Article Text


Savulescu’s objections to the future of value argument
  1. D Marquis
  1. Correspondence to:
 Professor D Marquis
 Department of Philosophy, The University of Kansas, Lawrence, Kansas, USA;


This essay is a response to Julian Savulescu’s objections to the future of value argument for the immorality of abortion published in the Journal of Medical Ethics, June 2002. Firstly, Savulescu’s claim that the future of value argument has implausible implications is considered. The author argues that the argument does not have these implications. Secondly, properties which, according to Savulescu, could underwrite the wrongness of killing and that are acquired only after implantation, are considered. It is argued that none of these properties is an adequate basis for the distinction between wrongful and permissible killing.

Statistics from

The future of value argument for the immorality of abortion claims that the best explanation for the wrongness of killing children and adults is that killing us deprives us of our futures of value. Our futures of value consist of all of the goods of life that we would have experienced had we not been killed. Fetuses have futures like ours. Therefore, (given some defensible assumptions) abortion is seriously wrong on almost all occasions.1

Julian Savulescu has offered numerous objections to the future of value argument.2 He points out that “abortion and embryo destruction prevent a future of value, but that does not make them wrong”. Savulescu notes that many actions prevent the existence of a future of value. An animal’s future has value, but we do not believe that it is seriously wrong to kill animals. There is value in works of art, but an artist may have good reasons for not creating a work of art if such creation would prevent her from adequately caring for her children. “On a future of value argument, killing a fetus is like failing to conceive a baby”. Failure to clone a skin cell prevents the existence of a future of value, yet we do not believe that it is wrong not to clone skin cells.

Savulescu’s objections to the future of value argument can be understood as a dilemma. Either the argument entails that actions, such as a successful attempt not to conceive a child, are as seriously wrong as the murder of an adult or child or it entails that many actions (such as abortion) are, at most, only minor wrongs (like, perhaps, choosing not to create a work of art) whose apparent wrongness easily can be overridden by other considerations.2 On the first alternative the argument entails conclusions that are absurd. On the second alternative the argument does not show that abortions are seriously wrong. On either alternative the future of value argument is unsuccessful in showing that abortion is seriously wrong on almost all occasions.

A response to these objections requires an explanation both of the nature of a future of value and of the motivation behind the future of value theory. Before this theory appeared on the philosophical scene the debate in philosophy over the abortion issue appeared to have reached a stalemate. Typically critics of abortion argued that as fetuses are clearly human beings, it is wrong to kill them, for we all agree that it is seriously wrong to kill human beings on almost all occasions. Abortion’s defenders normally argued that since clearly fetuses are not people, they fail to possess the crucial property that underwrites the serious wrongness of killing adults and children on almost all occasions. The future of value argument is based on the claim that neither abortion’s defenders nor abortion’s critics had offered adequate theories of the serious wrongness of killing (Marquis,1 p 189). If we do not have an adequate understanding of what makes it wrong to kill us—that is, individuals in cases where the serious wrongness of killing is uncontroversial—how could we understand whether it is seriously wrong to end the life of a fetus?

According to the future of value theory, killing adults and children is wrong because it deprives them of all of the goods of life that they otherwise would have experienced. This seems right because it makes killing a harm, and not only a harm, but one of the most serious harms that can be inflicted on someone. This fits with the attitudes of people who face premature death. It does not rely on an illicit inference from a biological property (being a human being) or from a psychological property (being a person) to a moral property (having the right to life). According to the future of value theory, fetuses can be victims of abortion in exactly the same way as adults or children can be victims of murder.


Because this theory offers an account of how killing harms a victim, an action that the theory claims is wrong will affect a victim (Marquis,1 p 189). Therefore, the future of value theory does not imply that it is wrong not to create things of value, for in such cases there may be no victim. It does not (apparently) imply that deciding not to conceive a child is wrong, for (apparently) there is no victim in this situation either. It does not imply that it is wrong to kill non-human animals. Fetuses have futures that are so much like ours that they contain everything that ours contain. The futures of (non-human) animals do not. The future of value theory does not tell us whether non-human animal futures are sufficiently different as to make it permissible to kill such animals. Note the discussion of this in my paper (Marquis,1 p 191).

