The Excessive Demands of Impartiality

Thomas Nagel

  • Moral Thinking: Its Levels, Method and Point by R.M. Hare
    Oxford, 250 pp, £11.00, December 1982, ISBN 0 19 824659 5

According to Professor Hare, most contemporary moral philosophers are benighted. They cannot get through their thick skulls the clear principles of moral reasoning which he has set out and developed in two previous book-length studies of ethical theory, The Language of Morals and Freedom and Reason, and which he spells out again and develops further in this one. Like a teacher of dim-witted children, he can’t always conceal his impatience at having to repeat himself, but the importance of the issues keeps him at it.

He finds the subject infested with ‘intuitionists’. These are philosophers who limit their moral thinking to one level – that of the rough, general, prima facie principles that they have been brought up to follow in everyday life, when they must act without having too much time to think and without perfect knowledge of the facts. Intuitionists take these principles, and the moral feelings associated with them, as the basis of moral thought, when actually they are derived from something else and can be justified or criticised only by moving to a different level, where moral intuitions are not relied upon.

This is the level of what Hare calls ‘critical thinking’, and its rules are determined entirely by the logic of the moral concepts – a logic that is discovered by investigating ordinary linguistic usage, and not by the exercise of any form of moral intuition or judgment. It is only by getting outside morality in this way, he believes, that we can hope to provide morality with a firm foundation instead of going around in circles. Once we do so, we shall discover that the unique moral position that it is possible to justify on the basis of the logic of the moral concepts is utilitarianism. It also turns out that utilitarianism, properly interpreted, is what underlies those common moral intuitions which intuitionists erroneously take to be basic, and which they use to try to refute utilitarianism.

Hare has come to this view gradually. He did not formerly believe that one could derive a single substantive moral position from the logic of the moral concepts – and for good reason. What is so implausible about this claim is its implication that all fundamental moral disagreements are in a sense illusory. Provided the parties to a disagreement are really making moral claims at all, they can reach different conclusions only because they disagree about the non-moral facts, or because at least one of them has made a logical error in applying the moral concepts. There are no disagreements that are moral all the way down. A libertarian who believes in the inviolability of individual property rights and a socialist who believes in promoting equality by progressive taxation will discover, if they consider carefully what they both mean by the words ‘right’ and ‘wrong’, ‘ought’ and ‘must’, that they are both committed to a single method of deciding which of their positions is correct – an empirical method that depends on which social system will most effectively satisfy the preferences of all the people affected by it.

How does Hare manage to extract this large moral rabbit from what looks at first like a rather small and empty linguistic hat? His theory about the logic of the moral concepts – universal prescriptivism – seems innocent enough. It says that a judgment about what ought to be done on a particular occasion commits one to prescribing that the same thing be done in any situation which is similar in all its universal characteristics – in particular, any situation in which the individual persons involved are different, or in which they occupy different roles.

This looks like something that the libertarian and the socialist could agree on. The libertarian who believes that he should not be taxed to provide subsistence for indigent single mothers also believes that if he were an indigent mother or one of her children, it would be wrong to tax others to feed him. And the socialist who favours taxing the rich favours it whoever the rich happen to be. The two of them can use the same moral vocabulary to disagree about what ought to be done because they differ over the importance to be assigned, universally, to liberty, equality, property and the satisfaction of individual preferences.

But Hare has come to believe that universal prescriptivism is much less hospitable to different moral convictions than this, because he interprets it in a way that permits each person to prescribe universally only one thing, given the facts of a particular situation, and it is the same thing that anyone else must prescribe. The only rational basis for a universal prescription, he believes, is to put yourself in the position of all the people affected by a given act and ask whether you would want it to be done or not if you had their preferences. The preferences which you recognise you would have, were you in a given situation, lead immediately and unavoidably to equally strong preferences which you now have, regarding what should happen if you were in that situation. You then add up all these hypothetical preferences, pro and con, allowing for their different weights; and the choice which has the greater weight of preferences on its side determines whether the act ought to be done or not, by determining whether you can prescribe universally that it should be done.

All this is supposed to be contained in the logic of the moral words used by all speakers of English, and other languages translatable into English. But it clearly isn’t. Not only would many people regard criticism of their moral views by this method as invalid, but even most of those who agree with Hare’s moral position would not regard those who reach moral judgments by a different method as misusing the language. Instead, they would hold that if someone believes, say, that each person may favour himself over others whether or not that will work out to everyone’s advantage in the long run, then he is in substantive moral disagreement with utilitarianism at a fundamental level. This is a disagreement not over logic or the facts, but over the method of determining what is to be universally prescribed.

The unsurprising truth is that Hare has smuggled a moral intuition into his analysis of moral language, and that is what gives him his method. It is a perfectly respectable intuition, the same one that Sidgwick saw to be the basis of utilitarianism: ‘I ought not to prefer my own lesser good to the greater good of another.’ But there are those who do not share it. The move from what one would want to happen to oneself in each possible circumstance to what one wants to happen universally is not a logical one, for there is more than one way of combining such preferences: simple maximisation of expected satisfaction, counting all the possibilities as equally probable, may not be the right choice even for an individual. But more important, the move from what one wants to happen universally to what one is prepared to prescribe should happen universally is not a logical one. An alternative method for arriving at universal prescriptions is to ask what any person would have reason to do in the given situation. One answer to this is the utilitarian answer, but there are others, including those which permit a person to give extra weight to his own concerns, and those which distinguish only some of the interests of others as those he should be concerned with.

