1. Kant’s ‘Concept of an Object of Pure Practical Reason’ and the End Independence of the Moral Law

As already noted, Kant’s remarks concerning the concept of an object of pure practical reason are what will allow me first to review the end independence of the moral law and then, in section 2, explain the physical conditions independence of the moral law. I thus begin by reviewing what little Kant says about these concepts of objects and the broader context surrounding his remarks.

Kant introduces us to the concept of an object of pure practical reason in the second chapter of the Critique of Practical Reason. He defines a concept of an object of practical reason as ‘the representation of an object as an effect possible through freedom’ (1996a: 186; KpV, 5: 57) and from this definition two points follow. First, we can speak more straightforwardly, as objects of pure practical reason are just those goals or ends that an agent freely decides to pursue or avoid. Second, we can make matters even easier by speaking simply of the objects or ends of pure practical reason, as opposed to the concept of an end or object of pure practical reason. Given the above definition and Kant’s tendency in this chapter to refer to both concepts of objects and objects without there being any discernible difference or attempt to differentiate between the two sets of terms, we can say that the former refer to the subjective grasp we have of our ends while the latter indicate their objective referent. And since this distinction is not relevant to an account of what these ends are, I will when discussing them use the terms ‘objects’ and ‘concepts of objects’ interchangeably.Footnote ²

Now while the language of ends is hardly foreign to Kant’s writings on morality, the account that follows Kant’s definition of the ends or objects of pure practical reason is by no means immediately intelligible. He proceeds to tell us these objects or ends are – somehow – given to us directly by the faculty of pure practical reason and its governing principle, the principle of autonomy. This principle is responsible for determining two objects for the will: ‘the good’ (das Gute), which is the object that pure practical reason demands that we pursue, and ‘the evil’ (das Böse), which is the object that pure practical reason demands that we avoid (1996a: 186; KpV, 5: 58). After laying out these points, Kant then spends the pages that follow explaining why the concepts of good and evil should not be confused with the sensibly grounded concepts of well-being (das Wohl) and woe (das Übel). Since the principle of autonomy does not trek in this empirical territory, it is not surprising that the concepts of objects that follow from this principle are also not at home here. But it remains unclear what purpose this discussion of good and evil as the objects of practical reason serves.

The most immediate response, I think, would be to argue that the concepts of good and evil are meant to organize the proposals we are inclined to act on into two general categories, namely those that pass and those that fail the permissibility test laid out by the principle of morality. Since Kant in the first Critique describes concepts as a means for organizing the intuitions provided by our senses, this description would not be terminologically baseless. But such a role for the objects of pure practical reason would be both unnecessary and uninteresting. If it is the principle of morality that rules out a proposal to act on some desire, then the deliberating agent has no real need for the concepts of good and evil, of something that ought or ought not to be the object of the will. To claim, then, that the good and the evil are a kind of shorthand for referring to the two possible results of applying the principle of morality is to render these objects superfluous and boring. Adding to the mystery is how these objects fit with Kant’s overall plan in the second Critique. As noted previously, he claims that his inquiry into pure practical reason must ‘begin with principles and then proceed to concepts, and only then, where possible, from them to the senses’ (1996a: 149; KpV, 5: 16). Thus we also need to understand both what it means to proceed from these concepts to the senses and why we must do so only ‘where possible’ (wo möglich).

The proposal I will now begin to explore is that we can fully understand these ends or concepts of objects, and how these concepts relate to the senses, only by embracing the theses of end independence and physical conditions independence.

