Moral Worth and Moral Hobbies

This paper argues that one promising recent account of moral worth, Julia Markovits’s Coincident Reasons Thesis, does not afford us the ability to draw a traditional distinction between actions that are performed from immediate inclination and actions that possess genuine moral worth (“Kant’s distinction”). I argue that a right-making reasons account of moral worth such as Markovits’s can in fact accommodate Kant’s distinction. To make this possible, we must go beyond the consideration of instrumental reasons and equip a right-making reasons account with a new, more general, sort of reason: a conditional reason. After introducing the notion of a conditional reason, I argue that one class of these reasons, hedonically conditional reasons, are crucial for understanding the reasons for which we pursue hobbies. Kant’s figure who pursues morally right action from immediate inclination might be understood as a moral hobbyist whose actions plausibly lack moral worth. This grounds an objection to Markovits’s own account as a sufficient condition on morally worthy action.

 
Consider the case of Donna Goldstein, about whom The Week reported, on December 25, 2015, that she

Could we, with any plausibility, entertain the doubt that there might be something less than morally praiseworthy about Goldstein’s actions in this case? Immanuel Kant has urged that we must be scrupulous when it comes to the moral evaluation of the Donna Goldsteins of this world. Some people, he infamously warns, are “so sympathetically attuned that, without any other motive of vanity or self-interest they find an inner satisfaction in spreading joy around them and can take delight in the satisfaction of others so far as it is their own work. But,” Kant continues, “I assert that in such a case an action of this kind, however it may conform with duty and however amiable it may be, has nevertheless no true moral worth . . .” (1996: 4:398). Morally worthy action, according to Kant, is performed “not from inclination but from duty” (1996: 4:398).

Kant contrasts a figure who, through sympathy, joyfully assists others with a different sort of philanthropist who has fallen on hard times such that he can no longer muster any sympathy for the plight of his fellow man. Christine Korsgaard projects a number of possibilities for this latter figure: “Maybe his wife gets cancer, or his child runs away, or his work, which once seemed promising, comes to nothing. And there is no pleasure in anything for him anymore. He is absorbed in his own sorrows and has no sympathy for anyone else”(2008: 185). Kant claims of such a downcast philanthropist that it is only when “no longer incited to it by any inclination, he nevertheless tears himself out of this deadly insensibility and does the action without any inclination, simply from duty” that his action first has moral worth (1996: 4:398).

This passage has long been a source of scandal for Kantian moral theory, insofar as it suggests that Kant may hold that, morally speaking, enjoying one’s good deeds is itself a liability and that the paradigmatic moral agent is an insensible yet dutiful grumpus.[1] Many Kantians now recognize that this is not the point of Kant’s discussion.[2] My aim in this essay, however, is not to advance an exegetical thesis about how best to understand Kant’s claims. Instead I here consider, in its own right, the distinction Kant purports to draw between acting from immediate inclination and acting from duty. What exactly does this distinction amount to: how might we understand acting from immediate inclination in contrast to acting from duty? Throughout, I’ll refer to the distinction between what we might characterize as acting from duty and acting from immediate inclination as Kant’s distinction.[3]

In particular, I wish to pursue the status of Kant’s distinction from the perspective of one promising contemporary account of moral worth: a right-making reasons account such as that advanced by Julia Markovits.[4] At a first pass, a right-making reasons account of moral worth claims that a morally right action has moral worth just in case the agent’s reasons for acting are the same as the reasons in virtue of which the action is indeed the morally right action to perform. In such a case, the agent’s reason for acting is not the fact that the action is morally right. Instead, the reasons for which the agent performs the action are the same as the reasons that make the action morally right.

My question is, does a right-making reasons account of moral worth allow us room to draw Kant’s distinction? And if so, what shape might that distinction take on a right-making reasons account? Put another way, if we dispense with the idea that in order for an action to possess moral worth an agent must act for a reason with moral content, can we nevertheless mark a difference between the sorts of reasons that distinguish an agent who acts from duty and one who acts from immediate inclination? I’ll argue that while Markovits’s right-making reasons theory cannot draw Kant’s distinction, there is, nevertheless, conceptual space to mark this distinction on a right-making reasons account. In order to do so, we need to recognize a new sort of a reason: a conditional reason. An agent’s reason is a conditional reason when the reason’s being the reason that it is depends upon some further fact obtaining, where this further fact is itself one of the agent’s reasons and serves to justify the agent’s taking of the conditional reason as the reason that it is.

In Section 1, I introduce the notion of a conditional reason. In Section 2, I focus on a class of conditional reasons that will prove important for drawing Kant’s distinction: noninstrumental reasons that are conditional on the agent’s pleasures, satisfactions, enjoyments, etc. I call this the class of hedonically conditional reasons. I suggest that we should understand hobbies in terms of hedonically conditional reasons. In Section 3, I argue that morally right actions that are performed for hedonically conditional reasons lack moral worth because, in such cases, morality is treated as a hobby.[5] This grounds an objection to Markovits’s Coincident Reasons Thesis, since it does not exclude the actions of the moral hobbyist from having moral worth. I suggest that we can draw Kant’s distinction as follows: the person who acts from immediate inclination (unlike the person whose action possesses genuine moral worth) treats morality as a hobby. He acts for a hedonically conditional reason. Finally, in Section 4, I consider and respond to the objections first that all and then that none of our reasons are conditional upon what we find pleasant.

1. Instrumental Reasons as a Species of Conditional Reasons

In this section, I introduce the notion of a conditional reason. I begin, however, with the distinction between instrumental and non-instrumental reasons upon which a full articulation of Markovits’s right-making reasons account relies. I argue that the category of instrumental reason is a species of a broader genus: the conditional reason. I go on to claim that, although Markovits’s account fails to accommodate Kant’s distinction, a right-making reasons account of moral worth equipped with the notion of a conditional reason is in a position to draw Kant’s distinction in a satisfying way.

According to Markovits, the Coincident Reasons Thesis states that, “my action is morally worthy if and only if my motivating reasons for acting coincide with the reasons morally justifying the action—that is, if and only if I perform the action I morally ought to perform, for the (normative) reasons why it morally ought to be performed” (2010: 205). I’ll follow Markovits in taking both normative reasons and motivating reasons to be facts:

To sum up, according to the Coincident Reasons Thesis, when an action has moral worth, the same fact serves as the justifying and the motivating reason for the action. And this fact justifies the action morally.[6]

Yet this sort of coincidence is hardly enough to suggest that an action has moral worth, for we can ask, to what extent must the motivating and the morally justifying reasons for some action overlap? To see the problem, consider the case of Cash, whose reason for jumping into the water is that a child is drowning. But there is more to the story of Cash. His reason for saving the child is that he’ll receive a reward. Cash has ulterior motives—not only does he take as his motivating reason the fact that morally justifies the action, but he takes other, morally irrelevant, facts as motivating reasons as well. Motivating reasons can form chains. In Cash’s case the “chain of motivating reasons diverges in its more fundamental links from the chain of normative moral reasons” (Markovits 2010: 227). In order for the Coincident Reasons Thesis to provide a plausible account of moral worth, therefore, Markovits needs to specify to what extent the chain of motivating reasons must overlap with the chain of morally justifying reasons.

