Charity or char: is the choice yours?

Joseph Raz, professor of the philosophy of law at Oxford University, has always written for two audiences: legal theorists, and philosophers. This is one for the philosophers. It explores the relationships between will, reason and value, and it will be of enormous interest to those working in ethics and those in the occupied territories within the philosophy of mind, known as the philosophy of action. It is an extended reflection on a small number of tightly connected themes, written from a standpoint of insight, intelligence and complexity of thought, and it deserves to be widely read and discussed. Although, as a collection of papers, it is resistant to simple summary, I will try to bring out the central themes.

The work begins by exploring the idea that sometimes we try to excuse our behaviour on the grounds, say, that "I was not myself that afternoon". The question is therefore, under what circumstances am I "myself" in this sense of owning or taking full responsibility for my actions? Rejecting solutions based on the idea of "identification", Raz suggests that the key idea is responsiveness to reason. We are ourselves when we see ourselves as rational creatures. Self-understanding and self-control, are, on this account, related in an illuminating way. Yet what is it to be rational; to be guided by reasons?

Rational decision theory has an answer. A rational person is a person who maximises expected utility, calculating probabilities of preference satisfaction from the array of available options, and who then chooses the path that offers the best prospects. Whatever the other merits of this theory, Raz believes, surely quite rightly, that it is untrue to the nature of reasoning. This is not what we do, and it is not even a sensible aspiration. Central to Raz's case is the idea that reasons for action can often be, as he puts it, incommensurate. Imagine I have received a modest salary increase and have to decide whether to raise my charitable donations, or to employ a cleaner. The reasons in favour of both are clear, but are also so different that it may seem impossible to rank one above the other. It does not seem right to say that the reasons in favour of either are equal. In such a case, all reason can do is to tell me which options are "eligible", and then I must simply make a decision. It follows that two sets of circumstances could be identical in all relevant respects, yet in the first case, I rationally choose one course of action, and in the second, I rationally choose the other. This, of course, would be regarded as merely inconsistent by the terms of standard formal decision theory, but for Raz, it is nothing out of the ordinary. If Raz is right, such formalisation would have to be considered simply as a construction made for particular purposes - perhaps predictive social theory - rather than as descriptively or normatively adequate.

It is part of Raz's general perspective that no sharp boundary between moral and non-moral reasons can be found, and his analysis pays particular dividends in his discussion of "moral particularism". Particularists attempt to undermine the pretensions of those moral theorists who hope to provide fundamental principles of morality. At its most ambitious, the goal is to provide a single principle that guides us in all possible circumstances. Classical utilitarianism, telling us always to maximise the balance of pleasure over pain, is a prime example. But the essence of moral theory is to assume that human morality can be encoded into a definite, and ideally rather small number of finitely specifiable principles. Moral particularism denies this, stating that there is no such set of principles. The tendency of Raz's thought would make one suspect that he is sympathetic to such an anti-systematic view. Although he does not go as far as calling himself a particularist (or indeed anything else), he does add considerable ammunition to their case.

Raz's analysis centres on an example made famous by Peter Winch; the behaviour of Vere in Melville's Billy Budd . Vere has to choose between the military code, which tells him to order Billy Budd's execution, and his private conscience, given that he believes Budd to be innocent. Vere sees this as a moral conflict and he decides that his duty is to follow the military code. Winch comments that although Vere did the right thing for him, Winch himself would have found it morally impossible to do the same thing. Winch claims that even in exactly the same circumstances, it would have been wrong for Winch to act in the way that was right for Vere. This seems both plausible and puzzling, and goes against the grain of the generalising tendency that morality is so often assumed to have. If we agree with Winch's assessment, perhaps we would look for some sort of special factor that differs between the cases and thus explains our differing judgements. Yet for Raz the case is straightforward. Moral reasons, like other reasons, may be incommensurate, as, no doubt, is thought to be the case here. Consequently, morality can tell us only which options are morally eligible. Beyond this, one simply has to make a decision. Vere made one fully understandable decision, while Winch would have found himself forced to make another. On Raz's understanding of reasoning, there is nothing in need of special explanation here. Only if we are in the grip of the idea that there must always be a morally or rationally best answer will we find this hard to understand.

How convincing is this approach to reasoning? Those who like the idea that philosophy should consist of systems, laws, principles, definitions, firm distinctions and generalisations will be frustrated by the lack of such things here. But as Raz remarks in another context: he cannot be blamed for this, for it is just how things are (at least according to him). Those prepared to take the trouble to engage with the arguments cannot fail to be enriched by the experience.

Jonathan Wolff is professor of philosophy, University College London.