Abstract and Keywords
The three major classical accounts of the morality of abortion are all subject to at least one major problem. Can we do better? This article aims to discuss three accounts that purport to be superior to the classical accounts. First, it discusses the future of value argument for the immorality of abortion. It defends the claim that the future of value argument is superior to all three of the classical accounts. It then goes on to discuss Warren's attempt to fix up her personhood account and David Boonin's attempt to fix up Tooley's desire account. Warren claims that her updated version of a personhood account is superior to any potentiality account, such as the future of value account. The article evaluates her claim. Boonin argues that his improved desire view both deals adequately with the apparent counterexamples to Tooley's original account and also is superior to the future of value account. The article evaluates his views as well.
Classical Arguments Concerning Abortion
Opponents of abortion choice typically have argued that because (human) foetuses are human, alive, and innocent, and because an abortion ends innocent human life, abortion is, unless special circumstances obtain, immoral. Call this argument ‘the human life argument’. Most supporters of abortion choice (in the philosophical world, at any rate) have argued that because (human) foetuses are not yet persons, and because what makes killing wrong is that the victim is a person, abortion is, unless special circumstances obtain, morally permissible. Call this argument ‘the personhood argument’. Both of these arguments are based on an understanding of why killing is wrong in cases where a consensus exists. Proponents of each argument argue that their understanding implies a conclusion concerning abortion.
Many discussions of abortion take the human life view of the wrongness of killing for granted and then go on to argue that foetuses either are or are not fully human or that foetuses either are or are not potential, but not actual, life. These issues are easy to resolve. The foetuses in which we are interested are all fully human, of course. (p. 396) They are not mosquitoes or mice. Because they exhibit growth and metabolism they are actually, not merely potentially, alive. When the argument is conducted within the human life framework those opposed to abortion choice easily win the argument.
Many other discussions of abortion take the personhood view for granted and go on to discuss whether or not foetuses are persons. This issue is also easy to resolve. In the sense of ‘person’ generally used by personhood proponents, no foetus is a person. When the argument is conducted within the personhood framework, those in favor of abortion choice easily win the argument. Thus, the far more important issue concerns which understanding of the wrongness of killing is correct. This point can be put in another way for readers who know a bit of logic. The syllogisms in terms of which we can understand the abortion debate are not really controversial because of their minor premises. The truth of their major premises is at the heart of the matter.
The human life and personhood arguments are alike in another respect. Their views of the wrongness of killing are subject to similar criticisms. The human life account is subject both to a theoretical difficulty and to a difficulty concerning cases. The theoretical difficulty concerns the connection between the biological properties of being human and being alive and the moral property of having a right to life. Critics argue that this connection needs to be established, not merely asserted. They argue that the connection is not self‐evident, that religious considerations are insufficient to establish it, and that no other argument establishes the connection. The human life account also faces a difficulty concerning cases because it seems to make too much killing wrong. It makes deliberately ending the lives of humans who will never be conscious (such as anencephalic newborns and persons in persistent vegetative state) wrong. It seems to make deliberately ending the life of a sperm or an unfertilized ovum (which are, after all, human and alive and not guilty) wrong. The human life account of the wrongness of killing seems to be both theoretically insufficient and too broad.
The classic personhood account comes in two versions: Mary Anne Warren's version and Michael Tooley's desire version. According to Warren the traits ‘most central to the concept of personhood’ are (1) consciousness and, in particular, the capacity to feel pain, (2) reasoning, (3) self‐motivated activity, (4) the capacity to communicate in a reasonably sophisticated way, and (5) the presence of self‐concepts (Warren 1979: 45). According to Warren, an individual need not have all of these traits to be considered a person, or even any particular one of them. However, an individual with none of these traits is plainly not a person. And even if an individual had only the capacity to feel pain, ‘it cannot be said to have any more right to life than, let us say, a newborn guppy’ (Warren 1979: 47). In Warren's view what makes a killing wrong is that the victim is a person. Since foetuses clearly aren't persons, abortion is morally permissible.
Warren's classic personhood account faces difficulties very much like the difficulties that face the human life account. The theoretical problem with Warren's account concerns the connection between the psychological traits in terms of which personhood is defined and the moral property of having the right to life. Neither the mere assertion by Warren that there is such a connection nor the general belief that there is such a connection is enough to establish an actual connection. From a theoretical perspective Warren's personhood view is no more compelling than the human life view.
Warren's classic personhood argument is also plagued by difficulties concerning cases. The claim that foetuses lack the right to life because they are not persons seems to imply that it is permissible to kill a newborn child because she is not a person. It also seems to entail, because being a person involves exhibiting certain psychological traits, that killing someone who is temporarily unconscious is morally permissible. Warren tries to avoid the former conclusion by arguing that infanticide is wrong because of the attitudes that adults have toward newborns (Warren 1979: 50–1). However, this has the consequence that if, for any set of infants, adults did not value those infants, then infanticide would be morally permissible. (Think of the infanticide of females in some cultures.) Thus, Warren's classic defense of abortion choice in terms of personhood suffers from problems that are the mirror image of the human life account's problems. The two accounts suffer from similar theoretical difficulties. Both have problems with cases. The human life account makes too much killing wrong; Warren's personhood account makes too little killing wrong.
Michael Tooley's personhood account is different from Warren's account. Tooley's account has the great virtue of avoiding the theoretical problem that plagues both the human life argument and Warren's classic personhood argument. Tooley stipulates the truth of ‘X is a person if and only if X has a serious moral right to life’ (Tooley 1972: 42). He then claims that ‘An organism possesses a serious right to life only if it possesses the concept of a self as a continuing subject of experiences and other mental states, and believes that it is itself a continuing entity’ (Tooley 1972: 45). Tooley defends this claim by arguing that an organism possesses a serious right to life only if she desires to live and she desires to live only if she possesses a concept of herself as a continuing subject of experience. Thus, the desire to live is basic to Tooley's account of the right to life. Accordingly, I refer to Tooley's personhood account as ‘a desire account’, both to characterize it and to distinguish it from Warren's account.
Here is how Tooley defends the moral importance of desires.
