Subjectivists about well-being claim that an object is good for someone if and only if this individual holds a certain type of pro-attitude toward this object. In this paper, I focus on the dispute among subjectivists that opposes those who think that the relevant pro-attitudes are actual to those who think that they are counterfactual under some idealized conditions. My main claim is that subjectivism should be stringently actualist, though our actual pro-attitudes may be criticized from an intrinsic perspective. To defend this claim, I first present three desiderata that a subjectivist theory of well-being should fulfil. Two of these desiderata result from the fact that a subjectivist theory of well-being should not be implicitly paternalist, while the other is that it should be able to play a normative role. I then show that several actualist theories that introduce light forms of idealization or other conditions that have a similar aim, fail to satisfy at least one of the antipaternalist desiderata. This gives some legitimacy to a very stringent version of actualism. I then describe three features of the kind of stringent actualism that I want to defend, which will explain the ability of what is good for someone to play its normative role. Finally, I show how these features allow us to deal with the classic objection that the objects of actual “defective attitudes” cannot be good for the holder of these pro-attitudes.
Les subjectivistes à propos du bien-être soutiennent qu’un objet est bon pour un individu si et seulement si cet individu possède un certain type d’attitude positive à l’égard de cet objet. Dans cet article, je me concentre sur le débat à l’intérieur du camp subjectiviste qui oppose ceux qui pensent que les attitudes positives pertinentes sont actuelles à ceux qui pensent qu’elles sont contrefactuelles sous une condition d’idéalisation. Ma thèse principale est que le subjectivisme devrait être rigoureusement actualiste bien que nos attitudes positives actuelles puissent être critiquées d’un point de vue qui leur est interne. Afin de défendre cette position, je présente d’abord trois desiderata qu’une théorie subjectiviste du bien-être devrait satisfaire. Deux de ces desiderata résultent du fait qu’une théorie subjectiviste du bien-être ne devrait pas être implicitement paternaliste, alors que le troisième résulte de ce qu’elle devrait pouvoir jouer un rôle normatif. Je montre ensuite que plusieurs théories actualistes qui introduisent des formes modestes d’idéalisation ou d’autres conditions ayant la même visée échouent à satisfaire au moins un des desiderata anti-paternalistes. Une version très rigoureuse de l’actualisme subjectiviste s’en trouve ainsi légitimée. Je décris ensuite les trois aspects de l’actualisme rigoureux que je souhaite défendre et qui permettront d’expliquer dans quelle mesure ce qui est bon pour un individu peut jouer un rôle normatif. Enfin, je montre comment ces aspects nous permettent de répondre à l’objection classique selon laquelle les objets d’« attitudes défectives » actuelles ne peuvent être bons pour ceux qui ont ces attitudes.
Corps de l’article
According to subjectivism about well-being, something is good for an individual if and only if this individual has a certain type of pro-attitude A toward this object under certain conditions C. Within this broad framework, there are many debates concerning the precise analysis that subjectivism should offer. For instance, subjectivists disagree about what type of pro-attitude is relevant. Some have suggested that it should be an attitude of desiring (Sidgwick 1907, Murphy 1999; Heathwood 2005) or of valuing; others, that it should be a second-order desire (Lewis 1989, Railton 1986a); still others, that it should be a belief about what is good for one (Dorsey 2012). However, I will not be concerned with this debate in the present paper. Rather, I will focus on the dispute between those subjectivists who think that the relevant pro-attitudes are actual pro-attitudes and those who think that they are counterfactual under some idealized conditions. More precisely, the aim of the paper is to defend a new version of actualist subjectivism that is stringently actualist while explaining how this stringent actualism can fulfil the normative role that what is good for someone plays. I will try to achieve this aim while staying as neutral as possible on all the other debates among friends of subjectivism.
The plan of the paper is the following: Section Two starts with the presentation and defence of three desiderata that a subjectivist theory of what is good for someone should fulfil. In a nutshell, the desiderata derive from two considerations that push in opposing directions. On the one hand, an important drive toward subjectivist approaches to well-being is to have a theory such that someone’s well-being is not detached from what matters to him or her. On the other hand, what is good for someone often plays a normative role: it is a tool to criticize our actual pro-attitudes. Putting these desiderata to work, I show in Section Three that some versions of actualism are not sufficiently actualist. This will be my plea for a very stringent form of actualism. However, the traditional problem with stringent actualism has been its inability to fulfil its normative role. In order to overcome this problem, I will present three features of the stringent actualism that I want to defend. One of these features opposes my stringent actualism to all other forms of stringent actualism. I will call this actualism Stringent Critical Actualism. In Section Five, I will finally show how the three features of Stringent Critical Actualism allow it to deal with the classic “defective” pro-attitude objection that the objects of some of our actual pro-attitudes may not be good for us. Section Six concludes the paper.
2. THREE DESIDERATA OVER SUBJECTIVISM
As mentioned above, the debate that opposes actualist to idealized approaches of subjectivism has appealed to two considerations. The first consideration is normative. We often take what is good for someone to depart from what that person desires, believes is good for him or her, etc. For instance, one may entertain desires whose fulfilment will be disappointing if not catastrophic. Hence, it seems that a subjectivist needs to be able to assess the pro-attitudes that we have de facto. Historically, this has been a strong argument against actualist versions of subjectivism.
