A. Ways of mattering morally

When we say that something counts morally in its own right, we are often said to be thinking of its intrinsic worth or value rather than its instrumental value. If it were morally right to treat animals well only because this would promote kindness between persons, then animals would count morally only instrumentally. That is, they should be treated well not because of what they are in their own right, but only because of the effects on others of treating them well. Philosopher Christine Korsgaard has argued, however, that the true contrast to mere instrumental value is having value as an end, not having intrinsic value.²

That is, if the animal counts morally in its own right, there is no further end that need be served by our treating the animal well in order for us to have a reason to treat it well. If something is an end (in this limited sense), this need not mean that it has value that can never be trumped, or that it can never be treated as a mere means. At minimum, it means only that its condition can provide a reason (even if an overrideable one) for attitudes or actions appropriate to its moral worth, independent of other considerations. Korsgaard argues that some things may be ends in virtue of their intrinsic properties, but others may be ends in virtue of their extrinsic properties. A thing's intrinsic properties are all its nonrelational properties.³ Its extrinsic properties are properties it has in virtue of standing in relation to other things. For example, Ronald Dworkin claims to have a theory of the intrinsic value of even nonsentient, nonconscious life, such as is found in an early embryo. However, he also says this value comes from the history of the embryo, in particular the investment that nature or God has made in it. This is not a theory of the intrinsic value of a life but of its extrinsic value, since it derives the value of the embryo from its particular history rather than from properties it has independent of this history.⁴ Further, an entity's ability to produce an effect is a relational property holding between it and the effect. It is possible, given what Korsgaard has said, that something could be worth treating as an end because it is capable of causing a certain effect, even if it never does.

A work of art or a tree may count morally in its own right in the sense that it gives us reason to constrain our behavior toward it (for example, not destroy it) just because that would preserve this entity. That is, independent of the pleasure or enlightenment that it does or will cause in people, a thing of aesthetic value gives us reason not to destroy it. In that sense, it counts morally. Nevertheless, this is still to be distinguished from constraining ourselves for the sake of the work of art or the tree. I do not act for its sake when I save a work of art, because I do not think of its good and how its continuing existence would be good for it when I save it. (Nor do I think of its exercising its capacities or performing its duties. Acting for the sake of its exercising capacities or performing duties might also involve acting for the entity's sake, though it need not involve seeking what is good for it.) Rather, I think of the good of the work of art, its worth as an art object, when I save it for no other reason than that it will continue to exist. By contrast, when I save a bird, I can do it for the bird's sake, because it will get something out of continuing to exist and it could be a harm to it not to continue. It seems that something must already have or have had the capacity for sentience or consciousness in order for it to be harmed by not continuing on in existence. This is because an entity must be able to get something out of its continuing existence, and a capacity for sentience or consciousness seems to be necessary for this. (I do not think either capacity is a necessary condition for us to be able to act for the sake of the entity, since each without the other is sufficient. Having the capacity is not the same as actually being sentient. It is also not the same as merely having the potential to be sentient, where the latter implies that an entity merely has the potential to have the capacity.) I shall say that an entity has (positive) moral status when in its own right, for its own sake, it gives us reason to help it and to refrain from destroying it.

On this account, a nonsentient, nonconscious embryo lacks moral status but could count morally in itself (e.g., could give us reason in its own right not to destroy it) because of its intrinsic and extrinsic properties, such as what its potential is. This is different from its merely having instrumental value because it will in fact give rise to a person that has moral status. For even if the embryo is not instrumental to there being a person (because it is deprived of an environment in which to develop), I believe its having the potential could still give it greater value than an embryo that lacks the potential. (Similarly, a Chippendale dining table may have value in itself and may have more value if it can also turn into a magnificent writing desk, though it will not.) Notice, however, that the instrumental value of an embryo might be greater if it will generate Beethoven rather than an ordinary person, even if the moral status of these two persons does not differ. Similarly, an embryo can have greater value in its own right if it has the potential to become an extraordinary human being (e.g., Beethoven) rather than an ordinary human being, even if these persons have the same moral status, and the embryo will not, in fact, generate anyone.