Both the second Critique and the Metaphysics of Morals offer evidence in favour of the end independence of the moral law. We can establish what position Kant takes in the second Critique regarding end independence simply by noting the following comment, which is by far the most unambiguous of his sparse references to end independence. The comment comes at the conclusion of his discussion of the concepts of the objects of pure practical reason, and it concerns the distinction between these concepts and the twelve pure concepts of the understanding discussed in the first Critique. The latter, he reminds us, can provide objects of knowledge only if they are given content by our spatio-temporal intuitions. Concerning the former, however, he writes:

Now one could try to read Kant as claiming simply that the directives of pure practical reason need not make any appeal to our capacity for pleasure in order to motivate us to follow them. But such a reading would ignore the clear evidence that Kant is not interested in motivational independence here. Rather, his focus is on the divergent conditions under which theoretical and practical concepts can provide our intellects and wills with concrete, meaningful objects. To say that the principle of freedom ‘at once’ (sogleich) provides concepts that ‘produce the reality’ (hervorbringen … die Wirklichkeit) of a certain disposition of the will is just to affirm end independence, or the idea that the purpose of these concepts is to direct the will’s attention to specific ends or goals and to do so without input from or regard for the particular desires and maxims of any given individual. And though he neglects to do so in the context surrounding this passage, Kant’s Tugendlehre can help us understand how he would argue for this notion of end independence.

While one of pure practical reason’s responsibilities is to place a formal requirement on our maxims, Kant argues that ‘ethics goes beyond this and provides a matter (an object of free choice), an end of pure reason which it represents as an end that is also objectively necessary’ (1996c: 513; MS, 6: 380). In other words, an agent no longer fulfils her duty simply when she acts only on those maxims that are ruled permissible by the universalizability test; rather, the moral law requires that an agent have a specific kind of end, namely an end that is also a duty. Kant explains:

Now it is not immediately clear why, without such ends, a Sittenlehre or ‘doctrine of morals’ would be impossible. The impossibility of morality would not seem to follow simply from the fact that the only ends an agent could consider acting on would be those that follow from her sensuous disposition. After all, an agent who is entertaining the prospect of suicide on grounds of self-love can use the criterion of universalizability to strike down any maxim that would warrant acting on a desire of this kind (cf. Potter Reference Potter1998: 103). It is also not the case that all commentators think this passage should occasion a discussion of much besides the motivational independence of the moral law. Nelson Potter, for example, thinks that ends that are also duties are simply ends that are sourced from sensibility but performed for the right reason; as such, the moral law is not providing ends that are independent of the senses but rather changing how we regard and pursue those ends that stem from our desires (cf. Potter Reference Potter1998: 114, n. 6).

Both of these concerns can be addressed by explaining how Kant argues for end independence in this passage, a task that has not been neglected by commentators. In fact, Korsgaard’s early work on Kant’s ethics accurately enough reconstructs the main thrust of his reasoning, even though I think there are some shortcomings in this account (and even though a central aim of this paper is to highlight deep problems with her later constructivist reading of Kant, in which the possibility of end independence has disappearedFootnote ³). In ‘Morality as Freedom’ Korsgaard reads Kant as claiming that, in order to withstand the temptations of self-love that are ineliminable in human life, there must be an end that does not stem from our desires.Footnote ⁴ Since our capacity to thwart inclination is deeply enhanced when we have some positive goal that such resistance furthers, Kant needs to establish the possibility of the end independence of the moral law; without this, our capacity for motivational independence would be threatened, as would, for example, the ability to strike down maxims of suicide based on their failure to be universalizable. And in this case, the very possibility of the categorical imperative would also be threatened.

This reconstruction is largely correct, but it is not without shortcomings. First, because it works by stressing a particular feature of human agency, it might lead to the incorrect conclusion that a struggle with self-love and the possibility of failure are integral features of subjection to the moral law. As such, it might seem as though end independence has a kind of instrumental value, that it is only important to have ends independent of sensibility because we need to shore up motivational independence. But this would get matters wrong or backwards in at least two important regards: it implies, first, that end independence is somehow non-essential and, second, that features of human agency explain the conditions and mechanics of agency more generally.