A second consideration has received much attention: it is the idea that what is good for someone must matter to him or her. Railton’s often-cited remark that “it would be an intolerably alienated conception of someone’s good to imagine that it might fail in any such way to engage him” seems to capture this idea. Let us call it the No-Alienation requirement. This requirement is certainly a main rationale that drives us toward subjectivist versus perfectionist or objective theories of well-being. The defect of the latter theories seems to be that by their very definition of well-being, they imply that some objects can be good for an individual while he or she is in fact indifferent to these objects. And this seems—at least to many—to pave the way to unacceptable forms of paternalism as it might be considered better to provide people with objects to which they are completely indifferent rather than with objects that matter to them for their good. Therefore, if subjectivist approaches to well-being are seen as superior because they satisfy a No-Alienation requirement, then this requirement becomes at least a prima facie constraint to which a subjectivist theory of well-being must conform. Notice, moreover, that the No-Alienation requirement as used by Railton or Dorsey, for instance, is also an argument against idealized forms of subjectivism. Indeed, such idealized versions of subjectivism have the same defect that traditional perfectionist and objectivist views have, since insofar as they also end up saying that what is good for someone may be indifferent to him or her as he or she actually is. And this underlines that the proper interpretation of the No-Alienation requirement has to be rather strict: it is a necessary condition for an object to be part of the good of a person that this object would matter to this person if he or she were considering it.
While these first two requirements are important, I want to suggest that we need to add what I will call a No-Rejection requirement. The idea is simple and is somehow symmetrical with the No-Alienation requirement: at least in some cases, it seems unacceptably paternalistic to act toward someone as if something that mattered to him or her counted for nothing. Why? Intuitively, the problem is that if something matters to someone, then it must count as part of that person’s well-being even if it may be true that this thing will in the end be disappointing. Consider, for instance, a young man who strongly wants to become a philosopher and who is enthusiastic about everything that concerns philosophy. However, suppose that he lacks a clear understanding of what a philosophical career consists in, and that, if he were able to consider it properly, he would not desire it anymore. Finally, suppose that he is presently unable to consider this career properly. If you tried to explain to him what the career consists in, he would just say that you were wrong, that it cannot be right, and would get upset and annoyed by your unwelcome explanations. I want to claim that in such a case his enthusiasm must be taken as implying that the philosophical career is good for him. Why? Because in doing so, we have the most natural ground to explain why it is wrong to act toward such people without taking this enthusiasm into account, which is that a certain part of what is good for them is neglected. In a nutshell, because some actions may be seen as unacceptably paternalistic in virtue of the fact that they neglect what matters to someone, we should accept a No-Rejection requirement on well-being, which precisely says that what matters to someone is good for this person.
Before going further, let me be clear about the status of the two last constraints that I have put forward. My point is not to argue directly that we must accept a certain theory of well-being because some forms of paternalism are wrong. My point is merely to argue that a subjectivist theory of well-being should allow for the rejection of certain actions as wrong and paternalistic because they do not take into account what matters to people. In other words, all things being equal, we should discard subjectivist theories of well-being that would be rejected by anti-paternalists on the grounds that these theories do not tie tightly enough what is good for someone to what matters to him or her. And I argue that the two constraints considered above should be accepted precisely as expressing this theoretical orientation.
I acknowledge that this point has little dialectical force against objectivists or perfectionists, who can respond that they are just trying to describe what well-being consists in, and that the legitimacy of paternalism is a different point. But my aim here is not to argue directly against these approaches to well-being. My point is merely that if one is tempted to reject these views because they are seen as potentially allowing some form of unacceptable paternalism, and if one is thereby tempted by a subjectivist approach, then the two antipaternalist requirements considered above express this orientation. In other words, I am only trying to offer arguments that should be seen as relevant to thinkers who are already within the subjectivist camp and who believe that certain antipaternalistic considerations may be relevant.
This being said, notice that the normative requirement seems at first sight to pull exactly in the opposite direction insofar as it emphasizes that a theory of well-being must offer the tools for an assessment of what presently matters to us. However, it does not follow that the three requirements are incompatible. In fact, the aim of this paper is precisely to argue that if one is attracted by a subjectivist approach to well-being, then one should accept all these requirements and therefore a stringently actualist version of subjectivism insofar as it is the only one which satisfies them all. If successful, such a program will offer a theory of well-being that could be a shared ground for discussions between paternalists and antipaternalists insofar as this theory does not exclude antipaternalists by adopting a theory of well-being that is unacceptable to them.
I therefore conclude that a subjectivist theory of what is good for someone must satisfy both a normative requirement and two antipaternalist requirements: the No-Alienation requirement and the No-Rejection requirement. The aim of this paper is to offer an actualism that satisfies all three desiderata. To justify this actualism, I will highlight, in the next section, that several actualist views of what is good for someone are mistaken because they are not stringent enough. These objections will pave the way for the Stringent Actualism that I wish to offer.
3. NOT SUFFICIENTLY STRINGENT ACTUALISMS
In this section, I show that we have to reject certain forms of actualism because they do not satisfy at least one of the two antipaternalist requirements. As a consequence, I will also show that ideal pro-attitude forms of subjectivisms must be rejected.