Hare cannot get his moral theory out of the logic of the moral concepts, and the illusion that he is doing so condemns the work to one kind of superficiality, for it prevents him from seeing his most important assumptions and therefore from defending them. This is a defect in the book even if his substantive moral position is correct.

But his defence of utilitarianism is partly independent of these claims about moral language. He presents a version of the view that emphasises the distinction between the ordinary rules that must be followed in the confusing and often emotionally fraught circumstances of everyday life, and the method of critical thinking that can be used in a cool moment, without constraints of time and information, to devise, evaluate or criticise those prima facie rules and settle conflicts between them. Only an omniscient being whose mind worked like lightning and who suffered from no emotional weaknesses could use the full method of utilitarian critical thinking for every action he had to perform. Real humans have to approximate conformity to utilitarianism by following rules that are simple, clear and learnable, and that do not permit much scope for self-deception. Hare also thinks they shouldn’t make excessively heavy demands on our motives – a curious point to which I shall return.

Hare says that many of the most popular arguments against utilitarianism – those which claim it would require some obviously wicked action, like murder or torture or the punishment of an innocent scapegoat, in sufficiently bizarre circumstances – depend on ignoring this pervasive distinction between the critical and intuitive levels of moral thought. I think he is right. A reasonable form of utilitarianism will require, not only the creation of stable institutions which promote the general welfare, but also the inculcation of consistent virtues, habits of conduct, instinctive reactions and feelings which will lead people to act on the whole for the best under ordinary circumstances of ignorance and temptation. We do not know the total outcomes of many of our acts in sufficient detail to permit utilitarian calculation to proceed act by act. It is much better that there should be an overwhelming revulsion against torture for any reason than that people should feel prepared to consider the possibility case by case. The utilitarian justification for training people to have such moral inhibitions depends on the real facts of human life, so it is no argument against utilitarianism that sound intuitions would prevent someone from doing the best thing on the basis of perfect knowledge in a bizarre situation of a kind that rarely or never actually occurs. I think it is very unlikely that Hare’s form of utilitarianism would in practice justify obvious moral atrocities (even though moral atrocities have often been defended by spurious utilitarian reasoning). The difficult theoretical question is whether Hare’s two-level account gives an adequate explanation of the rights, duties and prohibitions of common morality, or whether they have independent sources, not wholly derivable from the single intuition of moral impartiality that yields utilitarianism.

I have two doubts: first, whether Hare’s view gives the right explanation of common moral intuitions where it does endorse them; second, whether utilitarianism really coincides with common morality as much as he believes it does. In considering such requirements as veracity, loyalty, or fidelity to commitments we have made to others – requirements which demand that we ignore apparent general utility in particular cases – Hare asks us to believe that they have no sources of their own, no reasons coming from the particular circumstances and relations to which they apply. In other words, if, without regard to utility, I refrain from lying to someone or betraying someone who has trusted me, the ultimate justification lies in the general utility of a fairly inflexible disposition not to do such things. But even if such a disposition does have general utility, something else is going on. It is very implausible to claim that intuitive repugnance at personal betrayal is just an artefact of upbringing on principles warranted by utility. Hare’s view requires us to regard the sense of immediacy that these claims have as a kind of illusion, and all moral conflicts that seem to arise from them as soluble by rising to the level of utilitarian critical thinking. But without the support of his axiom that the sole method of moral justification is the appeal to pure impartiality, the diagnosis of illusion is unconvincing, and the operation of other, more personal reasons of a basic kind in ordinary morality seems likely.

In fact, personal reasons seem to cause ordinary morality to deviate substantially from even a two-level utilitarianism. This is my second doubt about Hare’s position. He softens the demanding character of utilitarian impartiality by an illegitimate device which makes it easier to explain the other aspects of common morality in terms of his theory. The device is to tailor the prima facie principles that people are asked to follow in everyday life to the natural partiality of ordinary people, toward themselves and those near them. A utilitarian will praise individuals of supererogatory virtue for doing more, but he will see no point in demanding of others, or of himself, a level of beneficence which would simply be rejected as too onerous. As Hare puts it, ‘If I were very saintly like Albert Schweitzer or Mother Teresa, I might have much more exacting principles than in fact I have.’

I would put it another way: if Hare were very saintly he might be more of a utilitarian. As it is, he is committed to prima facie principles for the conduct of life which often permit him to do what is wrong by utilitarian standards, simply because, like the rest of us non-saints, he is unwilling to swallow the more stringently impartial principles whose acceptance-utility, if they were accepted, would be higher. If he doesn’t think he’s doing anything wrong in behaving less selflessly than Albert Schweitzer, then he is giving his natural partiality a veto over the excessive demands of impartiality, and this is to fall away from a full utilitarianism – even allowing for differences of level.

I find Hare’s actual position more plausible than unadulterated utilitarianism, but to begin to settle the issue requires work in the foundations of ethics. These are not to be found in the analysis of moral concepts alone, though I agree with Hare that they are not to be found in the systematisation of our moral sentiments either. My own view is that the methods of moral thinking now available to us are fairly primitive – that while we have a grasp on what the basic kinds of moral reason are, including those on which utilitarianism is exclusively based, we don’t have a sound method for combining them to give a result when they conflict, as they do constantly in real life, when a variety of disparate claims can bear on a decision. I don’t know how to make the choice among different available interpretations of moral equality, nor do I think that ethical theory to date has exhausted the possibilities. Hare offers a single method, and this book presents it with the lucidity, economy and bite that make all his work so readable. But though it is worth careful study, it tends to confirm the view that in ethics theoretical simplicity is not a virtue.