1 Barry, Brian Political Arguments (London: Routledge and Kegan Paul 1965), 38–9Google Scholar

2 Cf. Barry, Brian The Liberal Tl1eory of justice (Oxford: Oxford University Press 1973)Google Scholar, chapters 3 and 11.

3 Allen Buchanan defends an account of the value of primary goods that is similar to the one I shall defend. Cf. ‘Revisability and Rational Choice,’ Canadian Journal of Philosophy 5 (1975) 395-408.

4 The fairness constraint could take the form of maximin; the result would be a moral theory that calls for maximizing the want satisfaction of whoever has the lowest level of want satisfaction. Maximin, however, attaches too much weight to an equal distribution and not enough to maximizing the sum total. The principle recommended by P. Weirich strikes a better balance between the two. Cf. Weirich, Peter ‘Utility Tempered with Equality,’ Nous 17 (1983) 423–39CrossRefGoogle Scholar.

5 If when we look we fail to see many frustrated wants, then, the want-regarding theorist will say, we are probably committing any of a number of different possible errors. For example, it is a mistake to infer that the downtrodden are content simply because they make no attempts to overthrow their oppressors. A slave's willingness to follow his master's orders doesn't show that he accepts slavery. It probably shows that he prefers following orders to the predictable consequences of disobedience, namely, punishment. Also, even if oppressed people compose the majority of a society's population, and they could, collectively, overthrow their oppressors, it needn't be rational for any individual to start up or join a subversive enterprise. Since it will be unlikely that the enterprise's success depends on any one individual's contribution, it pays for everyone to be a free-rider. For this reason too inaction doesn't show that people are indifferent to being oppressed.

6 I would not attempt to defend this claim in the way that R.M. Hare does. Contra Hare, I assume that the only way substantive moral principles (including utilitarianism) can be justified or refuted is by appeal to our shared intuitions. If we have clear intuitions about what it would be right to do in extraordinary or rare situations, then it counts against a moral principle that it cannot accommodate them. The same holds for intuitions about what it would be right to do in situations that violate very basic laws of nature. My hunch, though, is that the more fanciful the situation invented in a thought-experiment (i.e. the greater its departure from laws of nature, especially laws concerning the behavior and motives of human beings), the hazier and more uncertain become our intuitions about what it would be right to do in it.

7 Stability in the sense stipulated here needs to be distinguished from what Rawls meant by stability. Institutions are stable in his sense if they do not conflict with people's sense of justice. Both his and my senses of ‘stability’ differ from the ordinary one. Institutions are stable in the ordinary sense if they have a tendency to persist through time or are not easily subverted. Though institutions can be stable in this ordinary sense without being stable in either Rawls's or my senses, more often they won't be.

8 Barry (Political Arguments, 39) wrongly classifies hedonic utilitarianism as a want-regarding principle.

9 This assumption is in keeping with Barry's own position. He appears to reject those ideal-regarding principles which deny that only want satisfaction has intrinsic value on the grounds that they ‘can perhaps be sustained on a religious basis alone’ (Political Arguments, 41).

10 Some defenders of perfectionist accounts of the good believe that they can justify their ideals without appealing to our shared intuitions; intuitons, they think, can be wrong. They claim that a standard of human excellence or self-realization can be derived from the correct metaphysical account of man's true nature or end. I assume that the difficulties with this view are so numerous, grave, and obvious that not many people will find it attractive.

11 Cf. Rawls, John A Theory of Justice (Cambridge, MA: Harvard University Press 1971), 260–1Google Scholar.

12 This appears to be a tacit assumption made by several people, including 5. Lukes, ‘Alienation and Anomie,’ in Laslett, Peter and Runciman, W.G. eds., Philosophy, Politics, and Society Third Series (London: Basil Blackwell 1967) 134–56Google Scholar; also, Gray, John ‘On Negative and Positive Liberty,’ Political Studies 28 (1980) 507–26.CrossRefGoogle Scholar

13 An obvious candidate for such a criterion is: y is an alternative to x for person z if and only if learning of or becoming familiar with y would affect (i.e. weaken, strengthen, or eliminate) z's desire for x. Hence, ignorance of what Pepsi tastes like doesn't diminish the autonomy of one's preference for life in the countryside, because it is a brute fact of normal human psychology that tasting Pepsi won't affect such a preference. But there is a problem with this criterion. Suppose that Smith desires a, experiences alternative b, and then comes to desire b strongly and be indifferent to a. Suppose also that experiencing a, the object of his old desire, has no tendency to strenthen, weaken, or eliminate his new desire for b. Then a is not an alternative to b, according to the above criterion of what counts as an alternative, and Smith's past experience of a does not give his desire for b higher autonomy. That seems to me incorrect. I would want to say that his new desire for b has higher autonomy than his old desire for a did.

14 Elster, Jon ‘Sour Grapes - Utilitarianism and the Genesis of Wants,’ in Sen, Amartya and Williams, Bernard eds., Utilitarianism ami Beyond (Cambridge: Cambridge University Press 1980) 219–38Google Scholar; Brandt, Richard A Theory of the Good and the Right (Oxford: Oxford University Press 1981)CrossRefGoogle Scholar, chapter 6

15 Dworkin, Gerald ‘Autonomy and Behavior Control,’ Hastings Center Report 6 (1976) 23–8CrossRefGoogle ScholarPubMed. The distinction between first-order and second-order desires was first made by Frankfurt, Harry in ‘Freedom of the Will and the Concept of a Person,’ Journal of Philosophy 68 (1971) 5–20CrossRefGoogle Scholar.

16 Elster, 233

17 ‘Not surprisingly,’ because I took it to be a constraint on an acceptable account of autonomous desires that it have this consequence.

18 Cf. Sen, Amartya ‘Utilitarianism and Welfarism,’ Journal of Philosophy 76 (1979) 463–89CrossRefGoogle Scholar. Sen of course enlists these external desire prudes in a different cause, viz., to demonstrate the impossibility of a weak Paretian liberal.

19 The idea that pornography of the most explicit sort conveys or expresses values about sex is rejected (for reasons that I don't find convincing) by Schauer, Frederick Free Speech: A Philosophical Enquiry (Cambridge: Cambridge University Press 1982)Google Scholar, chapter 9, and defended by the ‘conservative’ and ‘radical liberal’ in Fred R. Berger's ’Pornography, Sex, and Censorship,’ Social Theory and Practice 4 (1977) 183-209.

20 I would like to thank Richard J. Arneson for the helpful comments he gave me on this paper and on the dissertation from which it is drawn.