All this does not mean, however, that an embryo's continued existence is good for it or that it is harmed by not continuing on. Similarly, an ordinary table might, by magic, be turned into a table that can develop into a person, and it may be good to be a person, but can a table be the sort of thing that is harmed by not getting the good of its transformation? It does not seem so. The person that would come from the embryo also cannot be harmed by never coming to exist. But we can act for the sake of a person who will develop from the embryo by doing things to the embryo not for its sake but for the sake of the person who will exist. (I shall return to this issue in Section IV below.) The fact that an embryo may have value in its own right in virtue of its extrinsic properties could account for why it might be wrong to use it for frivolous purposes. If so, the ground for objecting to such acts would be like the ground for objecting to making lampshades of the flesh of deceased human beings (who had died of natural causes). The flesh has no moral status (as I am using the term), but it has an extrinsic relation to once-living human persons who had moral status, and thus it may give us reason in its own right not to use it in certain ways. (The embryo's particular intrinsic and extrinsic properties, of course, differ from those of the dead flesh.)

Those things for whose sake we can act when we save their lives may or may not give us as much reason to save them as entities whose existence cannot be extended for their own sake. For example, if we had to choose whether to destroy the Grand Canyon or a bird (holding constant the number of people who would get pleasure or be enlightened by each), it could be morally wrong to choose the Grand Canyon. This illustrates how something like a bird can count morally because it can get something out of life, without its giving more reason to act, in the sense of how significant its continuing in existence would be, relative to other things. Sometimes the remarkableness of something (such as the Grand Canyon) or its uniqueness calls for more protection than something else's having moral status.

Not all things whose lives we could save (or not destroy) for their own sake are things it would be wrong not to save or to destroy (for example, dangerous dogs). But if we did have a duty sometimes to save or not to destroy such entities, this still does not imply that all the entities that have moral status are entities to whom we owe it to save or not destroy their lives. There is a difference between one's having a duty to do something and having a duty to a specific entity to do it. The latter is known as a directed duty, and typically it has a correlative that is a right or a claim possessed by the entity to whom the duty is owed against the person who owes it.⁵

Correspondingly, there is a difference between doing the wrong thing (for example, not fulfilling a nondirected duty) and wronging some entity by failing to perform the duty owed to her. The entity who is owed a duty is not necessarily the entity who is benefited or affected by the object of the duty. For example, if you owe it to me to take care of my mother, I am the right-holder, not my mother, even though the object of the duty is to benefit her, and you wrong me, not my mother, if you fail to help her. Arguably, the idea of respect for persons and the dignity of the person amounts to the idea that one owes it to a person to behave in certain ways, rather than simply that it would be wrong to treat her in certain ways because, for example, one owes it to God not to do so or because it would not maximize utility to do so and one has a general duty as a rational being (but not a duty owed to anyone) to maximize utility. Thus, just as only some entities that count in their own right are entities that have moral status (as I have defined it), so only some entities that have moral status have rights against us.

To what sort of entities is it possible to owe things or behaviors? Thomas Scanlon has argued that only entities capable of judgment-sensitive attitudes are entities to whom we can owe certain treatment.⁶

(Scanlon does not speak of rights as the correlatives of directed duties, but I believe the addition of rights talk in his system would be appropriate.)⁷ Entities capable of judgment-sensitive attitudes form attitudes or decide on actions on the basis of evaluating certain factors as reasons, that is, as considerations in favor of an attitude or action. For example, they do not just respond to aspects of their environment (as a cat would); they see these aspects as considerations in favor of or against action. Scanlon's view seems to be that if some entities can evaluate our conduct toward them (so that they can see a reason for us to act or not act in that way), then we may potentially owe it to them to act or not act in that way.⁸ He also seems to think that a creature capable of judgment-sensitive attitudes governs itself in the light of reasons, and thus it is only to self-governing creatures that we can owe things. (It is possible, however, to imagine that the capacity for judgment-sensitive attitudes does not go so far as to involve self-governance in the light of reasons. For example, a creature might take certain factors in the environment as true reasons to pursue food but might not be self-conscious and thus not self-governing. I am not sure what Scanlon would say if these two conditions were pulled apart.)