(p. 398) From this intuition Tooley generates the following argument. To have the right to life is to have the right to continue to exist. The right to continue to exist presupposes the desire to continue to exist. One has the desire to continue to exist only if one has a concept of oneself as a continuing subject of experience. No foetus has a concept of herself as a continuing subject of experience. Therefore, no foetus has a right to life. So abortion is morally permissible.
The basic intuition is that … in general, to violate an individual's right to something is to frustrate the corresponding desire. Suppose, for example, that you own a car. Then I am under a prima facie obligation not to take it from you. However, the obligation is not unconditional: it depends in part upon the existence of a corresponding desire in you. If you do not care whether I take your car, then I generally do not violate your right by doing so.(Tooley 1973: 60)
Tooley's desire account of the wrongness of killing is not subject to the theoretical difficulties that plague both the human life account and Warren's classic personhood account. Tooley's basic intuition is a plausible attempt to locate his abortion view within a general account of wrong action. However, he has noted that the desire account seems not to explain the wrongness of some cases of killing. Consider the case of an individual suffering from depression who says that she wishes she were dead, or, for that matter, who says sincerely that she 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 1972: 47–8). Apparently Tooley's account of the wrongness of killing implies that it is morally permissible to kill all of the above individuals. It is not. Thus, Tooley's account, like Warren's personhood account, permits too much killing.
On more than one occasion Tooley struggled with these apparent problems with his view (Tooley 1972: 48; 1973: 65–6; Purdy and Tooley 1974: 146). Tooley's problem was finding qualifications that rescue the desire account, that do not seem to be merely ad hoc, and that preserve abortion choice. Because of these difficulties Tooley eventually gave up his desire account (Tooley 1983: 109–112).
The three major classical accounts of the morality of abortion are all subject to at least one major problem. Can we do better? The purpose of this chapter is to discuss three accounts that purport to be superior to the classical accounts. First, I shall discuss the future of value argument for the immorality of abortion. I shall defend the claim that the future of value argument is superior to all three of the classical accounts. I shall then go on to discuss Warren's attempt to fix up her personhood account and David Boonin's attempt to fix up Tooley's desire account. Warren claims that her updated version of a personhood account is superior to any potentiality account, such as the future of value account. I shall evaluate her claim. Boonin argues that his improved desire view both deals adequately with the apparent counterexamples to Tooley's original account and also is superior to the future of value account. I shall evaluate his views.
Because of considerations of length, this chapter will not deal with arguments that purport to show that even if foetuses have a full right to life, even if they have the same moral status that we do, abortion is morally permissible because foetuses (p. 399) do not have the right to the life support that women must provide for nine months if foetuses are to survive.
The Future of Value Account
According to the future of value argument for the immorality of abortion the best explanation for the wrongness of killing is that killing deprives us of our futures of value. Our futures of value consist of all of the goods of life we would have experienced had we not been killed. Foetuses have futures like ours, for their futures contain all that ours contain and more. Therefore (given some defensible assumptions and qualifications), abortion is seriously wrong on almost all occasions (Marquis 1989).
What is a future of value? Consider the life of someone who lives to ‘a ripe old age’. The value of that life to the person who lives it consists of experiences that, at various times in her life, she values. These are the goods of her life. They are what make her life worth living. The list of these goods will vary somewhat from person to person. The list includes friendships, loves, absorption in various projects, aesthetic experiences, identification with larger causes seen as valuable, such as one's team winning a victory, and physical pleasures. One's future of value is the class of goods in one's future that occur later than a given time in one's life, if one does not die prematurely.
On the future of value account the wrongness of killing is based on the harm of killing. A present action cannot affect one's past. Strictly speaking, a present act of harming does not make another worse off in the present either, for the present is instantaneous, and harm, involving, as it does, causation, requires at least a small temporal interval for its effect to occur. A present act of harm affects the victim's future. It makes someone worse off in the future. To make someone worse off is to reduce that person's welfare, to reduce the quantity or quality of the goods in her future that she would otherwise have possessed. On the future of value account killing is wrong because it harms a victim.
Killing someone deprives her of all the goods of her future she otherwise would have experienced. Thus killing is a very serious harm. The future of value view accounts for why we regard killing as one of the worst of crimes. It coheres with why we believe that the premature death of a human being is one of the greatest misfortunes she could suffer. Thus the future of value account is plausible, not only because it coheres with our understanding of killing as a harm, but because it coheres with our understanding of premature death as a great misfortune.
We know that foetuses have futures of value because we were all foetuses once and their futures of value are the goods of our past lives, our present lives, and our future lives. (I am assuming that we are biological organisms. Jeff McMahan (p. 400) has offered interesting and important arguments for the falsity of this assumption; McMahan 2002.) Each of us has had first person experience with the future of value of one (former) foetus. Since we have friends and acquaintances, we know of the futures of value of many (former) foetuses. Since abortion deprives a foetus of a valuable future like ours, and since depriving an individual of a future like ours is what makes killing her seriously presumptively wrong, abortion is seriously presumptively wrong. Because ‘like’ is a vague predicate the precise scope of the future of value account of the wrongness of killing is indeterminate. Animal rights proponents may want to exploit this vagueness. However, because portions of the future of a foetus are identical to ours, human foetuses will clearly be within its scope.
An individual's future of value is based on her present potentiality. We have futures of value in virtue of being individuals who have the present potential to have experiences we will value in the future. This present potential is a function of our present biological nature. (In the case of persons from outer space whom it would be wrong to kill, this might not be so.) The future of value account should be distinguished from other potentiality arguments intended to deal with the morality of abortion. One might claim that what makes it wrong to kill a foetus is that it is a potential human being. This claim is incompatible with the future of value account because, according to the future of value account, what makes it wrong to kill a foetus is exactly the same property that makes it wrong to kill you and me. It certainly is not wrong to kill you and me because we are potential human beings. This is because we are not potential human beings; we are actual human beings.
One might claim that it is wrong to kill a foetus because it is a potential person. This claim is also incompatible with the future of value account because according to the future of value account what makes it wrong to kill a foetus is exactly what makes it wrong to kill you and me. But it is not wrong to kill you and me because we are potential persons, for the very simple reason that we are not potential persons. We are actual persons.
