Comments?
I'd love to hear!
Robert Bass
Department of
Philosophy
Coastal Carolina University
Conway, SC
29528
rhbass@gmail.com
Subjectivism about well-being1 (hereafter, just “subjectivism”) attempts to provide an account of what it is for one’s life to go well or, in particular circumstances, what makes one’s life go better. For present purposes, I will focus solely upon the latter.2 The attractive idea behind subjectivism is that what contributes to making one’s life better must somehow answer to one’s actual wants, desires, motivations and attitudes.3 presumably under some kind of idealization.4 The task for the subjectivist is to spell out what is involved in that idealization (and why that, rather than some other, is the correct idealization – for it is not to be assumed that all idealizations will yield the same results), what cognitive debilities and informational deficits are to be removed or circumvented, and thus to specify an epistemic standpoint from which the bearing of specific events and actions upon one’s well-being can be assessed.
The attractiveness of subjectivism derives, I think, from several sources. In the first place, it appears to fit readily into a naturalist world-view. What makes it the case that something contributes to a person’s well-being will just be ordinary – though perhaps very complicated – facts about it and its relation to that person. There will be no need to appeal to non-natural facts or mysterious faculties of intuition.5 Second, the kinds of cognitive capacities that seem likely to be required for the relevant idealization are those of ordinary and reasonably well-understood instrumental reasoning. Third, subjectivism seems to give us exactly the right sort of account about matters of taste. If I have a choice between, say, two flavors of ice cream (and there are no other differences between the two that I care about), the fact that I like one better than the other appears to be the only consideration that matters in deciding which is better for me. It is natural to hope to be able to extend this “taste model” to produce a more general account of well-being. Fourth, there is significant difficulty in developing a non-subjectivist account. It can appear that one must adopt a subjectivist account or nothing.6
Despite these attractions, I doubt the subjectivist’s task can be satisfactorily discharged. No doubt, the best response would be to work out a more plausible objectivist alternative, but, not having one in hand, I shall limit myself to explaining what I find unsatisfactory in the subjectivist program. I will begin by indicating why the problem of specifying the appropriate function of one’s attitudes to constitute one’s well-being is difficult. I do not take the difficulties that I shall point to at this stage to be decisive, however, so I will proceed by examining two accounts that I take to be representative of the best that can be done with the subjectivist program, Peter Railton’s and Connie Rosati’s. I shall claim that both fail to satisfy an intuitively acceptable adequacy condition for any subjectivist account, namely, they fail to identify a standpoint from which well-being or what contributes to one’s well-being can be distinguished from other things that might be desired or chosen or selected from the idealized standpoint.
The subjectivist sees well-being as some function of the attitudes of the agent whose well-being is in question. Whatever the function is, it will have to idealize in some way in order to be plausible. However, there is a prior question as to which attitudes the idealization is to take into account. It is simple but too short to say: the attitudes of the agent.
There are at least two reasons for this. First, among the motivationally salient attitudes of the agent may be attitudes that are specifically moral and related to (what the agent takes to be) obligations to act on behalf of others. Such attitudes may motivate self-sacrificial acts, which no one, including the agent, regards as contributing to her own well-being. Now, it is possible that such attitudes would be banished from consideration via the idealization process, but there doesn’t appear to be any necessity or even likelihood that this will occur. Moral motivations are not typically understood – and are not understood by those who have them – to involve inadequate information or defective reasoning about the circumstances. If anything, it may be all too clear that only a self-sacrificing option is morally acceptable. Consider, for example, a parent in an emergency situation in which she must abstain from food to give her children the best chance of surviving until rescue. There is no reason to suppose, in such a case, that she is poorly informed or irrational, yet it is enormously implausible to say that therefore starving to death (or risking it) is in her interests. So, any moral motivations that may lead to self-sacrifice will have to be excluded from the set of attitudes upon which the idealization is to operate.