3.1 Semi-actualist views
Let us first consider what I will call semi-actualist versions of subjectivism. I refer here to the approaches suggested by Sobel (2009) and Dorsey (2017) that have been constructed with the intent to satisfy the No-Alienation requirement. Their key point is to introduce a distinction between two roles that idealization plays in subjectivist theories of well-being. Dorsey presents the distinction as one between the two building blocks that constitute a subjectivist theory of what is good for someone. The first building block is what he calls the “Good-Value Link.” He states it as follows: “for any object, event, state, etc., O and agent x, is good for x only if O is valued, under condition c, by x” (Dorsey 2017, p. 200). He argues that it is with regard to this link that alienation must be avoided. Therefore, he defends an actual version of this link to the effect that, for any object, event, state, etc., O and agent x, O is good for x only if O is valued, under actual condition c. However, a subjectivist account of well-being is incomplete until one has offered an explanation of what it is to value something, hence the need for a second building block that explains what it is to value something. According to Dorsey, it is at this stage that idealization should be introduced. In his view, to value an object O cannot be reduced to the actual possession of a certain type of pro-attitude toward O. Rather, an individual x values an object O only if x would have a certain type of pro-attitude toward O under some idealization conditions. Why? As Sobel has suggested, taking desire as the relevant pro-attitude, the crucial point is that our actual desires may result from a misappreciation of their object. Hence, “desires that do not involve [a] perfect forecast are, in a sense, not actually for the option as it is but rather for the option as it is falsely imagined to be” (Sobel 2009, p. 345. Italics are mine). On the contrary, “the desire is truly for [O] when the desire is sustained or created in light of complete and accurate information about what [O] would be like” (ibid., p. 346). Along similar lines, Dorsey claims that “a necessary condition for x to value [O] is that x would take the relevant pro-attitude toward [O] given full consideration of the relevant range of imaginative conditions, i.e., ways [O] might be” (Dorsey 2017, p. 209). In addition, he claims that the pro-attitudes must also belong to the pro-attitudes of x were these pro-attitudes rendered minimally coherent. The main reason he offers for this second idealizing condition is that “when someone maintains incoherent pro-attitudes, it can be unclear, ultimately, what he values” (ibid., p. 204).
I contend that the distinction between the Good-Value Link and the Theory of Valuing Link is largely artificial when we consider the two antipaternalist requirements. In order to make the argument clearer, I will focus on the view that values should be explained in terms of desires. I believe that the argument offered here would apply mutatis mutandis to other analyses of values in terms of other pro-attitudes. To justify my contention, I will return to the young man who desires with great enthusiasm to become a philosopher. We have supposed that his desire is not for the career as it truly is. Nevertheless, if one tried to correct his understanding of what is in fact involved in embracing a philosophical career, one would feel that he somehow does not want to be informed. He would simply resist any change in his beliefs about the nature of a philosophical career. However, his present refusal may at some point vanish and he would then realize that he has no desire to become a philosopher. Able to fully consider a different career, he would in fact desire to become a political activist, a desire which he does not hold presently.
Now, let us see what semi-actualists would say about this case. First, insofar as the present desire would not persist if the young man were accurately informed, being a philosopher has no value for the young man. Second, full consideration would lead him to desire to be a political activist, so Sobel would claim that this is what has value for him. Dorsey, however, defends a slightly different view: on his view, for an object to be good for someone, it is necessary but not sufficient that this person desire this object while accurately informed. It is also necessary that the person presently desire this same object. It follows that, on Dorsey’s view, it is not good for the young man to embrace a philosophical career in virtue of the desire that he would have if accurately informed insofar as the young man is not also presently desiring this career.
How do these characterizations of what is good for the young man fare with regard to our two antipaternalist requirements? As we will see, neither satisfies both requirements. Start with Sobel’s view: first, he claims that the political career is good for the young man. But to accept that is just to violate the No-Alienation requirement insofar as the young man does not feel presently engaged by a political career. This is why Dorsey’s response is preferable. Turning now to the philosophical career, both Sobel and Dorsey claim that this career is not good for the young man insofar as he would not still desire it if he were accurately informed. Hence, both violate the No-Rejection requirement.
In addition, I would like to suggest that the whole idea of a full consideration of objects in order to reveal what is good for someone is misguided. Our desires, our pro-attitudes, and even our beliefs about what is good for us rely always on a partial understanding of their objects. In fact, a large part of our life consists precisely in moving toward or turning around the objects of our pro-attitudes and thereby discovering their nature from different perspectives, their consequences, and whether or not they continue to be of interest to us. Consider the simple case of wine tasting. Surely, the whole good of it is not to have the wines we like and will continue to like if we were experts. That could be boring. Part of the good that we derive from our experience of wines is progressively to acquire some knowledge about what we truly like, to be surprised about wines we thought we do not like (but in fact do like), and even to discover, interestingly, that we dislike some wines, or do not like them any more—or at least not as much as we used to like them, etc. In other words, the enterprise of acquiring knowledge about the objects of our desires through discoveries, surprises, and disappointments is part of what is good. And an important part of that is also to become less attracted to what seduced us at first. This seems to be a truism about the development of a culture in art. Hence, it is not true that what would be good for us would just be to have objects that we would still desire when fully aware of their exact nature. This would remove all the good we take in our first encounters with objects, in experiencing them, discovering them, and leaving them, in orienting and reorienting ourselves toward and away from what seems to us from moment to moment as having worth.
But this is not all. I want to claim that even if we put aside the good in acquiring knowledge and in dynamic experience, it is a mistake to assess the good of an object with reference to a unique, ideally informed point of view. To see this, consider the good of a certain friendship, a loved person, or a specific occupation. It seems inappropriate to assess how good they are for someone on the basis of a unique, ideally informed perspective. Consider the beginning of a love story or a new friendship. Certainly part of what is good lies in our ignorance of the other. To assess this good as if we knew the other perfectly would seem like assessing how the love story or the friendship would be a long time after it had started. Even this comparison is mistaken because, then, time will have modified these relationships. In other words, I would like to contend that it is a mistake to hold that one can assess how good an object is for a person at a given time through the assessment of what would be the pro-attitude of this person if he or she were ideally informed about this object. The only way we can assess how good something is for someone is through the good it offers at various moments, given the specific history that has related that person to the object. Therefore, the very idea of a complete and vivid consideration is not what we need to clarify what is good for us. What we need is our actual pro-attitudes toward an object as they may evolve through time while our knowledge improves—but also while what is relevant about these objects evolves. As I will argue in Section Four, this is not to say that there is no place for the remark that some objects are disappointing. For the time being, my point is merely that in designing a theory of value, we should not start by discarding badly informed pro-attitudes, and this, for two reasons. First, it seems very harsh to claim that the first music or wines or people that we loved were not good for us insofar as our experience has led us away from them. Second, the very idea of a complete and vivid representation of something is not what is relevant when we want to know why an object is or may be good for us in the present or in the future: our knowledge is always partial and, even when it becomes less so, it seems more proper to say that the object is good to us now in a different way than to say that we have reached the true way in which it is good to us.
