2. Kant on Happiness: Abandoning Natural Well-Being

Remarks on happiness, generally negative, appear early and often in Kant's body of work. However, the fact that Kant's remarks are mostly negative in character is not indicative of a shallow engagement with the issues involved, familiar caricatures of Kant as arch-rationalist to the contrary notwithstanding. Kant's model of happiness is far from a straw man.

In the first Critique, Kant describes happiness as “the satisfaction of all our inclinations” (Kant, Reference Kant, Guyer and Wood1998, A806/B834) and Kant develops this same basic definition across various later works. In the Groundwork, Kant lists “[p]ower, riches, honor, even health” as examples of inclinations we might want satisfied in our pursuit of happiness (Kant, Reference Kant, Gregor and Timmermann2012, 4: 393). Importantly, though, Kant concludes this list of familiar components of well-being with a non-specific inclination: “entire well-being and contentment with one's condition” (Kant, Reference Kant, Gregor and Timmermann2012, 4: 393, emphasis added). Kant leans into the last item on this list in both the second Critique, where he calls happiness “a rational being's consciousness of the agreeableness of life uninterruptedly accompanying his whole existence” (Kant, Reference Kant and Gregor2015, 5: 22, emphasis added) and The Metaphysics of Morals, where he calls happiness not only the “greatest” but the “whole desire in [an agent's] life,” and characterizes this as “continuous” and “complete” (Kant, Reference Kant, Denis and Gregor2017, 6: 480).

Clearly, Kant is not working with the shallowest possible conception of happiness. The straw man we might have expected would perhaps have simply followed every fleeting whim, perpetually in the present and happy only insofar as he is immediately gratified. But this is not what Kant suggests. First, even the specific separate inclinations Kant lists in the Groundwork (power, riches, honour, health) do not speak to a conception of happiness as particularly present gratification — the freedom to indulge one's whims that comes with status, money, and an able body would be more obviously immediate. Second, and more significant, Kant does not call happiness the satisfaction of any one or the other of these specific separate inclinations, but rather “the unification of all ends that are given to us by our inclinations into [a] single end” (Kant, Reference Kant, Guyer and Wood1998, A800/B828). For Kant, happiness involves inclinations rationally sorted into an ideal system of greatest extent, intensity, and duration (Kant, Reference Kant, Guyer and Wood1998, A806/B834) — it is something neither immediate nor simple.

And yet, Kant chooses to disregard concerns for happiness even so rationally construed. As early on in his body of work as he discusses happiness, we find Kant offering a moral psychological remark for thinking that happiness should not receive any special place in our conception of well-being: “[h]appiness alone is far from the complete good for our reason. Reason does not approve of it (however much inclination may wish for it) where it is not united with the worthiness to be happy, i.e., with morally good conduct” (Kant, Reference Kant, Guyer and Wood1998, A813/B841). Importantly, he goes on to admit that “morality alone … is also far from being the complete good” (Kant, Reference Kant, Guyer and Wood1998, A813/B841). These reflections lead Kant to posit a “highest good” of “virtue and happiness together” (Kant, Reference Kant and Gregor2015, 5: 110). This notion of the highest good is, properly speaking, Kant's full conception of well-being: well-being, according to Kant, has both a natural, empirical component (happiness) and a moral component (morally good conduct, worthiness to be happy). However, he predictably emphasizes the fact that, within this broader conception, the moral aspect remains “supreme” (Kant, Reference Kant and Gregor2015, 5: 110). This is what I refer to as the ‘moral shift’ in Kant's thinking about well-being.

Before considering this moral shift in more detail, the following terminological clarifications will help bring into relief what we have covered so far. According to Kant, a typical, subject-relative, prudential account of well-being is but one component of a two-part highest good; the second part is a moral one, and this satisfies our moral psychological need to feel worthy of happiness in order to be doing well. Within the Kantian framework, then (and thus the remainder of this article), ‘well-being’ is a general term referring to Kant's ‘highest good’ version of that concept; ‘natural’ and ‘moral well-being’ refer to those two parts of general well-being; and ‘happiness’ is synonymous with natural well-being. One more distinction needs to be made: I said that Kant betrays the moral shift in his thinking about well-being when he calls moral well-being the ‘supreme’ aspect of the highest good — this is because, in so exalting moral well-being, he is, I think, distinguishing between our concern for our well-being and our duties regarding our well-being, and asserting that the latter overrides the former. This last distinction will be made clearer as we continue, but for now we have what we need to understand the moral shift in Kant's thinking about well-being.