How some of Savulescu’s objections have gone astray can be explained. Many of Savulescu’s claims concerning the future of value argument are quite correct. It is true that abortion does prevent a future of value. It is also true that all deprivations of a future of value are preventions of a future of value. According to the future of value theory, all deprivations of a future of value are seriously wrong on almost all occasions. It does not follow, however, that, on the future of value theory, all preventions of a future of value are seriously wrong on almost all occasions. (This is because “All As are Bs” and “All As are Cs” do not entail “All Bs are Cs”.) What is needed for the wrong of killing is an individual who is deprived of a future of value. Mere preventions provide us neither with the requisite victim, nor with a deprivation.

Savulescu’s better objections to the future of value argument concern cases that seem to involve victims. He claims the argument implies it would be wrong to deprive a sperm and an unfertilised ovum (hereafter a UFO) of a future of value. He also claims it would be wrong to deprive any arbitrarily chosen human cell of a future of value by not cloning it. If Savulescu is correct, then, because of these implausible implications, the future of value argument must be rejected.

Consider first the sperm and UFO objection. The future of value of which I would be deprived by being killed is the valuable life of a later stage of me, of the same individual that I am now. Killing me deprives me only of my future of value, not your future of value, nor anyone else’s. Accordingly, if my parents had failed to conceive me, their inaction would have been wrong only if the sperm and the UFO that were my precursors were earlier stages of the same individual I am now. If that sperm and that UFO were earlier stages of me, then each of them would be the same individual as I. If each of them were the same individual as I, then, since identity is transitive, that sperm and that UFO were identical. They were not. It follows that the future of value theory does not imply that if my parents had failed to conceive me, their inaction would have been wrong. This argument can be generalised to show that the future of value theory does not imply that either contraception or decisions not to conceive are wrong.3

Savulescu’s failure to clone objection requires a more elaborate response. On the future of value view (and according to ordinary people) to kill someone is to deprive that individual of a future of value. To deprive someone of a future of value is to harm her. To harm her is to make her worse off than she otherwise would be or should be. Something about the victim is required to underwrite the truth of the claim about how things would have or should have gone for the victim if she had not been killed. Such a victim would have lived a longer life that she would have valued. Our conception of this longer life is well entrenched. It is based on our biological understanding of the natural history of a human organism, on the understanding we gain at an early age from contact with parents, grandparents, and older acquaintances of a natural and full human life span, and on a conception of a human life span found in literature. This comparison is what makes premature death a deprivation rather than a mere non-occurrence of the events of a life. Both “deprivation” and “harm” are implicitly comparative terms.

The nature of a skin cell (or other differentiated human cell) underwrites neither an entrenched notion of a natural human life span, nor an entrenched notion of any other life that is valuable to the skin cell. Therefore, a skin cell is not deprived of anything by a failure to clone. There are two reasons for this. Firstly, cloning involves destruction of a differentiated cell by removing its nucleus and inserting it into a germ cell from which the nucleus has been removed. Therefore, the differentiated cell has not been harmed by not being cloned, for if it were cloned, then it would have ceased to exist. Indeed, failing to clone it preserves it! Secondly, suppose that we waive the first point and suppose that cloning transforms a differentiated cell into an undifferentiated cell. The nature of differentiated cells is to perform the specialised tasks that differentiated cells perform. Their natural history is not affected by their not being transformed into something else. Compare killing a human being, whether that human being is a fetus, a child, or an adult. Her natural history is truncated by being killed. Therefore, the failure to clone objection can also be rebutted.

Savulescu might argue that all these responses fail to recognise the following distinction. He says: “There is a difference between killing or destroying something and preventing something from coming into existence. Preventing something coming into existence denies a future of value, as does destruction. But they are not the same”.2

Savulescu’s remark makes both too few distinctions and too many. I exist at present. My future of value does not. Therefore, they are different. Now ask what has been eliminated if someone kills me at the present moment. My killer has not destroyed the past stages of me, for no one can change the past. Whether the present stage of me exists or not does not (by itself) matter to me because the present is instantaneous. What matters is that my killer would eliminate future stages of me. He would prevent my future of value from occurring. Therefore, not all preventions are equal. Some preventions are what we would call “destructions” and some are not.