Therefore, I conclude that Sobel’s and Dorsey’s semi-actualist views face important objections. First, they both violate the No-Rejection requirement. Second, they both mistakenly consider that an object cannot be good for someone if this person does not have the relevant pro-attitude toward it while accurately knowing this object. Finally, and in addition, Sobel’s view violates the No-Alienation requirement.
3.2 Impure actualism
Let us now turn to more rigorous versions of actualism that do not involve any kind of idealization. In this subsection, I would like to show that even some such rigorous forms of actualism still do not satisfy the No-Rejection requirement. This will be done by focusing on the actualism defended by Heathwood (2005).
As an actualist version of subjectivism in terms of desires, this view claims that an object O is good for an individual A if and only if A actually desires O. This does not settle the supplementary question as to whether a desired object benefits the person who has this desire when the desire is fulfilled or when the person had the unfulfilled desire. Heathwood himself defends the claim that an object benefits the individual who desires it at the moment at which the desired object is obtained, a view that may be called after Dorsey (2013) the “time-of-object” view. On the contrary, an opponent to this view could defend a “time-of-desire” view (again, after Dorsey 2013), which claims that the object desired benefits the individual who desires it at the time at which he or she desires it and not at the time at which it is obtained. As it will appear, we won’t be concerned with this debate. Suffice it to say that in both cases it is claimed that it is impossible to benefit a desire if the desire is not fulfilled.
Finally, Heathwood introduces an additional condition for an object O to be good for someone, which he calls Concurrence: “the desired object O is desired at the same time that it is obtained” (adapted from Heathwood 2005, p. 490). For instance, “if I desire fame today but get it tomorrow when I no longer want it,” fame is not good for me (ibid., p. 490). Notice that Concurrence is compatible with both the time-of-desire and the time-of-object views. It just adds a condition for the fulfilment of a desire to benefit its possessor: that the possessor still has the desire when it is fulfilled.
All that being said, we can appreciate an interesting consequence of Concurrence. To do so, we can notice first that, at the very time at which a desire is satisfied, its object is obviously more vividly experienced. At least, if one still desires the object while one experiences it, then it seems that the object is not disappointing. Thus, if it is required, for an object to be good, that one still desire that object while the desire is being fulfilled, then it follows that if an object is good for someone, then it cannot be disappointing for him or her. It is therefore a strength of Heathwood’s view that it avoids the objection that the fulfilment of a desire may not be good for its possessor. Moreover, this objection is avoided without making use of any idealizing condition.
Unfortunately, the introduction of Concurrence, which itself presupposes that an object cannot be good for someone if it is not fulfilled, is unacceptable. This is because the object of a desire can benefit the possessor of the desire even if the desire is never realized. As it is well known, it is beneficial to someone who possess a desire to become quite sure that his or her desire is going to be satisfied in the future even if this desire is not satisfied in the future. Symmetrically, doubts about the satisfaction of a desire that concerns my child after my death diminishes my well-being. So it seems that it is an error to claim that the object of a desire brings no good if the desire is not fulfilled.
Now, there are two objections that one might raise. The first is to suggest that the experiences produced by our anticipations are just pleasures and displeasures. However, this response cannot be accepted within the subjectivist framework. We can understand these anticipations only in terms of the desire that we have for a certain object. Within this framework, it seems that the most plausible option is to slightly enlarge the desire-satisfaction view and to claim that someone benefits from a certain object that is desired either when the desire is fulfilled or when the possessor of the desire anticipates this fulfilment.
A second objection is to contend that this argument succeeds only for subjectivists who adopt a Knowledge condition that the fulfilment of a desire benefit its possessor only to the extent that he or she knows that it is fulfilled. But, as the argument goes, not all subjectivists adopt this Knowledge condition: many of them hold, on the contrary, that someone’s well-being is improved not by knowing that his or her desire is satisfied, but merely by having the desire satisfied, whether or not this fact is known. In response, I would like to show that an argument for the same conclusion can be offered even if the Knowledge condition is rejected. Let us start with a typical intuition put forward by those who reject the Knowledge condition. Consider Peter, who desires to be esteemed by his relatives and who believes falsely that he is. According to those who reject the Knowledge condition, his situation is bad insofar as his desire is not fulfilled. But now, suppose that, although his desire is not already fulfilled, it is close: for some reason, his relatives are collectively more positive about him. It seems that we have to say that his situation has improved. Maybe the situation is not as good as it would be if his desire were satisfied, but it is nevertheless much better than it would be if the desire were far from being satisfied.
Now, it may be said that we are making a category mistake: Peter’s situation is no better in absolute terms; rather, it is merely instrumentally better. But I am not sure about this objection. Consider situations in which someone got very close to the fulfilment of his or her desire, but didn’t fulfil it. We think that such a situation is worse than the mere absence of fulfilment. But certainly this situation is not instrumentally worse. Therefore, it seems that the situations in which one does not fulfil a desire are not all equal in terms of noninstrumental goodness-for. Therefore, even subjectivists who are against the introduction of a Knowledge condition should accept that someone’s well-being can improve or decline depending on that individual’s situation relative to the fulfilment of a certain desire. This is true of well-being independently of whether the desire is presently fulfilled and independently of whether the desire is fulfilled in the end.