Against this second implication, we must note that Kant occasionally but consistently cites instances of non-human rational agency in his moral writings.Footnote ⁵ Divine agents such as God and angels lack a sensible constitution and thus never experience susceptibility to inclination; their version of agency cannot be characterized by a capacity to enjoy motivational independence. Thus we can begin to highlight the importance of end independence by showing how it lets us understand how non-human agents participate in moral life – divine agents act morally by having certain ends, a capacity that exists despite the fact that such agents do not experience sensible desires that could provide the occasion to neglect these ends. Recognizing this point can also help us begin to see that, in human life, end independence likely cannot have a merely instrumental role. That is, we should resist making end independence a mere footnote to motivational independence. We should instead embrace the idea that human beings can have ends that in no way originate in sensible desires, and ends whose content and moral significance is irreducible to how they strengthen our efforts to resist impermissible desires or maxims.

Second, these points about both a human being’s vulnerability to temptation and Kant’s embrace of divine agency should make us consider other possible regards in which the moral law can assert its independence from sensibility. Though she does not phrase it this way, Korsgaard’s argument rests on the idea that if the only ends we could pursue were those ends that derive from our sensible embodiment, then sensibility would be afforded a kind of currency or voice that would ultimately undermine the motivational independence of the moral law. If this is correct, then we must at least consider the prospect that the same argument holds mutatis mutandis for physical conditions independence. If the moral law must at any point defer to the conditions of sensibility, and if moral ends can only be obligatory to the extent that they can be realized in the empirical world, this scenario might be as damaging to the possibility of the categorical imperative as is a scenario in which the only source of ends is our sensible desires.

Finally, if in fact we want to take Kant’s remarks about divine agency seriously, then we will need to continue filling out the conditions for the possibility of such agency. We have already established that this involves temporarily removing from view motivational independence and affirming that such agents can have ends even though they do not have sensible desires. And insofar as a divine agent such as God does not have a body and thus does not act in our spatio-temporal world, it seems that divine action must be action that neither takes place in nor is subject to physical conditions. To validate the possibility of physical conditions independence for both divine and, I will argue, human agents, we turn to the next section of this paper.

2. The Physical Conditions Independence of the Moral Law

I will now explore physical conditions independence and demonstrate the extent to which Kant frees the moral law from sensible constraints. Crucial to my account will be Kant’s claim that we must only proceed where possible from our concepts of the objects of pure practical reason to the senses, as I argue that this statement summarizes a view about how our pursuit and avoidance of the good and the bad in no way depend on the physical possibility of making these objects real.

To understand Kant’s view, we can begin by quickly reminding ourselves that there are two possible determining grounds of the will, or, two possible ‘fundamental maxims’Footnote ⁶ that an agent can choose between. A person can decide to have a will that is pure and guided by the moral law, or she can choose to have a will that is empirically grounded in self-love and the demand for sensible pleasure. Now, the choice of my fundamental maxim will determine not only whether I will be open to pursuing the ends provided by the principle of morality but also whether the ends I pursue will be subject to natural constraints. The little attention that Kant devotes to the objects of practical reason speaks to this very point. Concerning empirical practical reason he claims that:

And concerning pure practical reason he claims that:

To explain the divergent conditions that regulate these two types of practical reason, we should consider each remark carefully, starting with the latter explanation of the conditions governing the objects of pure practical reason. When Kant asks what conditions the adoption of such goals must meet, he distinguishes between two requirements: first, there is the possibility or impossibility of willing an action, and second, there is the question of our ability to make an object empirically real. Being an empirical question, the latter has no bearing on our assessment of what can count as a goal for an agent whose will is pure: as he says, a judgement concerning whether something is an object of pure practical reason is ‘quite independent’ (ganz unabhängig) of our ‘physical ability’ (physichen Vermögen) to make it real. What should be included under the heading of our physical abilities is left rather vague; we might wonder whether Kant considers irrelevant an individual person’s physical ability (e.g. a given person’s paralysis), the physical capacities that are typical of a species (e.g. the inability of human beings to jump higher than a few feet) or even what is possible within a system of nature (e.g. the ability to suspend oneself mid-air without any assistance). However, that Kant does not entertain arguments about whether physical possibility matters at some fairly specific level or a more general one seems good reason to think his view is that such possibility is überhaupt a non-issue for determining the objects of pure practical reason. Thus it is only the ‘moral possibility’ (moralische Möglichkeit) of an object – that is, the ability of an agent with a pure disposition to simply will the object – that allows some object to count as an object of pure practical reason.

