[Note to plagiarists: I have a high Google ranking. You will be very easily caught.]
What is the Best Conception of Well-Being?
Well-being is concerned with prudential value, or what is good for an individual. But what is good for an individual? Attempts to answer this question have given rise to three general types of theories. Mental-state theorists assess well-being against the quality of our conscious experiences – it’s good to be happy, and bad to suffer. Desire-satisfactionists claim that prudential value is determined by the agent’s own preferences, so people are well-off to the extent that they get what they want. Objective List theorists consider well-being to be a matter of meeting certain criteria, perhaps including happiness, achievement, understanding, and so forth. This essay will critically assess and compare these rival theoretical frameworks, highlighting the potential for convergence that emerges.
It would be very implausible to take well-being as being purely ‘objective’, in the sense of being independent of the subject’s contingent concerns and preferences. The attainment of objectively-prescribed goals is surely of little worth to someone who does not endorse or in any way appreciate those goals. We might avoid this problem by adding an ‘endorsement requirement’ to the goods on the objective list. But the converse problem still remains: individuals may find value in goods that are not mentioned on the list. Any distinctively ‘objective’ theory of well-being holds that welfare is determined from an evaluative viewpoint independent of the agent’s own. For example, someone who knew all the facts about their life, and looked upon them with absolute satisfaction, might nevertheless be judged ‘badly off’. This is not plausible. We should instead recognise that the substantive elements of well-being can only be provided by a subjective list – i.e. one that is responsive to the particular concerns of the subject. Anything less fails to do justice to the inherently perspectival nature of prudential value.
Mental-state or ‘hedonistic’ theories are in many respects diametrically opposed to the sorts of objectivism rejected above. Nevertheless, they fall victim to a closely related objection, as they too give insufficient weight to a subject’s particular concerns. This point is illustrated by Nozick’s thought-experiment involving a machine that can stimulate our brains so as to induce whatever experiences we please. If given the choice to permanently hook ourselves up to such an ‘experience machine’, few of us would assent. We care about more than just our mental states; we also want to achieve our goals in actual fact.
The mental-state theorist could respond by pointing out that, according to their theory, not all desires are relevant to well-being. So it should come as no surprise that we can desire things other than what is best for us. What we value – what is good to us – may not be what is good for us. The earlier objection implicitly assumes that an unrestricted desire-satisfaction theory of welfare is correct. But it is viciously question-begging to object to a rival theory on such an assumption! Once we distinguish distinctively prudential value from personal value generally, objections based on personal preference lose much of their force.
Nevertheless, the ‘experience machine’ objection might not be so easily avoided. It must yet be explained what property qualifies a mental state as being valuable. The variety of pleasant experiences open to us – from slaking an intense thirst, to quiet contemplation – seem to have no substantive quality in common, other than that we desire them. So the mental-state theorist should claim that well-being consists in satisfying the desires we have about our own conscious experiences. However, it seems arbitrary to restrict the scope of our desires in such a way. We have preferences not only about our mental states, but also about the external world. Why should only the former count towards our well-being? The ‘good to’ / ‘good for’ distinction is no help here. Even from a third-person perspective, solely concerned with what is good for another, it seems that secretly hooking someone up to an experience machine would not be what’s best for them.
This brings us to desire-satisfaction accounts, which are commonly divided into ‘actual’ and ‘ideal’ desire theories. Actualists are concerned with the fulfilment of whatever desires an agent happens to have, whereas idealists take as the relevant desires only those which would survive informed rational reflection. The latter account is widely accepted as superior, but I argue that there is in fact little substantive difference between the two, and what difference there is should lead us to favour an actualist position.
That there is little substantive difference between these two kinds of desire theories is a straightforward consequence of the instrumental conception of rationality. The process of ‘ideal deliberation’ presupposes the goal of satisfying the agent’s fundamental preferences. Endowing the agent with full information and perfect rationality would simply ensure that she recognises how best to realize these desires; it could not press upon her any new ultimate ends. It might be suggested that novel information could open her up to possibilities that she hadn’t previously been aware of and so a fortiori could not desire. But her attraction to the new possibility must be grounded in existing dispositions. As Railton writes: “Of course, as a fool I have no antecedent desire identifiable as a desire to lead a more reflective or more Socratic life. But, if my motivational set contained no potential positive sentiment that could be ‘recruited by’ the information I gain about the Socratic life, then how could my novel exposure to the Socratic have any tendency to engage me...?”