The first problem that we have raised against Concurrence and its implications leads to a second. Recall the young man who enthusiastically desires now to be a philosopher, although he will completely lose this desire when he is about to fulfil it. It follows from Concurrence that the object of his desire is not good for him, and that this is a violation of the No-Rejection requirement. Once again, we can see that there is something mistaken here by considering how we should act toward this young man. If we follow the subjectivists who adopt Concurrence, then the philosophy career is not good for the young man. Hence, we lack a reason to take this desire into account and this may allow actions that we might consider as unacceptably paternalistic. Certainly, one might argue that the desire could be taken into account on other grounds. But what other grounds? The grounds of respect will be too broad, and if we say that we should respect his desire, are we not saying that we should take into account what is good for him? For it is hard to think that respect tells us to take into account what is not good for someone. In other words, the simplest and most plausible way to say why we should take into account his desire is precisely to acknowledge that if a philosophical career presently matters to the young men, then it is presently good for the young man.
I believe that the lesson of the present assessment of Heathwood’s view is the following. Even if a condition such as Concurrence allows a stringent actualist to deny that the object of a disappointing desire is good for someone who desires it, the stringent actualist must reject this condition. This is because this stringent actualism fails to satisfy the No-Rejection requirement, and this in part because it fails to fully understand how we benefit from having desires. Does this leave open the possibility for an actualist to retain a condition in parallel with Concurrence, while allowing that one might benefit from the goodness of an object in virtue of a desire toward this object that will not be fulfilled or will be disappointing? Yes, it does. One might propose to replace Concurrence by Concurrence*, which would read as follows: “The object of a desire is good for the possessor of the desire only if the possessor possesses the desire at the moment at which he or she benefits from this object.” Such a clause seems to me plausible. However, Concurrence* does not have the supposedly theoretical advantage that made the initial attraction of Concurrence since the former does not imply that the objects of disappointing desires are never good for the possessors of these desires. I conclude from the assessment of Heathwood’s view that an actualist must avoid any idealization strategy, whether direct or indirect, and that we have at least good preliminary reasons to look for a very stringent and straightforward version of actualism.
4. STRINGENT ACTUALISM AND NORMATIVITY
From the previous section, it has appeared that subjectivism must be stringently actualist and must therefore merely assert the following:
An object, state of affairs, event, O is good for X if and only if X has actually a pro-attitude of type A toward O.
All this is good. However, one might wonder whether such a stringent actualism can satisfy the normative requirement. Is it faithful to the idea that the determination of what is good for someone, of his or her values, is a tool to improve his or her life? Or, as Railton puts it, is such a stringent actualism capable “of capturing important elements of the critical and self-critical character of value judgments” (Railton 1986b, p. 11)?
My aim in the next section is to show that a positive answer can be offered. In this section, I will present three features that allow a stringent actualism to play this normative role. The second and third features are shared by all actualisms, but I am going to argue that the first feature distinguishes the kind of stringent actualism that I want to defend from others. I will call this view Stringent Critical Actualism. Along the way, this presentation will allow me to show how this actualism accords with a set of intuitions about how we may help others and how people should think in order to improve their lives. The next section will put in practice the normative dimension of Stringent Critical Actualism and show how it deals with “defective” actual pro-attitudes that have been seen as counterexamples to stringent forms of actualism.
The first feature that I want to present has been rejected by most stringent actualists. They have argued that, if what is good is analyzed in terms of desires, then at least the underived or intrinsic desires cannot be assessed from an internal perspective. In particular, Murphy claims, “What does lie beyond criticism in terms of well-being, on the simple desire-fulfilment theory, is our deep, actual desires—whatever those happen to be” (Murphy 1999, p. 267). Hubin (1996) makes a similar claim. Although he is offering a theory not of well-being but of reasons in terms of desires, he is an actualist in claiming that an analysis in terms of counterfactual desires should be rejected in favour of an analysis in terms of actual desires. But what is relevant for us is that, in defending this reduction of reasons to actual desires, a key point in his view is that intrinsic motivations cannot be criticized.
With regard to theories of well-being with which I am concerned here, the thesis that intrinsic motivations cannot be criticized turns out to be false—for two reasons. First, if what is good for someone is at least in part to fulfil that person’s desires, then it is surely better to have more desires fulfilled. And from this, it follows that sets of desires can be assessed as more or less easy to fulfil, or as more or less coherent, in a way that allows the fulfilment of all or at least all of those that are of great importance. Hence, even if a desire taken in isolation cannot be assessed, it can be assessed given the abilities of its owner to fulfil it and given the full set of desires that its owner possesses. The argument extends easily if the relevant pro-attitudes are second-order desires, attitudes of valuing, beliefs concerning what is good for one, etc.
Second, my defence of stringent actualism in the previous section allows us to make a further criticism of our actual pro-attitudes or desires, and it is here that my view definitely departs from Heathwood’s and Murphy’s views. I have argued that even if a desire for an object will appear disappointing or is badly informed, the object is good for the person. Hence, I contend consistently both that a badly informed desire makes its object good, but that this lack of information is criticizable. It is preferable to have better-informed desires insofar as the lack of information may lead to disappointment. In other words, I concede to Sobel and Dorsey that our actual present pro-attitudes or desires may rely on a lack of consideration of their objects, even though I do not conclude, with them, that this is sufficient for their objects to lack value. So I depart from both kinds of actualism: I claim that the objects of our present actual desires are good for us, even if the desires or pro-attitudes will be disappointing. But the ground of this claim is not that our intrinsic desires cannot be disappointing or cannot be intrinsically criticized. On the contrary, I claim that we can criticize desires or pro-attitudes either because they lack coherence or because they rely on an insufficient or wrong understanding of their objects.
