Mind, Value, and Reality

John McDowell

• Publisher: Harvard University Press
• Year: 2001
• ISBN:
  9780674265936
• Topic: Filosofía
• Subtopic: Mente y cuerpo en filosofía

This volume collects some of John McDowell's influential papers, written at various times over the last two decades. One group of essays deals mainly with issues in the interpretation of the ethical writings of Aristotle and Plato. A second group of papers contains more direct treatments of questions in moral philosophy that arise naturally out of reflection on the Greek tradition. Some of the essays in the second group exploit Wittgensteinian ideas about reason in action, and they open into the third group of papers, which contains readings of central elements in Wittgenstein's difficult later work. A fourth group deals with issues in the philosophy of mind and with questions about personal identity and the special character of first-personal thought and speech.

---

PART II

---

REASON, VALUE, AND REALITY

---

ESSAY 4

---

Are Moral Requirements Hypothetical Imperatives?

1. In “Morality as a System of Hypothetical Imperatives”, Philippa Foot argues against the Kantian doctrine, and prevailing orthodoxy, that the requirements of morality are categorical imperatives. She notes that there is a distinction between a use of “should” in which a “should” statement needs withdrawing if the action in question cannot be shown to be ancillary to the agent’s desires or interests, and one in which that is not so; and that moral uses of “should” are of the latter sort. She argues, however, that this latter use of “should” does not mark a categorical imperative in the sense intended in the orthodox doctrine; for it is found equally in expressions of the requirements of etiquette. Defenders of the orthodoxy, she assumes, would deny that the requirements of etiquette are categorical imperatives, and would ground the denial on the thesis that it is possible, without irrationality, to question whether one has reason to conform to them. On this assumption, the orthodoxy amounts to the claim that such questioning is not possible with morality. But Mrs Foot insists that the claim is false: there is no irrationality in questioning whether one has reason to act as morality is alleged to require. On this construal of the orthodoxy, then, a categorical imperative is something that must, on pain of irrationality, be recognized as a reason for acting; and Mrs Foot’s thesis is that moral requirements are not categorical imperatives in that sense. She concludes that the requirements of morality exert a rational influence on the will only hypothetically; their influence is conditional on the presence of desires that are lacked by those who question whether they have reason to conform.

I want to agree that one need not manifest irrationality in failing to see that one has reason to act as morality requires, but I want to query whether it follows that moral requirements are only hypothetical imperatives.

2. The terminology calls for some preliminary comment. As Mrs Foot notes, Kant’s concern was not with imperatives on a strict grammatical construal of the classification. She concentrates on judgements expressible with the words “should” or “ought”; but I prefer to shift attention away from explicitly prescriptive or normative language altogether.

It seems plausible that if one accepts that one should do something, one accepts that one has a reason to do it. But the reason is not expressed by the “should” statement itself. The reason must involve some appropriate specific consideration that could in principle be cited in support of the “should” statement. Thus, if one does something because one thinks one should, then unless the thought that one should is merely accepted on authority, a more illuminating account of one’s reason will be available, citing the appropriate specific consideration that one takes to justify the view that one should act in that way. A formulation of the specific consideration will at least include a mention of what one takes to be relevant features of the circumstances in which the action is to be performed.

Now the fundamental difference that I think Kant was aiming at is one between different ways in which conceptions of circumstances influence the will; that is, between different ways in which they function in the explanation of behaviour in terms of the agent’s reasons. To a virtuous person, certain actions are presented as practically necessary—as Kant might have put it—by his view of certain situations in which he finds himself. The question is whether his conceptions of the relevant facts weigh with him only conditionally on his possession of a desire.

If we think of the requirements of morality as imposed by the circumstances of action, as they are viewed by agents, rather than by the associated “should” thoughts, we make it possible to defend the thesis that virtuous actions are dictated by non-hypothetical imperatives, without committing ourselves to the insane thesis that simply to say “You should . . .” to someone is enough to give him a reason for acting; as if, when he protested “But why should I?”, it was sufficient to reply “You just should, that’s all”.