Granted, the idealized deliberation process may lead the agent to reject some rogue desires, if they clashed with other, stronger, preferences. But these are ones that an actualist calculation would override in any case, since overall desire-fulfilment is maximized through their frustration. The only difference between the theories is whether the overridden desire should count for anything at all. I think it is clear that, although the agent is all-things-considered better-off having the rogue desires thwarted, she is nevertheless worse off in some respect. Any desire-frustration is bad in itself; it is just that this harm may be outweighed by the greater good of fulfilling other, stronger, desires. As such, the ‘actual desire’ theory gives the correct result in such cases, whereas an ‘ideal desire’ account would mistakenly claim that the frustration of rogue (non-ideal) desires does no harm whatsoever.
It might be objected that some things are intrinsically more worthy of desire than others, such that the value of fulfilling a desire will depend on the worth of its object. This sounds plausible because some things are more worthy of desire than others – but for instrumental rather than intrinsic reasons. ‘Mediocre’ objects are less able to elicit and sustain intense desires in humans than more ‘wonderful’ objects. As J.S. Mill observed: “it is an unquestionable fact that those who are equally acquainted with, and equally capable of appreciating and enjoying, both, do give a most marked preference to the manner of existence which employs their higher faculties.”
Mill judged the relative merit of two ‘pleasures’ by appeal to the majority opinion of those who are familiar with both. But such preferences might not hold universally, so we would do better to relativise such judgments to the individual. Here we could employ a counterfactual analysis: an individual would have been better off with alternative desires if that would have increased their level of desire-satisfaction. We can thus join the objectivist in claiming that the fool is missing out on something no matter his protestations to the contrary – but only if the alternatives really would prove superior from the fool’s own evaluative standpoint.
When weighing the relative ‘strength’ or importance of desires, we must take the structure of one’s desire set into account – mere ‘felt intensity’ will not do. As Parfit notes:
There are countless cases in which it is true both (1) that, if someone’s life goes in one of two ways, this would increase the sum total of his local desire-fulfilment, but (2) that the other alternative is what he would globally prefer, whichever way his actual life goes.
For example, becoming addicted to an otherwise ineffectual but readily available drug could maximize my brute desire fulfilment, since the daily cravings satisfied would be more intense than my persisting preference to live a life free of addiction. Nevertheless, it seems clear that the higher-order desire is more weighty. Desire theorists should recall that welfare assessments are properly made from the evaluative standpoint of the subject himself. To insist that someone would be better off addicted, regardless of their preferences, amounts to objectifying intensity of desire as the One True Good, rather than recognizing the sovereignty of the individual in deciding what is of value for him.
This raises technical questions about how the structural relations between desires should influence utility calculations. Desire fulfilment is always intrinsically (if not ‘all things considered’) better than frustration. But this does not mean that desire fulfilment always contributes positive utility; it could instead be that desire frustration yields negative utility. This neutral-negative combination is most plausible when the desire is itself unwanted, as in the addiction case above. We can then conclude that both (1) if addicted, I am better off satisfying my cravings than frustrating them; and (2) I would do even better to avoid the addiction in the first place. Of course, some desire-satisfactions do contribute positively. This is most plausible in cases where the desire is endorsed by an agent’s higher-order or global preferences.
When considering counterfactuals, we should assess hypothetical desires against their hypothetical structural relations. This allows us to reconcile our judgments that the fool would be better off with a desire to philosophize whereas I would not be better off craving the addictive drug. The difference is that, in the counterfactual situations, the former desire would be endorsed by the agent, whereas the latter one would not. I might be better off were I to become a willing addict, but given the description of the scenario, it is difficult to imagine any sane person genuinely having such a preference.
It was previously noted that mental state theories are too narrow, for we also have preferences about the external world that impact upon our well-being. However, the unrestricted desire account is too broad, for not all our desires are relevant to the quality of our life – as is highlighted by the ‘good to’/ ‘good for’ distinction. Instead, when assessing well-being, we should restrict our scope to people’s desires about their own lives. This would include desires about our mental-states, but also global desires about what sort of life we want to lead, which may in turn concern the achievement of goals we set ourselves in life.
