¹ Parfit, Derek, Reasons and Persons (Oxford, 1984), p. 157Google Scholar. Other philosophers have offered cases with the same structure. Cf. Brandt, Richard, A Theory of the Good and the Right (Oxford, 1979), p. 249Google Scholar; Bykvist, Krister, ‘The Moral Relevance of Past Preferences’, Time and Ethics, ed. Dyke, H. (Dordrecht, 2003), pp. 115–36Google Scholar (page references for this article are to the manuscript available online at: http://www.fil.lu.se/files/prodok8.pdf); Chris Heathwood, ‘Subjecive Desire Satisfactionism’ (manuscript), pp. 9–11.

² There are many views on how time should be taken to impact the question of what one rationally or prudentially ought to do at a time. For instance, Hare argues that the rational thing to do at a time is determined only by the desires one has at that moment. Hare, R. M., Moral Thinking (Oxford, 1981), pp. 104–5CrossRefGoogle Scholar. By contrast, Bricker argues that all of one's preferences, past, present and future, should count. Bricker, Phil, ‘Prudence’, Journal of Philosophy 77 (1980), pp. 381–400CrossRefGoogle Scholar. For more discussion, see Bykvist, Krister, ‘Prudence for Changing Selves’, Utilitas 18 (2006), pp. 264–83CrossRefGoogle Scholar; McKerlie, Dennis ‘Rational Choice, Changes in Values over Time, and Well-Being’, Utilitas 19 (2007), pp. 51–72CrossRefGoogle Scholar.

³ I will not directly address theories about what one ought, rationally or prudentially, to do at a particular time, which has been the focus of much of the literature to this point. However, several of the versions of Desire Satisfactionism I discuss here have clear analogues when it comes to the question of how to decide what to do at a time.

⁴ The contrasting view, Ideal Desire Satisfactionism, has many defenders. See e.g. Brandt, A Theory of the Good and the Right; Griffin, J., Well-being (Oxford, 1986)Google Scholar; Railton, Peter, Facts, Values, and Norms (Cambridge, 2003)CrossRefGoogle Scholar. However, this view faces the same problems because of the temporal nature of desires as Actual Desire Satisfactionism does. To keep the scope of this article manageable, I focus only on Actual Desire Satisfactionism. But presumably, the same sorts of solutions I discuss here are available to the Ideal Desire Satisfactionist as well. See n. 49 for more discussion.

⁵ Brandt, A Theory of the Good and the Right, p. 111; Parfit, Reasons and Persons, p. 117; Heathwood, Chris, ‘The Problem of Defective Desires’, Australasian Journal of Philosophy 83 (2005), pp. 487–504, at 489CrossRefGoogle Scholar.

⁶ Parfit, Reasons and Persons, p. 494.

⁷ Cf. Rawls's grass counterexample. Rawls, John, A Theory of Justice (Cambridge, Mass., 1971), p. 493Google Scholar; Parfit, Reasons and Persons, p. 500.

⁸ See Kraut, R. H., ‘Desire and the Human Good’, Proceedings and Addresses of the American Philosophical Association 68 (1994), pp. 39–54CrossRefGoogle Scholar; Brink, David O., Moral Realism and the Foundations of Ethics (Cambridge, 1989), p. 227CrossRefGoogle Scholar; Heathwood, ‘The Problem of Defective Desires’.

⁹ See e.g. Keller, Simon, ‘Welfare and the Achievement of Goals’, Philosophical Studies 121 (2004), pp. 27–41CrossRefGoogle Scholar.

¹⁰ Cf. Heathwood, ‘The Problem of Defective Desires’, p. 490.

¹¹ As with the generic form of Desire Satisfactionism presented earlier, we don't want to allow the concurrent satisfaction or frustration of one's instrumental desires to bear on one's welfare.

¹² This view is closely related to the position defended by Bricker in ‘Prudence’, although he formulates his view as an answer to the question of how one prudentially ought to act at a time, not to the question of how much welfare a life contains. Still, the views are quite similar. (McKerlie also mentions, but rejects, a view of this sort. See McKerlie, ‘Rational Choice’, pp. 53–7.)

¹³ Brandt, A Theory of the Good and the Right, p. 250.

¹⁴ Hare, Moral Thinking, p. 159.

¹⁵ Griffin, Well-Being, p. 16.

¹⁶ Heathwood, ‘Subjective Desire Satisfactionism’, p. 10.

¹⁷ Parfit Reasons and Persons, p. 157.