We can contrast with these comments those Kant made concerning the objects of empirically grounded practical reason. His treatment of the conditions regulating the pursuit of the objects an individual seeks when she has made self-love and an interest in sensible pleasure her overarching or fundamental maxim is, once again, unfortunately brief. Yet it is clear that Kant does not regard the ability to will such an object as a condition that can be divorced from the possibility that the object can be made real in the natural world. In other words, he appears to claim not only that both these requirements must be met with respect to the objects of empirical practical reason but also that these requirements are somehow inseparable when they concern these objects. Prior to setting as my goal the satisfaction of a particular desire, Kant says, I must have first inquired into the potential to actually achieve this satisfaction: ‘the physical possibility of it [i.e. the object] by the free use of our powers must precede our appraisal of whether it is an object of practical reason or not’.

Now although we are making progress in determining exactly what Kant is saying in the passages at hand, the explanation for these claims is far from clear. We first need to figure out Kant’s argument for the claim that empirical practical reason has an interest in the physical realizability of its objects that is lacking for pure practical reason. That is, why can the conditions of empirical realizability and willability not be distinguished when what motivates us is the expectation of pleasure? Second, with respect to the objects of pure practical reason, we need to ask why Kant thinks these objects can be willed even if they cannot be made real in the spatio-temporal world. Can we make sense of the idea that I can adopt a goal whose empirical achievement is not merely unlikely but simply unattainable?Footnote ⁷ Neither question, we will see, poses problems for Kant.

With respect to the initial question, we can begin by attending to Kant’s description of how practical reason functions when it is motivated by the desire for sensible pleasure. First, a cornerstone of Kant’s moral theory is the claim that action directed towards the attainment of such pleasure is not deterministic. An agent is governed by gratification only under the condition that she has made self-love her fundamental maxim. Second, when an individual is governed by self-love, we must note that the determining ground of her will is not pleasure tout court but rather the expectation of pleasure.Footnote ⁸ The drive for pleasure thus does not manifest itself simply as a sort of blind compulsion either towards or away from certain objects, with the will seeking out whatever falls under the heading of ‘pleasurable’. Rather, our interest in pleasure necessarily directs us to make empirically grounded assessments of actions based on whether they can reasonably be expected to provide satisfaction. Now it is uncertain how intelligent I am in my deliberations about what will bring pleasure, and it is of course also the case that I may end up completely disappointed with the results of those actions I take in the name of pleasure. I might conclude that baking is what will most likely bring me pleasure but later find that acquiring this skill has brought me no satisfaction. But, from the deliberative standpoint that precedes the determination of the will, what is relevant is the empirically grounded expectation that baking will bring me pleasure. And since it is this rational expectation that allows me to set learning to bake as my end, it runs contrary to empirically grounded practical reason to suggest objects that cannot be made materially real. A person whose fundamental maxim is self-love will be motivationally unable to pursue objects for which this possibility is entirely precluded; hence Kant’s claim that in these cases an appraisal of the physical possibility of an object conditions its ability to be an object of the will. Or, in other words, we could say that on Kant’s view empirically grounded reason (that is, reason grounded in the pursuit of pleasure) goes part and parcel with empirical exercises of reason (that is, reason orientated towards bringing into the natural world the objects of the will).

