1 Saylor, Thomas Reese ‘A New Legislative Mandate for American Food Aid,’ in Brown, Peter G. and Shue, Henry eds., Food Policy (New York: Free Press 1977), 202Google Scholar

2 Hayter, Theresa The Creation of World Poverty (London: Pluto Press 1981), 9Google Scholar. The Brandt Commission admits as much, saying, ‘It is a mark of the uneasy relations between North and South that even to speak of mutual interest can cause suspicion’ (Brandt, Willy North-South: A Programme for Survival [London: Pan 1980], 64Google Scholar).

3 Notice that this is different from saying that a person’s true motives might be empirically undecidable by external observers, as argued by critics of revealed preference techniques in economics, such as Sen, Amartya ‘Behaviour and the Concept of Preference,’ Economica 40 (1973) 241–59CrossRefGoogle Scholar. It is different again from saying that a person’s decision is dictated by unconscious forces, as argued in psychological critiques and sociological and philosophical accounts inspired by them, such as MacIver, Robert M. “The Imputation of Motives,’ American Journal of Sociology 46 (1940) 1–12CrossRefGoogle Scholar and Peters, R. S. ‘Motives and Motivation,’ Philosophy 31 (1956) 117–30CrossRefGoogle Scholar.

4 Economists face similar problems estimating the ‘consumer’s surplus,’ introduced by Marshall, Alfred Principles of Economics, 8th ed. (London: Macmillan 1920), bk. 3, ch. 6Google Scholar. If people would have been willing to pay substantially more for some goods than the price demanded, why on earth should they bother contemplating the question of just how much more they would have been prepared to pay? Enough to know they want to buy at the asking price. Such considerations led Henry Sidgwick to despair of the ‘moral judgment of motives’ altogether: The Methods of Ethics, 7th ed. (London: Macmillan 1907), bk. 3, ch. 12, sec. 3.

5 Notice that, just as my analysis suggests, talk of motives typically does arise principally in the justification (typically, after-the-fact) of one’s actions to others; see Peters and, more fully, Wright Mills, C. ‘Situated Actions and Vocabularies of Motive,’ American Sociological Review 5 (1940) 904–13.CrossRefGoogle Scholar

6 ‘Naively understood,’ because acts performed under different descriptions are not, in some sense, the ‘same act’ at all. But this external aspect of the act is all that outside observers can confidently monitor; and if as argued below agents are often anxious to keep their options open as to how to describe an act, even to themselves, then that is all anyone will ever have to work with. Philosophers anxious to describe an action in terms that the actor would recognize as his own must not foreclose arbitrarily the possibility that he is himself insistently ambivalent as regards act-descriptions, even in his own mind.

7 There are good reasons, of both a consequentialistic and Kantian kind, for such rule-bound behaviour. I put ‘habit’ in scare quotes, because the sort here in view is a deliberately adopted practice rule and hence is immune to the worries of Ryle, Gilbert The Concept of Mind (London: Hutchinson 1949), 110–13.Google Scholar

8 The convergence might be so strong, and any attempt at disentangling motives so superfluous, that the agent might be led to say that ‘it is not ... that I have two separate motives, self-interest and benevolence, for doing the same action. I have one motive, a desire to live in a certain way, which cannot be characterized as a desire for my good rather than that of others’ (MacIntyre, Alisdair ‘Egoism and Altruism,’ in Edwards, Paul ed., The Encyclopedia of Philosophy [New York: Macmillan and Free Press 1967] vol. 2, 466)Google Scholar.

9 For empirical evidence, see Nisbett, Richard E. and Wilson, Timothy DeCamp ‘Telling More Than We Can Know: Verbal Reports on Mental Processes,’ Psychological Review 84 (1977), 231–59CrossRefGoogle Scholar. Further complications arise from the fact that most socially interesting actions are not performed instantaneously, but rather as complexes or sequences across a possibly protracted period of time. It is entirely likely that, at different points over that period, different ones of his multiple, converging reasons for acting will weigh more heavily with the agent. So even if he could somehow pin down which was his ‘one true motive’ at each point in time, the question remains which in this long sequence of shifting motives should be called ‘the’ motive lying behand the action-complex as a whole. If I am right in my argument, though, the agent usually would not even get as far as pinning down one true motive for each moment in time.

10 We may, like Derek Parfit’s idealistic young Russian nobleman, hope our preferences do not change: ‘Later Selves and Moral Principles,’ in Alan Montefiore, ed., Philosophy and Personal Relations (London: Routledge and Kegan Paul 1973), 145-6. We may, more plausibly, have a meta-preference for consistency in our preferences over time so that our lives taken as a whole ‘add up to something.’ But those are reasons for the earlier self to hope that the later self does not abandon its previous projects. From the perspective of the later self, having decided to make a change after due calculation of all the relevant costs, there is no reason for regret in the change. Assuming payoffs should properly be assessed according to the values prevailing at the time when they occur, that is an end to the matter (see Robert E. Goodin, Political Theory and Public Policy [Chicago: University of Chicago Press 1982], ch. 3).