In order to achieve this, Markovits introduces the distinction between instrumental and noninstrumental reasons. A noninstrumental reason is a reason that is “provided by the ends of our actions that are worth pursuing for their own sake (in the case of justifying reasons) or that we pursue for their own sake (in the case of motivating reasons)” (Markovits 2012: 292). Instrumental reasons, by contrast, “are reasons provided by features of our actions (or their outcomes) that have instrumental value (or that we value instrumentally)” (Markovits 2012: 292). In these terms, the reason that Cash’s action of going into the water to save the drowning child lacks moral worth is that his noninstrumental motivating reason is that he will win a reward, while the noninstrumental justifying reason that morally justifies the action is simply that the child is drowning. A fuller statement of the Coincident Reasons Thesis, then, says that an action is “morally worthy whenever the noninstrumental reasons for which it is performed coincide with the noninstrumental reasons that morally justify its performance” (Markovits 2010: 230). If thus construed as a matter of noninstrumental reasons, the Coincident Reasons Thesis gains plausibility. This is because such an interpretation excludes actions performed from ulterior motives (such as Cash’s) from the category of the morally worthy. It also does not count actions in which an agent values for its own sake what is in fact only instrumentally valuable, morally speaking, as morally worthy. From the point of view of moral worth, what matters about the agent is that “she cares noninstrumentally about noninstrumentally valuable things” (Markovits 2012: 293).

One might be tempted to think that the same fact cannot serve as both an instrumental and a noninstrumental reason. But a moment’s reflection on the sort of case just discussed reveals that this is false. Supposing that the fact that saving the child will win Cash a reward really is a justifying reason for saving him, the fact that the child is drowning is both an instrumental and a noninstrumental justifying reason for Cash to jump in. Cases such as Cash’s, where the same fact serves as both an instrumental and a noninstrumental justifying reason, are crucial to the worries that arise about the moral worth of actions performed from ulterior motives. The question becomes, given that the same fact is a noninstrumental and an instrumental justifying reason, does it serve as a noninstrumental or an instrumental motivating reason for Cash? One way to say why Cash’s action lacks moral worth is that he does the right thing for the wrong reason. But if we think of reasons as facts, both Cash and the person whose action would have moral worth take the same fact as a reason. Markovits’s point is that he nevertheless doesn’t act for the same reason in the sense that this fact, considered as Cash’s motivating reason, is an instrumental reason. By contrast, the person whose action possesses moral worth acts for a noninstrumental reason, although this same fact may also be her reason.

This line of thought suggests that although reasons are facts, the same fact can constitute multiple reasons. How? The case of instrumental and noninstrumental reasons illustrates a straightforward answer: the justificatory structure in which a fact is situated constitutes that fact as the reason that it is. If we want to know whether two people acted for the same reason, it isn’t enough to look at an isolated fact that can be cited as a reason. That fact is the reason that it is on the basis of its standing in a particular web of justification with other facts. This comes out starkly in the case of the same fact constituting reasons of different kinds, e.g., instrumental and noninstrumental. But the same fact could constitute multiple reasons all of which are, say, instrumental reasons. Cash and Valor may both jump in the water because the child is drowning, but Cash’s reason is not the same as Valor’s if Cash is ultimately interested in the reward and Valor is interested in getting his name in the paper. This is a matter of Cash’s and Valor’s motivating reasons. As for justifying reasons, we can say that the same fact simply constitutes both justifying reasons.

In the rest of this section, I develop the notion of a conditional reason, and argue that instrumental reasons can be seen as a species of this larger genus. I use this notion of a conditional reason to argue that Markovits has not provided a sufficient condition on an action’s possessing moral worth. Some actions that meet the condition provided by the Coincident Reasons Thesis, may, I argue, nevertheless be performed for noninstrumental reasons that are conditional upon hedonic facts about the agent. At least some of these actions lack moral worth.

I propose that we understand conditional reasons as follows.

This specifies what it is for an agent’s motivating reason to be conditional. We can specify what it is for a justifying reason to be conditional in a similar way: a justifying reason is a conditional reason when that fact’s status as the justifying reason that it is is conditional upon some further fact, where the further fact itself is a justifying reason and serves to justify the status of the conditional reason as the reason that it is.

The category of conditional reasons includes instrumental reasons. Suppose, for instance, that Clark is chopping lemons because he is making lemonade. Among Clark’s reasons for chopping the lemons is the fact that one needs chopped lemons for lemonade. But this fact (that one needs chopped lemons for lemonade) serves as the reason that it does for Clark on the condition that he is making lemonade. This fact (that he is making lemonade) is itself one of his reasons for chopping lemons and serves to justify why the fact that one needs chopped lemons for lemonade is the reason that it is for him to chop them. The fact that one needs chopped lemons for lemonade is an instrumental reason for Clark, but this also renders it a conditional reason in the sense I have indicated above.

It is important that in order for a reason to be a conditional reason in the sense that interests me, the further fact that conditions the agent’s taking of some fact as that reason is itself one of the agent’s reasons and stands in a justificatory relationship to the conditional reason in question. In any particular case, an agent’s taking some fact to be a reason is itself brutely conditional upon any number of other facts, e.g., that the agent’s heart is still circulating blood around her body. Yet in most cases, the fact that an agent’s heart is circulating blood around her body will not be one of her reasons for taking any further facts as a reason. There may be many brute conditionality relationships that hold with respect to our reasons as a matter of psychology, physiology, chemistry, physics, etc. But these conditionality relationships don’t thereby render the reasons they condition conditional in the sense of Conditional Reason.[7]

Although the category of conditional reasons includes instrumental reasons, we need not think of it as exhausted by these. Some noninstrumental reasons will be conditional reasons. This occurs when a person values something for its own sake, yet on the condition that some further fact obtains, where this further fact provides the agent’s reason for valuing what she does noninstrumentally. Aristotle characterizes the virtuous agent as someone who does the right thing “to the right person, in the right amount, at the right time, with the right aim in view, and in the right way . . . ” (2000: 1109a). The agent’s noninstrumental reasons are provided by the aim that she has in view, that for the sake of which she acts. But if Aristotle is right, it isn’t enough for virtuous action that an agent have the right aim in acting—she must pursue this aim in relation to the right people, at the right time, to the right extent and in the right manner. The virtuous agent must be aware of the various facts of her circumstances that make pursing some particular aim appropriate in that situation—on account of her virtue, she takes these variables into consideration in choosing to act. These facts serve as the virtuous agent’s reasons, not because they explain what further aim she is pursuing in acting as she does, but by situating her aim with respect to a number of facts that condition the appropriateness of its pursuit. These facts too are reasons, considerations that speak in favor of acting as she does.[8]

2. Hedonically Conditional Reasons and Hobbies

With the concept of a conditional reason in hand, we’re now in a position to observe that when it comes to moral worth, it matters not just that the agent act for the right noninstrumental reasons (that she have the right aim in view), but also that these noninstrumental reasons serve as her reasons conditionally upon the right reasons. This shows up in a positive respect when thinking about Aristotle’s virtuous agent—her pursuit of some aim is conditional upon further considerations that serve as her reasons with respect to the appropriateness of pursuing that aim under the circumstances. But an agent’s reasons may also be conditional upon further considerations in ways that render the action less than morally worthy.

In Section 3, I’ll argue that the notion of a conditional reason can ground an objection to Markovits’s Coincident Reasons Thesis as a sufficient condition on morally worthy action. There I’ll make the case that the person who helps another for the sake of the other acts from inclination (and therefore performs an action lacking in moral worth) when the agent’s noninstrumental (motivating) reason is conditional on the fact that she finds helping the other (for the other’s own sake) to be satisfying, pleasant, enjoyable, etc. In this section, however, I first explore the sort of conditional reason that will be crucial for making this objection to the Coincident Reasons Thesis. I’ll call it a hedonically conditional reason. A hedonically conditional reason is a noninstrumental reason that is conditional on the agent’s pleasures, satisfactions, enjoyments, etc.