Scanlon thinks animals count morally in their own right and give us reasons to act for their sake. Hence, our conduct toward them can be right or wrong, independent of further considerations, but it cannot be owed to them and they cannot be wronged when we behave wrongly. This is because (he assumes) they are not capable of judgment-sensitive attitudes. Furthermore, he thinks that while we have a reason to help an animal in need, we can have the same reason to help a rational being in need plus an additional reason absent in the case of the animal: that is, we can owe it to the rational being to help him. On this account, the “greater” moral importance or value of rational beings (persons) gets fleshed out (in part) as the additional factor present in our relations with them: that is, we owe things to them or, as I would say, they have rights against us.

B. The bearing of personal identity on our treatment of persons

Having considered some different types of entities, I now want to consider whether and how we may take account of or ignore the facts of personal identity when we are dealing with persons.

There may be conflicts between satisfying interests or respecting rights and maximizing satisfaction of interests intrapersonally and interpersonally. (I will refer to interpersonally maximizing interests satisfied as “producing the greater good.”) Consider first the conflict between negative rights and great goods not protected by rights.⁹

It is often said that part of the reason why one person's right not to be paralyzed should take precedence over the greater good (when they conflict) is that others, not he, will receive the benefit of his sacrifice. Sometimes, however, it will be even worse to harm a person for his own greater benefit than it would be to harm him for the good of others. This is when we act against his will, for the charge of paternalism arises in the former case but not in the latter. In other cases, it seems that the reason for not imposing harm on one person is not that he will fail to be benefited, but rather that the greater good we produce is of the wrong sort. For example, it may consist in many small goods aggregated over many other persons, rather than a large benefit to any one other person. Consider a case in which someone has a negative right to not have his car damaged. I might nevertheless permissibly damage his car in the course of using it to rush someone else (who would otherwise die) to the hospital. If it were possible that, by damaging the car, I could produce a comparable number of additional years of life by giving an enormous number of people one minute each of additional life, would this be permissible too? It seems not. But is it really the separateness of the beneficiaries that stands in the way? Suppose that each of the beneficiaries of the small benefit faced as bad a prospect as the one person who would die (that is, suppose they were each on the point of death). Might not the small goods to many separate people then justify the imposition of harm on the one person?¹⁰

If we still think that the large, concentrated benefit of the life saved, and not the small benefits to each soon-to-die person, justifies the imposition in these cases, it is nevertheless true that aggregating some small benefits to people (each of whom would be very badly off) might sometimes justify a larger imposition on a person who would not be as badly off.

In the case just discussed, we have compared individuals one person at a time to see who will wind up worse off and who will get what, if we act one way rather than another. This procedure is called “pairwise comparison.” A proponent of pairwise comparison claims that the fact that potential recipients are separate persons—not only that the person undergoing harm is separate from anyone benefited—is relevant in deciding whether to perform an action that harms one person. By contrast, if we merely aggregate benefits over many people, then we ignore the distribution of the benefits over individual persons. (Notice that one person can be worse off than another through intrapersonal aggregation of harms, but aggregating over one life, even assuming constant personal identity, has problems of its own. For example, if one person will suffer a thousand headaches over the course of his life, but he will live seventy years and the headaches are interspersed evenly, he may be better off than someone who will have five hundred headaches over the course of a ten-year life, especially if they are bunched together.)

Arguably, the method of pairwise comparison is not adequate on its own. For sometimes it seems that a loss suffered by someone can be justified even though the benefit does not come to him but to others, each of whom would suffer less (if not aided) and benefit less (if aided) than he would. For example, suppose I refuse to save someone from death in order to save a million from paralysis. If aggregation in this case is permissible, it seems to diminish the significance of the separateness of persons in the way emphasized by pairwise comparison. Indeed, if I were to decide it was permissible to kill a person for the sake of saving the lives of a million people—a case in which each person would (if left unaided) suffer as great a loss as the one who is to be killed—this too would diminish the significance of the separateness of persons, at least as expressed by the requirement of pure pairwise comparison, since no one of the million would be any worse off (if not aided) than the one person would be if he died, nor would any one of the million receive a benefit (if aided) greater than the single person would receive if he were allowed to live.