Notice, in addition, that this view offers an original response to the objection of Enoch (2005) that any idealization of our actual pro-attitudes to define well-being relies implicitly on objective goods. Following Sobel, I claim against Enoch that some idealizations or criticisms that are intrinsic to desires or pro-attitudes make sense. However, as I have now argued at length, I do not follow Sobel in claiming that the objects of these idealized desires or pro-attitudes are the constituents of our well-being. To this extent, I agree with Enoch that if subjectivism is defensible, it must be stringently actualist. However, as already said, this is not because our actual pro-attitudes or desires have authority by default.
So much for the first feature of Stringent Critical Actualism. Let us then turn to the second feature that will allow it to fulfil the normative requirement. This is the simple fact that actualism does not preclude the assessment of what is good for someone and thereby of his or her pro-attitudes with respect to a variety of extrinsic norms and scales, by which I mean norms or scales that are not relevant to assess someone’s well-being. In particular, one can assess whether those things that are good for someone are in accordance with moral demands, whether these goods have intellectual or aesthetic worth, etc. Here my point is not to make a claim about these extrinsic norms. Nor is it to decide whether they are true norms or just scales that have no normative force. Nor is it even to explain which of these norms have precedence over others. Rather, my point is merely to insist that Stringent Critical Actualism is fully compatible with the possibility of assessing someone’s goods not only from an internal but also from numerous external perspectives.
The third and final important feature on which Stringent Critical Actualism may rely to satisfy the normative requirement derives from the temporality of pro- attitudes: insofar as pro-attitudes change over time, what is good for someone changes accordingly. Admittedly, this is an obvious point that applies to all subjectivist accounts. If it is held that O is good for x if and only if x has a certain type of pro-attitude toward O under certain relevant conditions, then a temporal change in the actual or counterfactual pro-attitudes of x is a temporal change in what is good for x. However, it is not clear that all subjectivists have adopted this temporal approach. The point is especially relevant with regard to idealized versions of subjectivism because the idealizing element may pull toward a conception of individual values as temporally stable. Suppose you define what is good for someone as what this person would desire having full information, if this desire belonged to a maximally coherent set of desires. In this configuration, it is tempting to suggest that what is good for someone at a certain time is also good at any other time. After all, if there is a set of desires that is somehow optimally rational, it seems that it is impossible to improve on it. Therefore, it seems that what is good for you at a given time is what is good for you at any other time. In any case, my point here is merely to emphasize that in contrast with idealized subjectivism, which may be tempted —even if not committed— to claim that our idealized set of pro-attitudes would be stable across time, actualist subjectivisms and thus Stringent Critical Actualism are committed to accepting that what is good for someone changes as his or her relevant pro-attitudes change. Hence, what is good for someone during his or her lifetime must not be restricted to what is good for that person at a certain given time. What is good for someone in his or her lifetime is the integration of what is good for that person at all the different times of his or her life, whichever view one may take on how these temporally different goods should be integrated.
These three features put us, I believe, in the best possible position to assess someone’s well-being and how beneficial it would be to attempt to improve it. First, it is possible to assess someone’s set of present actual pro-attitudes as providing less good than it could and to think of other sets of pro-attitudes that this same person could acquire and that would improve his or her well-being in the future. But second, insofar as someone’s present actual pro-attitudes constitute what is good for him or her at this present moment, they cannot be set aside as worthless. Third and finally, since changing one’s pro-attitude usually has a cost, this will also need to be taken into account. Thus, attempting to improve one’s life is not a decision that involves leaving non-goods for goods—in which case, very costly but small improvements may be taken as rational. Rather, what the decision has to balance is the costs involved in changing my present actual set of pro-attitudes and the benefits of having a new set of pro-attitudes that will improve my well-being.
Similar remarks apply to the assessment of actions intended to contribute to someone else’s well-being. Two questions will need to be answered. First, is it worthwhile to improve the set of pro-attitudes of this individual in view of the cost that the improvement itself may carry? Second, are we allowed to impose this cost, or even to interact in order to improve the pro-attitudes of the individual under consideration? The same considerations will apply if we consider a moral improvement of an individual’s set of pro-attitudes. Are the moral benefits really worthwhile? Are we allowed to impose some costs or to interfere to improve the morality of this set of pro-attitudes?
My aim is certainly not to answer the second question and its moral variation. These are important and further questions that go beyond our understanding of well-being. Nevertheless, I contend that Stringent Critical Actualism is the only subjectivist theory of well-being that offers a common ground to discuss the legitimacy or illegitimacy of paternalist actions and generally actions in favour of someone’s well-being.
5. THE DEFECTIVE PRO-ATTITUDE OBJECTIONS
It is now time to see how Stringent Critical Actualism satisfies the normative requirement when confronted with objects of defective pro-attitudes. Actualist versions of subjectivism in terms of desires have been criticized as unable to deal with so-called defective desires whose objects we do not intuitively want to take as good for someone. Among these defective desires, we find ill-informed desires such as the desire to drink this glass that contains petrol or the desire for a career that is grounded in an insufficient knowledge of what this career consists in; there are pointless or “quirky” desires, such as the desire to turn on radios as soon as one sees them (Quinn 1993) or to count the blades of grass in the lawn (Rawls 1971); there is also the desire of the former drug addict to fall back into addiction, etc. Obviously, the objection extends mutatis mutandis to other versions of actualism expressed in terms of other pro-attitudes. So, how does Stringent Critical Actualism deal with so-called defective pro-attitudes?