Against happiness-focused conceptions of well-being, Kant objects that even rationally informed, systematic conceptions of natural well-being remain problematically indeterminate. The first aspect of Kant's indeterminacy objection to the helpfulness of happiness for thinking about well-being has to do with the inter- and intra-personal inconsistency of ideas of happiness. Disagreement in matters of happiness is always possible, and thus happiness is an unhelpful target for practical reason. Kant explains this objection in the second Critique:

Happiness is essentially subject-relative: though all agents want it, each agent wants something different in wanting it. Indeed, even a single agent may want different things at different times, though all the while wanting only happiness (Kant, Reference Kant, Gregor and Timmermann2012, 4: 418). Admittedly, we probably see no problem making such an indeterminate concept an important aspect of our well-being — after all, we already saw Sumner require a certain subjectivity in any plausible account of well-being. We might, then, simply call this a non-problem for those unsympathetic to Kant's own highly objectivist leanings.

However, Kant does think that he has a more philosophically rigorous reason for calling our conceptions of happiness problematic. In the Groundwork, the indeterminacy objection is described as an instance of the general Kantian problem of empirical uncertainty: in calculating the course of action that would bring us the greatest happiness, we “only act according to empirical counsels … which experience teaches on average advance well-being most” (Kant, Reference Kant, Gregor and Timmermann2012, 4: 418). The probabilistic language reminds us of Kant's Hume-inspired scepticism about empirical principles: the ever-elusive nature of empirical ‘certainty’ is what leads Kant to call problems of happiness “completely insoluble” (Kant, Reference Kant, Gregor and Timmermann2012, 4: 418). In this light, the objection may seem less like a personal hang-up of Kant's: if the indeterminacy objection is a more general point about the limits of knowledge gleaned from experience, we might be more willing to grant it some credibility as a reason to be sceptical of the helpfulness of conceptions of happiness for thinking about well-being.

Admittedly, though, even this more principled line of argument will likely persuade only those already sympathetic to Kantian rationalism; the rest of us are unlikely to abandon concern for our natural well-being just because Kant has pointed out that this is empirically based. This is why I think that conceiving of Kant's moral shift as involving a distinction he makes between concerns for and duties regarding well-being is so helpful. Recall that Kant thinks the counsels of prudence are merely contingent, whereas the laws of morality are categorical. The Kantian point that moral principles command more strictly than empirical ones is familiar, and we can understand in light of this what Kant is doing in making the moral shift in his thinking about well-being. If there are any duties regarding well-being, then these will override any mere concerns for well-being; and, according to Kant, there are duties regarding specifically moral well-being; therefore, moral well-being overrides natural well-being. Understood against the backdrop of Kantian rationalism, this assertion of the supremacy of moral well-being becomes an instance of the ‘command trumps counsel’ remark we saw in the introduction to this article. Even if we remain unconvinced of the need for this moral shift, we can at least understand it — it is, after all, very typically Kantian. Moving forward, then, we should bear in mind that, when Kant makes prescriptions about first-personal well-being, he is referring to duties regarding this.Footnote ¹

Perhaps surprisingly, in light of this moral shift, Kant thinks that natural well-being figures in first-personal moral well-being in a novel way: Kant thinks that we have an indirect duty to secure our happiness for the sake of our moral well-being. This notion first shows up in the Groundwork: “[t]o secure one's own happiness is one's duty (at least indirectly); for lack of contentment with one's condition … could easily become a great temptation to transgress one's duties” (Kant, Reference Kant, Gregor and Timmermann2012, 4: 399). In the second Critique Kant clarifies a positive aspect of this argument: we have an indirect duty to secure our happiness “because happiness (to which belong skill, health, wealth) contains means for the fulfillment of one's duty” (Kant, Reference Kant and Gregor2015, 5: 93). A helpful explanation of the indirectness of the duty to secure one's own happiness comes in The Metaphysics of Morals, where Kant explains that, in the cases he has in mind, “the end is not the subject's happiness but his morality” (Kant, Reference Kant, Denis and Gregor2017, 6: 388). Morality is the ultimate aim of appropriate Kantian concerns for happiness, and thus remains the supreme element of well-being.

