We have been asking whether the act of conceiving a child can affect this child, for better or worse. If we answer “Yes,” the person‐affecting restriction makes no difference; principle (2) leads to the same conclusions as principle (1). We may therefore wish to answer “No” – but to this we have found objections.
The problem here can, I think, be solved. We can state the person‐affecting principle in a different form:
1 It is wrong to do what, of the alternatives, affects people for the worse.
We interpret (3) so that if people fail to receive possible benefits, they count as affected for the worse. If we adopt principle (3), we can afford to allow that conceiving someone is a case of affecting him. Since failing to receive benefits counts as being affected for the worse, principle (3) still tells us – like principle (2) – to do what benefits people most. But there is one exception. To the one benefit of receiving life (3) – unlike (2) – gives no weight. For when we fail to give this benefit, there isn’t an actual person who fails to receive it – who is thus affected for the worse. (I am now assuming, you remember, that we cannot affect possible people.)
Most of us, I claimed, think there is nothing wrong in not having children, even if they would have been very happy. But we think that having children who are bound to suffer is wrong. Principle (3) supports this asymmetrical pair of judgments. It supports our view that the Childless Couple did no wrong; but it also supports our view about “wrongful conception” – for the child here is an actual person affected for the worse.
In the move from (2) to (3), a natural principle is revised in a somewhat artificial way. But this revision does not seem to drain the principle of its plausibility. All the revision does is this. When we are choosing what to do, we are told to aim, not to achieve the outcome where people are better off, but to avoid the outcome where they are worse off. This procedure, adding up the “minuses,” seems to be just as general and as plausible as the other, adding up the “pluses.” So we are not, in moving to (3), “tailoring” our principles in an ad‐hoc way. And the justification for the move is that only principle (3) (combined with the assumption that conceiving is affecting) gives support to the asymmetrical judgments that we find plausible.5
So far, so good. But I shall now argue that the person‐affecting principle needs to be more drastically revised. This may drain it of its plausibility.
Consider the following case, which involves two women. The first is one month pregnant, and is told by her doctor that, unless she takes a simple treatment, the child she is carrying will develop a certain disability. We suppose again that life with this disability would probably be worth living, but less so than a normal life. It would obviously be wrong for the mother not to take the treatment, for this will disable her child. And the person‐affecting principle tells us that this would be wrong. (Note that we need not assume that a one‐month‐old foetus is a person, for there will be a person whom the woman has affected for the worse.)
We next suppose that there is a second woman, who is about to stop taking contraceptive pills so that she can have another child. She is told that she has a temporary condition such that any child she conceives now will have just the same disability; but that if she waits three months she will then conceive a normal child. It seems clear that it would be wrong for this second woman, by not waiting, to deliberately have a disabled rather than a normal child. And it seems (at least to me) clear that this would be just as wrong as it would be for the first woman to deliberately disable her child.
But if the second woman does deliberately have a disabled child, has she harmed him – affected him for the worse? We must first ask: “Could he truly claim, when he grows up, ‘If my mother had waited, I would have been born three months later, as a normal child’?” The answer is, “No.” If his mother had waited, he would not have been born at all; she would have had a different child. When I claim this, I need not assume that the time of one’s conception, or the particular cells from which one grew, are essential to one’s identity. Perhaps we can suppose that I might have been conceived a year later, if we are supposing that my parents had no child when they in fact had me, but a year later had a child who was exactly or very much like me. But in our case the child the woman would have if she waits would be as unlike the child she would have now as any two of her actual children would be likely to be. Given this, we cannot claim that they would have been the same child. (To argue this in another way. Suppose that I am in fact my mother’s first child and eldest son. And suppose that things had gone like this: she had no child when I was in fact born, then had a girl, then a boy. Can I claim that I, her first child, would have been that girl? Why not claim that I, her eldest son, would have been that boy? Both claims are equally good, and so, since they cannot both be true, equally bad. So, if she had waited before having children, I would not have been born at all.)6
The second woman’s disabled child is, then, not worse off than he would otherwise have been, for he wouldn’t otherwise have been. Might we still claim that in deliberately conceiving a disabled child, the woman harms this child? We might perhaps claim this if the child’s life would be not worth living – would be worse than nothing; but we have assumed that it would be worth living. And in this case being disabled is the only way in which this child can receive life. So the case is like that in which a doctor removes a person’s limb to save his life. It would not be true, at least in a morally relevant sense, that the doctor harmed this person, or affected him for the worse. We seem bound to say the same about my second woman.
I conclude, then, that if the second woman deliberately conceives a disabled rather than a normal child, she would not be harming this child. The first woman, if she deliberately neglects the treatment, would be harming her child. Notice next that in every other way the two acts are exactly similar. The side‐effects on other people should be much the same. These side‐effects would provide some person‐affecting grounds for the claim that the second woman’s act would be wrong. But it is obvious that if we judge the two acts on person‐affecting principles, the first woman’s act must be considerably more wrong. In her case, there are not just side‐effects – her child is seriously harmed. The second woman’s child is not harmed. Since this is the only difference between the two acts, the case provides a test for person‐affecting principles. The impersonal principle tells us to reduce misery and increase happiness, whether or not people are affected for better or worse. If there is any plausibility in the restriction to acts which affect people, it must be worse to harm someone than to cause equivalent unhappiness in a way which harms no‐one. The second woman’s act must, in other words, be less wrong than the first’s. It we think that it is not less wrong, we cannot accept the restriction to acts which affect people.
The acts which I have described are of course unusual. But this does not make them a worse test for the person‐affecting restriction. On the contrary, they are unusual because they are designed as a test. The two women’s acts are designed to be as similar as they could be, except in one respect. Each woman deliberately brings it about that she has a disabled rather than a normal child. The only difference is that in one case the disabled and the normal child are the same child, while in the other they are not. This is precisely the difference which, on the person‐affecting principle, matters. If we think that the two acts would be just as wrong, we cannot believe that it does matter.
Some of you may think that the person‐affecting principle survives this test. You may think: “Since the second woman doesn’t harm her child, what she does is less wrong.” But there are other cases where such implications seem harder to accept. Take genetic counseling. We could not advise the dominant carriers of diseases to accept genetic counseling for the sake of their children, for if they reject this counseling, and marry other dominant carriers, it will not