Savulescu also defends his criticism of the future of value argument by suggesting that at least one of several alternative accounts of the wrongness of killing is superior to it. Savulescu believes that there is a “property of killing” that fetuses and embryos lack, but that people have. Killing a human being is presumptively seriously wrong only if a human has acquired that property, which Savulescu calls “p”. What is p? According to Savulescu: “We need not settle on what p is. Property p surely exists”. Savulescu claims that at least one of four candidates for p is reasonable.

Savulescu believes that a property proposed by Mark Brown may be a reasonable candidate for p.4 Thus it may be the case that “Killing is wrong because it deprives a self conscious being of a self represented future of value”.2 He also believes that a property proposed by Michael Tooley5 and Peter Singer6 may be a reasonable candidate for p. Thus it may be the case that “Killing is wrong because it frustrates the desire to live of a self conscious being”.2 A third candidate for p is having a functioning brain.2 A fourth candidate for p is having the capacity for consciousness.2

Will any of Savulescu’s candidates survive critical scrutiny? Consider Brown’s candidate. In two essays in this journal I have offered what I still take to be good arguments for believing that Brown’s view is multiply ambiguous and that any version of Brown’s view is an inadequate account of a serious right to life.7,8 Savulescu offers no objection whatsoever to these arguments. Failing that, I see no reason to believe that Brown’s property is a reasonable candidate for p.

Consider now the Singer/Tooley view. The point of the search for p, of course, is to find a property the absence of which renders the killing of a human being morally permissible and which is absent in fetuses. The problem with taking a self conscious being’s desire to live as the basis for one’s right to life is that absence of this property permits too much killing. Tooley himself described such difficulties in his landmark 1972 essay. Consider the case of an individual suffering from depression who says that he wishes he were dead, or, for that matter, who says sincerely that he sees no point in living. Consider the case of someone who is not a self conscious being because she is temporarily unconscious and therefore not conscious of anything including her own self. Consider the case of an individual who “may permit someone to kill him because he had been convinced that if he allows himself to be sacrificed to the gods he will be gloriously rewarded in a life to come” (Tooley,5 pp 47–8). Killing such people is clearly wrong. Therefore, the Singer/Tooley candidate for p is not reasonable. Indeed, as the interested reader will see, Tooley has (on page 109–page 12 of his book) given up this view because of these problems with it.9

An attentive reader might wonder if the Singer/Tooley view could be salvaged by a minimal alteration so that our candidate for p is, instead of the actual desire to live, the conceptual capacity to desire to live. This alteration does appear promising, both because it underwrites a defence of abortion and because it is not subject to the above counterexamples. John Harris apparently has such a view in mind when he writes: “My suggestion then is that if we ask ‘which lives are valuable in the ultimate sense, which lives are the lives of persons?’, the answer will be ‘the lives of any and every creature whether organic or not, who is capable of valuing his/her or its own existence’”.10

Will this do? The plausibility of giving an account of the wrongness of killing in terms of one’s actual desire for, or actual valuing of, life is that we do believe that we ought to respect the desires, or the wishes, or the values, or the interests of others. The trouble is that we do not, in general, believe that is it wrong to deprive someone of something that she is merely capable of desiring or valuing, but does not in fact desire or value, or care about, or have an interest in. I am conceptually capable of desiring to keep the trash that I set out for the trash man each week, but that creates not the slightest presumption whatsoever that it is wrong for the trash man to deprive me of it. Tooley defends the intuitive plausibility of his view in terms of actual desires, not in terms of desires that one merely has the capacity to have (Tooley,5 p 44). When Harris offers a reason for his view, he talks of actual valuing, not the mere capacity to value. So Harris says: “Persons who want to live are wronged by being killed because they are thereby deprived of something they value. Persons who do not want to live are not on this account harmed” (Harris,10 p 307). Thus the problem with the minimal alteration of the desire view that we are considering is that the price that is paid for accommodating the counterexamples to the unaltered view is giving up the intuitive plausibility of the original view. Accordingly, there are reasons for rejecting both views.

