Walter Sinnott-Armstrong and Franklin Miller have presented an account of why killing is wrong that implies it can be permissible to kill certain human beings in order to use their organs for transplantation.1 Since I am going to criticise their arguments, I will begin by applauding their willingness to defend an unpopular position and by registering my agreement with their substantive conclusion about organ procurement. The criticisms I will offer are intended to be friendly in spirit; but they are also, of course, meant to be effective.

Sinnott-Armstrong and Miller (to whom, for brevity, I will subsequently refer as ‘the authors’) argue that the wrongness of killing is reducible to the wrongness of disabling, in particular, the wrongness of totally disabling. ‘The rule against disabling,’ they write, ‘fully explains all that is bad and wrong with…killing.’1

One might think that their view is just a slight modification of the familiar view that the wrongness of killing is reducible to the wrongness of harming. According to this view, killing is normally gravely wrong because it inflicts on its victim a particularly grievous harm by depriving him of all the good that his future life would otherwise have contained. If the authors' view were just a restatement of this harm-based account of the wrongness of killing, one might conclude that their emphasis on disability is needlessly provocative. But in fact their view is quite different. This is because disabling is only one form of harming. On might seriously harm a person—for example, by causing him to experience mild chronic pain for the remainder of his life—without in any way disabling him. He would retain all his abilities but would be less comfortable while exercising them. Because there are thus forms of harming that are not instances of disabling, a rule against disabling is a narrower constraint than a rule against harming.

This suggests that the authors would have done better to champion the familiar claim that a rule against harming explains all that is wrong with killing. One could, indeed, develop a variant of their argument that it is more economical to have a single prohibition of disabling rather than separate prohibitions of disabling and killing to show that it would also be more economical to have a single prohibition of harming rather than separate prohibitions of harming and disabling. Given a rule against harming, a separate rule against disabling may be otiose, even for their purposes. For a rule against harming unaccompanied by an additional rule against killing would permit the removal of organs for transplantation in all the cases in which their rule permits it (except, perhaps, in the case of a totally disabled individual who remains conscious and has predominantly pleasant experiences).

Yet for those who believe that killing a fetus is less morally objectionable than killing a normal, innocent adult person, the authors' narrower prohibition of disabling has what may seem to be an advantage. In his influential work on the morality of abortion, Don Marquis endorses the harm-based account of the wrongness of killing, claims that to kill a fetus is to cause it a grave harm by depriving it of the whole of a human life that it would otherwise have had, and concludes that abortion is normally seriously wrong.2 The authors' analysis of the wrongness of killing in terms of the wrongness of disabling seems to avoid this conclusion because abortion does not involve significant disabling, since fetuses have comparatively few abilities of which they are deprived by being killed. (The rule against harming and the rule against disabling coincide in implying that the killing of an animal is usually much less seriously wrong than the killing of a person, for the animal is deprived of less good life and of fewer abilities than the person is.)

One might point out that the authors repeatedly claim that what is wrong about killing is that it results in total disability. But if what is wrong about killing is that it causes the victim to be totally disabled, then perhaps the killing of a fetus is as wrong as the killing of an adult, since in both cases the result is that the victim is totally disabled. But this is clearly not the authors' view. If it were, they would be committed to accepting that the killing of an animal is as wrong as the killing of a person, since in both cases the victim is totally disabled. The claim the authors defend is instead that killing, and disabling generally, are wrong because they deprive the victim of valuable abilities. When this is conjoined with their further claim ‘that the wrongness of disabling varies with the degree of the disability caused’ (or, perhaps more accurately, with the value of the abilities of which the victim is deprived), their position does indeed imply that killing a fetus is in general less seriously objectionable than killing an adult, as the latter is deprived of more and greater abilities by being killed than the former is.

It would be premature, however, to conclude that the authors' rule against disabling is for this reason superior to the harm-based account of the wrongness of killing. For the harm-based account by itself does not lead to the conclusion that abortion is seriously wrong. It implies that conclusion only in conjunction with the assumption that a fetus is seriously harmed by dying. This assumption is true if the badness of death is a function only of the amount of good life the victim loses by dying, but that is a crude understanding of the badness of death. If it were correct, the worst possible death would be that which occurs immediately after someone begins to exist. If the common view that we begin to exist at conception were true, this understanding of the badness of death would imply that the worst possible death is one that occurs immediately after conception. That is hard to believe. It is more plausible to suppose that the badness of death is instead a function of two factors: the amount of good life that is lost and the strength of various psychological relations, such as desire, belief, memory and so on, that would have bound an individual at the time of death to himself at those later times at which the goods of his future life would have occurred. According to this view, since a fetus would be only very weakly psychologically connected with itself in the future, its death is less harmful than the death of an older child or adult, despite the fact that it is deprived of more good life. (I have elsewhere developed and defended such an account of the badness of death as well as an account of the wrongness of killing that is quite different from both the harm-based account and the authors' analysis in terms of disabling.3)

There are other differences between the authors' account of the wrongness of killing and the harm-based account. In some respects the authors' account is more egalitarian. Suppose, for example, that a 20-year-old and a 60-year-old have exactly the same abilities. If neither of them is killed, the 20-year-old will live another 50 years and the 60-year-old will live another 10. While the harm-based account implies that it would be more seriously wrong to kill the 20-year-old, the authors' account implies that the two killings would be equally wrong. (This assumes that the authors do not want to eliminate what is distinctive about their view by claiming that the 20-year-old, but not the 60-year-old, has the ability to live 50 more years, and that killing him would deprive him of that ability.) Similarly, suppose there are two 20-year-olds with equivalent abilities who, if not killed, will each live another 50 years. One of them, however, suffers from chronic pain that would make his subsequent life less good than the other's. Again the harm-based account implies that it would be less seriously wrong to kill the one who suffers from chronic pain, while the authors' view implies that killing one would be equally wrong as killing the other.