11 Williams, Bernard Problems of the Self (Cambridge: Cambridge University Press 1973), chs. 11 and 12CrossRefGoogle Scholar; Nozick, Robert Philosophical Explanations (Cambridge, MA: Harvard University Press 1981), 405–9Google Scholar

12 The classic proof that intransitivity is positively irrational is that someone with such preferences can be induced to make ‘Dutch book’ against himself: if he prefers A to B to C to A, then he would be willing to pay some sum of money to make each move, going around in a never-ending circle until he was broke (Davidson, Donald and Suppes, Patrick Decision Making [Stanford, CA: Stanford University Press 1957]Google Scholar, 2 and Raiffa, Howard Decision Analysis [Reading, MA: Addison-Wesley 1968], 78Google Scholar). That is supposed to be a reductio, but someone who really did prefer A to B to C to A might reckon it money well spent; he might, in effect, have a meta-preference for inconsistency of that sort. Even where he does not, it is unclear how understanding the motives underlying your own intransitive choices could have helped you avoid the disastrous consequences of pursuing them to your ruin. Mostly what is needed for that is simply knowing that you have intransitive preferences and what follows from that fact. Knowing why you do so is neither necessary nor even particularly helpful in avoiding their consequences.

13 Simultaneously preferring A to B and B to A, for example, would indeed render choice impossible. So would an incomplete preference ordering stymie choice, at least so far as it regards unranked options. See Goodin, Robert E. The Politics of Rational Man (London: Wiley 1976), ch. 2.Google Scholar

14 Lawyers such as Hart, H. L. A. (see his Punishment and Responsibility [Oxford: Clarendon Press 1968], ch. 5)Google Scholar, would emphasize that intentions are not strictly equivalent to motives. The crucial component of mens rea, in the criminal law, is intention rather than motive per se: what matters is simply that the criminal performed the act in question, fully intending its natural consequences, rather than why he performed it. Similarly, with double-effect moralities, the concern is with what the agent intends rather than with what motives led the agent to form one set of intentions rather than another. Now, our concern with intentions might derive from a deeper concern with an agent’s character–his ‘standing motivation’ as Brandt calls it. Thus, although judgments of legal guilt are independent of broader motivational enquiries, judgments of blameworthiness or excusability, and hence of appropriate punishment, typically turn on them; see George P. Fletcher, ‘Legality as Privacy,’ in Robert L. Cunningham, ed., Liberty and the Rule of Law (College Station, TX: Texas A & M Press 1979), 182-207 and Richard B. Brandt, ‘A Motivational Theory of Excuses in the Criminal Law,’ in J.R. Pennock and J.W. Chapman, eds., Nomos XXVII: Criminal Justice (New York: New York University Press 1985), 165-200. Still, if motive-based moral evaluation is usually infeasible, as I argue it is, then advocates of that model may themselves welcome a distinction between intention and motive as a valuable fall-back. Even where we cannot know why a person was trying to produce the results he was, we can often say with some confidence what the results were he was trying to produce.

15 Brennan, Geoffrey ‘Pareto Desirable Redistribution: The Case of Malice and Envy,’ Journal of Public Economics 2 (1973) 173–83CrossRefGoogle Scholar offers the delightfully perverse example of both altruistic and envious people agreeing, each for his own very different reasons, to a redistribution of income from the rich to the poor-the altruist acting out of sympathy for the poor, the envious out of hatred for the rich.

16 MacCallum, Gerald C. ‘Legislative Intent,’ Yale Law Journal 75 (1966) 754–87Google Scholar

17 Sen, Amartya K. ‘Isolation, Assurance and the Social Rate of Discount,’ Quarterly Journal of Economics 81 (1967) 112–24CrossRefGoogle Scholar

18 ‘Most favourable’ is understood as that assumption which leads us to treat that agent in the way we suppose he would prefer given the values we hypothesize he holds. Often this standard proves indeterminate: different postulates about a person’s motives identify different treatments as most favourable. Then the ‘most favourable treatment’ rule requires us to preserve his options, giving him (insofar as possible) things that would be useful whether he were an altruist or an egoist at heart, for example.

19 Ely, John Hart ‘Legislative and Administrative Motivation in Constitutional Law,’ Yale Law Journal 79 (1970), 1282 ff.CrossRefGoogle Scholar; Bennett, Robert ‘“Mere” Rationality in Constitutional Law: Judicial Review and Democratic Theory,’ California Law Review 67 (1979) 1049–103CrossRefGoogle Scholar

20 For helpful discussions, I am grateful to Lincoln Allison, Alistair Edwards, Vinit Haksar, David Hunter, David Miller, Max Neutze, Philip Pettit, Raymond Plant, Andrew Reeve, Hillel Steiner, Hugh Stretton, Richard Routley Sylvan, Michael Taylor, Jeremy Waldron, Hugh Ward and Albert Weale.