In the rest of this section, I provide a few examples to illustrate the category of hedonically conditional reasons. These examples are not particularly morally fraught. My aim in discussing them is simply to bring this sort of reason into view. I suggest that hedonically conditional reasons are characteristic of the pursuits that we think of as hobbies.

Suppose that Hazel asks Nelly, “Why did you plant that water oak in your yard?” Nelly might reply, “Because it had been pulled up and its roots were drying out.” This gives Nelly’s motivating reason. Nelly is, I think, plausibly justified in having planted the tree because its roots were drying out. So this also serves as a justifying reason for what she does.

But what sort of a (motivating and justifying) reason is it? Recalling the discussion from Section 1, the same fact can constitute multiple reasons, depending on the various justificatory webs in which such a fact is situated. The fact that the water oak’s roots were drying out might, for instance, be an instrumental reason. The plant might be, although not an opportunity for ornament, a means to secure the eroding soil on a bank. In this case, the fact that the oak’s roots were drying out would be an instrumental justifying reason to save the tree (in light of the further end of planting it and securing the soil), and were this Nelly’s reason, it would be an instrumental motivating reason. There might also be a noninstrumental reason to save the oak that is not hedonically conditional.[9]

These possibilities, however, are not exhaustive. Suppose Hazel were to press, “Well, yes, Nelly, the plant was going to die if its roots dried out. But why, Nelly, why did you save it?” Nelly might reply, “I just sympathized with its plight. It therefore brought me pleasure to save it and satisfaction to watch it now grow and flourish right here in my own yard.”

There are at least two ways that we might interpret this response. One way to interpret this answer is that Nelly is ultimately self-interested: she is engaged in a project of satisfying herself. Satisfying herself is her ultimate purpose in saving the water oak. Were this to be the case, the fact that the tree’s roots were drying out would turn out to be an instrumental reason for saving the tree. I will call a fact such as this an instrumental hedonic reason. If the life of the tree is a means to Nelly’s pleasure, then saving the tree would be a means for achieving the end of self-satisfaction.

A second way to interpret Nelly’s response is that it indicates that she acts for a hedonically conditional reason. In order to see this second possibility, notice that when an agent’s reason for performing an action is merely instrumental, then, roughly speaking, you would expect the agent to show little interest in the means if that means were no longer to bring about the end, but to retain an interest in the end, even if the purported means were to fail to bring it about. Imagine that Hazel, who fancies herself no novice when it comes to distinguishing instrumental from noninstrumental reasons, asks Nelly, “Supposing I give you a hard choice. Either you can believe that the tree is dead, when the tree is in fact flourishing, or you can believe that the tree is flourishing when in fact the tree is dead.” Nelly replies, “Wow, Hazel, that’s a devious question. But hands down I’d rather have the tree carry on flourishing, even if I had to believe it to be dead, as much as that would make me melancholy.”

If we imagine Nelly in this way, it strongly suggests that she is not engaged in the project of satisfying herself, and that the fact that the tree’s roots are drying out is not, for her, an instrumental reason. For if her end were her own satisfaction, why would she choose to take the horn of Hazel’s dilemma that would bring her comparatively so little satisfaction? Rather, she values the tree noninstrumentally. She values the life of the water oak for its own sake and not for the sake of the enjoyment, satisfaction, pleasure, etc., that rescuing it and watching it flourish brings to her. The fact that the tree’s roots were drying out is, for Nelly, a noninstrumental motivating reason—it is a reason that is provided by the end of her action that she pursues for its own sake. And insofar as she is justified in acting for this reason, it’s also a noninstrumental justifying reason.

Nevertheless, although Nelly’s reason is noninstrumental, it is conditional on the fact that she would find satisfaction in saving the tree. Among her reasons for saving the oak is the fact that saving it would bring her satisfaction because she sympathized with its plight. And, correspondingly, the fact that she sympathized with the plight of the tree is also among the justifying reasons that she has for saving it. On this way of understanding the case, the fact that the tree’s roots were drying out is a noninstrumental, hedonically conditional reason.[10] The reason is noninstrumental because it is provided by something that Nelly values noninstrumentally—the life of the oak. But it remains a conditional reason. Nelly has the noninstrumental reason that she has in virtue of the further consideration that she finds satisfaction in saving the oak.

The category of hedonically conditional reasons draws attention to the fact that our sympathies, our pleasures, our satisfactions, and what we enjoy can provide the conditions that rationalize the pursuit of other ends for their own sake. As Korsgaard points out, “Activities of various kinds might be thought to be good under the condition that we enjoy them and not good at all for those who, for one reason or another, cannot enjoy them, without forcing the conclusion that it is only for the sake of the enjoyment that they are valued” (1996: 264).

If there are hedonically conditional reasons, these are likely to be widespread. It seems clear that facts about what would bring us pleasure can serve as reasons. Once we have hedonically conditional reasons on the table, we can see, however, that the cases in which pleasure serves as the agent’s aim may be more circumscribed than we originally thought.[11] On occasion we may lie in the sun, get a massage, or eat a rich dessert in order to achieve the pleasure it affords us. But not all invocations of facts about pleasure as reasons for acting need be understood in this way. To suggest the sorts of reasons that might fall into this category, I provide two more examples. Generalizing on these cases, I claim that hedonically conditional reasons are essential for understanding the justificatory role that facts about pleasure play in the pursuit of hobbies.

Plausibly, Samantha and Carter each have a noninstrumental reason for acting that is conditional on certain of their own joys and satisfactions. In Samantha’s case, the noninstrumental reason that she is building the pumpkin trebuchet is because this sort of machine can hurl a pumpkin more than 1000 feet. Samantha is building the machine for its own sake. She isn’t building it for the sake of the satisfaction it will bring her to build it or to see it hurl a pumpkin on the day of the contest. Samantha needn’t be engaged in the further project of satisfying herself in constructing the machine. We can suppose of Samantha that given the choice between on the one hand, building the machine and having it hurl the pumpkin 1000 feet while believing that the machine had been destroyed or, on the other hand, believing that the machine had been constructed in such a way so as to successfully hurl the pumpkin while in fact it was not, Samantha would choose the former. For Samantha, the machine is important in its own right. Samantha’s reason for building the machine—that it can hurl a pumpkin more than 1000 feet—is, however, conditional.

When asked why she’s doing this, we can suppose that she would respond, “I just find the idea of it so enchanting.” In doing so, Samantha is giving us a reason why she is building it—a consideration that speaks in favor of doing so. And her noninstrumental reason (both motivating and justifying) for doing so (that it can hurl a pumpkin 1000 feet) is conditional upon the further reason (both motivating and justifying) that she finds the idea of such a machine enchanting.

Carter’s noninstrumental reason for placing the spirit baskets is because that improves the chance of the home team’s win. He values the home team’s win for its own sake. He doesn’t value it as part of his project of satisfying himself. Again, evidence of this is that given the choice between on the one hand, believing that the team lost while in fact the team won and, on the other hand, believing that the team won while in fact the team lost, he, a true fan, would choose the former. Yet Carter’s reason for helping the home team win is conditional, we can suppose, on his sympathy with the home team and his taking pleasure in its victory.