The future of value account should be thought of in the following way. In this universe there are many individuals. Some of them (such as virtually all other humans) have a present potential to have futures like ours and some of them (such as trees, rocks, and snails) do not. That present potential makes it wrong to kill individuals who are members of the former class.
The future of value account is not subject to the theoretical difficulties of the human life account and Warren's (1979) personhood account. The former account rests on an alleged connection between a biological property and a moral property and the latter account rests on an alleged connection between psychological properties and a moral property. In neither account is there an adequate defense of the connection between a property with value content and a purely natural property. The future of value account, like Tooley's desire account, is based on a connection between two properties with value content. Because the whole account is imbedded in an account of what harm is, the whole account is imbedded in a (p. 401) plausible general account of wrong action. Therefore, in this theoretical respect, the future of value account of the wrongness of killing is superior to those earlier accounts.
The future of value account handles cases better than any of the three classic accounts. Consider the problems that the human life account has with persons in persistent vegetative state and with anencephalic newborns. It is difficult to understand what is wrong with ending the lives of humans in either category. Ending their lives cannot deprive them of anything they would ever value. Accordingly, it is difficult to understand how they are harmed by their lives being ended. This being the case, it is difficult to understand how they are wronged by their lives being ended. Because of this anencephalic newborns have been intentionally allowed to die for centuries. The human life account cannot account for this. The future of value account easily accounts for this. The same considerations apply to humans who are in persistent vegetative state.
Warren's early account is plagued with the difficulty of accounting for the wrongness of infanticide. By contrast, according to the future of value account of the wrongness of killing, ending the life of a little baby is as obviously wrong as we all think it is. Warren's account also faces problems accounting for the wrongness of killing the temporarily unconscious. Since exhibiting psychological traits is necessary for being a person and since the temporarily unconscious do not exhibit psychological traits, her view apparently entails that killing the temporarily unconscious would not wrong them. This is, of course, a disaster for her view.
Strategies for repairing her view do not show promise. One strategy for fixing her view might be to allow that one is a person either in virtue of actually manifesting certain psychological traits or in virtue of being able to manifest certain psychological traits in the future. The trouble with this strategy is that foetuses are able to manifest psychological traits in the future, if allowed to live. The consequence of such an alteration would convert Warren's account to an anti‐choice view. No doubt she would not welcome that consequence.
Another strategy for fixing her view might be to allow that one is a person either in virtue of actually manifesting certain psychological traits or in virtue of being able to manifest certain psychological traits now. That move avoids the anti‐choice conclusion, but only at the price of not being able to account for the wrongness of killing the temporarily unconscious again. Accordingly this strategy would not work. A third strategy for fixing Warren's view might be to hold that the capacity to manifest psychological traits in the future as long as one has manifested such psychological traits in the past is the basis for the right to life. This move eliminates the unwelcome anti‐choice consequence without allowing us to kill the temporarily unconscious. However, sole justification for this particular wrinkle in her theory is that it both can account for the case of the temporarily unconscious and also can produce the pro‐choice conclusion that Warren desires. So such a wrinkle cannot justify her abortion view.
The future of value account also handles cases better than Tooley's desire account. In a nutshell, Tooley's desire account does not account for the wrongness of killing many persons who, for one reason or another, do not desire to live. The future of value account handles such cases nicely. Consider two classes of such persons: (1) persons who, perhaps because of a terminal illness, face a future of pain and suffering that cannot be alleviated and (2) persons who, because of depression, can be helped, through the aid of psychotherapy and psychotropic drugs, to live lives that they will later value. On the future of value account it is not wrong to kill persons in the former class, although getting the analysis just right demands some attention to detail. On the future of value account, it is wrong to kill persons in the latter class. This is exactly what most sensible people believe. The future of value account deals with Tooley's problem cases in the correct way. We can conclude, therefore, that the future of value account offers theoretical advantages over two of the three classical accounts and deals with cases better than any of them. Because it fits into a general account of harm, which, in turn, fits into a general account of wrongs, and because it, by comparison with alternative theories, handles cases well, it seems to be the basis for a correct account of the ethics of abortion.
Warren's Moral Status Account
In 1997 Mary Anne Warren offered an updated personhood defense of abortion choice (Warren 1997). Whereas her earlier view paid little attention to the connection between her psychological account of personhood and the right to life, her later view is anchored in a comprehensive account of moral status. Warren's account contains seven different principles, each of which underwrites some degree of moral status. Fortunately, evaluation of her defense of abortion rights does not require appraisal of her entire theory. Many of her principles do not concern the moral status of humans at all. They underwrite a level of moral status inferior to the status of humans. For our purposes, such principles can be neglected. The abortion issue turns on whether or not foetuses have the same full moral status as you and I. Because women have a strong presumptive right to control their own bodies, if foetuses have some, but less than full, moral status, then their interests would not be sufficiently strong to override the interests of women, who, after all, do have full moral status. The interests of a foetus could trump a pregnant woman's presumptive rights only if foetuses have the same full right to life as you and I have now.
Warren's AGENT'S RIGHTS Principle
In order to defend abortion choice in terms of the lack of full fetal moral status Warren needs some account (1) that explains why those humans whose moral (p. 403) status is not a matter of controversy do indeed have full moral status and (2) that does not include foetuses within the scope of that explanation. Warren's primary full moral status principle is what she calls ‘the Agent's Rights principle’. According to this principle
This principle is not controversial. Because it offers only a sufficient, but not a necessary, condition for the possession of full moral status, it cannot be criticized for excluding humans who are not moral agents. Warren defends this principle by appealing to the moral philosophy of Kant and to the Kantian moral philosophies of John Rawls and Alan Gewirth. It is reasonable to assume that such foundations for the Agent's Rights principle are sufficient to save it from the theoretical difficulties that beset the human life view or Warren's earlier personhood view. Warren's more recent view is a genuine improvement over her earlier view.