Second, a person may also have non-moral (or not especially moral) attitudes that are other-directed in the sense that they involve caring about the states of others. These need not even be especially nice. Desires for vengeance or more generalized attitudes of malice are among the kinds of states I have in mind. A person who is vengeful or malicious wishes for someone else to suffer or be harmed. For the moment, let’s consider someone who harbors the malicious desire that some particular other person suffer or be harmed. Once again, it is possible that the idealization process would eliminate such desires from consideration, but it seems unlikely that this would be so in every case. If they are not eliminated, however, then it will be easy to say what will satisfy them – harm or suffering to some other. But it doesn’t sound plausible to say that harm to the other, even if it is unknown to the person who harbors the desire,7 contributes to the first person’s well-being. Again, if we are to have a credible subjectivist account of well-being, such desires or attitudes will have to be excluded from the appropriate set for the idealization to operate upon.
Now, it might be thought that the common feature of these two kinds of cases is just that they involve other-directed desires or attitudes of some sort and that we could plausibly simplify by saying that the desires and attitudes to be subjected to idealization in determining a person’s well-being are just those that are not other-directed. Unfortunately, this is also too simple. There may be some – many – things that actually do contribute to our well-being that also essentially involve other-directed attitudes.8 A plausible example is being in a loving relationship with someone else. Our lives would be poorer if we never were in such relationships, but we cannot (genuinely) be in such relationships without having other-directed attitudes.
At this point, circularity threatens. The subjectivist is attempting to characterize a standpoint from which actions, events and states of affairs may be assessed for their bearing on a person’s well-being and to do so making use of nothing but the person’s attitudes under a suitable idealization. But it appears that, in order to do so, he will have to exclude some attitudes as acceptable inputs for the idealization. This would not in itself be troublesome if there were some independently motivated explanation for their exclusion. But there does not appear to be. It seems that, in order to explain what we will include or exclude among a person’s attitudes, we have to appeal to what we already believe about what contributes to a person’s well-being, the very notion we were trying to explicate.
In fact, I think this worry can be expressed in a more general form. If what the subjectivist is saying is that a person’s good or well-being is what would be identified as such under the suitable idealization,9 the question may be asked whether the added information available from the idealized standpoint explicitly includes information about what is good for that person. If it does, then, of course, there will be nothing surprising in the discovery that what is good for that person can be identified from the relevant standpoint. If, however, that information is not part of what is provided in constructing the idealized standpoint, it is not obvious that the idealized standpoint is a satisfactory position for judgment about what contributes to the person’s good.
I do not take these considerations to be decisive against subjectivism. It may be that there are satisfactory answers to them in a well-elaborated version of subjectivism. To consider that, we need to have some such account on the table. I shall take the versions defended by Peter Railton and by Connie Rosati as representative since I think theirs is among the best work that has been done in working out a subjectivist view.10
Peter Railton has worked out a version of subjectivism that involves (so far as I shall follow him11) two key notions and a proposal defined in terms of those notions that is intended to elucidate what it is for something to be in one’s interests or contribute to one’s well-being.
The first notion needed is that of subjective interests. For a given person, these include “his wants or desires, conscious or unconscious .... For me to take a subjective interest in something is to say that it has a positive valence for me, that is, that in ordinary circumstances it excites a positive attitude or inclination (not necessarily conscious) in me.” (Railton 1997, p. 142) Plausibly, he claims that subjective interests, so defined, can be viewed as secondary qualities that “supervene upon primary qualities of the perceiver [incliner], the object (or other phenomenon) perceived [inclined toward], and the surrounding context: the perceiver [incliner] is so constituted that this sort of object in this sort of situation will excite that sort of sensation [inclination]."12 (p. 142)
As was noted earlier (see note 4 above), subjective interests unmodified or unidealized are not an attractive candidate for specifying what contributes to one’s well-being. Railton’s second notion, accordingly, is an idealization of his first. He suggests that we define an objectified subjective interest in this way:
Give to an actual individual A unqualified cognitive and imaginative powers, and full factual and nomological information about his physical and psychological constitution, capacities, circumstances, history, and so on. A will have become A+, who has complete and vivid knowledge of himself and his environment and whose instrumental rationality13 is in no way defective. We now ask A+ to tell us not what he currently wants, but what he would want his non-idealized self to want – or, more generally, to seek – were he to find himself in the actual conditions and circumstances of A. (p. 142)
What A+ would want for A to want (seek, desire, etc.) if he were about to become A is what Railton calls an objectified subjective interest of A. This, too, he claims can be understood as a “complex, relational, dispositional set of facts,”( p. 143) and he terms this set of facts the objective interest of the person in question.