These examples suggest that we might think of hedonically conditional reasons—reasons that are conditional on further reasons to do with the agent’s satisfactions, pleasures, enjoyments, etc., as hobbyist reasons. In making this claim, I’m not claiming that all conditional reasons are hobbyist reasons. In particular, I’m not claiming that someone whose reason is conditional upon the further fact that, e.g., she is committed to some project, loves some cause, or cares about some person, is thereby pursuing something in the style of a hobby. The category of hedonically conditional reasons deals with a more circumscribed body of facts about what the agent finds, e.g., pleasant, enjoyable, satisfying, etc.[12]

By invoking hobbies, I don’t mean to suggest that any pursuit that is optional or any that isn’t moral must be a hobby. Reflecting on imperfect duties (or supererogation) should make us aware that the moral and the rationally compulsory (non-optional) are not coextensive, e.g., particular acts of beneficence may be rationally optional. But there are justifying reasons for performing such beneficent actions that don’t exhibit the structure of hobbyist reasons—these reasons for helping are not conditional upon further reasons to do with the satisfactions of the helper. The justifying reasons for helping have to do with the needs of the person who is helped and the claim that these needs make on others regardless of their sympathies. Moreover, many actions which, arguably, don’t fall within the scope of morality are, arguably, justified by reasons that don’t exhibit a hobbyist structure. Consider a woman who pursues a career in pure mathematics. Such a career is, we might suppose, both rationally optional and not supported by reasons that are moral in nature. The noninstrumental justifying reasons for dedicating her career to pure mathematics, reasons provided by the fact that knowledge and understanding are valuable for their own sake, are not dependent upon further justifying reasons to do with the mathematician’s satisfactions. Such a career, while rationally optional and not moral in nature, is not a hobby.[13]

The mark of a hobby in the sense that I’m interested in here is not, therefore, its contrast with morality or its optionality. Instead, the reasons we have for pursuing hobbies exhibit the two features I have been discussing. First, they are noninstrumental reasons. Although we may pursue hobbies with the further purpose in mind to enjoy ourselves, to pass the time, or to relax, we may, by contrast, pursue hobbies for the sake of the ends that these hobbies put us in touch with. In fact, hobbies’ potency for inducing relaxation and enjoyment may often depend on this feature, that they direct us toward ends beyond our own satisfaction. Hobbies provide us, therefore, with noninstrumental reasons. But hobbies, even when pursued with seriousness and passion, to some extent carry within themselves, in their own self-consciousness, a sense of their triviality—think of the true sports fan or the stamp collector. This is because the noninstrumental reasons provided by our hobbies are conditional upon further reasons to do with our own particular satisfactions and enjoyments. Our hobbies, therefore, in spite of the fact that they allow us to care about things well beyond ourselves, struggle to be any more grave or deep than our own brute satisfactions.[14]

3. An Objection to the Coincident Reasons Thesis

The Coincident Reasons Thesis is tailored to allow us to discriminate morally worthy action from action performed from an ulterior (morally irrelevant) motive. When an agent acts with an ulterior motive, she does not take the noninstrumental reason that morally justifies the action as her noninstrumental reason. She has a further aim or purpose and takes this reason as her own only instrumentally. She is, we could say, up to something else that the agent whose action has moral worth simply isn’t up to.

According to Kant, however, there is another case of action that we should hope to distinguish from morally worthy action—action performed from immediate inclination. This task is, however, more subtle. He writes, “it is easy to distinguish whether an action in conformity with duty is done from duty or from a self-seeking purpose. It is much more difficult to note this distinction when an action conforms with duty and the subject has, besides, an immediate inclination to it” (1996: 4:397). In this section, I argue that one way to understand the sorts of actions that Kant describes as performed from “inclination” is that they are performed for hedonically conditional reasons, and that plausibly these actions are, as Kant suggests, lacking in moral worth. This grounds an objection to Markovits’s Coincident Reasons Thesis as a sufficient condition on morally worthy action. I make a proposal for how a right-making reasons account of moral worth could, nevertheless, be modified so as to accommodate Kant’s distinction. The case that Kant uses to illustrate this distinction is, as we have seen, the case of the joyful do-gooder. She is “sympathetically attuned” and “finds an inner satisfaction in spreading joy around” herself (1996: 4:398). What could possibly be amiss with such a person according to a right-making reasons account such that her beneficent actions might lack moral worth? Before turning to the positive account of the example I offer below, it’s worth being clear about exactly why the Coincident Reasons Thesis as it stands won’t allow us to draw Kant’s distinction.

Supposing that the Coincident Reasons Thesis is correct, one possibility that we can rule out is the view, often attributed to Kant, that the person whose action possesses moral worth acts for the reason that “this is the morally right thing to do.” Barbara Herman, for instance, interprets Kant as saying that “For an action to have moral worth, moral considerations must determine how the agent conceives of his action (he understands his action to be what morality requires), and this conception of his action must then determine what he does” (1981: 375). On this view, the agent whose action possesses moral worth is one whose “maxim of action has moral content” (1981: 371).[15] Paulina Sliwa has recently defended a view along these lines in its own right, not as an interpretation of Kant, which she calls the Rightness Condition: “A morally right action has moral worth if and only if it is motivated by concern for doing what’s right (conative requirement) and by knowledge that it is the right thing to do (knowledge requirement)” (2015a: 2).

These sorts of views would indeed allow one to discriminate between actions that possess genuine moral worth (the motivating reasons for which would include moral content) and actions performed from immediate inclination (the motivating reasons for which would lack such content). According to the Coincident Reasons Thesis, by contrast, even in the case of morally worthy action, the agent’s reasons need not possess moral content. As Markovits puts it, on her view, “My motivating reason for performing some action [in the case of morally worthy action] will not be the duty-based reason ‘that the moral law requires it’ but the reasons for which the moral law requires it” (2010: 205).

It wouldn’t be impossible both to hold the Coincident Reasons Thesis and to claim the person who acts in a morally worthy way must act for the reason that the action is morally required. But this would require one also to subscribe to the view that the fact that some action is morally required is itself a normative reason for performing the action. Markovits rejects this possibility:

When we say that the action of saving the child is morally right we are more plausibly reporting the fact that there are (other) reasons that speak in favor of saving the child than attempting to record a reason that speaks in favor of doing so.

A better way to draw Kant’s distinction given the tools of the Coincident Reasons Thesis would be to claim that acting from immediate inclination is a matter of acting for the reason that “I find satisfaction in helping others.” Suppose we were to ask Goldstein why she was giving away toys to poor children and she were to reply, “Because it brings me a great deal of satisfaction to do so” in addition to saying something such as, “Because these children had no presents for Christmas.” Perhaps the first sort of answer (were it true) would indicate that she was acting from immediate inclination, while the second sort of answer on its own (were it true) would indicate that she was acting from duty.

Recall that according to the Coincident Reasons Thesis, an action is “morally worthy whenever the noninstrumental reasons for which it is performed coincide with the noninstrumental reasons that morally justify its performance” (Markovits 2010: 230). In order to show that Goldstein’s action lacks moral worth, the Coincident Reasons Thesis would, therefore, have to interpret the answer, “Because it brings me a great deal of satisfaction to do so,” as demonstrating that Goldstein doesn’t act for the noninstrumental reason that morally justifies the action. Helping the children would thus have to be merely instrumental in Goldstein’s project of satisfying herself.