But is this restriction enough? Kagan argues that we should distinguish between a person and their life, such that “it might be one thing for a person to be well-off and quite another for that person's life to go well.” Further, he makes the plausible claim that for something to genuinely benefit a person, “it must make a difference in the person”. This suggests that success in life can only affect one’s well-being if it enters their experience. It might seem that this leads us back to mental-statism. But perhaps whether an experience is veridical, is a feature that is relevant to our well-being.
Let us define the word ‘enjoyment’ so that it is not merely synonymous with pleasure, but rather is directed at some worldly object. So, for me to enjoy my success, it must be the case that I really am, in fact, successful. Otherwise my pleasure is misdirected; I am not really enjoying success, but rather am misled into feeling pleasure at an illusion. ‘Enjoyment’ is thus understood as pleasure felt in accurate appreciation of the way the world is. It is enjoyment of an object, and thus depends upon external reality. And let us understand ‘disappointment’ analogously, as a sadness that corresponds to the way the world really is, as opposed to resting on a mistaken belief.
Given these technical definitions, we could say that a person’s well-being depends not on their pleasure and pain, but rather, their enjoyment and disappointment. This melding of hedonist and desire theories captures the most important features of each. In particular, it escapes the ‘experience machine’ objections to hedonism, recognizing that illusory pleasures are lacking in a crucial respect, whilst meeting the plausible requirement that what's good or bad for a person must impact upon that person. Thus, while unknown successes may make one’s life go better, they do not (intrinsically) affect one’s well-being.
We are left with several varieties of personal value. Well-being is generally understood as the value of a life for the person, and I have argued above that a ‘veridical’ mental-state theory (equivalently: an ‘experienced’ desire-satisfaction theory) is the best account of this species of value. Unrestricted desire-satisfactionism, by contrast, gives an account of broad personal value, or what is good to a person. The restricted version of this, which I earlier advanced, might best be understood as accounting for the value that their life has to the person. This in turn can be compared to more objective standards concerning achievement and excellence, which I would understand as assessing the perfectionist value of a life. However, being independent of the subject’s evaluative viewpoint, this species of value is quite distinct from well-being. It is not obvious which – if any – of these varieties of value should serve as the standard of ‘utility’ for consequentialist moral theories. Certainly, restricted notions of well-being seem too narrow. There are goods beyond those that are good for our person or even for our lives. Thus I share Kagan’s conclusion: “If well-being is limited in its extent, then it may also be limited in its significance.”
 D. Parfit, Reasons and Persons, p.502, suggests something along these lines.
 I intend the qualifier ‘absolute’ to mean that there are no close possible worlds where they achieve greater subjective satisfaction. Without this technicality, even subjectivists might be committed to the seemingly ‘paternalistic’ judgment. See the discussion of counterfactuals, later in this essay.
 L.W. Sumner, ‘Two Theories of the Good’, p.5.
 R. Nozick, Anarchy, State, and Utopia, pp.42-45.
 S. Darwall, Philosophical Ethics, p.123.
 Parfit, p.493.
 Cf. J. Griffin, Well-Being, p.9: one may prefer “bitter truth to comfortable delusion…not because it would be morally better, or aesthetically better, or more noble, but because it would make for a better life for me to live.”
 C. Heathwood, ‘The Problem of Defective Desires’, pp.1-4.
 But cf. J. Griffin, ‘Replies’, p.282: “The relevant desires have to be ‘informed’ or ‘rational’ in a particularly strong sense – namely, formed in proper appreciation of the nature of their object”. The idea that certain objects are intrinsically more deserving of desire than others is tantamount to the objectivism that desire theories aim to avoid. As such, I will not be discussing such ‘strong’ conceptions of rationality in this essay. (But see below for more on ‘worthy desires’.)
 P. Railton, ‘Taste and Value’, p.55. Note that I follow Railton in understanding desires/preferences as dispositional rather than occurrent mental states. See also H.E. Baber, ‘Adaptive Preference’, pp.3-5.
 By ‘rogue’ desires I mean to include both ill-informed desires, e.g. a desire for food that turns out to be poisoned, and also irrational desires, e.g. a desire to avoid the dentist, despite knowing that the later toothache will be worse than the pain of the dentist’s drill. See Heathwood, pp.7-11.
 Heathwood, p.14
 J.S. Mill, ‘Higher and Lower Pleasures’, pp.62-63.
 Parfit, p.496.