The third step in Railton’s construction is to define what contributes to a person’s well-being in terms of objective interests: “X is non-morally good for A14 if and only if X would satisfy an objective interest of A.”(p. 143)
I believe that this account – as an account of well-being – faces at least one major problem.15 It is a problem that was alluded to earlier with respect to the question of which attitudes an idealization is to proceed from. Railton has done nothing to exclude the relevant other-directed attitudes from the set of subjective interests upon which the account of objective interests and therefore of well-being is to be based.
At first sight, it may appear that he has done something in that direction since he claims that what he’s defining is what is “non-morally good” for a person. But, in fact, when the subjective interests were introduced, no restriction was placed upon them. If having a subjective interest in something just means “that it has a positive valence for me, that is, that in ordinary circumstances it excites a positive attitude or inclination (not necessarily conscious) in me” (p. 142), then it seems that any moral attitudes that I have may perfectly well qualify as being in my subjective interests.16 Then, they will be part of the motivational base upon which objective interests are constructed and, unless it is assumed that A+ is an egoist or that he wants the A he is about to become to be an egoist (and, if he is or does, how does he tell what contributes to his well-being?), these moral attitudes may survive as part of what it is non-morally good for A to have or exercise or satisfy. This is a bit too neat to qualify as a reconciliation of morality and self-interest. The problem, in short, is that well-being becomes all-encompassing and that nothing that wouldn’t be defeated simply by fuller information or by better reasoning about the best way to satisfy one’s desires is excluded.
In “Internalism and the Good for a Person,"17 Connie Rosati argues for what she calls a two-tiered internalism. Unlike Railton’s, her proposed idealization does not necessarily18 appeal to the perspective of a character possessed of unlimited knowledge and reasoning capacity. She, rightly I think, seems suspicious that the perspective of such a being might be inadequately suited to the actual person whose good is in question.
I shall not go into detail with regard to the argument she develops for her proposed idealization, but shall only sketch what seems to be an important idea behind it. One of the central intuitions behind subjectivism is “that a person’s good must suit her, that it cannot be alien to her.” (Rosati 1996, p. 301) Railton, in common with many other subjectivists, endorses this thought, but when he turns to theory construction, allows a person’s actual attitudes to serve as inputs to an idealization process but does not allow for any input from the agent as to what the appropriate idealization itself is to be. Rosati’s alternative might be described as an attempt to be a subjectivist “all the way down.”
Briefly, her proposal is that the appropriate standpoint for judging whether something can be a part of a person’s good is itself one that would be selected by the person under certain conditions. The relevant conditions, she terms “ordinary optimal conditions”:
We need a way to rule out alienated conditions, without allowing the appropriateness of conditions to depend upon how they now strike a person, whatever her present state might be.
We can do so by requiring that a person regard counterfactual conditions as appropriate not in her present state, but under ordinary optimal conditions. These would include that a person not be sleeping, drugged, or hypnotized, that she be thinking calmly and rationally, and that she not be overlooking any readily available information. This list ... is not intended to be exhaustive. By ‘ordinary optimal conditions’ I mean whatever normally attainable conditions are optimal for reflecting on questions about what to care about self-interestedly. (p. 305)
Thus, her position is that
[S]omething X can be good for a person A only if two conditions are met:
1. Were A under conditions C and contemplating the circumstances of her actual self as someone about to assume her actual self’s position, A would care about X for her actual self;
2. Conditions C are such that the facts about what A would care about for her actual self while under C are something A would care about when under ordinary optimal conditions. (p. 307)
Two points should be noted here. First, this seems to come closer than Railton’s account did to ruling out the kinds of other-directed motivations that left him without an account (specifically) of well-being, for she specifies that ordinary optimal conditions are to be “optimal for reflecting on questions about what to care about self-interestedly.” (Rosati 1996, p. 305) Involved, I think, is the plausible thought that though we might be less than adequate at specifying in detail what will make our lives go better, we may still be able to recognize or (reliably) imagine greater expertise than our own.19 Though we may be in the dark, we are not totally in the dark. Nonetheless, there is nothing in any admission conditions that Rosati imposes on the motivations or attitudes that are to be considered from the relevant counter-factual standpoint that would ensure that no disruptive other-regarding attitudes will be granted a role in determining what contributes to one’s well-being. At most, the standpoint she tries to identify may be helpful toward providing an account of well-being; it cannot do everything by itself.