Moral agents have full and equal basic moral rights, including the rights to life and liberty. (Warren 1997: 156)
A moral agent is ‘an individual who is capable of using reason to discern and follow universal moral laws’. (Warren 1997: 156)
Warren's HUMAN RIGHTS Principle
Warren's Agent's Rights principle plainly needs supplementation. It does not underwrite the full moral status of young children and the mentally disabled. Because she is well aware of this, Warren adds to her account what she calls ‘the Human Rights principle’. According to this principle:
Although in philosophy it is customary first to determine just what a claim means and then to ask whether or not it is true, this procedure is not apt for analysis of Warren's Human Rights principle. Warren first defends the Human Rights principle as it applies to already born humans and then discusses whether or not the principle should be extended to protect foetuses. Adequate appraisal of her defense of her view requires following the path of her argument. It is also worth noting that Warren does not believe that the Human Rights principle is literally true. It entails that 3‐year‐olds have the liberty rights of moral agents. This is, as Warren realizes, absurd. Warren actually believes that in the case of children ‘their interests carry the same moral weight as do those of other human beings’ (Warren 1997: 164). My appraisal of her defense of the Human Rights principle will take this ‘moral equality of interests’ interpretation for granted.
Within the limits of their own capacities and of [the Agent's Rights principle], human beings who are capable of sentience but not of moral agency have the same moral rights as do moral agents. (Warren 1997: 164)
Because Warren's defense of the Human Rights principle as it applies to children after birth is not extensive it is possible to quote it in full: (p. 404)
Much of Warren's defense of her Human Rights principle seems quite implausible. Is the dependency of children on moral agents sufficient for treating children as the moral equals of moral agents? House plants are dependent on moral agents. We do not think that this implies that we should treat them as our moral equals. Is the education of children by moral agents a good reason for giving children the same rights as moral agents? Adolescents learn to drive from persons who have the right to drive. It does not follow that such adolescents have the right to drive. Is the emotional abhorrence of rejecting the rights of children a good reason for treating them as if they have the same rights as moral agents? Many people in our society find sexual relations between two males emotionally abhorrent. Surely that is not a good reason for being opposed to gay rights. Do instinct and culture constitute solid grounds for the truth of the Human Rights principle? If they do, then the instinct and culture of the Christian conservative opponent of abortion rights constitute solid grounds for outlawing women's right to choose. No doubt Warren would not welcome this conclusion. Thus many of the considerations to which Warren alludes do not support the truth of her Human Rights principle.
in reality human beings become moral agents only through a long period of dependence upon human beings who are moral agents already. During this period of dependency we learn language, and all of the other mental and behavioural capacities that make moral agency possible. In Annette Baier's words, ‘A person is best seen as one who was long enough dependent upon other persons to acquire the essential arts of personhood. Persons essentially are second persons, who grow up with other persons.’
For this reason, it is both impractical and emotionally abhorrent to deny full moral status to sentient human beings who have not yet achieved (or who have irreparably lost) the capacity for moral agency. If we want there to be human beings in the world in the future, and if we want them to have any chance to lead good lives, then we must at least value the lives and well‐being of infants and young children. Fortunately, instinct, reason, and culture jointly ensure that most of us regard infants and young children as human beings to whom we can have obligations as binding as those we have to human beings who are moral agents. (Warren 1997: 164–5)
I suspect (although I am not sure) that Warren would choose to go to the bank with something like the following argument in defense of her Human Rights principle: It is necessary to award full rights to those humans who will become moral agents or who are mentally disabled in order to respect the rights of the moral agents those humans will become or once were. This is an interesting argument that deserves some analysis. I shall confine my attention to children.
It is worth noting that, strictly speaking, this argument reduces the Human Rights principle to a corollary of the Agent's Rights principle. We are asked to respect the rights of (grant the rights to?) children in order to respect the rights of the moral agents those children will become. So the question becomes: Is Warren's Agent's Rights principle sufficient to underwrite our obligations to children? Warren says little about this. However, it is hard to believe that we should treat any child as having a full right to life because the moral agent the child will become has a full (p. 405) right to life. After all, if the child is killed, there is no agent the child will become, and therefore killing the child does not violate the rights of that agent. On Warren's account killing such a child would be a victimless crime. Furthermore, it is certainly reasonable to think that some children are such poor prospects for becoming productive members of society that, if they have no right to life as children, it would not be wrong to kill them. The best argument in defense of Warren's Human Rights corollary is implausible.
This failure is fatal to Warren's theory. One typical problem for pro‐choice moral status theories, as we have already seen, is that they are too narrow. They make too little killing wrong. In this respect Warren's (1997) theory is no improvement over her earlier theory and over Tooley's desire theory. Indeed, it is a step backward. Warren's (1979) theory could not account for the wrongfulness of infanticide and so she had to offer weak arguments against killing little babies. Warren's (1997) theory, by finding the locus of full moral status only in the nature of moral agents, offers us a theory of moral status that seems far more theoretically defensible than her (1979) view. The price that she pays for this advantage is that she has trouble accounting for the wrongness of killing not only infants, but any young children.
The inadequacy of Warren's theory should be judged in the context of the alternatives to it. On the future of value account, killing young children is wrong because it deprives them of the goods of their future. Thus, killing young children is wrong for the same reason that killing adults is wrong. No special explanation is needed to deal with the wrong of killing young children. Warren's account of the wrongness of killing must be judged a failure.
The HUMAN RIGHTS Principle and Non‐Sentient Foetuses
There are, however, additional problems with Warren's view. Let us suppose that the argument of the preceding section is entirely wrong and that the Human Rights principle has been justified. The foetuses in which we are interested are, after all, human. Accordingly, there seems to be a reason for extending the Human Rights principle to include the unborn. Warren has arguments for rejecting this extension. Her arguments for not extending the Human Rights principle to foetuses before they are sentient are different from her arguments for not extending the Human Rights principle to foetuses after they are sentient.
Let us consider first her arguments for not extending the Human Rights principle to foetuses before they become sentient. Warren offers two arguments for such a restriction, a direct argument and a reductio argument. Her direct argument for the exclusion of non‐sentient foetuses is that ‘prior to the initial occurrence of conscious experience, there is no being that suffers and enjoys, and thus has needs and interests that matter to it’ (Warren 1997: 204). Presumably Warren believes that because non‐sentient foetuses have no interests and needs, they fall, like rocks, outside the scope of serious moral concern.