There is, however, another way of taking account of the separateness of persons besides pure pairwise comparison. I call this “balancing of equals.” Imagine cases where there is a conflict of interests and we must decide whether to aid a smaller or a larger group (of non-overlapping persons). A person on one side of a conflict is balanced out against another on the other side, and the side with the greater number is helped, at least when each stands to lose and gain as much. This is a form of substitution of equals; for in a conflict between persons A and B, we think, from an impartial point of view, that we accomplish as much if we save B as if we save A, and hence we accomplish even more if we save B plus C. We think this even though A does not benefit if B is saved (because they are separate persons) and even though, from A's partial perspective, the outcome in which she survives is better than the one in which B survives. The balancing-of-equals approach can still respect the separateness of persons by not weighing against A's loss of twenty years of life a combination of B's loss of ten years and C's loss of ten years. That is, we can first engage in pairwise comparison to check to see if a person will suffer the greater loss and/or receive a greater benefit and only balance out those who stand to lose or gain as much. The combination of pairwise comparison and balancing allows us to justify saving a greater number rather than a smaller number of people in a conflict situation. However, if we allow an aggregation of losses suffered by several persons to outweigh another individual's greater loss (for example, saving a million from paralysis rather than one person from death), we still move beyond anything that balancing entails.

In cases involving negative rights, balancing is ruled out more frequently, even when the greater good we would produce by transgressing one person's negative right involves minimizing violations of comparable rights in others (each of whom stands to lose as much). For example, we may not kill one person to harvest his organs in order to save ten people from death. What is being relied on here to reinforce the moral significance of the separateness of persons and rule out balancing, I believe, is a claim on the part of a person to what would be taken from him (e.g., his property or his liberty) that is much stronger than anyone else's claim to it. By contrast, in cases where I must choose when to aid, each party may have the same (or no) claim on my assistance.¹¹

However, another suggestion that has been made to explain the cases involving negative rights reminds us of yet another way in which the separateness of persons has been thought to play a role in ethics. Some have said that my transgressing a right in order to prevent others from doing so produces the worse state of affairs from my perspective. The emphasis in this account seems to shift from the significance of the separateness of the potential victim and the potential beneficiaries to the significance of the separateness of a particular agent from other agents. This suggests that if another agent were to transgress someone's negative right for the sake of minimizing comparable rights violations, I would have no reason to stop him. For example, suppose B will violate A's right not to have his arm amputated in order to save many others from having their arms amputated. Do I have a reason (even if not a duty) to put aside my activities to stop the violation of A's right, rather than let it be the cause of the greater good? I think I do.¹²

Also, if it were most important that I not be involved in violating a negative right, should I minimize the number of such acts that I commit? May I kill one person now to stop a threat I started yesterday that will soon kill five other people, making me the killer of five people? I think not. These cases suggest that it is a mistake to interpret the constraint on transgressing one person's right as expressing an agent's concern with his own agency.

So far, I have been characterizing some ways in which the separateness of persons is taken account of in certain moral theories. We can summarize these ways as follows: (1) not substituting one person for another, period; (2) not substituting one person for another unless there is an equal loss and/or gain to be sustained by each; (3) not substituting one person for another when the person has a claim to control what makes him separate; (4) an agent not substituting for other agents.

C. Substitutability and ways of mattering morally

Now let us consider creatures who have a good but are not rights-bearers, and also entities of value in their own right but not ones for whose sake we can act. Let us see whether there is any principle analogous to the separateness of persons that can be applied when dealing with such entities. Would it be permissible to destroy one non-rights-bearing animal in order to prevent the destruction of others? Arguably, yes. (Notice that the sacrificeability of one animal for others is consistent with the fact that it is the sort of being that has a life of its own and could have been benefited if we had not harmed it.) Does this further imply that we may be utilitarians and simple maximizers with respect to animals? No, because it may still be wrong to aggregate small losses to many animals against a big loss to one animal at least when the former would not be as badly off (if not helped) as the latter. It might also be wrong to sacrifice a higher animal (for example, one with a greater degree of intellectual and interactive capacities) to save a greater number of lower ones. (In these respects, Robert Nozick seems to have been wrong in suggesting deontology for people and utilitarianism for animals.)¹³

Would it be permissible to destroy one embryo to save others from being destroyed, at least when the latter will develop into people? I believe so. Would it be permissible to destroy one great artwork to prevent several other comparable artworks from being destroyed? I believe so. Would it be permissible to destroy one human embryo to allow many monkey embryos to develop? I believe so.¹⁴

Would it be permissible to destroy one great art work for the sake of several less significant ones? Interestingly, this seems truly misguided.