My claim is, therefore, that the only resources that the proponent of the Coincident Reasons Thesis has to accommodate Kant’s claim that actions performed from immediate inclination lack moral worth is to classify such actions as performed for (what I called in Section 2 above) instrumental hedonic reasons. On this approach, acting from immediate inclination would be a subspecies of acting from an ulterior motive in which the further purpose would be the project of bringing about some kind of pleasure, enjoyment, or satisfaction on the part of the agent. As Markovits points out, “people who act rightly just because it makes them feel good or just because failing to do so makes them feel bad act for reasons that are fundamentally selfish—self-directed—and not (significantly) morally relevant” (2010: 224).

While it seems plausible that actions performed for instrumental hedonic reasons lack moral worth, I doubt that this can be quite what Kant had in mind when discussing immediate inclination, because he describes the sympathetic person who acts from immediate inclination as acting without any “motive of vanity or self-interest” (1996: 4:398). Moreover, Kant draws from his discussion the lesson that we can’t distinguish between the person who acts from inclination and the person who acts from duty on the basis of their purposes: “an action from duty has its moral worth not in the purpose to be attained by it” (1996: 4:399). If the person that Kant intended to pick out as acting from immediate inclination had as her aim to satisfy herself, she would have a different purpose than the person whose action possesses moral worth. Korsgaard points out that Kant sees his two figures as sharing the same purpose: “a person who does a beneficent action from immediate inclination and a person who does one from duty have the same purpose—namely, to help someone” (2008: 178).

The Coincident Reasons Thesis, then, cannot accommodate Kant’s distinction. On the one hand, if we claim that the person who acts from duty acts for a reason with moral content (while the person who acts from inclination does not), this is not in keeping with the spirit of a right-making reasons account of moral worth. On the other hand, if we claim that the person who acts from inclination has her own satisfaction as her aim (while the person who acts from duty does not), we abandon the idea that the two agents share the same purpose and that the difference between them lies elsewhere.

I propose that the category of hedonically conditional reason is a better tool for capturing what Kant describes as acting from inclination. Thus, in order for a right-making reasons account of moral worth to accommodate Kant’s distinction, we need to equip it with the resources to describe a hedonically conditional reason. Here, therefore, is my suggestion for an additional condition on morally worthy action: If a morally right action possesses moral worth then “the noninstrumental reasons for which it is performed coincide with the noninstrumental reasons that morally justify its performance” (Markovits 2010: 230) and the agent’s motivating reasons are conditional only upon those reasons that the morally justifying reasons are conditional upon.[16]

Like any right-making reasons account of moral worth, applying this theory in a particular case to make a proper judgment of moral worth requires that one know what the morally right action is and what reasons morally justify it. Claiming that certain actions lack moral worth, then, requires us to make substantive judgments on these matters. Conditional reasons are a powerful tool because they allow us to recognize more fine-grained distinctions between the sorts of reasons that agents have and the sorts of reasons that can morally justify actions. Given the account of conditional reasons developed above, and, in particular, the account of hedonically conditional reasons, we can identify an interesting sort of moral failure that would allow us to draw Kant’s distinction.

Let us return to the case of Donna Goldstein and apply this condition of moral worth to her case. Perhaps unsurprisingly, the description we’ve been given of her so far doesn’t determine whether her action is performed from immediate inclination. She tells us that she feels good to be doing what she can to help needy children and that doing so makes her very happy, but she doesn’t tell us that this is why she is doing it. We don’t know whether the fact that helping the children brings her joy is among her reasons for helping them. Imagine that our reporter had asked her, “So I see that helping these needy children brings you great joy. Is that why you’re doing it?” She might have replied, “Heavens no! I’m doing it for the children, so that they can be happy on Christmas morning and can have some of the things that they need.” Suppose our reporter, scrupulously interested in making fine discriminations when it comes to moral worth, were to press, “I know you’re doing it for the sake of the children. But why do you have reason to do something for their sake? Just because you enjoy it?” We might imagine Goldstein replying, “Not at all. I’m not helping them because I enjoy it.” Were this answer truthful, her noninstrumental reasons for acting (that giving Christmas gifts will make the children happy) would not be conditional on finding satisfaction in the well-being of the children. In this sort of case, Goldstein would be acting in a morally worthy way.

Suppose that Goldstein’s relationship to the children were instead like Nelly’s relationship to the water oak. Suppose that she values the children for their own sake (not instrumentally), but only conditionally upon the further reason that their well being brings her satisfaction. Although she doesn’t help them in this case in order to feel good, among her reasons for helping is the fact that she enjoys helping them. In such a case, when asked, “I see that you are helping the children for their own sake, but why do you have reason to do something for the sake of the children? Because you enjoy it?” She would say, “Yes—I’m helping them because I find satisfaction in their well being.” In this sort of case, where Goldstein is offering a further hedonic justification for helping the children for their own sake, Goldstein’s actions would lack moral worth. Helping the children would be, for her, a sort of hobby.

Is Goldstein thus described an exemplar of a recognizable form of moral shortcoming? The moral hobbyist does strike me as a stock psychological profile. Although such people are not self-interested, in the sense of aiming at their own well being or even their own good feelings, there is a sense in which their justificatory appeal to what feels good to them lends their seeming moral pursuits an air of triviality. Facts about their own delight and their own satisfaction, to return to Kant’s description, feature among the reasons that such people do what they do. Were such people to cease to take pleasure in the things that they do, as Kant imagines with the downcast philanthropist, they would no longer be able to act for the reasons that they do.[17] Like the person who acts from ulterior motives, therefore, the moral hobbyist’s actions are only accidentally morally right, not a reflection of the fact that she is acting from the reasons that morally justify the action.

Let me emphasize the difference between this account of Kant’s distinction and another account of the distinction (with which it might be confused), one which we might call a dispositional account. On the dispositional account of Kant’s distinction, the person who acts from inclination performs the helping action only on the condition that she enjoys helping. The person who acts from duty performs the helping action regardless of whether she enjoys helping, so her action is unconditioned by her own enjoyment. So a person who performs a morally right action acts from duty, on such an account, if and only if the performance of the action isn’t conditional upon the agent enjoying it. According to such a dispositional account, what is conditioned in the case of the person who acts from inclination is not her reason for acting (as on my proposal for Kant’s distinction), but the action itself.

The dispositional account of Kant’s distinction is problematic because it counts some actions that are intuitively performed from inclination as being performed in such a way that they possess moral worth and vice versa. Consider the case as Kant describes it: a person performs helping actions from immediate inclination, and then, through some misfortune, is rendered insensible to the suffering of others. Nevertheless, the person carries on helping others, now, rather than from inclination, in such a way that the action is morally worthy. According to the the dispositional account, Kant’s description of the case according to which the person first acts from immediate inclination and then later in a morally worthy way must be mistaken. This is because even at first the person is such that were he not to enjoy helping he would nevertheless help. This is precisely what is demonstrated by the fact that he carries on helping even though he stops enjoying it. And so his performance of the action is not conditioned by his enjoyment, even in the first instance. Therefore, on the dispositional account, he never acted from inclination. Yet this misses the point of Kant’s story. Intuitively, the person Kant describes is acting for two different sorts of reasons—first from inclination and then in a morally worthy way. So the dispositional account counts the person who seems to be acting from inclination as in fact acting in a morally worthy way. To capture the point about moral worth that we’re after we need to look, not at counterfactual statements about what the agent would do under different circumstances, but at the justificatory web in which his reason is situated in his actual circumstances.