For pure practical reason, however, physical possibility is a non-issue. Pure reason directs the will’s attention to certain objects by means of its foundational principle – the principle of autonomy – and determination of the will in accordance with this principle does not require that we seek only those objects that can be made empirically real. To begin to understand why, we can cite a point from my earlier discussion of end independence. Kant’s understanding of this moral principle as one that governs both human beings and divine agents that lack a sensible constitution makes it, on the one hand, rather intuitive that there should be objects of the will that are not the hoped-for products of empirical activity. Furthermore, in Kant’s discussions of maxims that fail to satisfy the categorical imperative, a concern with empirical realizability is not alluded to. What makes a maxim fit or unfit from a moral point of view is the state of the agent that results from willing it, not whether the end embedded in the maxim is one that proves empirically unrealizable. There are several cases in which a maxim drives the will into conflict. It might do so by virtue of some internal contradictions within a given maxim, as is the case with the ‘contradictions in thought’ that Kant discusses in the first two examples within Groundwork, II (suicide and false promises). Conflict also ensues either when a proposed end is at odds with the goals and maxims we have already adopted, or, Kant often suggests, when this end proves incompatible with our ‘true needs’ or the interests that are appropriate to rational beings such as ourselves.Footnote ⁹ Finally, conflict or ‘moral impossibility’ arises when the adoption of a maxim is driven by a doctrine of self-exceptionalism that is incompatible with the universality demanded by the principle of autonomy. Pure practical reason requires that we avoid these sources of intrapersonal conflict, but Kant offers no indication that pursuit of an empirically unrealizable end necessarily pushes the will into such conflict. Thus while the constraint to pursue only those objects that can be made empirically real follows directly from a consideration of a will grounded in self-love, this constraint does not appear to emerge when we consider what is required by a will that takes the moral law and the principle of autonomy as its determining ground.

We can also note that if empirical realizability is considered a moral requirement, then Kant would be left vulnerable to one of the alleged dissemblances (Verstellungen) that Hegel outlined in the Phenomenology of Spirit. For Hegel, Kant is unable to decide whether the mark of a morally worthy action is having pure duty as one’s motive or securing a settled, empirical success. Kant’s emphasis on the purity of the will pushes him to see the former as the characteristic of morally good action, but Hegel claims that, since Kant also thinks action must seek an accomplishment (Vollbringung) in nature,Footnote ¹⁰ Kantian morality vacillates between two accounts of what characterizes a morally right action. However, if my thesis regarding physical conditions independence is correct, then Kant is quite clear that our moral disposition can be meaningfully engaged even when we are not trying to make ends empirically real. In other words, if material realizability is not a constraint on the adoption of ends, then the alleged dissemblance vanishes from view.

But despite this initial defence, more should be done to illustrate and defend my claim that a will without materially real objects can be said both to act and to act morally. I now turn my attention to these tasks.

3. Possible Criticisms and Further Discussion of Physical Conditions Independence

To begin this section, it might help to note another argument levelled against Kant by Hegel. In the Phenomenology Hegel discusses the figure of the ‘Beautiful Soul’, an individual so content with moral ideals that he never engages in any attempt to translate them into action. One might worry that if there exist objects of pure practical reason not underwritten by potential empirical realizability, then a person can maintain his commitment to these without ever treating them as realizable or trying to approximate them in the phenomenal realm. In other words, it might seem that once we let go of the requirement of physical realizability we will also let go of any claim we have on agents to tangibly pursue – and not merely intellectually contemplate and admire – the objects of pure practical reason.

However, this charge can be avoided if we note that, even though divergent conditions constrain the exercise of empirical and pure practical reason, they are compatible in at least the following sense. The categorical imperative need not comply with empirical reason’s interest in securing a materially realizable end, but there is no reason why we should assume that pure practical reason is at all hostile to the task of making objects real in this manner. Simply put, what is important to Kant is that pure practical reason has a say in our ends and that we then let these ends speak for themselves. If the right way to relate to a given end is to commit ourselves fully to finding the means for its empirical realization, then we ought to do so. If a given end requires that we attend to it in the manner that a divine agent would – namely by having a proper regard for it and understanding of its moral significance – then we ought to do so. If an end demands that we acknowledge it as an empirically unattainable ideal but nonetheless strive to pattern our behaviour after it, then we ought to do so.Footnote ¹¹

