Problems of utilitarianism II: Consequentialism versus absolutism

But this basic problem of defining and measuring utilities is by no means the only challenge that faces the utilitarian. Let us put to one side the question of how to measure utility and simply suppose that it can be solved.

But how are we to judge whether what seems immoral really is immoral? In developing our moral views in a philosophical way, we take into account our metaethical views. But, as we have seen, metaethics does not, by itself, settle substantial moral questions. When we consider a substantive moral theory such as utilitarianism, we have to require not only that it be consistent with our metaethics, but also that it be consistent with our existing basic moral beliefs. No amount of philosophical argument is likely to persuade us to give up our deepest moral beliefs. We might find ourselves changing some of our moral views as a result of reflection, not merely in order to make them logically consistent, but because, for example, we see that certain principles that we have held in the past would lead, once universalized, to horrible consequences. But, in the end, there will be a kind of movement back and forth between the moral beliefs we start with, and moral theory. I shall discuss this process in a little more detail in 6.12. For the moment, let us just proceed in this way.

Consider the simple and familiar moral principle that one should not lie. Utilitarians, because they are consequentialists, are not likely to accept this principle. They will say that sometimes telling a lie may have better consequences for human utility than telling the truth.

An absolutist will say, on the other hand, that lying is always wrong. It does not follow that the absolutist will never tell a lie. For, an absolutist can say, though lying is always wrong, some things are a good deal worse than lying. Thus, suppose Theresa lives in a total­itarian state. She is helping to hide an opponent of the regime, who risks being tortured if he is caught, though all he has done is to speak out against torture and oppression. Suppose a police officer comes to the door asking whether she is hiding that person. Even if Theresa is an absolutist about lying, she does not have to tell the truth. To do this would not only be a betrayal of trust but lead to the suffering of a noble individual.

Theresa will say not that lying, in these circumstances, is right but that it is the lesser of two evils. Fate has dealt her a choice between principles. Her view, as a moral agent, is that lying in this situation is obviously the lesser evil.

But what is the content of Theresa's judgment that it is wrong to lie even in this case? She would certainly agree that in this case, she ought, all things considered, to lie. The difference between Theresa and the consequentialist here is not in the actions they carry out but in the attitude they take to them. Theresa will regret having to lie. The utilitarian will not. In this sort of case, many people will agree with the utilitarian that Theresa has the wrong response. She simply has nothing to regret, they will say. If Theresa agrees with this, we shall have no answer to the question what the content is of her judg­ment that the act was wrong.

It is because many people believe that it is simply right to lie in such cases that they find the consequentialist position very plausible in the case of lying.

a)     the practice of truth telling contributes to human happiness in most cases—which is why we all begin by thinking of lying as wrong; but also,

b)     individual lies are justified if telling the truth would lead to more harm than good.

Indeed, a consequentialist can argue that feelings of regret, such as Theresa may feel, can themselves be given a consequentialist justi­fication. Hare says:

Nobody who actually uses moral language in his practical life will be content with a mere dismissal of the paradox that we can feel guilty for doing what we think we ought to do.

And he suggests a number of reasons why a consequentialist should actually want us to have such reactions. First of all, he takes it for granted, surely correctly, that such feelings help us keep to our prin­ciples. Without them, many of us would be constantly slipping into doing what we believed was wrong. So the feelings are essential. Now, we could try to develop a sophisticated set of feelings that went exactly with our moral beliefs. But to do that, we should have to attach the feelings, so to speak, to very complex principles. Once we start on this process with our principles, Hare argues, we will end up with moral principles of tremendous complexity. We start with a principle that says, “One ought never to do an act that is G” (where G is, say, “a lie”); then we consider Theresa's problem. So, as Hare says, we modify our principle. Instead of reading “One ought never to do an act that is G,” it now reads “One ought never to do an act that is G, unless it is necessary to avoid an act that is F.”

Here F might be “the betrayal of a noble individual.” Reflection on other cases will soon have us adding that even if it is necessary to do G to avoid F, we should not do so in circumstances H unless—as another case might make us think—it is also I. And so on.

But once we get to principles of this complexity, it is hard to get our feelings attached to them in the right way.

Nevertheless, there are cases where most people think that con- sequentialism about lying is simply wrong. Suppose, for example, Ben is dying of a rare disease. Someday soon he will just drop dead, and nothing he or anyone else does can change that. Jane, a utilitar­ian doctor, might well feel that she should just not tell him, because it will only make him unhappy. Yet many of us think that, in these circumstances, Ben would have a right to know that he was going to die.

This sort of case is a more challenging problem for the utilitarian because it suggests not only that our moral feelings do not fit utili­tarian principles but also that our moral judgments do not fit them either. We can give a utilitarian explanation of why we might want to have nonutilitarian feelings, but it would be just inconsistent to give a utilitarian explanation of why utilitarian principles were wrong.

The intuition that we cannot accept consequentialism as a moral theory is even stronger in cases where more is obviously at stake: in cases, for example, which involve killing people against their will. Jonathan Glover, a British philosopher, has suggested just such a case. He asks us to consider a man in prison.

His life in prison is not a happy one, and I have every reason to think that over the years it will get worse. In my view, he will most of the time have a quality of life some way below the point at which life is worth living.

This case constitutes an objection to utilitarianism, indeed to most forms of consequentialism. It looks as though the consequentialist will here have to agree that I should kill the prisoner, for the conse­quences of doing so will be better for him. But Glover suggests that the consequentialist might argue that drawing this conclusion ignores two important considerations.

First, such a killing may have many side effects that have so far not been mentioned. Thus, for example, the man's family might regret his death, even if they knew that his life would have been unpleasant. And, for another example, if it came to be known what you had done, this would have a terrible effect on the morale of other prisoners in the prison. They might well fear that you would make such a judgment about them. This is especially likely to worry them because of the second consideration that the consequentialist may say we have ignored: namely, that it is not, in fact, very easy to predict what the future course of a person's mental states will be. As Glover says: “If a man wants to go on living, although this does not force me to accept that his life is worth living, I would have to be very optimistic about my own judgment to be sure that he is wrong.”

But drawing attention to these considerations does not really allow us to accept the utilitarian's claims. For we can simply con­struct a case where these considerations do not apply. Suppose we were sure that no one would find out, sure that the prisoner had no family, and sure about his current and future mental states.

The view that it would be quite wrong to kill the prisoner will be defended by any philosopher who believes that it is a central moral principle that we should respect a person's autonomy. Respecting people's autonomy means placing a very great deal of weight in our

decisions about them on what they themselves judge to be impor­tant. For “autonomy” means, in essence, the capacity for self-rule. (There's the word “nomos” again, from the Greek for “law”: an autonomous person is bound by his or her own laws.) Kant expressed this idea when he said that it followed from his universal­izability principle that we should never treat people merely as means but always as ends in themselves. To kill the prisoner is to regard his utility as more important than his wishes, and thus, in a sense, to treat him as a means to the end of maximizing utility.

If Kant was right, and we must respect people's autonomy, then consequentialism—the claim that we should always judge actions simply by their consequences—must be wrong. Even if we think that it is generally a good thing to maximize utility, the application of this principle must be subject to constraints. In particular, we may maximize people's utility only in contexts where this is consistent with respecting their autonomy. In the end, consequentialists can­not explain why many of us regard respect for other people's auton­omy as important.

5.11