A lack of space precludes a full discussion of the long list of defective pro-attitudes that may be relevant. I will not consider here desires that are considered as defective in virtue of extrinsic norms (such as moral norms), insofar as they are not relevant to the assessment of well-being. I do not want to rehearse a point here that most subjectivists have made. That being said, Stringent Critical Actualism can also rely on responses that other actualists have already offered in order to deal with some cases of defective desires that I will briefly consider first. Second, I will consider the cases in which Stringent Critical Actualism has an original answer, and especially the case of pro-attitudes that rely on insufficient information about their object.
To begin, let us consider the objection raised by quirky desires or pro-attitudes. There are two responses that stringent actualists may offer. The first is to underline that the relevant pro-attitude in the subjectivist analysis of what is good is not the pro-attitude of desiring, but either a cognitive attitude or a conative pro-attitude, such as intrinsically desiring or valuing. If it is the pro-attitude of believing that something is good for one, the objection collapses since we do not have incomprehensible beliefs, such as the belief that it is good for me to turn on all the radios I come across. Similarly, the objection collapses if the pro-attitude is an intrinsic desire or an attitude of valuing. Indeed, if we consider our intrinsic desires, it seems that their objects are never seen as completely external to us. Similarly, if an object is valued, then the desire for this object cannot be an event completely detached from us.
Let us then turn to the case of the former drug addict. According to Stringent Critical Actualism, it is good for the former drug addict to have his or her drug because the drug is the object of, say, a desire. But it is often argued that it cannot be good for the drug addict to have his drug. My response is in two parts. First, I see no reason to deny that Stringent Critical Actualism is giving the proper response. I do not see why we should deny that satisfying the desire for the drug is good for the drug addict to some extent. Nevertheless, I admit that the fulfilment of this desire could be bad, all things considered for the drug addict. There is nothing problematic here insofar as this good may certainly be inferior to other goods for him or her that are incompatible with this addiction. Moreover, it is certainly better for the drug addict to have a future set of desires that excludes this desire, since fulfilling it will always be in conflict with his or her other desires. Hence, because Stringent Critical Actualism can assess a set of desires as offering someone a life that could be improved, it can make sense of the intuition that satisfying this desire will be bad for the drug addict on the whole and that it would be better for him or her not to have this desire at all. Nevertheless, admitting all this does not require us to deny that the drug is presently and actually good for the drug addict. The point is that he or she should merely reject this good in order to live a better life.
Finally, let us turn to cases of ill-informed pro-attitudes or desires. In the context of this paper, it is crucial that Stringent Critical Actualism deal properly with such cases since we have rejected as inappropriate views that have been designed precisely to deal with them. I believe that in order to manage them, we should distinguish two types of cases.
The first type of case is that where one holds a mistaken instrumental belief or a mistaken belief that a certain object specifies a certain more general one. This is the case where I want to drink this glass of liquid because I am thirsty, but ignorant that it contains petrol. Are stringent actualists forced to concede that drinking petrol is good? Certainly not. The Stringent Critical Actualist can claim, as other actualists have, that an object is good for someone only if this individual has an underived or intrinsic desire or pro-attitude toward this object. So, it is easy to deal with these cases insofar as the person in question has no underived desire to drink petrol, but rather an underived desire to drink something that will quench his or her thirst, and this excludes petrol.
This response is, however, insufficient when we consider the second type of case, the case that involves a deeply erroneous pro-attitude because it is far from obvious that such a desire is derived. Let’s go back to the young man who desires to become a philosopher, although the desire relies on an erroneous, partial, or abstract but well-entrenched understanding of this career, and who would be disappointed if he were to fulfil his pro-attitude. Is the object of this pro-attitude good for the young man? According to Stringent Critical Actualism, it is, and I think that this is the correct response. It may even be an important good that presently contributes heavily to the well-being of his life. He thinks a lot about it, is excited by the prospect when he discusses philosophical topics, and he takes pleasure in viewing himself as a philosopher in the future. Hence, I believe that insofar as being a philosopher is a good for him from which he presently benefits, this good may demand some attention from us. On the other hand, Stringent Critical Actualism can appeal to temporal considerations: if it is true that he will be disappointed by such a career and will thus not desire it anymore if he were to become a philosopher, then being a philosopher will not be good for him at some point in the future. Hence, the balance of his present and probable future goods may justify actions intended to discourage him, maybe by indirectly allowing him to discover on his own what the job that he presently values so much would truly be like, or by offering alternative possibilities whose value might be more lasting.
The stringent actualism that I have offered here relies on two points that have been insufficiently highlighted. The first is that the No-Alienation requirement is not sufficient to protect a theory of well-being from being implicitly paternalistic. We must add a No-Rejection requirement whose precise characterization should be the object of further work. In any case, when this requirement is taken into account, then it offers a strong rationale for adopting an actualist approach to subjectivism against any idealizing condition or any other condition (such as Concurrence) that goes in the same direction. The second point is that if the No-Rejection requirement is taken as a main rationale for being actualist, then we must concede that pro-attitudes, and especially desires, are intrinsically assessable: they may rely on a faulty representation of their object or they may be incoherent and lead thereby to a life endowed with less well-being. Given these possible assessments, a stringent actualism is not stuck with a view that claims the pro-attitudes cannot be intrinsically assessed. And, therefore, it can satisfy the normative requirement.
Sidgwick (1907); Brandt (1996); Railton (1986a and b).
Railton (1986b, p. 9).