Warren's view that each being that lacks interests and needs falls outside the scope of serious moral concern certainly seems right. What Warren requires for her argument is the claim that if an individual has no interests and needs that matter to it, then she has no interests and needs. Is this true? It all depends upon what counts as mattering. On the one hand, if something can matter to an individual only if she is conscious of it, then no non‐sentient foetus can have interests and needs that matter to it. However, on this interpretation of the ‘mattering’ claim we are not justified in inferring that no non‐sentient foetus has interests and needs. Plainly there can be things that a young child needs or that are in her best interest that do not matter to her at all on this interpretation of ‘mattering’. On the other hand, if something can matter to an individual whether or not she is conscious of it, then Warren's argument fails to establish that non‐sentient foetuses must fall outside the scope of serious moral concern. This difficulty with Warren's view can be seen less abstractly if one considers persons who have been anaesthetized for surgery. In one clear sense nothing matters to them; in another sense, many things matter to them. They certainly have interests and needs.
Warren seems to be aware of this difficulty. She says that persons who are unconscious ‘have not lost the capacity for sentience; they are simply not exercising it at present, or not at present able to exercise it’. Surely it is true that in one clear sense of ‘capacity’, non‐sentient foetuses lack the capacity for sentience and anaesthetized persons have the capacity for sentience. The problem is that it does not follow from this that non‐sentient foetuses lack interests and needs. Neither does it follow that, in one clear sense of ‘mattering’, nothing matters either to non‐sentient foetuses or to anaesthetized persons. Accordingly, the difficulty with Warren's direct argument for excluding presentient foetuses from the Human Rights principle remains.
Warren also claims that ‘that the potential of the presentient foetus to become a human being is enough to give it full moral status is subject to a reductio argument’. Her reductio is that, if this were so, then an unfertilized ovum (hereafter UFO) would also have full moral status because of its potential to become a human being. This, of course, is absurd (Warren 1997: 206).
One problem with this reductio argument is that it is directed against the wrong target. Those who are anti‐choice typically do not hold that early abortions are wrong because presentient foetuses are potential human beings. They claim that foetuses before they become sentient are human beings, just very young and undeveloped and non‐sentient ones. Nevertheless, because the human life account is unsound, the claim that a foetus before it becomes sentient is a human being does not imply that such a foetus has full moral status. The argument that it does have full moral status must be a potentiality argument of some kind. This leaves open the possibility that Warren's argument against attributing moral status on the basis of potentiality may have some force.
The claim that a foetus prior to sentience has a future of value is a claim about that foetus's potential. Is this claim open to the reductio that, if it were true, then UFOs would have a future of value? According to the future of value view, that killing you would deprive you of your future of value makes it wrong to kill you. An individual now has that sort of future of value just in case a later phase of that same individual would have a future it would value at later times and it has a future like ours. Accordingly, that earlier phase of that individual and those later phases of that individual must be phases of the same individual. That a UFO does not meet this condition can be shown by the following argument. Suppose the UFO that was my precursor was the same individual as I. If this were so, then there is just as good reason to hold that the sperm that was my precursor was the same individual as I. Since identity is transitive, it follows that the UFO that was my precursor and the sperm that was my precursor were the same individual. This is false. Therefore, the supposition is false. Accordingly, Warren's reductio argument is unsound (Stone 1987).
If one believes that the issue is whether a foetus prior to sentience is a potential human being, this point might be missed. According to Warren: ‘In any case, the identity debate seems irrelevant to whether ova or zygotes are potential human beings. If an entity may develop into a human being, then surely it is a potential human being—even if the developmental process would alter it so greatly that we might reasonably wonder whether it has remained the selfsame entity’ (Warren 1997: 207).
The reason the identity debate is entirely relevant is because the argument Warren needs is not about whether an entity can develop into a human being, but rather about whether an individual has a future of value. Thus, both Warren's direct argument and her reductio argument for excluding non‐sentient foetuses from the scope of the Human Rights principle are flawed.
The HUMAN RIGHTS Principle and Sentient Foetuses
The analysis in the previous section shows that the sentience restriction on the Human Rights principle is unjustified. One might be tempted to conclude that, as a consequence, Warren's Human Rights principle commits her to an anti‐choice view through and through. Notice, again, her Human Rights principle:
If the sentience condition is removed, then the Human Rights principle might be thought to entail that, because foetuses are human beings, they have the same rights as moral agents. Moral agents have the right to life. Therefore, foetuses have the right to life. The claim that foetuses have the right to life implies (given some (p. 408) defensible assumptions) that killing them is wrong. So apparently Warren's Human Rights principle corrected for her error concerning non‐sentient foetuses supports an anti‐choice view. Is this correct?
Within the limits of their own capacities and of [the Agent's Rights principle], human beings who are capable of sentience but not of moral agency have the same moral rights as do moral agents. (Warren 1997: 164)
Warren rejects the above line of reasoning. According to her the Agent's Rights principle limits the Human Rights principle. She says:
Unlike presentient foetuses, women are moral agents, with the rights to life, liberty, and the responsible exercise of moral agency. These rights are undermined when women are denied the freedom to decide whether and when to have children, and how many of them to have. (Warren 1997: 210)
Therefore, even with the sentience condition removed, Warren's argument, if sound, would justify abortion at any time during pregnancy.
Birth … ends the infant's complete and necessary dependence upon the woman's body, thus removing the potential conflict between her moral rights and the infant's rights under the Human Rights principle, and bringing the latter principle fully into play for the first time. For these reasons, birth is still the most appropriate point at which to begin fully to enforce the moral rights that the Human Rights principle accords to sentient human beings. (Warren 1997: 218)
Is Warren's argument sound? Warren argues that foetuses don't have the right to life because, if they did, their right to life would conflict with the liberty rights of moral agents. The trouble with this view is that virtually no one has thought that liberty rights have such a broad scope that they include the liberty right to kill other human beings. Indeed, even libertarians believe that a moral agent's liberty rights are limited by the rights of other human beings not to be harmed. But if this is so, then there could be no conflict between a moral agent's liberty rights and the human rights of foetuses. Hence, Warren's argument fails.