This broad sacrificeability of one entity for the sake of others is in sharp contrast with the way it is permissible to treat a self-governing being. Yet it is also in contrast to the way we should treat some non-rights-bearing entities. For such sacrificeability is not true of entities whose value is purely symbolic, like flags, nor for holy entities. These things have no rights, and no good for whose sake we act, yet there are reasons not to destroy one for the sake of others that are analogous to reasons that apply to persons. For example, it may sometimes make no sense for it to be permissible to burn one flag to prevent many other flags from being burnt impermissibly. This is the case if it is their status as inviolable that we are concerned with in trying to save the many. For the permissibility of burning one to save others denies that very inviolability. If this is true of purely symbolic entities, but not of embryos, this suggests that the value of the embryo does not reside merely in its role as a symbol of human life. An embryo may serve merely as a symbol only when we are not interested in its future development. Hence, we may think that it makes sense to destroy one embryo for the sake of the future development of other embryos, but not for the sake of their survival as mere symbols of life in a freezer.

In Section II, I have considered different types of entities falling into increasingly narrow categories: those that count in their own right, those that have moral status, and those that can have claims against us. I have argued that the moral nonsubstitutability of the entities that can have claims against us is stronger than the nonsubstitutability of those that merely count in their own right or that merely have moral status. This is true even though those in the latter class can have separate points of view and interests of their own. In the next section, I attempt to apply these conclusions about nonsubstitutability to the issue of reproductive cloning.

A. The embryo of a future person

I have argued that a cloned person (just like a noncloned person) could not permissibly be harmed in order to minimize comparable harms to other persons, cloned or uncloned. Now I want to consider whether an embryo that will in fact develop into a person is also protected against less-than-lethal interventions that will negatively affect the future person and that are undertaken for the sake of benefiting other future persons by affecting other embryos from which they will develop.

Previously, I have argued against what I shall call the View:²⁵

We have as strong a duty not to do things to an embryo that will result in harm (or fail to prevent harm) to the person that the embryo will definitely develop into as we have not to harm (or fail to prevent harm to) that person when he is already in existence. Notice that the View is compatible with thinking that a pregnant woman is not obligated to do (or refrain from doing) certain things that a woman would have to do (or refrain from doing) for an offspring outside her body. This is because more of an imposition on her may be involved if she must do these things when the embryo is in her body than when it is outside, and the duty she has may not license such an imposition.

One premise in an argument (call it the Argument) for the View²⁶

that I have considered elsewhere is that, if we have duties to an embryo while it is still only an embryo, it is because of the person it will develop into. Those who make the Argument think that this premise implies the view that duties that exist while there is only an embryo should be as strong and of the same general type as duties that we have to the person once he has developed (at least if we are certain that the embryo will develop into a person and that what we do to it as an embryo will affect the person in the same way).

Recently, philosopher Elizabeth Harman has characterized the Argument as follows (speaking of early fetuses, whom she claims have no moral status but are capable of being harmed):

In place of the Argument, Harman offers what she calls the Actual Future Principle: “An early fetus that will become a person has some moral status. An early fetus that will die while it is still an early fetus has no moral status.”²⁸

(I assume Harman would apply the principle to embryos, too.) Notice that Harman's principle takes the view that the fetus does not merely give rise to a person. Harman thinks that it becomes a person in the sense that there is one human being from conception through personhood and the fetus that will become a person is that human being's earliest stage. Notice also that, as stated, the Actual Future Principle does not fully support the View. This is because it does not claim that the fetus's moral status is the same as the status of the person it will become and therefore that we have as strong a duty not to do things to an embryo that will result in harm to the person as we have not to harm the person when she is already in existence. Thus, Harman's principle may not strictly imply that the duty not to do things to the fetus that will harm (or not prevent harm to) the person it will become is as strong as the comparable duty to the person when she already exists. Nevertheless, it is not clear why Harman should not hold the View, for she also says: “The Actual Future Principle recognizes the moral status of early fetuses that will become persons; it is precisely these early fetuses in which persons can be said to be already present.”²⁹ If a person is already present, then the fetus's moral status should be the same as the status of the person.