On the other hand, the dispositional account also will count some people whose actions, intuitively, possess moral worth as acting from inclination. Suppose we imagine someone who helps another simply because that person needs help. The fact that she enjoys helping the person isn’t among her reasons for doing it. It seems perfectly plausible to assume that such a person’s action is morally worthy. But suppose that it is true of her that were she to have become depressed, she wouldn’t even have gotten out of bed. She wouldn’t, under those circumstances, have helped the person. On the dispositional account, this counterfactual renders it true even in the present circumstances in which she can get out of bed that her action is lacking in moral worth. This seems needlessly ungenerous.[18] Everyone suffers from practical irrationality at various times—weakness of will, sloth, etc. But why think that the fact that this happens—or might happen—to a person need cast aspersions on those cases when a person does what is required by duty for the right reasons? If it needs to be true of a person, in order for her action to possess moral worth, that no matter what temptations she faces or how terrible she is feeling, she would nevertheless perform the action in question, then it threatens to make morally worthy action impossible for imperfect beings like us. Again, capturing the point about moral worth requires us to attend, not to whether she would act according to the reasons that she has under different circumstances, but to the reasons that she acts upon in the actual circumstances. The dispositional account does not offer a plausible understanding of Kant’s distinction. But the view I’ve put forward here does not commit one to claims about what the agent would or would not have done based on counterfactuals about her psychology. Rather, it draws the distinction between acting from inclination and morally worthy action on the basis of whether the agent’s reasons are rationally conditioned by other reasons to do with the agent’s satisfactions, pleasures, and enjoyments.

We are now in a position to see why Kant’s distinction might be intuitively important. One way to think about what it is for an action to be justified morally is to see that action as a matter of rendering what is owed to oneself or another in virtue of the sorts of beings that we are. When we act for moral reasons, we recognize the value in ourselves and in others that is independent of our own satisfactions, pleasures, and enjoyments. To act “from duty,” then, is a matter not only of valuing others and myself noninstrumentally, but is also a matter of taking that noninstrumental value not to be conditioned by what happens to satisfy me. To act in accordance with morality “from inclination,” then, is a matter of treating our actions that are performed from a recognition of the noninstrumental value of ourselves and others like a hobby. When we pursue our hobbies we can act for the sake of sorts of things that our hobbies put us in touch with. Still, even when we act for the sake of the plant, the pumpkin chucking trebuchet, or the home team, it is with an awareness that our own satisfactions are ultimately the source of the (noninstrumental) value that we find there. We are self-consciously the source of the value we find in these projects because of the pleasure we take in them.

The reason that morally right actions performed from immediate inclination lack moral worth, then, is because they amount to treating our moral concern for ourselves and others as a hobby. There may be genuine reasons to treat our interest in ourselves and others as we would treat hobbies, just as there may be genuine instrumental reasons to save a drowning child (e.g., in order to get one’s name in the paper). As we have seen, there may be many reasons, and many sorts of reasons, to perform some particular action. And the same fact may even be situated in various justificatory structures so as to constitute different sorts of reasons to perform that action. The intuitive idea behind a right-making reasons account of moral worth is that an action has moral worth when the agent performs the action for the right reason, that is, for the reason that morally justifies the action. One way to act for the wrong reason is, ultimately, to have the wrong purpose, to act from an ulterior motive. There is, however, another way in which morally right actions can lack moral worth. Performing an action for the sake of a person as one would perform an action for the sake of one of the ends of one’s hobbies is not a matter of performing an action with moral worth.

4. Reasons Conditional on Our Satisfactions: All or Nothing?

One might think that in order to evaluate the claims that I have made about the existence of hedonically conditional reasons, we would need to settle a meta-normative dispute about the role of pleasure in having a reason. In fact, this isn’t the case. In this section, I explain why this is so: the claims I have made about hedonically conditional reasons are substantive normative claims that should be compatible with whatever turns out to be the most compelling meta-normative account of reasons.

Consider the following account of reasons which we might call the “pleasure theory”:

One might think that if the pleasure theory of reasons is correct, then all of our justifying reasons are in fact conditional on our pleasures. If this were so, then it would be impossible for me to draw a distinction between certain justifying reasons that are conditional upon our pleasures and others that are not. One might also object to my drawing such a distinction if, on the other hand, one were to think that the pleasure theory of reasons is false because a person’s pleasures are always irrelevant to the justifying reasons that the person has so that none of our justifying reasons are conditional on our pleasures. One might take this position if one were to hold that finding something pleasant usually involves “thinking, or imagining, that there is something to be said for doing it. And from the point of view of the agent it is the considerations that are seen as” the grounds for finding something pleasurable, rather than the fact of the pleasure, that provides reasons for action.[20]

In order to evaluate the merits of these objections, however, we need to be clearer on the register in which we should hear the pleasure theory of reasons. Such a theory of reasons is ambiguous. It can be read, on the one hand, as a normative claim about what reasons we actually have. It can be read, on the other hand, as a reductive meta-normative claim about what constitutes a reason. When read in the first way, as a normative claim,

In my discussion of moral worth above, I’ve relied on the intuition that, considered as a normative view, the pleasure theory is false. For instance, I don’t think that in order for Goldstein to have a reason for buying Christmas gifts for needy children, doing so must contribute to something that would bring her pleasure. She may have a reason for doing so that ultimately appeals to the fact that this will bring joy to children on Christmas, not to the fact that she would enjoy their joy. Although I haven’t offered an argument for this normative view, I think it’s quite plausible and certainly more plausible on its face than the normative claim that the only thing we have reason to do is to promote our own pleasure.

I’ve also denied the normative view that facts about what we take pleasure in can never serve as reasons. Again, however, this should be fine, because a normative theory that denies pleasure any reason-giving role isn’t particularly credible. Some philosophers deny that the fact that I desire to do something provides me with a reason to do it, but these philosophers recognize that pleasure can nevertheless provide reasons.[21] Although it may often be true that I take pleasure in doing something because I see that there are reasons for doing it that are independent of my taking pleasure, this isn’t always the case. For instance, I may lie on a rock in the sun in the middle of a river simply because I would find doing so pleasant, not because of some further consideration about lying in the sun in light of which I take pleasure in doing so. As a normative matter, it seems easy enough to imagine a case in which such a fact justifies my lying in the sun on the rock.

What if we read the pleasure theory of reasons as a meta-normative claim that attempts to give a reductive account of what it is to act for reasons? On the face of it, this shouldn’t stand in the way of drawing the normative distinction that I’ve drawn because one desideratum for a meta-normative theory of reasons is that it be able to account for the normative reasons that we take ourselves upon reflection to have.

The most sophisticated version of a meta-normative pleasure theory of reasons would be one based on the work of Mark Schroeder.[22] Schroeder’s discussion makes clear that holding a meta-normative pleasure theory does not, in itself, require one to deny that an agent may act for reasons that have nothing to do with facts about her own pleasures. One might simply claim that such a fact is necessary to constitute her reason as a reason, but it doesn’t itself serve as one of her reasons.[23] And it should be the ambition of any such theory to account for what Schroeder calls the “strong modal status” of agent-neutral reasons such as the reason Goldstein has for helping the needy children: for anything in particular that Goldstein finds pleasant, she would have a reason to help the needy children even if she didn’t take pleasure in that particular thing.[24]

Discussing this meta-normative account of reasons in any depth lies beyond the scope of my argument. My aim in this section is to make two simple points. First, I wish to be clear that the distinction upon which I am relying rests on a (quite plausible) substantive claim about the kinds of normative reasons that we have. Second, it isn’t obvious that this normative claim is incompatible with any particular meta-normative view about reasons, even one that claims that our reasons are constituted in part by our pleasures.