That the will or the faculty of desire can take a range of different practical stances in regard to its objects has been at least alluded to by other Kantian commentators. For example, when discussing how feeling figures into Kant’s theory of action and motivation, Jeanine Grenberg briefly notes, but does not argue for, the possibility of a will that has ends or objects but that is in no way attempting to bring about their physical realization. For Grenberg, a will cannot count as determined to action unless it makes some ‘movement’ towards its objects, but this movement ‘could be anything from a mere turning of attention to some object or state of affairs to a wish for the existence of the object, to actual movement toward achieving an object of desire’ (Grenberg Reference Grenberg2001: 163). Now such liberation from natural constraints will inevitably lead to intense debates about the best way to attend to some end. A case in point is the post-Occupy Wall Street discussion concerning how, and even whether, one should politically advocate for ideals concerning income equality, as we may recall the movement’s insistence that we occupy everything and demand nothing. But in addition to affirming the latitude agents should have when deciding how to respond to something of value, such debates nicely demonstrate the surprisingly diverse philosophy of action that Kant endorses: while certain ends or objects of pure practical reason may be compatible with their empirical realization, action does not require bodily movement, action might not take place in the physical world and action does not need to be governed by an imperative to achieve a material accomplishment. The will can be determined to act, and rightly described as acting, both when an agent is engaged in a task that she thinks can never be brought to full empirical fruition and when an agent does the difficult work of turning her attentionFootnote ¹² to an object that is best respected by abandoning any attempt to make that object empirically real. It is only in light of this nuanced account of a good will and its independence from physical constraints that we can explain why we must move on only ‘where possible’ from the objects of pure practical reason to the senses.

These points explain why we are not left in the position of the Beautiful Soul, and they also explain why physical conditions independence does not prevent us from acknowledging that certain (but by no means all) ends in fact demand empirical realization. For instance, the highest good requires the promotion of virtue and happiness, with happiness being not mere Stoic contentment but concrete sensible gratification or enjoyment.Footnote ¹³ As such, even though the highest good is not grounded in self-love or a fundamental commitment to maximal pleasure, attentiveness to physical conditions figures appropriately into the deliberations and articles of belief of an agent as she attends to the highest good. Furthermore, consider Kant’s brief discussion in the third Critique of how a good person will take an interest in natural beauty because such occurrences can be read as offering a ‘trace’ or ‘sign’ that nature’s products can correspond to our intentions (2000: 180; KU, 5: 300). This point likewise fits with my account insofar as I am not trying to argue that there is any necessary opposition between the moral possibility and empirical realizability of some end.Footnote ¹⁴ Again, since there will always be cases where we may or ought to bring about the empirical realization of our ends, we will continue to attend to how nature can provide an ‘incidental confirmation’ (beiläufige Bestätigung) that the ‘ideas created by pure practical reason’ can be made natural ends (2000: 301; KU, 4: 445).Footnote ¹⁵

Finally, we must clarify how physical conditions independence can be reconciled with remarks from Kant that might seem especially at odds with my view. Chief among these is another element of the third Critique, particularly Kant’s claim that ‘to will something and to have satisfaction in its existence, i.e., to take an interest in it, are identical’ (Kant Reference Kant2000: 94; KU, 5: 209), a remark that is reiterated at other points in the text and especially in the First Moment of the judgement of taste, including at 5: 204, 205 and 207. Passages along these lines, particularly with their references to existence (Dasein in 5: 207 and 209; Existenz in 5: 204 and 205), make it seem as though the will, insofar as it is practically engaged and not enjoying the disinterestedness of aesthetic experience, is invariably committed to seeking the material reality of its object or ends.

However, further attention to these passages shows that they do not pose any difficulties for my reading of Kant.Footnote ¹⁶ Before I address them, it will help to note that Kant’s moral writings in no way suggest that the concept of an interest (Interesse) should lead us to think of any necessary connection between willing and having as one’s aim a materially existing end. In the second Critique, an interest is defined as ‘an incentive of the will insofar as it is represented by reason’ (1996a: 204; KpV, 5: 79), a definition that allows Kant to highlight how the human will can be determined to action in a sense-free manner and by means of moral concepts alone. This focus is not abandoned in those sections of the third Critique where interests are discussed, and in fact matters of when and how judgements are guided by concepts are the central concern of these passages. Kant does not investigate how willability relates to the empirical realizability of objects, and this can be seen from even a brief review of these passages.