A problem that arises for the Actual Future Principle that does not arise for the Argument is that the former implies (when combined with Harman's view that it is possible to harm early fetuses, and presumably embryos) that we may not do things to a fetus that harm it, even when the harm is short-lived and will not affect the person whom the fetus will definitely become. This is because we have a prima facie duty not to harm a person, even when the harm is short-lived and will not go on to affect the person in the future. I think (a) it is either not true that it is impermissible to cause a short-lived harm, or (b) if it is impermissible to cause such a short-lived harm (e.g., pain), this will be true even if the embryo/fetus will not develop into a person.

I shall put this issue aside, for I am interested to show that the View is wrong and that if the Actual Future Principle implies the View, it too is wrong. Hence, either the Actual Future Principle does not imply the View and we need to know why it does not, or, I will argue, the principle is wrong. With respect to the Argument, I claim only to show (as I have tried to do elsewhere) that while its first premise may be true, it does not imply the View. That is, we can agree that any duties we have to treat a fetus in a certain way exist only because of the person it will develop into, but this does not imply the View.

Here is an example which, I believe, shows that it is permissible to affect a future person by doing something to the fetus (or to the embryo) from which he develops, though it is not permissible to affect the person in the same way by doing something to him once he exists. Suppose a woman has given a fetus genes that will result in a person with an IQ of 160. She decides this is too smart, not for the good of the person who has the high IQ, but for the good of the family. As a result, she takes a drug during early pregnancy to reduce the future person's IQ to 140.³⁰

This is a case of causing a person to be worse off than he would otherwise have been (I shall assume). I believe this is permissible (for reasons to be given below). But it would not be permissible, I believe, for the woman to give her child, once it exists, a pill that reduces its IQ from 160 to 140 or alters its genes so that the child will have an IQ of 140 rather than 160 in the future. What is the difference between (1) affecting the person by affecting the fetus and (2) directly affecting the person himself? A fetus is not the sort of being that is entitled to keep a characteristic that it has, such as a genetic makeup that will generate a 160 IQ. This is because it is not the sort of being that can be the bearer of rights (to retain anything). It lacks moral status (as defined in Section II) and lacks additional properties that would make it a rights-bearer, in part, because it is not sentient or conscious. In addition, the person who will develop from the fetus will not fall below an acceptable level of intelligence if he has only a 140 IQ, so he (as a person) is not owed a 160 IQ by his parent. (An IQ of 140 is already far above the minimal standard that, it might be argued, is owed to the people we create. I will here merely assume and not argue for this claim, that persons do owe their offspring certain things.)³¹ These two facts are crucial to the permissibility of taking back from the fetus IQ points that the parent gave it. But since a child is already a person (I assume), it is entitled to keep a beneficial characteristic it has, even if doing so raises the child far beyond the standard it is owed. Hence, I believe it is impermissible to give the IQ-reducing pill to the child, even if doing so would not cause his IQ to fall below the minimum owed to one's child.³² By contrast, suppose we owe a good chance of an IQ of at least 100 to people we create. In this case, doing something easily avoidable in early pregnancy to a fetus that results in a person who develops with an IQ below 100 may well be as impermissible as doing to the later child something that lowers its IQ to below 100.

Because the fetus is not yet a person (or does not yet have other properties that make it an entity that is entitled to keep what is given to it)—even though it will become a person—the act of taking away characteristics the fetus has (which will impact on the person that will be) is no different from not giving it those characteristics to begin with. And one would have a right not to give a future person that one created genes sufficient for a 160 IQ. Analogously, if a parent puts money into a bank account that will belong to someone when he exists as an adult, the fact that he will exist as an adult, by itself, does not imply that it is impermissible for the parent to take back the money before the person reaches the age at which he can claim his bank account. (This is true even if the parent has not earned her own money, but has inherited it from others as people inherit their genes.)