5. Conclusion

Kant’s figures of the sympathetic helper and the downcast philanthropist lend themselves to a kaleidoscopic variety of interpretations. One traditional portrait of these two figures portrays the sympathetic helper as relatively spontaneous and unreflective, by contrast with the downcast philanthropist who is, finally, forced by his sorrows to a life of reflection and governance by reason. On such interpretations, it is sometimes even difficult to see the sympathetic helper who acts from inclination as acting for any motivating reason at all.[25]

The portrait of these figures that I have been painting, however, is one according to which the sympathetic helper clearly acts for a reason where that reason is something like, “Well, that person needs help.” The sympathetic helper has no further end in helping. But the sympathetic helper does have a further reason for helping. The sympathetic helper is reflectively conscious of her feelings of satisfaction as the source of the noninstrumental value of the ends that she pursues for their own sake; she values the other noninstrumentally for the reason that doing so brings her pleasure, satisfaction or enjoyment. Her relation to moral ends is that of the hobbyist. Far from being a reflective ingénue, the sympathetic helper herself may be in the grip of a philosophical picture that forces such a mediated relationship to the noninstrumental value of others, a picture according to which any noninstrumental reason for action must, for its rational vindication, have a basis in a further reason to do with the agent’s own satisfactions. Although the figure who helps “from duty” has been charged with having “one thought too many,” a right-making reasons account of moral worth that makes use of the notion of a conditional reason flips this charge on its head. On the view I have been sketching, it is the sympathetic helper acting from inclination who, because she takes her own enjoyment as a reason, is herself liable to the charge of having one thought too many.

Acknowledgments

This paper benefited from the insightful and constructive comments of an audience at the Auburn University Philosophical Society and two anonymous Ergo referees. I am also grateful for discussions of this material with Stina Bäckström, Keren Gorodeisky, Arata Hamawaki, and Guy Rohrbaugh. As ever, I am indebted to Tom Lockhart for conversation, philosophical and otherwise.

References

• Anscombe, G. E. M. (2000). Intention. Harvard University Press.
• Aristotle (2000). Nicomachean Ethics (Roger Crisp, Ed. and Trans.). Cambridge University Press. https://doi.org/10.1017/CBO9780511802058
• Arpaly, Nomy (2003). Unprincipled Virtue. Oxford University Press.
• Arpaly, Nomy and Timothy Schroeder (2013). In Praise of Desire. Oxford University Press. https://doi.org/10.1093/acprof:oso/9780199348169.001.0001
• Ebels-Duggan, Kyla (2011). Kantian Ethics. In Christian Miller (Ed.), Continuum Companion to Ethics (168–189). Bloomsbury Academic.
• Gauthier, Jeffrey A. (1997). Schiller’s Critique of Kant’s Moral Psychology: Reconciling Practical Reason and an Ethics of Virtue. Canadian Journal of Philosophy, 27(4), 513–544. https://doi.org/10.1080/00455091.1997.10717484
• Herman, Barbara (1981). On the Value of Acting from the Motive of Duty. The Philosophical Review, 90(3), 359–382. https://doi.org/10.2307/2184978
• Kant, Immanuel (1996). Practical Philosophy (Mary J. Gregor, Ed. and Trans.). Cambridge University Press.
• Korsgaard, Christine (1996). Two Distinctions in Goodness. In Creating the Kingdom of Ends (249–274). Cambridge University Press. https://doi.org/10.1017/CBO9781139174503.010
• Korsgaard, Christine (2008). Kant and Aristotle on Morally Good Action. In The Constitution of Agency (174–206). Oxford University Press. https://doi.org/10.1093/acprof:oso/9780199552733.003.0007
• Markovits, Julia (2010). Acting for the Right Reasons. Philosophical Review, 119(2), 201–242. https://doi.org/10.1215/00318108-2009-037
• Markovits, Julia (2012). Saints, Heroes, Sages, and Villains. Philosophical Studies, 158(2), 289–311. https://doi.org/10.1007/s11098-012-9883-x
• Scanlon, T. M. (2014). Being Realistic about Reasons. Oxford University Press. Schroeder, Mark (2007). Slaves of the Passions. Oxford University Press.
• Sliwa, Paulina (2015a). Moral Worth and Moral Knowledge. Philosophy and Phenomenological Research, 91(3), 1–26.
• Sliwa, Paulina (2015b). Praise Without Perfection: A Dilemma for Right-Making Reasons. American Philosophical Quarterly, 52(2), 171–185.
• Stratton-Lake, Philip (2000). Kant, Duty and Moral Worth. Routledge.

Notes

1. Friedrich Schiller captures what seems alarmingly off-kilter about Kant’s discussion of moral worth with his joke (as cited in Gauthier 1997: 513):

   “Gladly I serve my friends, but alas I do it with pleasure. Hence I am plagued with doubt that I am not a virtuous person.” “Sure, your only resource is to try to despise them entirely, And then with aversion to do what your duty enjoins you.”

   For a discussion of Schiller’s critique of Kant, see Gauthier (1997).

2. One standard interpretation that avoids this difficulty is described by Kyla Ebels-Duggan: “the problem isn’t that [the sympathetically inclined person] enjoys helping, but rather that he takes his enjoyment as his exclusive or overriding reason for action. And that is not what a good person would do; the virtuous person knows that sometimes she ought to help whether she wants to or not” (2011: 171).

3. By “Kant’s distinction”, I mean to point to a distinction between, on the one hand, actions performed from inclination which are lacking in moral worth and, on the other hand, actions motivated in some other way such that they do possess moral worth (what Kant refers to as “from duty”). I don’t mean for the second part of the distinction, acting “from duty,” to pick out cases in which the agent performs the action for reason that “duty requires it.” Rather, I’ll use “from duty” as a catchall phrase to describe actions motivated in such a way that they possess genuine moral worth, whatever that way turns out to be.

4. I won’t here defend a right-making reasons account of moral worth. For a presentation of the appeal of such an account see Markovits (2010). For related accounts of moral worth, see Stratton-Lake (2000), Arpaly (2003), and Arpaly and Schroeder (2013). For an illuminating critical discussion of right-making reasons accounts, see Sliwa (2015b).

5. This way of putting the point is inspired by Korsgaard who says that “the agent who consciously employs the principle of self-love . . . seems to choose beneficence as one might choose a hobby” (2008: 183). Korsgaard doesn’t develop this thought in much depth, because she doesn’t take it to be a promising avenue for analyzing Kant’s distinction. As far as I can tell, Korsgaard interprets Kant’s sympathetic do-gooder as not acting for a reason at all, in the sense that he cannot provide a rational justification in response to the question, why are you doing that? On her view, the person who acts from immediate inclination is simply “moved” (2008: 181) by the thought that the other needs help; he is “allowing his choices to be governed by his natural inclinations, and so is simply following where nature leads” (2008: 184); he acts “impulsively” (2008: 183); he is unreflective and lacks a “justification” for what he does such that his true reason for helping could only be given by “theoretical self-scrutiny” (2008: 183). I will develop Korsgaard’s suggestion that the person who acts from inclination and acts for a reason (not in an impulsive manner as one might act from anger) treats morality as a hobby in a somewhat different direction that she takes it. In her brief mention of hobbies, Korsgaard seems to think of our engagement with them as a matter of being “covertly selfish” or of looking to oneself, as if one’s ultimate purpose in the pursuit of hobbies must be to experience pleasure (2008: 181, 183). I’ll argue, by contrast, that our hobbies can put us in touch with things that we value for their own sake, not for the sake of the pleasure that they bring us.