¹⁸ A future-directed desire is one whose object is in the future relative to the time at which one holds the desire. More precisely, for any two times (intervals), t₁ and t₂, if S desires, at t₁, that p obtains at t₂, and t₂ is later than t₁, then S's desire is future-directed. Present-directed and past-directed desires can be defined in analogous ways.

¹⁹ Bykvist, , ‘Comments on Dennis McKerlie's “Rational Choice, Changes in Values over Time, and Well-Being” ’, Utilitas 19 (2007), pp. 73–7, at 74–5CrossRefGoogle Scholar.

²⁰ Bykvist, ‘Comments’, p. 74.

²¹ Heathwood, ‘Subjective Desire Satisfactionism’, pp. 10–11.

²² Bykvist's reason for thinking this is that we should ‘respect the autonomy of person-stages’ (Bykvist, ‘Comments’, pp. 74–5). Heathwood thinks that allowing the satisfaction of long-gone desires to count towards welfare makes the view ‘in some cases, as paternalistic as any objective list theory’ (Heathwood, ‘Subjective Desire Satisfactionism’, p. 11).

²³ Thanks to an anonymous reviewer for this journal for pressing me on these responses.

²⁴ See Bricker ‘Prudence’, p. 389; Bykvist ‘Moral Relevance of Past Preferences’, p. 21; Heathwood, ‘Subjective Desire Satisfactionism’, p. 11; Parfit, Reasons and Persons, p. 151.

²⁵ One might also argue that our desires have other conditions built into them. Perhaps the atheist's desires are all conditional on his not having them merely because he is a confused and addled old man. If this condition does not obtain when it comes to the desire adopted late in life to have a priest come and pay him a visit, then not receiving such a visit would not in fact entail any desire frustration for him. In this way, one might also try to defuse the current objection. However, to preclude such a response, it seems we can just describe the case in such a way that the atheist's desire late in life for a priest is not conditional in such a way. The problems that such cases pose for Weak Concurrentism thus cannot be so easily side-stepped, I think.

²⁶ Bykvist, ‘Moral Relevance of Past Preferences’, p. 21.

²⁷ Cf. Parfit, Reasons and Persons, p. 157.

²⁸ Hare, Moral Thinking, p. 159.

²⁹ Griffin, Well-Being, p. 16.

³⁰ Heathwood, ‘Subjective Desire Satisfactionism’, p. 10.

³¹ Hare, Moral Thinking, pp. 101–4, where he suggests that it is possible to define “greatest happiness” . . . as the maximal satisfaction of now-for-now and then-for-then preferences. The happiest man is then, in this sense, the man who most has, at all times, what he prefers to have at those times.’

³² Cf. Heathwood, ‘Subjective Desire Satisfactionism’, p. 29; Bykvist, ‘Moral Relevance of Past Preferences’, pp. 23–4.

³³ For several examples of this sort, see Velleman, David, ‘Well-Being and Time’, Pacific Philosophical Quarterly 72 (1991), pp. 48–77CrossRefGoogle Scholar.

³⁴ Not only is Strong Concurrentism unable to account for the ways in which past-directed desires are capable of enhancing one's welfare, but it is, of course, also unable to account for the fact that such desires can diminish one's welfare. This is nicely illustrated by a case of Bricker's: ‘A man in his youth sets out various goals for himself, and, in the course of his life, succeeds in attaining them all. But as he enters old age, he looks back upon all his earlier activity with disgust and regret; he now believes that he has wasted his youth upon vain pursuits. How shall we evaluate this man with respect to prudence?’ (Bricker, ‘Prudence’, p. 383). Bricker argues that, when it comes to the total amount of desire satisfaction his life contains, things go worse for this man because he comes to regret the goals he worked so hard to attain in his youth. There would have been a greater degree of correspondence between the man's actual life and his desires, aims and goals if he had not come to regret, as an old man, the pursuits of his youth. Thus in so far as one favours a desire satisfactionist approach to welfare, one should think that this man is harmed in terms of welfare.

But Strong Concurrentism cannot yield this result. After all, the desires that this man forms in his old age are all past-directed. And the frustration of past-directed desires can have no impact on a person's welfare according to this view; it only lets one's present-directed desires count. Thus Strong Concurrentism is unable to give the result, which desire satisfactionists should all accept, that this man is harmed by coming to desire as an old man to have spent his youth otherwise than he actually did.

³⁵ Thanks to Fred Feldman for pointing out this response.

³⁶ Recall that I will eventually argue that, by the Desire Satisfactionist's lights, Weak Concurrentism actually gives the correct result about all these cases. My point now is just that people like Heathwood and Bykvist who wouldn't like Weak Concurrentism's implications about cases like that of the dying atheist should not be satisfied with Discount Concurrentism either.