To summarize: from Kant's duties-first perspective on well-being, natural well-being matters only indirectly, only insofar as it influences our moral well-being. At this point, the easy move for commentators is to bite the familiar Kantian bullet of harsh objectivism. Jens Timmermann, for instance, remarks that, though an indirect duty to secure one's own happiness completes “a strong and articulate conception of what makes a distinctly human, rational life go best …, a Kantian theory of the ‘good human life’ … would be highly objectivist” (Timmerman, Reference Timmermann2006a, p. 528, emphasis added). Timmermann continues: “[t]here is nothing individual or personal about this ideal,” as it ultimately has everything to do with agents’ rational moral activity (Timmerman, Reference Timmermann2006a, p. 528). The language of ‘conception’ and ‘theory’ is helpful: though Kant's full conception of well-being has a significant subject-relative component, his theory of how we ought to pursue our own well-being does not lean into this prudential dimension. So much for Sumner's intuition about subject-relative well-being!

It is perhaps not obvious, though, that all agents will require exactly the same things to be able to function well morally, to have moral well-being. Could this be an in-road for individuality in moral well-being? In the next section, I will consider this possibility.

3. Indirect Duties and Individuality: Subject-Relativity in Moral Well-Being

Alice Pinheiro Walla (Reference Pinheiro Walla2015) has argued that there are cases in which the indirect duty to secure one's own happiness becomes direct. In arguing for this possibility, I think that Pinheiro Walla is unpacking an extreme case of the intuition that different agents might need different things to be morally well.

Before addressing Pinheiro Walla's argument, let me say a little more about Kantian indirect duties. In quoting Kant's explanation of the indirectness of the duty to secure one's happiness above, I suggested the most obvious interpretation of indirect duties: these are indirect because they ultimately aim at the end of moral well-being (Kant, Reference Kant, Denis and Gregor2017, 6: 388; Timmermann, Reference Timmermann2006b). Examples of other indirect duties include those to conscience, pleasure at moral behaviour, sympathy, sensitivities to animals and nature, and other related ‘moral endowments.’Footnote ² Pinheiro Walla leans into the language of moral endowments in noting that indirect duties “involve some natural feature of human beings that we can neither create nor manipulate at will, although these natural features may allow cultivation” (Pinheiro Walla, Reference Pinheiro Walla2015, p. 35). Kant himself confirms that “[t]o have these predispositions cannot be considered a duty” in any straightforward sense because, when they are had, they are had naturally (Kant, Reference Kant, Denis and Gregor2017, 6: 399). To be clear, Pinheiro Walla agrees that these moral predispositions support moral well-being; she disagrees, however, that this is the fact in virtue of which they get their status as indirect duties.Footnote ³

Pinheiro Walla's specific interpretation of the indirectness of indirect duties will explain why she emphasizes the importance of the moral endowments for indirect duties. On Pinheiro Walla's telling, indirect duties are duties to do things that cannot be done from duty in the special Kantian sense (Pinheiro Walla, Reference Pinheiro Walla2015, p. 34). In the Groundwork, Kant distinguishes between an action's being done in conformity with duty and its being done from duty (Kant, Reference Kant, Gregor and Timmermann2012, 4: 398). Kant's most famous illustration of this distinction is the case of the naturally sympathetic “friend of humanity” who finds “an inner gratification” in beneficence, and acts beneficently so as to feel that moral pleasure (Kant, Reference Kant, Gregor and Timmermann2012, 4: 398, 4: 460–461). Though beneficent actions so motivated conform with duty, they are not done from duty; moral endowments — such as sympathy — are inclinations “on the same footing as other inclinations” (Kant, Reference Kant, Gregor and Timmermann2012, 4: 398), and any action done from inclination is not done from duty. This is why Pinheiro Walla flags the importance of the moral endowments for understanding indirect duties: indirect duties are indirect precisely because they involve moral endowments, which, as inclinations, provide motivation over and above the motive of duty.