Warren's argument won't do for another reason. The claim that the liberty rights of moral agents always trump the alleged rights of humans who are not yet moral agents is incompatible with parenting as we know it. This is the deep truth behind the remark attributed to the rabbi who claimed that life begins not at conception, and not at birth, but when the kids leave home and the dog dies. Warren is fond of claiming that while, one is pregnant, one is not free to hand off one's child to another, but after birth, it is possible for someone else to raise it. However, social attitudes being what they are, we are not free to hand off our children after birth. There are strong reasons having to do with social stability why this is so.
Accordingly, we may conclude that although Warren's updated (1997) personhood defense of abortion choice is more elaborate than her earlier account, it is even less successful. This is because her view does not adequately meet the challenge that anti‐choice moral status views must face. She has failed to show that her account of moral status is sufficiently broad to account for the wrongness of killing in those cases in which there is a consensus that it is wrong. Worse, even if we (p. 409) suppose (falsely) that she had shown that her view is not too narrow, she would not have shown that either non‐sentient foetuses or sentient foetuses lack full moral status. Thus, her account would fail to support the abortion choice position that she plainly wishes to defend. Her updated personhood view is doubly unjustified.
David Boonin's Improved Desire Account
Michael Tooley's desire based abortion choice view has been updated by David Boonin. Boonin has recently developed a sophisticated version of a desire account of the wrongness of killing that, he claims, (1) deals in a satisfactory way with the difficulties to which Tooley's (1972) desire account is subject and (2) is superior to the future of value account (Boonin 2003). Boonin calls the desire account a present desire account. He characterizes the future of value account as a present or future desire account, as follows:
If an individual P has a future‐like‐ours F and if either (a) P now desires that F be preserved, or (b) P will later desire to continue having the experiences contained in F (if P is not killed), then P is an individual with the same right to life as you or I. (Boonin 2003: 63)
Boonin's present desire account drops clause (b) from the antecedent of the above conditional thus:
Plainly the present desire account allows for abortion choice (unless there is another sufficient condition for having the same right to life as you or I), while the present or future desire account does not.
If an individual P has a future‐like‐ours and if P now desires that F be preserved, then P is an individual with the same right to life as you or I. (Boonin 2003: 64)
Boonin proposes two strategies for dealing with Tooley's apparent counterexamples to the present desire account. Consider, first, the temporarily comatose adult whom we believe it is wrong to kill, but who, because of lack of awareness, not only does not desire to continue to exist, but is incapable of desiring to continue to exist. Boonin argues that, although such an adult lacks an occurrent desire to live, she continues (under ordinary circumstances) to have a dispositional desire to continue to live. He goes on to argue for the continued existence of beliefs and desires while one is comatose on the grounds that such beliefs and desires do not have to be reacquired when one becomes conscious. Furthermore, it is reasonable to believe that a woman desires that her husband not commit adultery even when she is thinking of an issue associated with her profession. Accordingly, we have good reasons for accepting the reality of dispositional desires. Boonin's account of and defense of a dispositional desire strategy for dealing with the alleged temporarily (p. 410) unconscious adult counterexample to the present desire view seems reasonable (Boonin 2003: 64–70).
Consider now persons who lack a present occurrent or dispositional desire to live, but who would have, in the future, a desire to live. Boonin notes that a person's actual desires may be formed ‘in conditions of great emotional distress’ or when there is ‘a lack of full and accurate information’ (Boonin 2003: 70, 71). In these cases the present desire one ‘would have had if the actual desire had been formed under more ideal circumstances’ is morally important (Boonin 2003: 70). Boonin calls such desires ‘ideal desires’. The person who is depressed because of mental illness or because of great emotional trauma or who wants to die because of misinformation concerning her future will (typically) have an ideal desire to live, even though she lacks an actual desire to live. Accordingly, a present ideal dispositional desire version of a desire account of the wrongness of killing is not vulnerable to any of the Tooleyan counterexamples. Because foetuses have no desires at all, whether occurrent, dispositional, or ideal, abortion choice is morally permissible (Boonin 2003: 73). Boonin offers three arguments for his view that his present ideal dispositional desire account of the wrongness of killing is superior to the present or future desire, that is, the future of value, account.
Boonin's Parsimony Argument
According to Boonin his present ideal dispositional desire account is preferable to the present or future desire account because it is more parsimonious. On the present desire account we need to appeal only to one property of an individual to explain the wrongness of killing; on the present or future desire view, an appeal to two properties is required (Boonin 2003: 66–7, 73).
Boonin's argument is successful only if the future of value argument's lack of parsimony is due to the nature of the future of value account itself rather than to Boonin's rendering of it. In short, Boonin's appeal to parsimony would fail if the future of value account can be stated in a way that is as parsimonious as Boonin's account.
It usually is stated in just such a fashion. It makes reference only to the value of one's future, not to the value of one's present or past. Accordingly, the lack of parsimony that Boonin finds in the future of value account is really a function only of Boonin's statement of that account of the wrongness of killing, not of the account itself. Because there is no good reason to include present desires in the statement of the future of value account, other than for the purpose of rejecting the account on grounds of parsimony, I shall discard the unwieldy locution of present or future desires and refer to the account Boonin rejects as a future of value account.
Boonin's Salience Argument
Boonin claims that his present ideal dispositional desire account of the wrongness of killing is more ‘salient’ than a future of value account. He means that his account (p. 411) of the wrongness of killing ‘enables us to account for the prima facie wrongness of killing by understanding killing as one instance of a more general category of acts that are prima facie wrong: acts that fail to respect the desires of others’ (Boonin 2003: 74). The future of value account, by contrast, makes the wrongness of killing the suicidal ‘an anomaly’ (Boonin 2003: 76).