Kant’s references to interests in the third Critique occur primarily in sections 2–4 of the First Moment, and these sections are devoted to differentiating the satisfaction that belongs to judgements of taste from the satisfaction that belongs to both the agreeable and to the good, whether the good is the useful or the morally good. Now in the second section, which is concerned with judgements of taste, Kant does at least linger over the idea that such judgements are disinterested insofar as they are indifferent to, say, the material existence of a palace that I have judged as beautiful. And concerning judgements of agreeability in the third section, Kant does make comments that are suggestive of an interest in materially real objects, as in this case he claims that pleasing sensations always create a ‘desire for more objects of the same sort’ (2000: 92; KU, 5: 207). But the following three points must be noted.

First, no parallel case about interests and physically existing objects is made in Kant’s discussion of moral engagement with some object. In the case of moral judgements about the good, the sole point Kant stresses is that we can find something esteemed only insofar as we rely on a determinate concept to guide our wills, an emphasis which fits quite well with the second Critique definition of interest. Kant does not require or even come close to arguing for the idea that we can only be guided by a determinate concept if we are using that concept to bring about an empirically real end or state of affairs. Moreover, although I have argued that empirically grounded reason necessarily attends to the empirical realizability of ends, this point is not relevant to the technical definition of an interest and it does not factor into Kant’s explanation in the third Critique of why pleasure in the useful is ‘interested’ in the technical sense. As Kant makes clear in the fourth section of the First Moment, what makes us interested with respect to the useful is simply the fact that the useful pleases insofar as determinate concepts are needed to discern the instrumental purposes for which something is good or useful. In short, then, Kant’s text gives us little reason to think that he intends to abandon the definition of interest in the second Critique and replace it with an account whereby I can never be interested in something unless I am aiming to make a thing empirically real.

Second, Kant’s remarks about interests and existence do not make allusion to empirically or materially real objects. He never qualifies Objekt or Gegenstand in any manner that would lead us to think that he has any such narrow range of possible objects in mind, and insofar as I have already shown that there can be objects of the will that are not empirically situated, we have no cause to make this narrowing move on Kant’s behalf. But what is perhaps more noteworthy about Kant’s language is that, in the fourth section of the First Moment, his two remarks about interests and existence at 5: 207 and 209 offer an opportunity to clarify his earlier comments. While we have already noted how the latter points out that ‘to will something and to have satisfaction in its existence, i.e. to take an interest in it, are identical’, the former reads:

At 5: 209, the ‘something’ Kant refers to is the generic etwas, which might weakly support my claim that he does not attempt to restrict to what is materially realizable what can count as an object of the will and what we can be interested in. The evidence at 5: 207 is more suggestive, though, insofar as Kant makes it quite clear that we are interested insofar as the will is orientated towards and able to find satisfaction not only in existing objects but also Handlungen. This point, which recalls Kant’s few remarks about interests in the Groundwork,Footnote ¹⁷ meshes quite well with the emphasis in this paper on how we need a broader account of the different possibilities for moral engagement, an account which does not assume that we can only act in pursuit of or be interested in empirically plausible ends.

Third, while I have pointed out that Kant’s remarks about interest in the morally good do not include any claims about empirically real existence, with a bit of reconstructive work his account can actually clarify how the ends of pure practical reason can be said to exist even when they are not empirically viable. Paul Guyer’s discussion of interest in the third Critique notes that the beautiful can provide pleasure simply insofar as we simply perceive, for example, a palace; the mere representation of it delights. There are a few reasons why this is the case and why it makes some sense to characterize such experiences as indifferent to the existence of the palace: the pleasure we feel is not the product of those empirical causal laws that explain how material objects and our perceptions of such objects behave, the material existence of the palace does not figure into aesthetic judgements regarding its beauty, and in formulating an aesthetic judgement we are not concerned about the implications – moral, utilitarian, economic, etc. – of the palace beyond how it pleases.Footnote ¹⁸ However, to point out in contrast to these claims that attention to the good involves attention to its existence is not to say that moral law only commands the pursuit of empirically realizable ends. Rather, it is to stress that the good provokes from an agent a kind of attention and engagement not called for by aesthetic experience. When an agent determines her ends under the concept of the good, she creates connective ties between her and her ends that go well beyond the mere contemplation of aesthetic delight.Footnote ¹⁹ In actively attending to the moral law, her own virtue and struggle towards moral progress,Footnote ²⁰ and to questions regarding what sort of practical stance does best justice to the objects of her will, an agent becomes tied to something demanding and something beyond herself. In other words, she becomes bound to something real, something that has Dasein or Existenz – just not something that is necessarily empirically real or physically possible.