Suppose now that the fetus is not in the woman's body, dependent on her services in carrying it. Instead, it is growing in a mechanical gestation device. Yet imagine that the world is such that if the woman engages in an activity such as exercising (no matter where), this has the effect of altering the fetus so that the person who develops will have an IQ of 140 rather than 160. Would it be permissible for her to exercise? First, consider the variant in which she has given the fetus its intelligence genes. Here I think it is permissible for her to do what causes the drop in IQ, since she is taking back what she has given to an entity not yet entitled to retain this thing. However, it would be prima facie impermissible for her to engage in the same activity when it would affect a child of hers who is in possession of genetic material that will bring about the higher IQ, material he is entitled not to be deprived of. Now consider the variant in which it is the father who gave the fetus the intelligence genes. It is true that the fetus is not entitled to keep these genes, and on the basis of his donation, the father may remove them, but that does not mean that just anyone may take an action that would remove them (or alter their effects on the child's intelligence). Since the woman in this case has neither given the relevant genes nor is contributing any efforts necessary for the fruition of the genes, I think that she may not do those things that lead to a drop in IQ to 140. (Nor may any other entity, such as a government, without the father's permission, do things to lower the IQ to 140, other things being equal.)

What about the permissibility of a parent not rendering assistance to a fetus to prevent a natural change that will lead the person who develops from the fetus to have a 140 IQ rather than a 160 IQ? If the change would occur in the fetus before it is entitled to keep the traits, I believe that one need not make as great an effort to stop the reduction in IQ as one should make once one is the parent of the child who has the 160 IQ or the genetic traits that will lead to the 160 IQ. A parent's duty to help a child keep a beneficial trait that the child already has can be stronger than a parent's duty to see to it that her offspring comes to have such a trait by helping the embryo retain the genetic material.³³

But now consider an in-between case.³⁴

Suppose the parent is not physically able to remove the genetic material at the fetus stage, and (as I have argued) not permitted to take an action that would remove IQ points from the child-person. May she give to the fetus a drug that will have a delayed reaction in childhood (like a slow bomb), eliminating the child's genetic material that would lead it eventually to have a 160 IQ rather than a 140 IQ? I do not believe this is permissible. I also do not believe that anything I have said implies that it is permissible. For it involves doing something at time t₁ that will remove something at t₂ when there is a person to whom that item belongs. (That is, the person will actually suffer a loss, rather than merely living with the effect of an earlier loss.)

Recall the View as I set it out at the beginning of this section: We have as strong a duty not to do things to an embryo that will result in harm (or fail to prevent harm) to the person that the embryo will definitely develop into as we have not to harm (or fail to prevent harm to) that person when he is already in existence. Suppose that my arguments have shown that the View is incorrect. If the View is implied by the Actual Future Principle, then the latter is incorrect.

Now, imagine again that a mother has given a fetus genetic material that would give the person it will develop into a 160 IQ. Imagine further that if the mother takes from the fetus some genetic material, this makes the fetus develop into a person with a 140 IQ instead, and the mother can then transfer the material into two other fetuses, thereby raising their IQs from 130 to 140 each. What I have said above, I think, implies that doing this would be permissible. The mother would be morally free to equalize beneficial traits among future persons by affecting their embryos, even though she thereby makes some person worse off than he would otherwise be. However, I do not think it would be permissible for her to (safely) take from a child (already a person) some characteristic that will or does give him a 160 IQ, leaving him with a 140 IQ, so that she can transfer the material into two other children, raising their IQs from 130 to 140 each.

When modifying the genome of an embryo is contemplated, it is usually in a context where we can improve the life of the person who will develop from the embryo. I have raised the possibility that it might be permissible, for good reason, to alter an embryo in a way that will cause the person who develops to be worse off than he might otherwise be. It might be noted that in the real world there would be no need to take something from one embryo in order to place it in another embryo, as a means of improving the latter, as was true in my hypothetical case. And while making one person worse off without thereby improving others could increase equality between people, this sort of “leveling down” to achieve equality is often taken to be a counterexample to the value of achieving equality. However, taking something away from one embryo without transferring anything to others could still improve the lives of persons who will develop from the other embryos. For example, it might make it more likely that they will win their share of competitions with the other person. Then there could be an increase in their absolute well-being and the charge of “leveling down” would be defeated.