Second, I think Rosati is aware of this. She does not actually claim to be providing a full subjectivist account of well-being (or, in her terminology, a full internalist account of what is good for a person). If her definition is examined, it turns out only to claim to offer necessary conditions. She does not place a great deal of emphasis upon this, but she never says that she is providing a set of necessary and sufficient conditions.20
Thus, by being more modest in her aspirations, her account is more successful than Railton’s. However, it is not more successful in the sense of succeeding in specifying what contributes to a person’s well-being in subjectivist terms but rather in the sense that it does not attempt the task and so, does not fail.
Throughout this paper, I have been pressing variations of the same argument, that subjectivism about well-being does not in fact pick out something plausibly understood as the well-being of the agent. Rather, insofar as it appears to make progress in that regard, it does so by presupposing understandings of well-being.
I do not see that there is any satisfactory way to avoid this conclusion. It seems to me that the entanglement of other-directed attitudes in our motivations is so complete that we cannot expect to disentangle them thoroughly. We can have self-interested reasons for acquiring genuinely altruistic motivations toward others, other-directed attitudes the satisfaction of which contributes to our own well-being and even a thorough fusion and commingling in which there is no telling, apart from particular circumstances, how the attitude bears on our well-being.
If I am correct, there is no hope of having a satisfactory subjectivist account of well-being, but it may be that something can be salvaged from various attempts to provide one. It may be that the attempts can be reinterpreted as working towards a subjectivist account of what matters or what is of value in general.21
References
Broome, John. 1993. Can a Humean be moderate? In Value, welfare, and morality, edited by R. G. Frey and Christopher W. Morris. Cambridge: Cambridge University Press.
Hampton, Jean, and Richard Healey. 1998. The authority of reason. Cambridge ; New York: Cambridge University Press.
Korsgaard, Christine. 1997. The Normativity of Instrumental Reasoning. In Ethics and practical reason, edited by G. Cullity and B. Gaut. Oxford: Clarendon Press.
Railton, Peter. 1997. Moral Realism. In Moral discourse and practice: some philosophical approaches, edited by S. Darwall, Allan Gibbard and Peter Railton. New York: Oxford University Press.
Rosati, Connie S. 1996. Internalism and the Good for a Person. Ethics 106 (January):297-326.
Sobel, David. 1994. Full Information Accounts of Well-Being. Ethics 104 (July):784-810.
Sumner, L. W. 1995. The Subjectivity of Welfare. Ethics 105 (July):764-790.
1. I shall generally speak of “well-being,” but what I am referring to can be equally well designated as what is in one’s interests or what is good for one.
2. Powerful criticism of subjectivism as a theory about what constitutes one’s interests or contributes to one’s well-being in a whole life may be found in Sobel 1994.
3. I shall employ “attitudes” as a general term for this collection of motivationally salient psychological states. I shall take it that their motivational salience may be indirect or counter-factual. It would be an unfortunate result for a theory of well-being if it did not allow that surprises could be good for us, even if there is nothing we can directly do to arrange to be surprised. “In this sense, to motivate is to prompt or elicit a proattitude – such as desiring, liking, being glad of, caring about, and so on .…” (Rosati 1996, p. 301)
4. Almost no one is attracted to the idea that what makes one’s life go well or better is constituted by what we actually want, prefer, etc. – that is, without any idealization. It is too easy to describe cases in which such attitudes are infected by ignorance or mistakes in reasoning. (Does drinking a poison make my life better if I want wine and believe the poison to be wine?)
In addition, since what one does not know may influence one’s well-being, I will, in the following, limit the idealizations that I consider to those that involve (or may involve) the use of better or more complete information than is available to the agent.