In light of this, the remark in the Groundwork that to secure one's own happiness is at least an indirect duty (Kant, Reference Kant, Gregor and Timmermann2012, 4: 399) is important to Pinheiro Walla. The duty to secure our happiness can be indirect, and most of the time it is, because Kant thinks that we are naturally inclined to secure our happiness (Kant, Reference Kant, Gregor and Timmermann2012, 4: 399) — i.e., most of the time we secure our happiness in conformity with, but not from, duty. We can, however, imagine cases in which we do not want to secure our happiness, especially where this is construed as radically long-term the way we saw Kant characterize happiness above. Pinheiro Walla focuses on Kant's example of the gout sufferer, who can reasonably prefer short-term enjoyment even if it makes long-term well-being — or even longer life — less likely. In such a case, the gout sufferer's duty to secure his own (long-term) happiness becomes direct: he has to do it for the sake of his moral well-being, only now, in the absence of a supporting inclination, it must be done just from duty. Pinheiro Walla's analysis of the gout sufferer's case is similar to the way Kant treats cases of suicide. Obviously, the argument is similar — sometimes we need to care for ourselves and prolong our lives just from duty, whether we want to or not. But there is another, more subtle similarity: Kant analyzes suicide as the “annihilat[ion of] the subject of morality in one's own person,” a “root[ing] out [of] the existence of morality itself from the world, as far as one can” (Kant, Reference Kant, Denis and Gregor2017, 6: 422–423); likewise, if we completely disregard our natural well-being, we might lose the very possibility of moral well-being. So, the indirect duty to secure our own happiness is particularly important for our moral well-being because it is in a certain sense the foundation of any moral well-being — and, importantly, there are situations in which this duty can come to the forefront as a direct duty.

Importantly, Pinheiro Walla thinks that this is the only case in which duties regarding well-being are so altered: “the indirect duty to secure one's happiness differs from the other indirect duties” in that it sometimes becomes direct (Pinheiro Walla, Reference Pinheiro Walla2015, p. 37). This is perhaps surprising — after all, the other indirect duties are supposed to be significant for moral well-being, too, and Pinheiro Walla acknowledges this. Pinheiro Walla may well be right that the other indirect duties never become direct the way the duty to secure one's own happiness does; however, this does not mean that the other indirect duties in service of moral well-being are never reprioritized for agent-specific reasons.

Kant himself offers a remark to suggest this idea in a discussion about sensitivity to animals. Kant imagines that the butcher's profession “dulls his shared feeling of [animals’] suffering and so weakens and gradually uproots a natural predisposition that is very serviceable to morality” (Kant, Reference Kant, Denis and Gregor2017, 6: 443). Earlier, we saw Kant say that agents should not be blamed for not possessing moral endowments: these are natural, and, as Pinheiro Walla emphasized, we cannot create these in ourselves at will. However, Kant clearly thinks that we can be held responsible for losing moral endowments: these can be uprooted through desensitization and neglect. Surely, we should expect agent-specific reprioritizations of indirect duties in the face of such threats to moral endowments and, in turn, moral well-being — imagine that Kant's paradigm butcher becomes so desensitized to the sufferings of animals that he begins to lose his sensitivity to the sufferings of persons! It seems appropriate to say that the butcher's occupation makes it so that the duty to develop some sensitivity to animals is more important to him than it is to others. More generally, the butcher may need to be more intentional than most about finding ways to exercise his sympathetic moral endowments — given the specifics of his situation, his personal moral well-being requires this.