6. For a further discussion of ignorance and facts about beliefs as motivating reasons, see Markovits (2010: 221–223).

7. Mark Schroeder distinguishes two senses in which we can answer the question why some consideration is a reason for someone to do something. Consider the question: Why is the fact that some poor children have no presents for Christmas a reason for Goldstein to spend the money she won on these children? One sort of answer to this question seeks to provide whatever further justifying reasons there are that speak in favor of taking the fact that the poor children have no presents as a reason. Another sort of answer to this question is merely explanatory and can cite anything that might be relevant to a full explanation of why Goldstein has such a reason. Some of the conditions cited by this second sort of explanation will not themselves be part of the reason that Goldstein has for spending her money on these children. These will be what Schroeder calls background conditions—since they will not function in the foreground of Goldstein’s deliberations. In these terms, in order for some fact to count as a conditional reason in the sense of conditional reason, it isn’t enough for its status as a reason to be conditional upon some fact as a background condition. See Schroeder (2007: Chapter 2).

8. Thinking about the reasons for which Aristotle’s virtuous agent acts is helpful for grasping the general notion of a conditional reason. My suggestion below will be that we can use this general notion to spell out one particular sort of conditional reason that can be employed to draw Kant’s distinction. Kant’s distinction is likely not the only application for conditional reasons when it comes to a right-making reasons account of moral worth.

9. I didn’t countenance this possibility in an earlier draft of this paper and am grateful to an Ergo referee for drawing my attention to it. Although such a reason may seem bizarre or weak, this doesn’t demonstrate that there is no such reason at all. See Schroeder’s argument that negative existential intuitions about reasons may be misleading (2007: Chapter 5).

10. There may be other ways to interpret Nelly’s response that she saved the oak because it brought her satisfaction. For instance, she may be providing a causal explanation of why she acts on the noninstrumental reason she does (not, as I describe, providing a further reason). This is a genuine possibility, one that might plausibly be distinguished from the case as I’m fleshing it out if we suppose that Nelly, in further conversation, insists that in citing her satisfaction she’s just explaining why she saved the tree, not really giving us her reason for doing so. Such a case would be like my citing my drowsiness as an explanation for my steering the conversation towards trivia, not as my reason for doing so. Again, I’m indebted to an Ergo referee for this point. When, however, the situation is described in such a way that Nelly takes herself to be giving a justification when discussing the fact of her satisfaction, I see no reason to dismiss the idea that her motivating reason is hedonically conditional.

11. Since all instrumental reasons are conditional reasons, if a person is engaged in the project of satisfying herself, then her instrumental reasons will be conditional and conditional on facts about her own pleasures. I’ve called these sorts of reasons instrumental hedonic reasons. I’m reserving the term hedonically conditional reason to pick out noninstrumental reasons conditional on the agent’s pleasures.

12. I’m indebted to an Ergo referee for urging me to clarify this point. On the view I’ll articulate in Section 3, whether or not actions performed for reasons that are conditional on further reasons about the agent’s cares, concerns, loves, commitments, values, etc., are lacking in moral worth depends on substantive normative judgments about whether these reasons do indeed morally justify the actions such agents perform.

13. One may disagree with my assessment of this particular case, or of the ones discussed above, but hopefully the examples make clear what is at stake in arguing that any particular reason is or is not a hobbyist sort of reason.

14. In my view, one theoretical benefit of recognizing conditional reasons, and in discussing hedonically conditional reasons in particular, is that it allows for this analysis of hobbies and the sorts of reasons we have for pursing them. Some may not find this account of hobbies compelling. This strikes me as an interesting opening for discussion, since, as far as I am aware, there is almost no contemporary philosophical work on hobbies.

15. See Kant (1996: 4:398): "For the maxim lacks moral content, namely that of doing such actions not from inclination but from duty.”

16. The claim of this section is that there are actions that are performed for hedonically conditional reasons that are lacking in moral worth. This further condition is, however, strong enough to rule out other actions as lacking in moral worth, depending on the correct substantive account of moral reasons for action. Not all conditional reasons are hedonically conditional reasons. Consider the following example: someone might help another person for that person’s own sake, for the further conditional reason that it doesn’t cost too much to help. Depending on the correct normative judgment about the situation (and what reasons in fact morally justify helping), it could turn out that such an action is lacking in moral worth. Even more closely related to hedonically conditional reasons are reasons conditional upon facts about what the agent would find painful. Under certain conditions, an action might lack moral worth if it is conditional upon such facts. It is one benefit of discussing moral worth in conjunction with conditional reasons that it will allow us to raise these sorts of questions. I won’t attempt to answer them here, and will focus on the case of hedonically conditional reasons. These are enough to demonstrate that Markovits’s account of moral worth isn’t sufficient.

17. Such a person might still have reason to do the right thing. But acting on this remaining reason would be acting for a different reason than the reason that was hedonically conditional.

18. See Markovits’s argument at (2010: 211–215).

19. My discussion here mirrors Scanlon’s treatment of the “desire theory” of reasons. See Scanlon (2014: 3). Since I’m not considering noninstrumental reasons conditional on our desires but on our pleasures, I’ve modified the discussion to reflect this point.

20. Compare this view to what Scanlon says about the problems with a normative desire theory: “Having a desire to do something typically involves thinking, or imagining, that there is something to be said for doing it. And from the point of view of an agent it is the considerations that are seen as desirable, rather than the fact of having the desire, that provide reasons for acting” (2014: 88). I don’t mean to suggest that Scanlon would endorse the view that there are no hedonic reasons. He acknowledges that the fact that one would enjoy doing something can be a reason for one to do it. Sometimes “that it would be pleasant to do so” is itself a consideration that speaks in favor doing something and serves to explain why one desires to do it. See Scanlon (2014: 47–50).

21. Again, see Scanlon (2014: 47–50). Anscombe points out that a person who simply says, “I want a saucer of mud” is likely to be asked what for, and such a person won’t be able to provide a satisfying answer to this question by appealing merely to the fact that he wants it. Instead he needs to characterize the object as “somehow desirable” (2000: 70–71). Yet Anscombe also claims that “‘It’s pleasant’ is an adequate answer to ‘What’s the good of it?’ or ‘What do you want that for?’” (2000: 78).

22. Schroeder suggests that one possible incarnation of the Humean theory of reasons would explain all reasons in this way in terms of pleasure (2007: 3).

23. Such facts about the agents own pleasures could serve as background conditions on her having the reasons she does, not as themselves further reasons for actions. See Footnote 7.

24. See Schroeder (2007: 106). This ambition could be fulfilled in the way that Schroeder suggests if the actions in question were such that they contributed to anything that a person could possibly take pleasure in. Such agent-neutral reasons would apply to a person only insofar as she experienced some pleasure, but would be independent of the particular sources of her pleasure. Compare to Schroeder’s discussion at (2007: 108–117).

25. See the discussion of Korsgaard’s interpretation of these two figures in Footnote 5.