilitarianism will apply only to our dealings with persons and only so far as our actions satisfy or frustrate their preferences. Classical Utilitarianism will apply to our dealings with both persons and nonpersons so far as our actions affect their experiences of pleasure or pain. Thus killing a happy person will be doubly wrong. So far as it will frustrate his wish to live it will offend against Preference Utilitarianism; so far as it deprives the world of his future happiness it will, like killing a happy dog or other nonperson, offend against Classical Utilitarianism. This combination may appear tolerably clear so far, but unfortunately Singer’s account of Preference Utilitarianism is so sketchy and incomplete in this book that it is impossible to judge how it will work in practice on many issues, and on one major point quite central to his outlook Singer seems to provide no argument. Thus he insists that persons are not interchangeable or exchangeable and are not mere receptacles for impersonal experiences of pleasure and pain; so they are not replaceable and the creation of a new person with preferences which may be satisfied or thwarted cannot compensate for the death of a person killed against his will.

I do not think, however, that the non-replaceability of persons for Preference Utilitarianism can be established by anything that Singer says. Preference Utilitarianism is after all a form of maximizing utilitarianism: it requires that the overall satisfaction of different persons’ preferences be maximized just as Classical Utilitarianism requires overall experienced happiness to be maximized. So although a person’s wish to live is a reason for not killing him it is not a conclusive reason. Singer himself states that a person’s wish to live might be in conflict with the preferences of other persons and may be outweighed by them. So, to take a crude example, if the death of one person who wished to live was necessary to save the lives of ten others who wished to live, Preference Utilitarianism would support the killing of that one person. The fact that in some sense each person’s life plans are different from those of others would not affect the calculation of what in such cases would maximize overall preference satisfaction, for presumably in such calculations conflicting preferences would be weighted only by number and intensity, not by their content.

If preferences, even the desire to live, may be outweighed by the preferences of others, why cannot they be replaced by new preferences created to take their place? The fact that preferences may be outweighed shows that for Preference Utilitarianism persons are in a sense mere receptacles, not indeed for experiences of pleasure or pain, but for preferences the frustration of which in one person may be justified by the satisfaction of the preferences of other persons. Persons in their full concrete individuality may indeed be unique and, as Singer states, not interchangeable or exchangeable. But Preference Utilitarianism with its goal of maximizing overall preference satisfaction can only take notice of persons so far as they have preferences which may be thwarted or satisfied and there is nothing to show that such preferences evaluated for the purposes of Preference Utilitarianism by reference to numbers and intensity cannot be replaced by others as well as outweighed by others.

If there is nothing in the nature of a preference inconsistent with its replacement with another, then the only reason for regarding this as excluded by Preference Utilitarianism must be that Preference Utilitarianism cannot take a Total View form but must always take a Prior Existence form so that the creation of additional new persons whose preferences are satisfied would not count toward the goal of maximizing overall Preference Satisfaction. Singer hints at such a theory but does not argue for it, and it seems that he could not do so successfully; for an argument similar to that which led him to his second thoughts about Total View Classical Utilitarianism will apply to Preference Utilitarianism. Just as it would be plainly a bad thing deliberately to create a miserable being, so it would be plainly a bad thing deliberately to create a being most of whose preferences would be thwarted. If the first is a reason for accepting the Total View on the grounds that it follows that it must be a good thing to create a happy being, so must the second be, on the grounds that it follows that it must be a good thing to create a being in whose life Preference Satisfaction will predominate.

Singer indeed stands in need of a good argument for rejecting Total View Preference Utilitarianism, for its acceptance would license in principle, side effects apart, the painless killing of innocent persons in order to replace them by others in whose lives total net Preference Satisfaction would be greater. No doubt the argument from side effects in the form of terror and grief that would be generated by such practices would outweigh the advantages offered by this version of the Brave New World, but few would be willing to embrace and even less willing to teach to their children an ethic, however acceptable in other ways, that did not consider the killing of an innocent person with such replacement to be wrong, apart from its side effects. So the addition of Preference Utilitarianism to the repertoire of moral principles does nothing, I think, to meet the familiar criticism that unless it is constrained by quite independent principles such as a theory of individual rights might provide, utilitarianism cannot yield an acceptable account of the moral wrong of taking an innocent person’s life, even though Preference Utilitarianism might help to show why it is worse to take the life of a person than any other form of sentient life.