These more egalitarian implications of the authors' account may make it seem more plausible. Yet it is only contingently egalitarian in these cases. It is not an egalitarian view of the wrongness of killing. This emerges when we consider its implications for the killing of the disabled. Because the authors claim that the wrongness of killing is reducible to the wrongness of disabling, and because they also claim that the wrongness of disabling varies with the importance or value of the abilities of which a person is deprived, their view implies that the fewer abilities a person would lose by being killed, the less seriously wrong it would be to kill him. Thus, the implication of their view that they showcase is that it is not wrong to kill an individual who has no remaining abilities and, a fortiori, not wrong to kill a totally disabled individual to take his organs. Although they do not discuss this, their view might also imply that it can be permissible to kill an individual who has very few remaining abilities if his organs are necessary to save the life of a person who, if saved, will have a great many abilities. This is because the killing of such an individual would involve relatively little disabling, and the extent to which a killing is morally objectionable is a function of the value of the abilities of which the victim is deprived. Just as it might be permissible to inflict a minor disability on an innocent person to save the life of another, so it might be permissible to kill an individual who retains only minor abilities for the same purpose.

What this means, however, is that according to their view, the more disabled a person already is, the less seriously wrong it would be to kill him. Just as killing a fetus is less seriously wrong because the fetus has fewer abilities to lose, so killing a severely disabled person is less seriously wrong for the same reason. This is a highly inegalitarian implication, one that very few people would be willing to accept.

The authors are aware of this problem, though they do not state it quite so starkly. They offer, in quick succession, five strategies of rebuttal. But none of these strategies is likely to help. The fifth is to deny that the objection is really an objection by conceding that their view has this initially counterintuitive implication but then showing that it is our intuition rather than the implication that should be rejected. It would be interesting to see how such an effort might proceed, but they do not offer any proposals. Their third suggestion is that abilities and disabilities are difficult to measure so that in practice ‘we cannot really know which person has more ability.’ But while that may well be true for a great range of people with disabilities, it is not true of those with multiple and profound physical and mental disabilities. And in any case it leaves the practical implications of their view dependent on our continued inability to measure degrees of disability with accuracy. But there is no reason to suppose that our current epistemic situation is unalterable.

This leaves strategies one, two and four, all of which involve the abandonment of their view that ‘the rule against disabling…fully explains all that is bad and wrong with…killing’ and that ‘the wrongness of disabling varies with the degree of disability caused.’ The first suggestion is that ‘the value of a person does not vary after abilities…pass a minimal threshold that is above plants.’ That avoids the implication but only by explaining the wrongness of killing by reference to the invariant value of persons rather than in terms of the wrongness of disabling. The second suggestion is that ‘equality and justice’ make ‘it morally wrong to treat people differently even if they have different abilities.’ Again this makes the wrongness of killing at least in part a matter of equality or justice rather than solely a matter of disabling. Finally, the fourth suggestion is that ‘our moral duty is to honor (or not to disrespect) the value of abilities…and that the prohibition on disrespect ensures equality and justice.’ Again, however, this involves abandoning the earlier claim that the wrongness of disabling varies with the value of the ability of which the victim is deprived.

Admittedly, of these three suggestions, the first and fourth do seem to support the authors' claim that it is permissible to kill the totally disabled. The first does so on the assumption that the totally disabled have no abilities above those of plants. The fourth does so because killing the totally disabled cannot violate a duty to respect abilities if they have no abilities to respect. According to the second suggestion, however, equality or justice might make it wrong to kill a totally disabled individual—since that would certainly involve treating him differently—despite his being different from others in his abilities.

Whether the authors' first and fourth suggestions are ultimately plausible is a question they do not consider. Nothing they say supports either view, since both views are quite different from the one they defend. I doubt, however, that either view is promising. Unless the first view draws an arbitrary distinction between human beings and other animals, it will imply that killing an animal is as seriously wrong as killing a person. The fourth view seems to make a fetish of abilities. It is the bearers of abilities—persons—not the abilities themselves, that ought to be the objects of respect.

My own view is that the authors are entirely right that it can be permissible to kill a living human organism that has irreversibly lost the capacity for consciousness in order to use his or her organs for transplantation. But the explanation I would give is different and rather more complicated than their own.4 The hypothetical case of an individual who has become totally disabled but remains conscious and experiences continuous pleasure raises different and perhaps rather difficult issues, though for practical purposes they are probably irrelevant.