3. When we explain an action in terms of the agent’s reasons, we credit him with psychological states given which we can see how doing what he did, or attempted, would have appeared to him in some favourable light. A full specification of a reason must make clear how the reason was capable of motivating; it must contain enough to reveal the favourable light in which the agent saw his projected action. We tend to assume that this is effected, quite generally, by the inclusion of a desire. (Of course a reason that includes a desire can be specified elliptically, when the desire is obvious enough not to need mentioning; as when we explain someone’s taking an umbrella in terms of his belief that it is likely to rain.) However, it seems to be false that the motivating power of all reasons derives from their including desires.

Suppose, for instance, that we explain a person’s performance of a certain action by crediting him with awareness of some fact that makes it likely (in his view) that acting in that way will be conducive to his interest. Adverting to his view of the facts may suffice, on its own, to show us the favourable light in which his action appeared to him. No doubt we credit him with an appropriate desire, perhaps for his own future happiness. But the commitment to ascribe such a desire is simply consequential on our taking him to act as he does for the reason we cite; the desire does not function as an independent extra component in a full specification of his reason, hitherto omitted by an understandable ellipsis of the obvious, but strictly necessary in order to show how it is that the reason can motivate him. Properly understood, his belief does that on its own. Thomas Nagel (in The Possibility of Altruism, pp. 29–30) puts the point like this:

That I have the appropriate desire simply follows from the fact that these considerations motivate me; if the likelihood that an act will promote my future happiness motivates me to perform it now, then it is appropriate to ascribe to me a desire for my own future happiness. But nothing follows about the role of the desire as a condition contributing to the motivational efficacy of those considerations.

This passage is quoted in part, and its thesis endorsed, by Mrs Foot in “Reasons for Action and Desires”.

Why should the reasons that move people to virtuous behaviour not be similar to the reasons that move them to prudent behaviour? To explain an action we regard as virtuous, we typically formulate a more or less complex characterization of the action’s circumstances as we take the agent to have conceived them. Why should it not be the case, here too, that the agent’s conception of the situation, properly understood, suffices to show us the favourable light in which his action appeared to him? If we credit him with a suitable desire, then, as before, that need be no more than a consequence of the fact that we take his conception of the circumstances to have been his reason for acting as he did; the desire need not function as an independent component in the explanation, needed in order to account for the capacity of the cited reason to influence the agent’s will.

4. There may seem to be a difficulty: might not another person have exactly the same conception of the circumstances, but see no reason to act as the virtuous person does? If so, adverting to that conception of the situation cannot, after all, suffice to show us the favourable light in which the virtuous person saw his action. Our specification of his reason must, after all, have been elliptical; a full specification would need to add an extra psychological state to account for the action’s attractiveness to him in particular—namely, surely, a desire.

We can evade this argument by denying its premise: that is, by taking a special view of the virtuous person’s conception of the circumstances, according to which it cannot be shared by someone who sees no reason to act as the virtuous person does.

This may seem problematic. But if one concedes that a conception of the facts can constitute the whole of a reason for prudent behaviour, one is not at liberty to object to the very idea that a view of how things are might not need supplementing with a desire in order to reveal the favourable light in which someone saw some action; and a view with that property surely cannot be shared by someone who sees no reason to act in the way in question. If we allow this for prudence, why should we not allow it for morality too?

Suppose someone was incapable of seeing how a fact about the likely effect of an action on his own future could, on its own, constitute a reason for the action. On some suitable occasion, he might be unmoved by such a fact. It would not be wrong to say that an ordinarily prudent person, in parallel circumstances, would differ from him in having a certain desire. But according to the concession, the desire is not a further component, over and above the prudent person’s conception of the likely effects of his action on his own future, in the explanation of his prudent behaviour. It is not that the two people share a certain neutral conception of the facts, but differ in that one, but not the other, has an independent desire as well, which combines with that neutral conception of the facts to cast a favourable light on his acting in a certain way. The desire is ascribable to the prudent person simply in recognition of the fact that his conception of the likely effects of his action on his own future by itself casts a favourable light on his acting as he does. So the admitted difference in respect of desire should be explicable, like the difference in respect of action, in terms of a more fundamental difference in respect of how they conceive the facts.