Similar points apply to part of Singer’s discussion of infanticide. He argues movingly and convincingly, with sufferers from spina bifida in mind, that it would be better painlessly to kill babies suffering from an incurable defect that would condemn them to a life of misery rather than allow them to linger on in pain to an early death. But he goes further and presents an argument based on Total View Utilitarianism that even where an inherited defect such as hemophilia would still leave the baby with the prospect of a worthwhile life in spite of much misery, it should be permissible, if adoption is not available, to kill the baby painlessly at the request of parents unwilling to take on the burden of its upbringing, but anxious to have another child with better prospects of a happy life if they could be relieved of their burden. But such an argument from the Total View cannot be confined to physically defective babies. In principle only the arguments from side effects would count against the killing of a normal baby in order to replace it with a child whose prospects of happiness were greater. Moreover on the Total View the secret killing of a normal happy child without replacement, by parents unwilling to be burdened with its upbringing, would be no greater moral wrong than that done by parents who abstain from conceiving a child for the same reasons.

6.

There are I think a number of other doubts that remain to be clarified before Singer’s combination of Classical and Preference Utilitarianism can be fully assessed, and some of these affect his discussion of abortion. He has some excellent pages in which he carefully examines and accepts the point frequently made in conservative arguments against abortion that, so far as being human is concerned, there is no morally significant dividing line between the zygote and the later fetus, the newborn child and the adult: all are human if being human means merely being a member of our species. But this, Singer says, settles nothing, since membership of a species is no more morally significant than membership of a particular race. What is significant, he claims, is that the fetus is not a person and so has no claim to life such as he believes Preference Utilitarianism recognizes in a person who desires to live. Fetal life has indeed a value but only the same value as the lives of nonhuman animals at the same level of consciousness and capacity for pleasure or pain.

At this point in the argument one would expect Singer to apply his general view derived from Classical Utilitarianism that it is wrong, other things being equal, to kill any existing being whose life is likely to contain or can be brought to contain more pleasure than pain. This applies to any being whether or not it is a person but Singer does not in fact consider this objection to abortion, which would involve weighing the future expected happiness of the fetus if it were allowed to live against the interests of the mother who wants the abortion. Instead Singer focuses on the question whether the fetus life has “any intrinsic value,” and decides that since up to eighteen weeks a fetus is probably not capable of feeling, an abortion at that point terminates an existence of “no value at all.” He concedes that an abortion at a later stage will end a life of some intrinsic value “but a woman’s serious interests will normally override this.”

This approach is very puzzling and seems inconsistent with much that Singer says elsewhere. It is not clear what the intrinsic value of a life at a given time can mean to a Classical Utilitarian or how it can be assessed without raising the question of the life’s expected future happiness which he himself actually raises in discussing infanticide. Of course it may be that the fact that the mother wishes for an abortion will often be evidence that the fetus’s life after birth will be unhappy. But for a utilitarian it is plain that such expectation of happiness, and not the value of the fetus’s life judged by the state of its development at the time of the abortion, is the relevant consideration to be weighed against the interests of the mother.

A wider objection to Singer’s treatment of abortion is that he sees only the interests of the mother as competing with the life of the fetus. This is to proceed as if only Classical Utilitarianism were applicable to the question of abortion. But how can Preference Utilitarianism be ignored here? The mother is a person though the fetus is not and her preference for abortion as distinct from her interests must be considered. But if they are to be considered we need to know, but are not told by Singer, how preferences are to be weighed against claims based on happiness. Does the satisfaction of a preference of a person always have priority over a nonperson’s claim based on happiness whether in the present or in the future? If so there would be very little to argue about over abortion; for the mother’s demand would settle the moral question. There are however many further difficulties. If Preference Utilitarianism is relevant why should only the mother’s preference be considered? If overall Preference Satisfaction is to be maximized will not all who have preferences in the matter, including the anti-abortion lobby, have to be taken into account? Unfortunately Singer has not said enough about Preference Utilitarianism to answer any of these questions.

7.

I have not been able to do justice to the many excellent features of this book besides those which I have discussed. Among these are Singer’s discussion of the moral basis of human equality and his demonstration of the irrelevance of differences in intelligence or IQ-tested qualities either between races or individuals to the question of how we should treat beings with similar needs and interests. There is too a thoughtful discussion of Malthus-type arguments that large-scale relief of starving communities may exacerbate the problem by encouraging growth in population. Singer’s book is packed with admirably marshaled and detailed information, social, medical, and economic, and has a splendid appendix of notes and references to further reading. The utility of this utilitarian’s book to students of its subject can hardly be exaggerated.