³⁷ I don't claim to know what discount rate it really is best to use, but it doesn't matter because a similar problem case can be constructed no matter what discount rate is chosen.

³⁸ I'm not alone in having doubts about temporally biased formulations of Desire Satisfactionism. Bykvist (‘Moral Relevance of Past Preferences’, pp. 19–20), for one, argues for temporal neutrality. Bricker (‘Prudence’, pp. 383–4) also seems to favour temporal neutrality.

³⁹ Bykvist, ‘Moral Relevance of Past Preferences’; Bykvist, ’Comments’.

⁴⁰ Cf. Bykvist, ‘Comments’, p. 73.

⁴¹ Bykvist, ‘Moral Relevance of Past Preferences’, p. 29.

⁴² An anonymous reviewer has pointed out that it may sound odd to say that a desire has ‘full inside support’ if it concerns a time after one's death or before one's birth. These times, after all, are ones when the person whose desire it is has no ‘inside’ or consciousness in the first place.

⁴³ This, of course, is a controversial view. Mental state theorists about welfare (e.g. hedonists) will not be able to accept that things that happen after your death can have any impact on your welfare.

⁴⁴ Bykvist, ‘Moral Relevance of Past Preferences’, p. 29.

⁴⁵ Bykvist, ‘Moral Relevance of Past Preferences’, p. 24.

⁴⁶ Bykvist, ‘Comments’, p. 74.

⁴⁷ Bykvist emphasizes that he does not mean his talk of person-stages to commit him to any particular view about personal identity, e.g. the denial of the view that a person is ‘wholly present’ at any given moment. He says that ‘my use of “person-stage” was only meant to be a convenient way of referring to a person as he is at a particular time’ (Bykvist, ‘Comments’, p. 74).

⁴⁸ For similar reasons, Bykvist-ian Desire Satisfactionism cannot account for Bricker's case of the old man who comes to regret the activities of his youth. We rejected Strong Concurrentism on the grounds that it failed to imply that the man's welfare is decreased by the fact that he comes to develop, in his old age, a past-directed desire not to have spent his youth as he did. Bykvist-ian Desire Satisfactionism cannot yield this result either. After all, the old man's past-directed desire does not have full inside support. It is not matched by any present-directed desires that he held in his youth. So the frustration of the old man's past-directed desire does not reduce his welfare.

⁴⁹ As discussed above, some philosophers point out that a desire satisfactionist might try to appeal to the idea of desires that are conditional on their own persistence. See e.g. Bricker, ‘Prudence’, p. 389; Bykvist, ‘Moral Relevance of Past Preferences’, p. 21. However, as we saw, this strategy will not work across the board. Not all desires are conditional on their own persistence: ‘all preferences that express ideals seem not to be conditional on their own persistence’ (Bykvist, ‘Moral Relevance of Past Preferences', p. 21). Thus the problems concerning desire change will still arise for desires in cases that deal with ideals.

Another strategy that the desire satisfactionist might appeal to is the move to ideal desire satisfactionism. The idea, as Bykvist explains it, is that one could argue that ‘we will not have any intertemporal conflicts between rational preferences, since rational preferences cannot change over time: If I rationally want p at one time I will always rationally want p’ (Bykvist, ‘Moral Relevance of Past Preferences’, p. 21). However, I can see no principled reason to think that one's ideal preferences must remain unchanged. Why think that, if an ideal version of myself, at t₁, would desire something, then an ideal version of myself at a later time, t₂, must also desire it? I know of no reason to make this supposition. Sometimes our desires seem to change not because we gain new information, but simply because our tastes, concerns, interests or priorities change. I suspect that desires might change for no special reason at all. Or perhaps desires might change because of a baseline preference for variety in life. Accordingly, I think there can be cases of ideal desire change, as well. If I'm right, we can construct versions of cases like that of the dying atheist in which all the relevant desires are ideal ones. Ideal desire satisfactionism, therefore, would have to deal with the issue of desire change, just like actual desire satisfactionism.

⁵⁰ After all, we can divorce the fact of a desire's being satisfied from any positive affect that might accompany it. For more discussion of such issues, see e.g. Heathwood's treatment of the ‘dead sea apples’ objection (Heathwood, ‘The Problem of Defective Desires’, pp. 493–4).

⁵¹ I think that something analogous can be said also when it comes to self-concern, or self-love. While I remember to some extent who I was as a child, I experience who I am right now with a much greater level of vividness. So of course I'm going to care more about doing what's good for my current self than what would be good for my self as a child. But this doesn't show that it's always the case that the total amount of welfare contained in my life would be maximized by doing what is good for my current self.