It is not clear that we really can make sense of the idea of someone who is otherwise rational but cannot see how facts about his future can, by themselves, constitute reasons for him to act in various ways. But to the extent to which the idea does make sense, it seems to be on just the lines we should expect: we picture him as someone with an idiosyncratic view of what it is for a fact to concern his own future. Perhaps he thinks of the person involved in such a fact as some future person, connected with the one who is currently deliberating by links of continuity and resemblance that are too tenuous, in his view, for it to be anything but arbitrary for the current deliberator to pay special attention to that future person’s welfare. What is special about a prudent person is a different understanding of what it is for a fact to concern his own future. He sees things otherwise in the relevant area; and we comprehend his prudent behaviour by comprehending the relevant fragment of his world view, not by appealing to the desire that is admittedly ascribable to him. That is to be understood, no less than the behaviour is, in terms of the world view.

Why should it not be similar with explanations of virtuous behaviour in terms of the virtuous person’s conceptions of situations in which he acts?

5. So far I have responded only ad hominem to qualms about the idea that a conception of how things are might constitute, on its own, a reason for virtuous action. That is how it was conceded to be with prudential reasons, and there is no obvious argument that the possibility, once granted, should be restricted to prudential considerations. But presumably someone with sufficiently strong doubts about the case of morality will be encouraged to doubt the whole idea, and suppose that it cannot be so even with prudential reasons; he will not be impressed by the thought that, if granted there, the possibility cannot be dismissed out of hand for the case of morality.

I suppose the general doubt is on these lines. A view of how things are is a state or disposition of one’s cognitive equipment. But the psychological states we are considering are to suffice, on their own, to show how certain actions appeared in a favourable light. That requires that their possession entails a disposition of the possessor’s will. And will and belief—the appetitive and the cognitive—are distinct existences; so a state that presents itself as cognitive but entails an appetitive state must be, after all, only impurely cognitive, and contain the appetitive state as a part. If such a state strikes its possessor as cognitive, that is because he is projecting his states of will on to the world (a case of the mind’s propensity to spread itself upon objects). The appetitive state should be capable in principle of being analysed out, leaving a neutrally cognitive residue. Thus where it appears that a conception of how things are exhausts an agent’s reason for acting in a certain way, an analysed and less misleading formulation of the reason will be bipartite: it will specify, first, a neutral conception of the facts, available equally to someone who sees no reason to act in the way in question, and second, a desire, which combines with that conception of the facts to make the action attractive to its possessor.

This paper is primarily addressed to those who are vulnerable to the ad hominem argument. In their view, since the line of thought I have just sketched falsifies the workings of prudential explanations of behaviour, it simply cannot be generally right. In the rest of this section I shall make some remarks, not ad hominem, about the general issue; but a proper discussion is impossible here.

There is room for scepticism about whether it is acceptable to discount the appearances in the way the objection urges. Explanation of behaviour by reasons purports to show the favourable light in which an agent saw his action. If it strikes an agent that his reason for acting as he does consists entirely in his conception of the circumstances in which he acts, then an explanation that insists on analysing that seemingly cognitive state into a less problematically cognitive state combined with a separate desire, while it will show the action as attractive from the standpoint of the psychological states it cites, is not obviously guaranteed to get the favourable light right. If one accepts an explanation of the analysing sort, one will not be baffled by inability to find any point one can take the agent to have seen in behaving as he did; but what leaves one unpuzzled is not thereby shown to be a correct explanation.

The analysis will nevertheless seem compulsory, if the objection seems irresistible. If the world is, in itself, motivationally inert, and is also the proper province of cognitive equipment, it is inescapable that a strictly cognitive state—a conception of how things are, properly so called—cannot constitute the whole of a reason for acting. But the idea of the world as motivationally inert is not an independent hard datum. It is simply the metaphysical counterpart of the thesis that states of will and cognitive states are distinct existences, which is exactly what is in question.

If a conception of a set of circumstances can suffice on its own to explain an action, then the world view it exemplifies is certainly not the kind of thing that could be established by the methods of the natural sciences. But the notion of the world, or how things are, that is appropriate in this context is a metaphysical notion, not a scientific one: world views richer than that of science are not scientific, but not on that account unscientific (a term of opprobrium for answers other than those of science to science’s questions). To query their status as world views on the ground of their not being scientific is to be motivated not by science but by scientism.

6. It is not to be denied that behaviour that is in fact virtuous can in some cases be found unsurprisi...