5. I do not mean to be discounting non-mysterious intuition (which I think is probably indispensable in most areas in philosophy) – by which I simply refer to revisable beliefs for which we cannot, at present, see any further reason. I also do not mean to suggest that non-natural properties or mysterious intuitions provide the only possible ways of avoiding subjectivism.
6. Wayne Sumner’s argument in “The Subjectivity of Welfare” might be interpreted – perhaps by someone less sanguine about the success of the subjectivist program than Sumner himself – as pointing in this direction: Everything else fails; therefore, it’s subjectivism ... or nothing. (1995)
7. I think it obvious that malicious desires may be so structured that what would satisfy them is actual harm to the other, whether or not the malicious person knows about that harm. To elicit an expression of such a desire, the malicious person might be asked which he prefers of two hypothetical cases: one in which the object of his malice is unharmed (and he does not believe the person to be harmed) and one in which the object of his malice is harmed (and he does not know or find out about the harm). Since he will believe the same thing in both cases, if he prefers the second case, it must be because of the actual harm done, not (solely) because of phenomenological accompaniments of believing the other to have been harmed.
8. By saying that other-directed attitudes are essentially involved, I mean that these things which do contribute to our welfare would not be what they are, and so, would not contribute to our welfare (at least not in the same way) if they did not involve other-directed attitudes.
9. It would be too strong to say that it is what would be chosen under the suitable idealization unless one thinks that suitable idealization would make egoists of us all. (An alternative might be to say that it is what would be chosen ceteris paribus under the suitable idealization, but filling in the details of the ceteris paribus clause would also involve having some prior account of what contributes to one’s well-being.)
10. I cannot, of course, consider all attempts to work out a subjectivist view. But showing that problems infect the best attempts will help to build the case that it is subjectivism rather than (just) certain statements of it that is untenable.
11. Railton is not seeking just to work out a subjectivist position, but to do so as part of working out a version of naturalistic moral realism (1997). For present purposes, I am concerned only with the subjectivism, not with the further uses to which he puts it.
12. Railton refers to perception and sensation because he is drawing a parallel with sensory secondary qualities. I have inserted cognates of “incline” in brackets in the relevant places, hoping to capture somewhat more accurately what it means to think of subjective interests as secondary qualities.
13. I find persuasive the arguments that several authors have urged that we cannot adequately understand instrumental reasoning without invoking normative notions which cannot themselves be accounted for fully in terms of instrumental reasoning. For arguments of this sort, see Broome 1993; Hampton and Healey 1998; Korsgaard 1996. However, I shall not pursue those issues further here.
14. What is non-morally good for an agent seems to be the same as what contributes to his well-being.
15. It may face others as well. I am sympathetic to the notion that the idealization is so extreme that we simply have no idea – and there may be no fact of the matter – as to what A+ would want A to want were he about to be transformed into A, but I will leave those questions aside.
16. So may the other kinds of other-directed concerns discussed above.
17. Rosati 1996.
18. It might allow such a perspective if that were chosen as the relevant counter-factual situation for assessing what contributes to a person’s well-being under ordinary optimal conditions.
19. It is an interesting question why Rosati did not pursue the apparently natural extension of her line of reasoning. If, under ordinary optimal conditions, I can select a counter-factual vantage point from which to judge what contributes to my good, why might I not from that superior vantage point select a further counter-factual vantage point from which to judge – and so on? Perhaps, the answer would be that that doubly counter-factual vantage point might not be sufficiently responsive to the actual person’s attitudes unless the initial selection of a counter-factual standpoint was understood as a standpoint from which a further selection would be made.
20. Given her remarks at the end of the article about the possibility and plausibility of anti-realism about well-being, she could hardly have consistently claimed to provide sufficient conditions. (Rosati 1996, pp. 325-326)
21. I could not claim to be especially optimistic that it will succeed there, but I see fewer obstacles in the way than with a subjectivist account of well-being alone. (On the other hand, it is a larger project. Even if I am right in discerning fewer obstacles, it may be that, all things considered, it is as difficult or more so.)