There are reasons for thinking that neither of Boonin's claims is true. Consider the second. Let us, following Boonin, discuss the case of Hans, who ‘has been dumped by his girlfriend and has plunged into a deep depression. He can think about nothing else and has no desire to go on living’ (Boonin 2003: 70). The future of value account makes killing Hans wrong for the same reason it is wrong to kill almost all other human beings. To kill Hans is to make him worse off than he otherwise would have been. To make him worse off than he otherwise would have been is to harm him. To harm him greatly is prima facie wrong. Boonin's claim that the future of value account makes the wrongness of killing the suicidal anomalous is plainly false.
Now consider Boonin's first claim. Boonin seems to be committed to the view that an account of wronging others in terms of failing to respect their desires is salient. There is something to be said for such a view. Of course, it is presumptively wrong to fail to respect the desires of others. The desires we should respect are the actual desires of others. In the case of Hans, Boonin holds that we should not respect his actual desires. Therefore, Boonin's account of the wrongness of killing Hans is anomalous with respect to the actual desire account of wrongness that meets his test of salience. Boonin has the matter of anomalies and the contrast between his present desire view and the future of value view exactly backwards.
Worse is to come. According to Boonin it is wrong to kill Hans because killing him fails to respect his ideal desires. A nice feature of a present actual desire view is that it is not difficult to determine what the actual desires of others are. One just asks them. Ideal desires are harder, because they are hypothetical. We cannot determine what they are by asking the relevant other. How then does one determine their content?
On this score Boonin is not especially helpful. He says, ‘I am no more prepared to offer a full‐blown theory of ideal desires than I am to offer a full‐blown theory of dispositional desires. But again I do not believe that anything like this is necessary’ (Boonin 2003: 78). Boonin justifies this lack of an account of the central notion in his theory of the wrongness of killing by noting that it is as difficult to determine that Hans would have a future of value as it is to determine what Hans's ideal desire concerning his future would be (Boonin 2003: 78).
It is not, of course, difficult to have good reasons for believing that certain persons who are suicidal have futures of value. Such persons are like other persons who have been suicidal and who have been treated with psychotropic drugs or with psychotherapy, and whose lives have improved. Psychiatrists have a great deal of data concerning such persons. To the extent that they are committed to (p. 412) scientific medicine, they will regard gathering such data as one of their professional responsibilities. Our judgment that many persons who are suicidal have a future of value has a sound empirical basis. It is indeed no more difficult to determine what a person's present ideal desire is in these cases than it is to determine whether or not such a person has a future of value. This is because someone's present ideal desire is nothing more than the desire of a rational and fully informed individual in that individual's situation, and a rational, fully informed desire about one's own future is based on the judgment that a rational and prudent individual would make about the value of her future. (Tricky issues arise in contexts in which duties to others are involved, of course.) Therefore, Hans will have an ideal desire to live just in case a rationally, fully informed individual in Hans's situation will judge that his future life has positive value as compared with death. A rationally, fully informed individual in Hans's situation will judge that Hans's life has positive value as compared with death just in case Hans has a future of value. Boonin's claim that the present ideal desire view and the future of value view will pick out (with one rare exception) the same range of (postnatal) cases is quite true. It is true because the present ideal desire view, when explicated in an obvious way, is parasitic upon the future of value view! The present ideal desire view turns out to be concerned, in the final analysis, neither with one's desires, as contrasted with one's future welfare, nor with one's present as compared to one's future. Thus, although Boonin's claim that the future of value view fails the salience test is incorrect, his claim that the present ideal desire view is salient is correct. It is salient because it is the future of value account stated in the language of desires.
Boonin's Counterexample Argument
Boonin believes that there is a counterexample to the future of value view that is not a counterexample to the present ideal desire view. He asks us to imagine a case of a person whose future will contain many of the sorts of experiences other humans take to be valuable, but, because of an irreversible chemical imbalance in her brain, does not now and never will value any of those experiences. Boonin claims that the future of value theory implies that it would not be wrong to kill this person. He is correct. Boonin claims that the present ideal desire view entails that this person ‘has the same right to life as you and I’. Is this true?
The ideal desire view asks us to imagine what a fully informed rational person would presently desire in this chemical imbalance case. So let us imagine that we ourselves are in this chemical imbalance situation, but that there is a drug that can relieve us of the imbalance only temporarily, only briefly, and only once so that we can survey our future rationally and with full information. It is hard to imagine why such a person would have the present desire to live after the drug was no longer effective. Boonin's analysis of this alleged counterexample is incorrect.
Boonin's argument fails in another respect. Suppose his analysis were correct and our chemically imbalanced individual did have the ideal desire to live. Why should this show that the future of value theory is false? Because, says Boonin, ‘virtually every critic of abortion will agree that [this individual] has a right to life’ (Boonin 2003: 77). Boonin's criterion of truth is not correct. If it were, then we would have a (too easy) proof for the existence of God! We may conclude that none of Boonin's arguments for the superiority of the ideal desire account over the future of value account of the wrongness of killing is successful.
Why Boonin's Account Is Arbitrary
Boonin's ideal desire account of the wrongness of killing suffers from another problem. At least two versions of an ideal desire account are possible. On Boonin's version, having an ideal desire at a time requires having some actual desire at that time. An ideal desire is an actual desire that is corrected, when necessary, to account for imperfect information and imperfect rationality. Because no preconscious foetus has an actual desire, no preconscious foetus has an ideal desire to live, and because no preconscious foetus has an ideal desire to live, no preconscious foetus has the right to life. However, the Boonin version of the ideal desire account could be altered so that the actual desire requirement is dropped. Then foetuses could have ideal desires also. So just as Hans, were he fully rational and fully informed, would have the desire to continue to exist, a preconscious foetus, were she rational and fully informed, would have (because she, like Hans, has a future of value) the desire to continue to exist. Why prefer Boonin's version of the ideal desire account to the ‘Marquis version’ of the ideal desire account? The characterization of the second version is apt because, of course, such a version of the ideal desire account is just the future of value account in ideal desire clothing.