B. Future generations and the Non-Identity Problem

These results bear in an important way on our responsibilities to future generations. Take the imaginary case in which we know that certain particular people will definitely exist in a hundred years. Suppose we engage in activities today that will affect the environment in such a way that the air quality will not be as good in one hundred years, though it will still be above the level that we owe to future generations. Suppose, further, that there is no person in existence yet whose environment that future environment is. I believe we may engage in such activities. But if we were (somehow) transported one hundred years hence, it could well be impermissible to engage in the same activities that reduce to the same degree the air quality that the persons then living are already enjoying. In fact, it is not necessary that these people (or the children in my previous cases) actually already be in possession of the better environment (or the more advantageous trait, whether it is the higher IQ or the genetic trait that will lead to it). Given that they are already persons, if their prospects as persons are for acquiring such a superior environment (or advantageous trait), they should not be deprived of these prospects by an event that occurs once they are persons.³⁵

This means that we are permitted to act in ways that reduce the superior air quality (but not below an acceptable level) before future persons who will be affected by this exist. By contrast, we may not be permitted to have the same effect on the environment somewhat delayed, so that it occurs when the future persons already exist and impacts an environment they can claim as theirs. If we wish to do right by future generations, therefore, it will be very important to know what level of environmental quality they are owed and, surprisingly, when the alteration to the environment will occur relative to the existence of the persons affected by such an alteration. We cannot merely say that there is a duty not to make any future environment worse than it would otherwise be without our action.

Finally, let me consider the possible bearing of these issues on what is known as “the Non-Identity Problem.” Derek Parfit famously argued that sometimes, at least, it seems not to matter morally whether we are affecting the same (identical) person for the worse or just making someone worse off than some separate (nonidentical) person would have been.³⁶

Moral principles that tell us not to make persons worse off than they would have been, or to make persons better off than they would have been, are called “person-affecting principles.” The moral insignificance of nonidentity is a problem for those who think that all moral principles are person-affecting. Moral principles that tell us not to make there be people who are worse off than other people would have been (or to make persons who are better off than other persons would have been) are called “non-person-affecting principles.” This is not because there are no effects on persons in the latter cases, but rather because no person is better off than he would otherwise have been, given that the better-off person is someone else. So, for example, suppose that if we behave in a certain way now, we will affect the environment so that future people who would otherwise have existed anyway are worse off than they might have been. In this case, we affect people for the worse. Alternatively, suppose that if we behave in a certain way now, we will affect the world so that different people will exist in the future than would otherwise have existed, and they will live worse lives than those other people would have lived, due to changes our behavior also makes to the environment. In this case, we do not affect anyone so that he is worse off now than he would have been. Parfit claims that, morally speaking, it does not matter whether we affect existing persons for the worse or create worse-off persons.

Let us assume that the worse life in these examples is still a life worth living, and also that it is a life that is good enough to meet the standards that a responsible parent could be held to in creating new people. Then we cannot argue that people are entitled not to have the worse life because they are entitled not to be in certain states, whether by our making them worse off than they would have otherwise been or by our doing something that leads to the existence of worse-off people (than would have otherwise existed). That is, we cannot argue that person-affecting principles do not have special weight because people simply have a right not to be in the worse state. It is just the comparative worseness of one set of people that speaks against the act that produces them, just as it is the comparative worseness of the way in which a given person's life will go that speaks against acting in a way that makes him worse off.

How might what I have said against the View bear on the Non-Identity Problem? I have argued that sometimes the way in which we affect someone for the worse can make a moral difference, not just the fact that we affect him for the worse. If we affect him by doing something to him or to some trait or resource he is entitled to (only because he has it and it is or will be beneficial to him), this can have greater moral significance than if we affect him similarly by doing something to a trait or resource he is not yet entitled to because he does not yet exist as a being who can have entitlements. Standard cases in which we affect people for the worse are like the former. This leaves it open that person-affecting cases of the latter sort are not morally more significant than non-person-affecting cases, while cases of the former sort are more significant. Choosing to perform easily avoidable acts that lead to a given offspring having fewer resources may be no worse than choosing to perform easily avoidable acts that lead to creating a worse-off person rather than a different better-off person. Yet both sorts of acts may be less morally problematic than performing an act that will take from someone what he is entitled to keep (or not provide him with help he is entitled to get), resulting in his having fewer resources. If the person-affecting cases that are used to illustrate the Non-Identity Problem do not involve entitlement based on personal possession, then they do not compare the stronger form of person-affecting principles with non-person-affecting principles. This means that the argument for the moral equivalence of non-person-affecting and person-affecting actions is crucially incomplete.