Admittedly, the case of the butcher does not seem as problematic as the case of the gout sufferer or the suicide: only a part of the butcher's moral well-being is uprooted, not the very possibility of his moral well-being. Moreover, any time the butcher ‘practices’ being motivated by his sympathetic feelings, his motivation is indirect in Pinheiro Walla's sense — i.e., it is not being done from duty in that special Kantian sense, but rather in conformity with duty from the motivation of his feeling of sympathy, however insignificant his endowment of this may be. However, I think that this is only a problem if we want to stick to Pinheiro Walla's description of agent-specific alterations to duties regarding well-being as occurrences of special direct duties. If we simply want to acknowledge that the point of indirect duties is to draw attention to the importance of the moral endowments for moral well-being, it will be enough to note that the specifics of agents’ situations can alter the priority of these indirect duties insofar as they can significantly affect agents’ moral endowments. The butcher's duty to cultivate his sensitivity to animals does not need to become direct in Pinheiro Walla's sense in order for it to figure more prominently in our conception of what is important for the butcher's moral well-being.

So, even though Kant's theory of first-personal well-being makes duties primary, it is not for this reason ‘highly objectivist’ — at least, not on a fine-grained analysis. Duties regarding our moral well-being, though they always remain supreme, are sensitive to agents’ specific situations and the effects these have on agents’ moral endowments; thus, moral well-being will require different things of different agents in the fulfilment of their indirect duties. Though this may not be respecting agents’ personal concerns for their well-being — their personal ideas of happiness — it is nevertheless respecting their individuality, and this is certainly a more subject-relative aspect of Kantian well-being than we might have expected to find.

There is, however, an aspect of Kantian well-being that can (indeed, must) respect agents’ personal ideas of happiness, too — namely, its other-regarding aspect. I will consider this in the next section.

4. Beneficence, Respect, and the Well-Being of Others

In The Metaphysics of Morals, Kant famously divides duties of virtue into self-regarding and other-regarding duties — more specifically, these are duties aimed at “one's own perfection” and at “the happiness of others” (Kant, Reference Kant, Denis and Gregor2017, 6: 385). Within the first branch of this division, there are duties regarding both one's natural and one's moral perfection (Kant, Reference Kant, Denis and Gregor2017, 6: 386–387, 6: 444–447). Though natural and moral perfection are not the same things as natural and moral well-being,Footnote ⁴ these duties — unsurprisingly, given what we have seen thus far — remain ultimately aimed at moral well-being. But it is not so with duties aimed at the happiness of others. As the return of the language of happiness should signal, other-regarding duties of virtue will be concerned with exactly what Kant has been so keen to reject thus far: personal conceptions of natural well-being, in all their previously problematic indeterminacy.

To understand why Kant does not think that our duties regarding the well-being of others must be as exclusively morally focused as those regarding our own well-being, we will need to look to that all-important Kantian ethical notion of respect. Famously, Kantian respect is due to all agents simply by virtue of their humanity, their rational capacity to autonomously set ends for themselves (see, e.g., Kant, Reference Kant, Gregor and Timmermann2012, 4: 428–429; Kant, Reference Kant, Denis and Gregor2017, 6: 392). “The principle … of the respect [agents] owe one another,” Kant says, tells agents “to keep themselves at a distance from one another” (Kant, Reference Kant, Denis and Gregor2017, 6: 449). To express it colloquially, we are to give others space; more precisely, respect requires that we allow all agents to exercise their capacities to set and pursue their own ends. Because working on moral well-being just means learning to set good ends and to choose good means to these, respect ensures that the project of moral well-being is each agent's own, not to be interfered with. Kant is explicit about this: “[t]he happiness of others includes their moral well-being …, and we have a duty, but only a negative one, to promote this” (Kant, Reference Kant, Denis and Gregor2017, 6: 394Footnote ⁵). “[T]he [moral] perfection of another human being,” Kant says, “consists just in this: that he himself is able to set his end in accordance with his own concepts of duty” (Kant, Reference Kant, Denis and Gregor2017, 6: 386); thus, Kant does not just say that we need not involve ourselves exclusively in the moral aspect of the well-being of others — he forbids that we involve ourselves in this at all.Footnote ⁶