4. Conclusions and Implications

In arguing that the demand for empirically realizable ends is not one that runs throughout Kant’s practical philosophy, I have essentially argued that there is an important divide between Kant’s theories of empirical and pure practical reason. This result will likely come as a disappointment to some Kantians, particularly those who find much to admire in the recent work of constructivist thinkers like Christine Korsgaard who, in the words of Barbara Herman, uses a Kantian theory of empirical practical reason as a kind of guiding thread or a way to ‘bootstrap’ ourselves into moral theory (Herman Reference Herman2007: 154). But, as should now be clear, this paper casts doubt on the viability of constructivism, both in terms of its general strategy and its particular execution.

Considered generally, I have defended Kant’s view that when practical reason is grounded in a commitment to pleasure we must seek only empirically realizable ends, but when reason is governed only by the principle of autonomy, this constraint disappears. In other words, an analysis of pure practical reason can proceed only by attending to its unique determining ground, and the workings of the empirical practical reason cannot be used to expose the workings of the pure practical reason. Yet a constructivist approach, at least in the form advocated by Korsgaard in texts such as Self-Constitution (2009), is contingent upon the existence of just this connection. Both on her direct reading of Kant and in her independent work that is meant to serve as an extension of his account, the requirements of pure practical reason are generated by way of the exercise of empirical practical reason: when an agent determines herself to pursue any end whatsoever – even an end that is on its face contrary to moral concerns – that agent will see that if she attends to the practical requirements that underwrite the successful pursuit of that end, then she will ultimately have to embrace moral principles.

I have offered, however, an alternative to this approach by revealing the plausibility of a view whereby the formal requirements governing practical activity differ depending on the substantive, fundamental maxim that underlies an agent’s pursuits. Here, the question of whether a given action is rational cannot be answered unless we first know what kind of end an agent is seeking and whether this end is given to us by our sensible desires or by the principle of morality. For example, while it may initially seem that pursuing an empirically unrealizable end is contrary to reason, we have learned that this is not the case with respect to certain ends, such as the objects of pure practical reason. In other words, what appears to be an intuitive and uncontroversial requirement of practical reason may not be so upon inspection.

More particularly, my account also poses problems for the most recent incarnation of Korsgaardian constructivism. Although I can only outline in the broadest details the challenge that my reading of Kant creates, it rests on the fact that the overarching argument advanced in Self-Constitution treats a commitment to the empirical realizability – or as she puts it, the ‘efficacy’ – of one’s ends as a feature of practical reason that holds regardless of its determining ground. Korsgaard’s strategy in this work is one she compares explicitly to Plato’s in the Republic, as it involves examining several different types of lives or practical guises and then comparing them, unfavourably, to a life governed by the moral law. What allows her to assert the practical shortcomings of these various guises – such as the democratic soul who is led by the theory that the life of a wanton is best, or the timocratic soul who is more interested in the maintenance of his moral ideals than translating them into materially effective, concrete action – is their inability to have an efficacious strategy for the empirical realization of their goals. Efficacy is in Korsgaard’s work a requirement that must govern the pursuit of any end whatsoever, and this requirement also, on her view, is connected necessarily to the more explicitly moral requirement that we act autonomously.Footnote ²¹ But if the argument of this paper is correct, Korsgaard cannot claim the historical Kant as the forerunner of these constructivist arguments in Self-Constitution.