A similar point is emphasized by Dorsey (2017, p. 196).
I say “potentially allowing” because other considerations can lead a theory of right conduct to reject as wrong any action that would not take into account what matters to someone while denying that what matters is good for him or her. However, it is unclear what such considerations could be. Moreover, it is hard to believe that neglecting what matters would not be relevant to the the assessment of the action as wrong.
In fact, Dorsey insists at several places that he does not want to commit himself to the idea that these two idealizing conditions are the only ones. Others may also be legitimate.
Heathwood (2005) also defends what has been termed, after Rabinowicz and Österberg (1996), a satisfaction view and not an object view within subjectivist approaches. This is a debate about the bearers of the property good for. On the object view, it is the object of our desires that is good for us. On the satisfaction view, what is good for us is to have desires satisfied. In this latter case, the relata of the property good-for is the state that conjoins a desire for an object and the object itself. In any case, this distinction is not relevant for us here. It is orthogonal to all the other distinctions made in this summary of Heathwood’s view.
In fact, Heathwood accepts a Knowledge condition. Thus, I would not need to consider this objection if he were my only target. However, insofar as my aim is to show that an actualist defending Concurrence is mistaken, I want to extend my critique to an actualist subjectivist who would adopt Concurrence but reject a Knowledge condition.
Heathwood (2005, p. 492) holds the same view.
In fairness to Heathwood, I must underline that he acknowledges this first kind of criticism, although he takes it to be external (Heathwood 2005, p. 494).
See, for instance, Murphy (1999, p. 268) for this claim.
Certainly, it might be argued that some of these norms are relevant and thereby intrinsic to the assessment of well-being. However, I stick here to the subjectivist tradition, which denies the relevance of these norms as determining how well someone’s life is going. The aim of this paper here is not to defend subjectivism against alternative views such as perfectionism.
At least, in most cases. In fact, desires may change due to aging or to hormonal factors in such a way that the most coherent and fully informed set of desires that could be derived from someone’s present actual desires would be different from the set of most coherent and fully informed desires that could be derived from the same individual’s past actual desires.
Smith (1994) has even argued that this set of desires would be the same for all humans.
For this term and a discussion of this specific class of defective desires, see Bruckner (2016).
See especially Heathwood (2005); Hubin (1996); Murphy (1999).
The strategy has been adopted by Hubin (1996), Murphy (1999), and Heathwood (2005).
Here I partly join Mendola (2009), who is dubious about the possibility for actualists to use the distinction between underived and derived desires. While I agree with this criticism in the kind of case we are now considering, it seems to me that in the first type of case, considered above, there is no difficulty in seeing them as derived.
- Brandt, Richard B., A Theory of the Good and the Right, 2nd ed., Oxford, Clarendon Press, 1998.
- Bruckner, Donald W., “Quirky Desires and Well-Being”, Journal of Ethics & Social Philosophy, vol. 10, no. 2, 2016, p. 1-34.
- Dorsey, Dale, “Desire-Satisfaction and Welfare as Temporal”, Ethical Theory and Moral Practice, vol. 16, no 1, 2013, p. 151-171.
- Dorsey, Dale, “Idealization and the Heart of Subjectivism”, Noûs, vol. 51, no. 1, 2017, p. 196–217.
- Dorsey, Dale, “Subjectivism without Desire” Philosophical Review, vol. 121, no 3, 2012, p. 407-442.
- Enoch, David, “Why Idealize?”, Ethics, vol. 115, no. 4, 2005, p. 759-787.
- Fletcher, Guy, “A Fresh Start for the Objective-List Theory of Well-Being”, Utilitas, vol. 25, no. 2, 2013, p. 206-220.
- Heathwood, Chris, “The Problem of Defective Desires”, Australasian Journal of Philosophy, vol. 83, no. 4, 2005, p. 487-504.
- Lewis, David, “Dispositional Theories of Value”, Proceedings of the Aristotelian Society, Supplementary Volumes, vol. 63, 1989, p. 89-174.
- Mendola, Joseph, “Real Desires and Well‐Being”, Philosophical Issues, vol. 19, no. 1, 2009, p. 148-165.
- Murphy, Mark C., “The Simple Desire‐Fulfillment Theory”, Noûs, vol. 33, no. 2, 1999, p. 247-272.
- Quinn, Warren, “Putting Rationality in its Place”, in R. G. Frey and C. Morris (eds.), Values, Welfare, and Morality, Cambridge, Cambridge University Press, 1993, p. 26-50.
- Rabinowicz, Wlodek et Österberg, Jan, “Value Based on Preferences”, Economics and Philosophy, vol. 12, no. 1, 1996, p. 1-27.
- Railton, Peter, “Moral Realism”, The Philosophical Review, vol. 95, no. 2, 1986a, p. 163-207.
- Railton, Peter, “Facts and Values”, Philosophical Topics, vol. 14, no. 2, 1986b, p. 5-31.
- Rawls, John, A Theory of Justice, Cambridge, Harvard University Press, 1971.
- Rosati, Connie S., “Persons, Perspectives, and Full Information Accounts of the Good”, Ethics, vol. 105, no. 2, 1995, p. 296-325.
- Sidgwick, Henri, The Methods of Ethics, 7th ed., Cambridge, Harvard University Press, 1907.
- Smith, Michael R., The Moral Problem, Hoboken, NJ., Blackwell Publishing, 1994.
- Smith, Michael R., “Internal Reasons”, Philosophy and Phenomenological Research, vol. 55, no. 1, 1995, p. 109-131.
- Sobel, David, “Subjectivism and Idealization”, Ethics, vol. 119, no. 2, 2009, p. 336-352.