 Griffin, WB, p15. “The desires I feel most intensely could be satisfied by your constantly imperilling my life and saving me only at the last moment, whereas I should clearly prefer peace to peril”.
 Parfit, p.498. Note that ‘global’ desires are about (some part of) one’s life considered as a whole.
 Ibid., p.497.
 I should note that the ‘structural’ considerations I’ve presented yield something quite similar to an ‘ideal desire’ account. The key difference, as I see it, is that I hold all of an agent’s actual desires to be relevant to his well-being. It is just that whether they are relevant in a positive or negative sense will depend on the agent’s structurally ‘higher’ (and to this extent ‘idealized’) preferences.
 Parfit, pp.494-5, notes that we should exclude desires that are only superficially about one’s own life – e.g. the desire to be someone whose children’s lives go well (independently of your own efforts) – whilst including similar ones such as the desire to be a good parent and so have given one’s children a good start in life.
 S. Keller, ‘Welfare and the Achievement of Goals’, argues that achieving a goal, i.e. having it realized due to one’s own efforts, makes an intrinsic contribution to one’s welfare. But a better explanation is that we desire such achievement. The benefit to our welfare depends upon our preference, not just the objective achievement. Someone who had no such preference would be just as well-off whether they authentically achieved their goals or had them realized by a wish-granting-machine (cf. Nozick, p.44). To suggest otherwise is to confuse welfare with the perfectionist value of a life.
 S. Kagan, ‘The Limits of Well-Being’, p.182 (fn7).
 Ibid., p.186.
 I should note that this idea is a development of Griffin, WB, pp.18-19, though his account of enjoyment was purely subjective. Mine is meant to be factive: if S enjoys that P, then P is true. L.W. Sumner makes a similar proposal in ‘Something in Between’.
 This would include pleasure and pain, since one might take a mental state as the object of enjoyment or disappointment.
 This is what is valuable for a life from the person’s evaluative standpoint. It is this first-person perspective (and the possible lack of ‘impact upon the person’) which makes it value to the person rather than for the person, even if it is good for their life (since it does, at least, impact upon that).
 It is thus misleading to follow Parfit in calling the restricted desire account “Success Theory”, since what matters is not ‘success’ in any objective sense (as that would yield perfectionist value – cf. note 20 above), but rather, the realization of whatever life the agent wants to live. This is an unusually subjective sense of ‘success’, which might lead some to confuse it with perfectionist value.
 Kagan, p.189.
Baber, H.E., ‘Adaptive Preference’, http://www.sandiego.edu/~baber/research/adaptivepreference.pdf [Presented at the APA Pacific Division Meeting, March 2004.]
Darwall, S., Philosophical Ethics. Boulder: Westview Press, 1998.
Griffin, J., ‘Replies’ in R. Crisp and B. Hooker (eds.), Well-Being and Morality. Oxford: OUP, 1999.
Griffin, J., Well-Being, Oxford [Oxfordshire]: Clarendon Press, 1986.
Heathwood, C., ‘The Problem of Defective Desires’, http://www-unix.oit.umass.edu/~cheathwo/PDD.pdf [forthcoming in the Australasian Journal of Philosophy.]
Kagan S., ‘The Limits of Well-Being’ in E.F. Paul et al. (eds.), The Good life and the Human Good. Cambridge: Cambridge University Press, 1992.
Keller, S., ‘Welfare and the Achievement of Goals’ in Philosophical Studies 121:1 (2004) 27-41.
Mill, J.S., ‘Higher and Lower Pleasures’ in J. Glover (ed.) Utilitarianism and its Critics. New York: Macmillan, 1990. [Excerpted from chp. 2 of Mill’s Utilitarianism.]
Nozick, R., Anarchy, State, and Utopia. New York: Basic Books, 1974.
Parfit, D., Reasons and Persons. Oxford [Oxfordshire]: Clarendon Press, 1984.
Railton, P., ‘Taste and Value’ in R. Crisp and B. Hooker (eds.), Well-Being and Morality. Oxford: OUP, 1999.
Sumner, L.W., ‘Something in Between’ in R. Crisp and B. Hooker (eds.), Well-Being and Morality. Oxford: OUP, 1999.
Sumner, L.W., ‘Two Theories of the Good’ in E.F. Paul et al. (eds.), The Good life and the Human Good. Cambridge: Cambridge University Press, 1992.