Could the time at which a human organism begins to possess p be the time at which fetal brain function begins? Savulescu defends this candidate by asserting that: “Death is defined currently in terms of brain death” and “If we cease to exist when our brain dies, we only begin to exist when our brains start to function”.2 This is not an occasion for a discussion of the correct definition of death. Assume Savulescu is right about that (although I do not think he is). Even so, there are at least two problems with this candidate for p. The first is that it is possible to argue that our brain begins to function so early in fetal development that this candidate for p will justify very little, if any, abortion. A little cellular specialisation at the cephalic end of the embryonic neural tube arguably could qualify. The second is that death is defined is terms of the irreversible loss of function. Nothing corresponds to this at the beginning of life. Plainly the mere absence of function is not going to do, for the mere absence of function is not sufficient for death.

Savulescu’s final candidate for p is the acquisition of consciousness. Consciousness begins at about 20 weeks of gestation. On the one hand, this is considerably later than the time at which minimal brain function begins. For abortion’s defenders such a p has the happy consequence of justifying virtually all abortions that are actually performed. On the other hand, this candidate for p cannot be justified by Savulescu’s argument based on the nature of brain death. The most serious problem, however, concerns people who are temporarily unconscious. People who are temporarily unconscious are people who not only are not conscious (think of people who are sleeping!) but who cannot be brought to a state of consciousness. People who cannot now be brought to a state of consciousness are those who (now) lack the capacity for consciousness. Since temporarily unconscious people clearly have the right to life, the capacity for consciousness (much less consciousness) clearly is not a necessary condition for the right to life. It is worth noting that, of course, fetuses are merely temporarily unconscious.

Accordingly, none of the candidates for p offered by Savulescu is a property on the basis of which we can distinguish those who have the right to life from those who do not. I have also shown that Savulescu’s criticisms of the future of value argument are unsuccessful. Does this analysis, then, throw us into the clutches of the opponents of abortion?

Hardly. For one thing, there may be other successful criticisms of the future of value argument that have not been discussed in this essay. I do not think so, of course, but there is nothing in this essay to show that no criticism of the future of value argument is successful. In addition, there may be successful candidates for p that Savulescu did not offer. Furthermore, I have not addressed the important arguments that purport to show that even if fetuses have the same full bodied right to life that you and I have, pregnant women do not have the obligation to provide them with life support.11 Finally there is another prochoice option to which Savulescu refers that I have not yet discussed.

Savulescu endorses Jeff McMahan’s account of when we began to exist. According to McMahan, we began to exist when the consciousness that is causally connected in a particular way with our present consciousness began. Therefore, we began to exist at about 20 weeks of fetal gestation.12

It is important to distinguish McMahan’s view from the property p view. According to the property p view, we are (I suppose) biological organisms who existed from the time of conception or, perhaps, implantation. At one stage in our early history we acquired a property p that underwrites a serious right to life. McMahan denies this. He denies that we are human organisms. On McMahan’s view, we are essentially conscious beings. We cannot exist in the absence of consciousness. We are only that part of the brains closely associated (but not identical) with us that are the physical basis for our continuing consciousness. Thus I did not begin to exist at all until at least 20 weeks after conception. Since the biological organism that I believe I am began to exist either at conception or at implantation, and since if I were that biological organism I would have all of the properties of that biological organism, it follows that I am not a biological organism (and you are not either, reader). Thus, the preconscious biological organism that was my precursor did not have my future of value because that biological organism was not an earlier phase of me. Therefore, even if the future of value argument were correct, it would have been morally permissible for my mother to have an abortion in her first 20 weeks of pregnancy with the biological organism out of which I was created (McMahan,12 pp 88–94).

What are we to say of this view? McMahan’s view is based upon an analysis of the metaphysics of personal identity. Discussion of this view would take us farther into some very abstruse issues than, I suspect, readers of this journal would wish to go, and, in any event, far beyond the confines of a short response to Savulescu. I believe there are difficulties with McMahan’s view that are sufficient to reject it, but, clearly, this is not an argument, but only a promissory note. My scientific prejudices make it very difficult for me to share McMahan’s view that I am not a biological organism. Furthermore, I suspect that readers of this journal with at least as much background in biology as I will also find it difficult. Finally, I note that Savulescu’s presentation of McMahan’s view is ambiguous concerning whether the McMahan view is a property p view or the robust dualistic view it actually is. Perhaps this is because Savulescu himself is unwilling to embrace the radical dualism that McMahan defends.

I conclude that Savulescu has not shown that the future of value argument is defective and also has not shown that there is a viable alternative to it.


I thank John Harris for helpful comments.


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.