The moral shift having been ruled out for reasons of respect, Kantian beneficence can only work toward the natural well-being of others: to be beneficent is to make “the happiness … of others … one's end” (Kant, Reference Kant, Denis and Gregor2017, 6: 452, emphasis added). If we were not relieved enough to see this all-too-human concept make a return to Kant's theory of well-being, Kant goes on to make an anti-paternalistic clarification: “I cannot do good to anyone in accordance with my concepts of happiness …, thinking to benefit him by forcing a gift upon him; rather, I can benefit him only in accordance with his concepts of happiness” (Kant, Reference Kant, Denis and Gregor2017, 6: 454). This clarification is important to Kant because it ensures that beneficence will not clash with the negative duty of respect discussed above (paternalism in beneficence involves subtly replacing others’ ends with one's own, and thus implies disrespect); but it should be important to us because it ensures that, when thinking about the well-being of others, we are thinking about something essentially up to them. To summarize: when thinking about the well-being of others within a Kantian framework, respect tells us both that we are forbidden from interfering with their moral well-being and that, when we would attempt to further their natural well-being, we must do so according to their own conceptions of happiness. It would seem, then, that Kant's theory of the well-being of others is prudentially informed and decidedly subject-relative — perhaps Kant can do justice to Sumner's intuition after all!

We might be surprised, though, at this apparently un-Kantian result. We might wonder, for instance, why Kant's indeterminacy objections do not crop up again in this discussion of respecting conceptions of happiness. We have seen that respect is due to all agents on account of their being rational end-setters — and the emphasis here has traditionally been on rational. Recall, though, that Kant thinks that conceptions of happiness are unhelpful in thinking about well-being because these are based only on inductive empirical reasoning, which is always merely probabilistic. Why must we respect agents’ conceptions of happiness if we know that these are impurely rational? Clearly, Kant thinks that we must.

I do not think that Kant makes a mistake in requiring us to respect others’ imperfectly rational conceptions of happiness — indeed, if imperfections of rationality warranted withdrawal of respect, then Kantian respect would never get off the ground.

As Michael Cholbi (Reference Cholbi and Timmons2020) has recently observed, the imperfection of human rationality is borne out by much recent empirical research. In light of this, he considers conceptions of Kantian respect that could withstand these “naturalistic headwinds” (Cholbi, p. 182). Cholbi points out that we cannot get equal respect for all persons if we only grant respect to agents insofar as they are successful in exercising their rational agency. If, for example, one agent tends to have a very accurate conception of what would make her happy, but another agent tends to go wrong about this, we would, on a success model, be warranted in respecting the mistaken agent less — but this is unacceptable to the Kantian. Similarly, Cholbi doubts that we could define a satisfactory minimum capacity for rational agency. If set too high, such a criterion might force us not to respect certain cognitively impaired persons’ ideas about what would make them happy, an intuitively unsatisfying result.Footnote ⁷ Cholbi's preferred model of Kantian respect is what he calls aspirational constitutivism: persons are owed respect because they aspire to govern their lives according to reasons they set for themselves. This model satisfactorily downplays the importance of reason in its emphasis on personal reasons. Indeed, even the agent who explicitly rejects rationally ordered conceptions of happiness in favour of spontaneity and immediate gratification is owed respect on this model: in “an attempt to put reason in what the denier views as its proper place,” such an agent is nevertheless setting an end and choosing a means, and thus acting on a reason (Cholbi, Reference Cholbi and Timmons2020, p. 196). On this model, then, even an awareness of the irrationality of agents’ conceptions of happiness does not warrant a withdrawal of respect: we are imperfectly rational in all sorts of ways, but this does not change the fact that we aspire to act according to reasons (even if not reason), and thus we — and our conceptions of happiness — remain worthy of respect.

So, even though Kantian duties regarding the well-being of others are tempered by respect, this does not block their being concerned with others’ imperfectly rational conceptions of happiness — indeed, respect demands that they must be concerned with this empirical element of others’ well-being. Thus, Kant's theory of the well-being of others, too, is far from being ‘highly objectivist’: in respecting the indeterminacy of others’ conceptions of happiness, it respects the subject-relativity of prudential good.