Boonin emphasizes that his preferred account is the Boonin, not the Marquis, version. That is no argument. Boonin notes the differences between the Boonin and the Marquis versions of the ideal desire account (Boonin 2003: 80–3). That is no argument either. Perhaps Boonin regards the following as an argument: ‘There is no desire that a rock would have under more ideal circumstances, for example, because a rock does not have any desires to begin with. But it follows from this that a particular ideal desire can meaningfully be attributed only to someone who has at least some other actual desires’ (Boonin 2003: 80). If this is a claim about Boonin's version of the ideal desire theory, this is true, but does not establish the superiority of Boonin's version. One would not attribute even ideal desires to rocks because there is no present or future stage of the rock that it would later value or disvalue. Foetuses are quite different. Hypothetical desires can be attributed as easily to foetuses as to Hans. There are reasonably clear criteria for the attribution in both cases. Accordingly, even if Boonin had shown that an ideal desire account of (p. 414) the wrongness of killing were superior to the future of value account, he still would not have established that his ideal desire account of the wrongness of killing should be accepted. Hence, he would not have adequately defended abortion choice.
We may safely conclude that, although Boonin's ideal desire version of Tooley's desire account of the wrongness of killing is not vulnerable to counterexamples, Boonin fails to show that his theory is better than the future of value theory. Indeed, if we eliminate an arbitrary feature of Boonin's theory, Boonin could not have shown that the ideal desire theory is preferable to the future of value theory. With this elimination, the ideal desire theory is the same as the future of value theory.
Summary and Conclusion
A brief survey of the analysis of this chapter will be useful. Think of individuals whom we all agree it is presumptively seriously wrong to kill. If pressed, we might say that it is wrong to kill them because they are human beings or because they are persons. Each reason has different implications for the ethics of abortion. There are difficulties, as we have seen, with each reason. A number of philosophers have attempted to resolve the abortion controversy by finding reasons for the wrongness of killing embedded in broader, and clearly relevant, moral considerations. There are at least three such accounts. The desire account is based on the moral truth that, in general, impeding the self‐regarding desires of others is wrong. The future of value account is based on the moral truth that, in general, inflicting harm on others is wrong when we would regard it as wrong to inflict a harm of that sort on us. The moral agency account is based on the moral truth that, in general, failing to respect the will of rational moral agents is wrong. The purpose of this chapter was to appraise these three accounts, all of which are intended to support particular views of the ethics of abortion.
The future of value account seems to deal adequately with cases in which there is a consensus that killing is wrong and also with cases in which few object to ending intentionally a life. It supports the view that abortion is immoral. Is there an account that is at least as plausible that underwrites abortion choice? According to the desire account, because foetuses do not desire to live, they lack the right to life. The desire argument strategy fails because many adults who do not desire to live plainly do have the right to life. David Boonin has attempted to plug the holes in this Tooleyan actual desire account by proposing to replace it with an ideal desire account. However, Boonin's strategy faces, as we have seen, many difficulties. In general, our duty not to interfere with the desires of others is a duty not to impede their actual desires, not their ideal desires when those ideal desires are different. Furthermore, the ideal desire view, in order to deal adequately with consensus cases, turns out to be parasitic on the future of value view. In addition, Boonin preserves (p. 415) abortion choice only by adopting an ad hoc restriction to avoid attributing ideal desires to foetuses.
Mary Anne Warren's (1997) defense of abortion choice shares a difficulty with the actual desire account: Her Agent's Rights principle, although it clearly embodies a generally acknowledged principle of respect, fails to account for the wrongness of killing many individuals whom we all believe it is wrong to kill. Warren attempts to deal with this gap by adding a Human Rights principle to her account. Because her Human Rights principle does not fit easily into a general theory of wrong action her defense of this principle is inadequate. It seems more like an ad hoc attempt to deal with the problems with her view. Worse, without utterly arbitrary restrictions to the Human Rights principle, the principle supports the claim that abortion is immoral.
Interestingly, Warren's and Boonin's views are similar. Neither a ‘bare bones’ desire view nor a respect for moral agency view account for the wrongness of killing all those who clearly have the right to life. Hence, strategies must be found for expanding both views. Unless arbitrary restrictions are placed on both expansions, both views will support the view that abortion is immoral. Hence, both Warren and Boonin, in the final analysis, must restrict their accounts in arbitrary ways.
The analysis of this chapter supports the conclusion that, of the alternatives considered in this chapter, the future of value view is superior and abortion is immoral. There are, of course, other alternatives to, and objections to, the future of value view that have not been discussed in this chapter. Furthermore, the important view that, even if foetuses have the right to life, women do not have the obligation to provide them with uterine life support also has not been considered.
Boonin, D. (2003), A Defense of Abortion (Cambridge: Cambridge University Press).Find this resource:
McMahan, J. (2002), The Ethics of Killing: Problems at the Margins of Life (New York: Oxford University Press).Find this resource:
Marquis, D. (1989), ‘Why Abortion Is Immoral’, Journal of Philosophy, 89: 183–202.Find this resource:
Purdy, L., and Tooley, M. (1974), ‘Is Abortion Murder?’, in R. Perkins (ed.), Abortion: Pro and Con (Cambridge, Mass.: Schenkman).Find this resource:
Stone, J. (1987), ‘Why Potentiality Matters’, Canadian Journal of Philosophy, 17: 815–30.Find this resource:
Tooley, M. (1972), ‘Abortion and Infanticide’, Philosophy and Public Affairs, 2: 37–65.Find this resource:
—— (1973), ‘A Defense of Abortion and Infanticide’, in J. Feinberg (ed.), The Problem of Abortion, 1st edn. (Belmont, Calif.: Wadsworth), 51–91.Find this resource:
—— (1983), Abortion and Infanticide (Oxford: Clarendon Press).Find this resource:
Warren, M. A. (1979), ‘On the Moral and Legal Status of Abortion’, repr. in R. A. Wasserstrom (ed.), Today's Moral Problems, 2nd edn. (New York: Macmillan), 35–51.Find this resource:
—— (1997), Moral Status: Obligations to Persons and Other Living Things (Oxford: Clarendon Press).Find this resource:
Thanks to Katrina Elliott, Ben Eggleston, and Bonnie Steinbock for helpful comments.